R. v. Charles, 2024 SCC 29

SUPREME COURT OF CANADA
Citation: R. v. Charles, 2024 SCC 29			Appeal Heard: January 18, 2024 Judgment Rendered: September 25, 2024 Docket: 40319
Between:   Yves Caleb Jr. Charles Appellant   and   His Majesty The King Respondent   Official English Translation   Coram: Karakatsanis, Côté, Rowe, Martin, Kasirer, Jamal and Moreau JJ.
Reasons for Judgment: (paras. 1 to 80)	Moreau J. (Karakatsanis, Martin and Jamal JJ. concurring)
Joint Dissenting Reasons: (paras. 81 to 109)	Côté and Kasirer JJ. (Rowe J. concurring)

 

 

Note: This document is subject to editorial revision before its reproduction in final form in the Canada Supreme Court Reports.

 

 

Yves Caleb Jr. Charles                                                                                   Appellant

v.

His Majesty The King                                                                                 Respondent

Indexed as: R. v. Charles

2024 SCC 29

File No.: 40319.

2024: January 18; 2024: September 25.

Present: Karakatsanis, Côté, Rowe, Martin, Kasirer, Jamal and Moreau JJ.

on appeal from the court of appeal for quebec

                    Criminal law — Evidence — Admissibility — Hearsay — Principled exception to hearsay rule — Trial judge admitting witness’s out‑of‑court statement into evidence — Whether statement could be admitted into evidence under principled exception to rule against hearsay.

                    Following an incident involving the accused and the complainant at the school they attended, the accused was charged with assault with a weapon, using an imitation firearm in the commission of an assault, and uttering threats. At trial, the complainant testified that he had entered a washroom at the school and had felt something on his hip while washing his hands. When he turned around, he saw that it was a pistol being held by the accused, used to threaten him. Two other students were present at the time. One of them, whom the Crown called as a witness at the accused’s trial, claimed on being questioned that he had no recollection of the events. The Crown therefore requested a voir dire, seeking to admit into evidence an out‑of‑court statement that had been given by that witness to the police investigators the day after the events. Accompanied at the time by his mother, the witness, who had been arrested and taken into custody in connection with the same incident, was questioned by the police investigators for about an hour and provided a statement in writing. Among other things, the witness admitted in his statement that he was in possession of two pellet pistols. The police conducted a search and recovered the pistols at the witness’s residence, in the location indicated in his statement.

                    The trial judge admitted the witness’s out‑of‑court statement into evidence. He found that the only likely explanation for the statement was its truthfulness as to its material aspects, given the circumstances in which the statement was made and the seizure of pistols, which the trial judge considered to be corroborative evidence. Following the trial, the accused was found guilty of the three counts. The trial judge accepted the complainant’s version of events, which was supported for the most part by a surveillance video and the witness’s statement. A majority of the Court of Appeal upheld the trial judge’s decision to admit the statement into evidence and dismissed the accused’s appeal from his convictions. In addition to being of the view that the trial judge’s conclusion was justified, the majority pointed out the striking similarity between the complainant’s testimony and the witness’s statement, which, according to the majority, tended to confirm that the statement was sufficiently reliable.

                    Held (Côté, Rowe and Kasirer JJ. dissenting): The appeal should be allowed, the convictions quashed and a new trial ordered.

                    Per Karakatsanis, Martin, Jamal and Moreau JJ.: The trial judge erred in finding that the witness’s out‑of‑court statement had the required indicia of reliability and in admitting the statement into evidence at trial. The results of the search subsequently conducted at the witness’s residence do not meet the criteria for corroborative evidence set out in R. v. Bradshaw, 2017 SCC 35, [2017] 1 S.C.R. 865. Moreover, the circumstances surrounding the statement do not support a finding that threshold reliability is established. Finally, the majority of the Court of Appeal should not have relied on the complainant’s testimony, tendered outside of the voir dire, to establish the threshold reliability of the witness’s statement.

                    Hearsay evidence is presumptively inadmissible. Under the principled exception, hearsay can exceptionally be admitted into evidence when the party tendering it demonstrates that the twin criteria of necessity and threshold reliability are met on a balance of probabilities. At the threshold reliability stage, one can only rely on corroborative evidence if it shows, when considered as a whole and in the circumstances of the case, that the only likely explanation for the hearsay statement is the declarant’s truthfulness about, or the accuracy of, the material aspects of the statement. In the absence of a connection between the corroborative evidence and the aspect sought to be proved, the evidence is quite simply of no assistance in determining whether that specific aspect is true or accurate; it merely corroborates the declarant’s credibility, the accused’s guilt or one party’s theory of the case, which is not sufficient. The combined effect of the corroborative evidence and the circumstances of the case, not the evidence taken in isolation, must rule out plausible alternative explanations for the material aspects of the statement. The absence of leading questions, inconsistent statements, promises of benefit or a criminal lifestyle simply points to an absence of factors that, if present, would detract from an otherwise trustworthy statement.

                    In this case, the Crown was required to show that the search results confirmed the accused’s role in the events if it was seeking to use the witness’s statement to establish the accused’s degree of involvement as well as the use of the weapon. There is no connection between the discovery of the pistols and the accused’s degree of involvement. The location of the pistols therefore cannot serve to show that threshold reliability is established through that aspect of the statement alone. Aside from establishing the truth of that aspect of the statement, that evidence is not capable of ruling out plausible alternative explanations for the events.

                    In addition, the witness’s statement raises particular reliability concerns. Because the witness is an accomplice, there is a very real danger that he tried to shift his responsibility onto the accused in his statement. It was to the witness’s advantage to provide an account that limited his participation to possession of the weapons, avoiding the charges that involved a greater degree of participation. In the absence of external evidence confirming that the accused played the primary role in the washroom, the circumstantial guarantees cannot overcome the dangers presented by the witness’s statement. Indeed, the absence of a criminal lifestyle is not at all clear. Moreover, the presence of the witness’s mother is not actually an indicium of reliability, because it is possible that the witness did not want his mother to know his degree of involvement, which could have motivated him to lie. The temporal proximity between the statement and the events is also not a useful factor in assessing the specific danger raised by the statement, namely, that the witness lied. Lastly, the witness’s consultation with counsel does not make it possible to exclude the risk that he tried to minimize his responsibility. Nor are the indicia of procedural reliability reassuring, because the usual substitutes for the traditional safeguards are absent: there was no recording of the statement or the interview that preceded it, the witness was not under oath and was not given a warning by the investigators concerning the need to tell them the truth and the consequences associated with lying, and the defence was deprived of any opportunity to cross‑examine the witness. In short, the combined effect of the corroborative evidence and the circumstances does not overcome the specific hearsay dangers raised by the out‑of‑court statement.

                    As for the complainant’s testimony, that evidence cannot be considered in analyzing the threshold reliability of the witness’s statement. Despite the fact that the complainant testified before the trial judge rendered his decision on the voir dire, his testimony was not part of the voir dire. On appeal, the appropriate mechanism for considering the complainant’s testimony that was not formally tendered in the voir dire is the curative proviso in s. 686(1)(b)(iii) of the Criminal Code, which cannot be applied in this case because the trial judge’s admission of the statement is not a harmless or trivial error and it is not clear that the evidence pointing to the accused’s guilt is so overwhelming that any other verdict but a conviction would be impossible.

                    Per Côté, Rowe and Kasirer (dissenting): The appeal should be dismissed. There is no reviewable error in the trial judge’s decision to admit the witness’s out‑of‑court statement into evidence under the principled exception to the rule against hearsay. However, there is agreement with the majority that, on appeal, the appropriate mechanism for considering the complainant’s testimony is the curative proviso.

                    The trial judge made no reviewable error in taking the discovery of the weapons into account as corroborative evidence in the assessment of threshold reliability. Given the charges laid and the burden they entailed for the prosecution, there is indeed a logical connection between the aspect of the statement pertaining to the presence of the weapon in the washroom, corroborated by the discovery of the weapon at the witness’s residence, and the aspect of the statement pertaining to the handling of that same weapon by the accused at the same time and in the same location. Such a logical connection makes it possible, in the assessment of threshold reliability, to consider corroborative evidence that does not relate to all of the material aspects of a statement. Threshold reliability can be established through the combined effect of the corroborative evidence and the circumstances that constitute indicia of reliability, as those circumstances make remedying the insufficiency of the corroborative evidence possible.

                    In the absence of an error in principle that tainted the analysis at first instance during the consideration of certain circumstances in the assessment of threshold reliability, the trial judge’s decision was owed deference by the Court of Appeal. The relevance of the circumstances depends on the specific dangers associated with the hearsay in question and thus on the facts of the case. It would be wrong to categorize, objectively and independently of the facts of the case, the circumstances that are neutral or secondary and those that are more important. It is true that circumstances that in essence simply point to an absence of factors that, if present, would detract from an otherwise trustworthy statement do not provide a circumstantial guarantee of trustworthiness. However, those circumstances are relevant. While such circumstances are not sufficient on their own to establish threshold reliability, such circumstances considered in conjunction with others (for example, corroboration even insufficient on its own), may lead to the conclusion that the statement has all the attributes required for acceptable threshold reliability. The fact that the witness spoke with counsel, the fact that his mother accompanied him when he made his statement and the fact that his mother was made aware of the tenor of the witness’s rights are indicia of reliability that the trial judge could validly consider.

Cases Cited

By Moreau J.

                    Applied: R. v. Bradshaw, 2017 SCC 35, [2017] 1 S.C.R. 865; referred to: R. v. Khelawon, 2006 SCC 57, [2006] 2 S.C.R. 787; R. v. Conway (1997), 36 O.R. (3d) 579; R. v. Youvarajah, 2013 SCC 41, [2013] 2 S.C.R. 720; R. v. B. (K.G.), [1993] 1 S.C.R. 740; R. v. Starr, 2000 SCC 40, [2000] 2 S.C.R. 144; R. v. Mohamed, 2023 ONCA 104, 423 C.C.C. (3d) 308; R. v. Couture, 2007 SCC 28, [2007] 2 S.C.R. 517; R. v. L.T.H., 2008 SCC 49, [2008] 2 S.C.R. 739; R. v. Khan, 2001 SCC 86, [2001] 3 S.C.R. 823.

By Côté and Kasirer JJ. (dissenting)

                    R. v. Le, 2019 SCC 34, [2019] 2 S.C.R. 692; R. v. Youvarajah, 2013 SCC 41, [2013] 2 S.C.R. 720; R. v. Couture, 2007 SCC 28, [2007] 2 S.C.R. 517; R. v. Bradshaw, 2017 SCC 35, [2017] 1 S.C.R. 865; R. v. Hall, 2018 MBCA 122, [2019] 1 W.W.R. 612; R. v. Burns, 2016 SKCA 67, 337 C.C.C. (3d) 523; R. v. Allary, 2021 SKCA 110; R. v. U. (F.J.), [1995] 3 S.C.R. 764; R. v. Khelawon, 2006 SCC 57, [2006] 2 S.C.R. 787; R. v. Khan, [1990] 2 S.C.R. 531; R. v. Larue, 2018 YKCA 9, 434 D.L.R. (4th) 155, aff’d 2019 SCC 25, [2019] 2 S.C.R. 398; R. v. Bernard, 2018 ABCA 396, 80 Alta. L.R. (6th) 258; R. v. L.T.H., 2008 SCC 49, [2008] 2 S.C.R. 739.

Statutes and Regulations Cited

Canada Evidence Act, R.S.C. 1985, c. C‑5, s. 9(2).

Criminal Code, R.S.C. 1985, c. C‑46, s. 686(1)(b)(iii).

Authors Cited

Lederman, Sidney N., Michelle K. Fuerst and Hamish C. Stewart. Sopinka, Lederman & Bryant: The Law of Evidence in Canada, 6th ed. Toronto: LexisNexis, 2022.

Paciocco, David M., Palma Paciocco and Lee Stuesser. The Law of Evidence, 8th ed. Toronto: Irwin Law, 2020.

                    APPEAL from a judgment of the Quebec Court of Appeal (Doyon, Cournoyer and Bachand JJ.A.), 2022 QCCA 1013, 82 C.R. (7th) 373, [2022] AZ‑51869771, [2022] J.Q. n^o 7437 (Lexis), 2022 CarswellQue 9888 (WL), affirming the convictions of the accused for assault with a weapon, using an imitation firearm in the commission of an assault and uttering threats. Appeal allowed, Côté, Rowe and Kasirer JJ. dissenting.

                    Emmanuelle Rheault, for the appellant.

                    Marianna Ferraro and Mathieu Locas, for the respondent.

English version of the judgment of Karakatsanis, Martin, Jamal and Moreau JJ. delivered by

                   Moreau J. —

TABLE OF CONTENTS
Paragraph
I.      Overview	1
II.    Facts	6
III.   Decisions of the Courts Below	18
A.    Court of Québec (Judge Dupras)	18
(1)      Voir Dire Ruling	18
(2)      Decision as to Guilt	23
B.    Quebec Court of Appeal, 2022 QCCA 1013, 82 C.R. (7th) 373	27
(1)      Bachand J.A., Dissenting	28
(2)      Doyon and Cournoyer JJ.A.	33
IV.   Issues	39
V.    Analysis	41
A.    Standard of Review	41
B.    General Principles Concerning the Admissibility of Hearsay Evidence	43
C.    Use of the Search Results in the Threshold Reliability Analysis	49
D.    Threshold Reliability	64
E.     Use of the Complainant’s Testimony in the Threshold Reliability Analysis	76
VI.   Conclusion	80

 

 

 

I.               Overview

[1]                             This case concerns the admission into evidence, during the appellant’s trial, of an out‑of‑court statement in writing of a Crown witness, K.A., who claimed on being questioned by the Crown that he had no recollection of the events forming the subject matter of the charges against the appellant. The appellant submits that the trial judge erred in determining that the out‑of‑court statement had the indicia of reliability required by R. v. Khelawon, 2006 SCC 57, [2006] 2 S.C.R. 787, and R. v. Bradshaw, 2017 SCC 35, [2017] 1 S.C.R. 865, for admission into evidence. A majority of the Quebec Court of Appeal upheld the trial judge’s decision to admit the statement into evidence. The dissenting judge would have excluded the statement and ordered a new trial.

[2]                             The central issue to be determined is whether the trial judge erred in finding that the witness’s out‑of‑court statement had the indicia of reliability required according to the principles set out in our jurisprudence. This determination affords us an opportunity to reaffirm the principles laid down in Bradshaw.

[3]                              I agree with the dissenting Court of Appeal judge that the trial judge erred in admitting the witness’s statement in writing into evidence at trial. The results of the search subsequently conducted at the witness’s residence do not meet the Bradshaw criteria for corroborative evidence. Because the Crown sought to use the witness’s statement to establish the appellant’s role in the events, it was required to show that the search results confirmed that aspect of the statement. As for the circumstances surrounding the statement, they do not support a finding that threshold reliability is established.

[4]                             Moreover, the majority of the Court of Appeal should not have relied on the complainant’s testimony, tendered outside of the voir dire, to establish the threshold reliability of K.A.’s statement. On appeal, the appropriate mechanism for considering the complainant’s testimony is the curative proviso in s. 686(1)(b)(iii) of the Criminal Code, R.S.C. 1985, c. C‑46. However, that provision cannot be applied in this case. It is therefore unnecessary to decide the discrete question of the self‑contained nature of the voir dire at the trial stage, especially since the trial judge expressly declined to consider evidence not tendered in the voir dire, by agreement of the parties.

[5]                              I would accordingly allow the appeal, quash the convictions and order a new trial.

II.            Facts

[6]                             The appellant was charged with assault with a weapon, using an imitation firearm in the commission of an assault, and uttering threats.

[7]                             The circumstances at the heart of this case unfolded on February 24, 2016, at the school attended at the time by the appellant, K.A. and the complainant. At the appellant’s trial before a judge alone, the Crown called K.A., the complainant, a police officer and a rehabilitation counsellor who worked at the school. The appellant testified in his own defence.

[8]                             The complainant testified that he had asked the appellant to stop bothering his girlfriend. The appellant initially became angry but subsequently calmed down. Later, the complainant entered a washroom at the school and, while washing his hands, he felt something on his hip. When he turned around, he saw that it was a pistol being held by the appellant. Two other students, K.A. and a person named Fares, were also present. When asked by the complainant whether the pistol [translation] “was a real one”, the appellant replied: “Do you want to see if it’s a real one? I think I have a bullet left in it” (A.R., vol. II, at pp. 305‑6). K.A. and Fares later caught up with the complainant and told him that the firearm was a fake.

[9]                             The day after the events, K.A. was arrested and taken into custody for possession of a firearm for a dangerous purpose, possession of an imitation firearm, carrying a concealed firearm, uttering death threats and assault with a weapon. K.A. was informed of his rights and consulted counsel.

[10]                         Accompanied by his mother, K.A. was questioned by the police investigators for about an hour and provided a statement in writing. In his testimony at the voir dire on the admissibility of K.A.’s statement, the police investigator who had taken notes during the interview could not guarantee that he had taken down all of the questions asked or, for that matter, everything said during the interview. No recording (video or audio) of the interview was made, nor was the statement made under oath. The police investigators did not warn K.A. of the possible consequences of giving a false statement.

[11]                         In his written statement, K.A. admitted that he was in possession of two pellet pistols belonging to Fares. According to K.A., while he was with Fares and the appellant in the washroom, he was asked to give one of the pistols to the appellant. K.A. was not aware of the issues between the appellant and the complainant. The appellant pointed the weapon at the complainant while uttering threats against him, then tried to wipe his prints off the weapon before returning it to K.A. A short time later, K.A. and Fares found the complainant and told him that the appellant was not serious and was only joking.

[12]                         K.A.’s statement also described the pistols and where they were located in his residence, namely, in a drawer of his dresser. The police then conducted a search and recovered the pistols in the location indicated by K.A. K.A. subsequently pleaded guilty to a charge of carrying a weapon for a purpose dangerous to the public peace.

[13]                         When called by the Crown as a witness at the appellant’s trial, K.A. claimed to have no recollection of the events. The Crown then requested a voir dire, seeking to admit into evidence the recording of K.A.’s guilty plea before the Youth Division of the Court of Québec. Since the complainant had already lost part of a work day, the trial judge decided to adjourn the voir dire in order to allow him to testify. When the voir dire resumed, the Crown advised the trial judge of its intention to adduce into evidence as well the out‑of‑court statement given by K.A. to the police investigators on February 25, 2016.

[14]                          The trial judge inquired as to whether all of the trial evidence would be tendered in the voir dire. The parties agreed that the trial judge could consider K.A.’s demeanour while testifying but that the other trial evidence would not be tendered in the voir dire.

[15]                          The trial judge admitted K.A.’s out‑of‑court statement into evidence at the trial, but not his guilty plea.

[16]                          The appellant testified during the trial that on February 24, 2016, K.A. — a person unknown to him — offered to show him an iPad electronic tablet. The appellant later saw K.A. and another person in the washroom. K.A. showed him the iPad and said that he also had something else to show him. K.A. then took out a firearm. The appellant held it for a few seconds before returning it to K.A. The complainant, who was also in the washroom, asked whether the pistol was “a real one”. The appellant answered that it was not; however, the individual with K.A. said that there was still a bullet in it.

[17]                          Following the trial, the appellant was found guilty of the three counts.

III.         Decisions of the Courts Below

A.           Court of Québec (Judge Dupras)

(1)          Voir Dire Ruling

[18]                         At the outset of his reasons, the trial judge noted that the evidence adduced at the trial had not been tendered in the voir dire, except for the evidence concerning K.A.’s demeanour at the trial. The trial judge relied on the Ontario Court of Appeal’s decision in R. v. Conway (1997), 36 O.R. (3d) 579, as well as similar remarks made in Bradshaw, in determining that he had to confine himself to the evidence tendered during the voir dire.

[19]                         The trial judge had no difficulty in finding that necessity — the first criterion for the admissibility of hearsay — was met given that K.A. claimed during the trial to have no recollection of the events.

[20]                          He also found that the criterion of reliability was met, highlighting the following circumstances:

•                         K.A. was read his rights from a form used specifically for minors.

•                         K.A.’s mother was present at all times while he was being questioned and while the statement was being written, and her son’s rights were explained to her.

•                         K.A. consulted counsel before giving his statement.

•                         K.A. admitted, without hesitation, his responsibility in relation to the events.

•                         The police investigators’ questions were not leading.

•                         There was no evidence of inconsistent statements.

•                         K.A. did not have a criminal record and there was no evidence as to moral character, criminal lifestyle, past dishonesty or interest in the outcome of the trial.

•                         K.A. provided his statement the day after the events occurred.

•                         K.A. did not attempt to diminish his criminal responsibility, as he even made assertions that lessened the appellant’s responsibility. This distinguished the situation at hand from those in which an accomplice attempts to evade responsibility by blaming someone else.

•                         There was an intrinsic structure to what was said, and K.A. seemed to adhere to an inherent logic, particularly in minimizing the intent that might be inferred from the appellant’s conduct.

[21]                          The trial judge also considered the seizure of pistols at K.A.’s residence to be corroborative evidence. He noted that those items had been seized by consent, which demonstrated K.A.’s willingness to cooperate fully with the authorities, at the risk of incriminating himself.

[22]                         The trial judge identified the honesty of the declarant as the specific hearsay danger raised. However, given the circumstances and the corroborative evidence, he found that the only likely explanation for the statement was its truthfulness as to its material aspects.

(2)          Decision as to Guilt

[23]                         The trial judge began his analysis by assessing the probative value of K.A.’s statement. He pointed out that there were certain discrepancies between that statement and a surveillance video (“Exhibit P‑6”) which showed the individuals concerned in a hallway beside the washroom. Noting that he had not yet viewed Exhibit P‑6 at the time he assessed threshold reliability, the trial judge found [translation] “that once it is placed in the general context of the evidence as a whole, there are elements of [the statement] whose probative value must be adjusted downward” (para. 45, reproduced in A.R., vol. I, at p. 49).

[24]                         The trial judge concluded that he did not believe the appellant’s version and that it did not raise a reasonable doubt.

[25]                         Relying on Exhibit P‑6 and the complainant’s testimony, as well as the portions of the appellant’s testimony that confirmed the complainant’s testimony, the trial judge found that the Crown had established the appellant’s guilt beyond a reasonable doubt. The probative value of certain passages of K.A.’s statement was compromised, but other passages remained useful. In particular, the statement had been proven truthful with respect to the location of the weapon or weapons. With regard to the [translation] “crux of this case, the criminal use of the weapon by the accused”, the trial judge noted that K.A.’s statement generally supported the complainant’s account (para. 68). When K.A.’s comments were properly situated in the context of the evidence as a whole, they [translation] “compel[led] the recognition of [their] definite probative value” (para. 69). The trial judge accepted the complainant’s version of events, which he considered to be supported for the most part by Exhibit P‑6 and K.A.’s statement.

[26]                         The trial judge therefore convicted the appellant of the three counts and entered a conditional stay of proceedings on the second count.

B.            Quebec Court of Appeal, 2022 QCCA 1013, 82 C.R. (7th) 373

[27]                         Before the Court of Appeal, the appellant argued that the trial judge had erred in considering the location of the pistols as corroborative evidence and in finding that the circumstances surrounding the statement were sufficient to establish threshold reliability. While the dissenting judge would have allowed the appeal on the basis of those arguments, the majority rejected them. The Court of Appeal unanimously rejected a second argument concerning s. 9(2) of the Canada Evidence Act, R.S.C. 1985, c. C‑5, which the appellant did not pursue before this Court.

(1)          Bachand J.A., Dissenting

[28]                          Bachand J.A. found that the search results were not relevant in determining whether the statement was admissible to establish the appellant’s role in the incident. Corroborative evidence may be used in the analysis of threshold reliability only when that evidence, considered as a whole and in the circumstances of the case, shows that the only likely explanation for the statement is the statement’s truthfulness or accuracy regarding its material aspects. In particular, Bachand J.A. cited Karakatsanis J.’s majority reasons in Bradshaw stating that the function of corroborative evidence is to mitigate the need for cross‑examination on the point that the statement is tendered to prove.

[29]                          However, there was no connection between the search results and the issue of whether the appellant had handled the weapon and used it to threaten the complainant. Therefore, if the Crown adduced the statement to prove that fact, the search results could not be used to establish the statement’s admissibility. This remained true even if the Crown also wished to adduce the statement to show that a weapon had been used. The search results could then be used to establish the reliability of the statement in relation to this second aspect, but not in relation to the first.

[30]                         As for the circumstances surrounding the statement, Bachand J.A. determined that they did not provide sufficient guarantees. The trial judge relied on certain circumstances that, in light of the case law, had to be regarded as having very limited weight. K.A. had reasons to lie, since he was suspected of having committed offences in connection with the events of February 24, 2016. The statement [translation] “certainly had the effect of diminishing his own role while attributing responsibility for the offences mainly to the appellant” (para. 45). The trial judge gave little weight to K.A.’s interest in not being identified as the person who had used the pistol and threatened the complainant. It could not reasonably be said that the circumstances of the statement [translation] “compel[led]” the conclusion that cross‑examining K.A. would have added nothing to the process (para. 48).

[31]                         With regard to the complainant’s testimony, Bachand J.A. pointed out (in a footnote) that it was not part of the evidence relied upon for the purposes of the voir dire.

[32]                         Finally, Bachand J.A. found that the curative proviso in s. 686(1)(b)(iii) could not apply.

(2)          Doyon and Cournoyer JJ.A.

[33]                         The majority dismissed the appeal, finding that the dissenting judge was [translation] “unduly pars[ing] the evidence and the jurisprudence”, which resulted in “the negation of the key principle that favours a flexible approach, an assessment made on a case‑by‑case basis, in light of all the circumstances” (para. 53). That assessment [translation] “is first and foremost a matter for the trial judge because the trial judge is in the best position to determine the extent to which the dangers of hearsay evidence are present” (para. 53).

[34]                         The situation differed from that in Bradshaw. Here, K.A. had not made prior inconsistent statements and had participated in a police interview within hours of the incident. His credibility was not damaged at the time of his statement, and the evidence corroborated one of the two material aspects of the statement. The evidence could not be split up to limit corroboration to one aspect: [translation] “At that stage of the trial, the scope of the statement was much broader than that described by [the dissenting judge], and the discovery of a weapon, according to the [trial] judge, corroborated the entire statement” (para. 59), which was not an error. In addition, the trial judge did not rely primarily or exclusively on corroboration to admit the evidence. Even though each factor was insufficient in itself, it was open to the trial judge to consider the factors as a whole in order to find that threshold reliability had been established.

[35]                         The majority summarized the evidence concerning the circumstances as follows, at para. 64 of its reasons:

     [translation] A young person, who had consulted counsel and who understood the gravity of the situation and his duty to tell the truth because of the caution and the presence of a parent, gave a short, very coherent statement shortly after the events. He was accompanied by his mother (in accordance with the Youth Criminal Justice Act to ensure the integrity of the statement and even its reliability: R. v. L.T.H., 2008 SCC 49, [2008] 2 S.C.R. 739, at para. 38) and made a written statement, which he gave and signed without coercion or even any leading questions. In the statement, he completely incriminated himself. In addition, he consented to the police going to his residence to seize the weapons, which demonstrated a genuine desire to cooperate with the police by telling the truth.

[36]                         Considered as a whole, these factors could justify the trial judge’s conclusion.

[37]                         The majority was also of the view that K.A. had not shifted his responsibility onto the appellant’s shoulders. There was nothing to suggest that K.A. had an interest in making a statement that would prevent him from being identified as the person who had used the pistol and uttered threats.

[38]                         Finally, the majority pointed out the striking similarity between the complainant’s testimony and K.A.’s statement. The majority noted that K.A. had not discussed the content of his statement with the complainant, which tended to confirm that the statement was sufficiently reliable.

IV.         Issues

[39]                         The appellant raises only one issue: Did the Quebec Court of Appeal err in law in upholding the decision to admit K.A.’s statement into evidence on the basis of the principled exception to the hearsay rule?

[40]                         The reasons of the majority of the Court of Appeal raise a second issue: Did the majority of the Court of Appeal err in relying on the complainant’s testimony to find that the threshold reliability of K.A.’s statement had been established?

V.           Analysis

A.           Standard of Review

[41]                         Both issues are subject to a correctness standard. The admissibility of hearsay evidence is a question of law. However, as my colleagues in dissent have correctly noted, an appellate court must accord deference to the findings of fact underlying an admissibility ruling. It must also be borne in mind that “a trial judge is well placed to assess the hearsay dangers in a particular case and the effectiveness of any safeguards to assist in overcoming them” (R. v. Youvarajah, 2013 SCC 41, [2013] 2 S.C.R. 720, at para. 31). Thus, “absent an error in principle, the trial judge’s determination of threshold reliability is entitled to deference” (para. 31).

[42]                         The question of whether it was open to the majority of the Court of Appeal to rely on the complainant’s testimony to find that the threshold reliability of K.A.’s statement was established is also a question of law. The correctness standard therefore applies to this question as well.

B.            General Principles Concerning the Admissibility of Hearsay Evidence

[43]                         Hearsay evidence is presumptively inadmissible (see, e.g., Bradshaw, at paras. 1 and 21). Its presumptive inadmissibility is due to the fact that it is often difficult to assess the truth of a statement made outside the courtroom. In Bradshaw, Karakatsanis J. explained that, generally, “hearsay is not taken under oath, the trier of fact cannot observe the declarant’s demeanor as she makes the statement, and hearsay is not tested through cross‑examination” (para. 20). However, “[t]he truth‑seeking process of a trial is predicated on the presentation of evidence in court” (Bradshaw, at para. 19), and “our adversary system is based on the assumption that sources of untrustworthiness or inaccuracy can best be brought to light under the test of cross‑examination” (Khelawon, at para. 48). It is “mainly because of the inability to put hearsay evidence to that test” that such evidence is presumptively inadmissible (Khelawon, at para. 48; see also Bradshaw, at para. 1).

[44]                         The admission of hearsay may therefore “compromise trial fairness and the trial’s truth‑seeking process” (Bradshaw, at para. 20). It is possible that the statement has been “inaccurately recorded, and the trier of fact cannot easily investigate the declarant’s perception, memory, narration, or sincerity” (Bradshaw, at para. 20, citing Khelawon, at para. 2). There is thus a risk that such evidence “may be afforded more weight than it deserves” (Bradshaw, at para. 21, quoting Khelawon, at para. 35).

[45]                         That being said, in some circumstances, hearsay evidence “presents minimal dangers and its exclusion, rather than its admission, would impede accurate fact finding” (Khelawon, at para. 2 (emphasis in original), quoted in Bradshaw, at para. 22). Over time, the case law therefore developed categorical exceptions to the exclusionary rule and, ultimately, a more flexible approach. Under the principled exception, “hearsay can exceptionally be admitted into evidence when the party tendering it demonstrates that the twin criteria of necessity and threshold reliability are met on a balance of probabilities” (Bradshaw, at para. 23, citing Khelawon, at para. 47). To establish the threshold reliability of a statement, a party may demonstrate its procedural or substantive reliability.

[46]                         Procedural reliability is established when there are adequate substitutes for testing the truth and accuracy of the statement “given that the declarant has not ‘state[d] the evidence in court, under oath, and under the scrutiny of contemporaneous cross‑examination’” (Bradshaw, at para. 28, quoting Khelawon, at para. 63). Triers of fact must have “a satisfactory basis . . . to rationally evaluate the truth and accuracy of the hearsay statement” (Bradshaw, at para. 28). Substitutes for the traditional safeguards “include a video recording of the statement, the presence of an oath, and a warning about the consequences of lying” (Bradshaw, at para. 28, citing R. v. B. (K.G.), [1993] 1 S.C.R. 740, at pp. 795‑96). Some form of cross‑examination of the declarant, such as preliminary inquiry testimony, is usually required (Bradshaw, at para. 28).

[47]                         Substantive reliability is established when the statement is inherently trustworthy. To determine whether this is the case, trial judges may consider the circumstances in which the statement was made as well as the evidence that corroborates or conflicts with it. The standard is a high one (Bradshaw, at para. 31). That being said, it is not necessary for reliability to be established with absolute certainty. Rather, judges must be satisfied that the statement is “so reliable that contemporaneous cross‑examination of the declarant would add little if anything to the process” (Khelawon, at para. 49, quoted in Bradshaw, at para. 31). In other words, the evidence must be “sufficiently reliable to overcome the dangers arising from the difficulty of testing it” (Bradshaw, at para. 26, quoting Khelawon, at para. 49). As Karakatsanis J. explained in Bradshaw, at para. 31:

      Substantive reliability is established when the statement “is made under circumstances which substantially negate the possibility that the declarant was untruthful or mistaken” ([R. v. Smith, [1992] 2 S.C.R. 915], at p. 933); “under such circumstances that even a sceptical caution would look upon it as trustworthy” (Khelawon, at para. 62, citing [J. H. Wigmore, Evidence in Trials at Common Law (2nd ed. 1923), vol. III], at p. 154); when the statement is so reliable that it is “unlikely to change under cross‑examination” (Khelawon, at para. 107; Smith, at p. 937); when “there is no real concern about whether the statement is true or not because of the circumstances in which it came about” (Khelawon, at para. 62); when the only likely explanation is that the statement is true ([R. v. U. (F.J.), [1995] 3 S.C.R. 764], at para. 40).

[48]                         In the criminal context, “the threshold reliability analysis has a constitutional dimension because the difficulties of testing hearsay evidence can threaten the accused’s right to a fair trial” (Bradshaw, at para. 24). By ensuring that only hearsay that is necessary and reliable is admitted, “the trial judge acts as an evidentiary gatekeeper. She protects trial fairness and the integrity of the truth‑seeking process” (para. 24).

C.            Use of the Search Results in the Threshold Reliability Analysis

[49]                         Recall that, to determine whether “corroborative evidence is of assistance in the substantive reliability inquiry”, a trial judge should

1.                      identify the material aspects of the hearsay statement that are tendered for their truth;

2.                      identify the specific hearsay dangers raised by those aspects of the statement in the particular circumstances of the case;

3.                      based on the circumstances and these dangers, consider alternative, even speculative, explanations for the statement; and

4.                      determine whether, given the circumstances of the case, the corroborative evidence led at the voir dire rules out these alternative explanations such that the only remaining likely explanation for the statement is the declarant’s truthfulness about, or the accuracy of, the material aspects of the statement.

      (Bradshaw, at para. 57)

[50]                         The Crown argues that the seizure of the pistols at the location specified in the statement constitutes extrinsic corroborative evidence that is relevant to establishing the statement’s substantive reliability. It submits that K.A.’s statement serves to establish two material aspects: (1) the use of the weapon and (2) the appellant’s degree of involvement. For the purposes of analysis, let us assume that this is indeed the case. The Crown urges us to find that, since the search results corroborate the first aspect, they can serve to corroborate the statement in its entirety. According to the Crown, this is because it is not necessary for evidence to corroborate all material aspects of a statement: it is sufficient for it to corroborate just one aspect.

[51]                         As for the appellant, he argues that there is no connection between the discovery of the pistols during the search and the aspect of the statement pertaining to his role in the events and that, as a result, this evidence cannot be used to establish the reliability of the statement.

[52]                         With respect for the contrary view expressed by my colleagues in dissent, there is no connection between the discovery of K.A.’s pistols and the appellant’s degree of involvement. In no way can this evidence confirm that it was the appellant who handled a pistol in the washroom, and not K.A. (despite the fact that it was K.A. who was in possession of the weapons the next day). That aspect of the statement puts in issue not merely the handling of a weapon, but rather the handling of a weapon by the appellant. Consequently, it is necessary to determine how the principles set out in Bradshaw apply where several material aspects of a statement are not connected with each other.

[53]                         I reject the Crown’s arguments on this point, as they conflict with the logic underlying the analytical framework established in Bradshaw. In my view, evidence cannot serve to corroborate the aspects of a statement with which it is not connected, even when the evidence confirms another material aspect of the statement in question.

[54]                         At the threshold reliability stage, “not all evidence that corroborates the declarant’s credibility, the accused’s guilt, or one party’s theory of the case, is of assistance” (Bradshaw, at para. 44). Accordingly, one can “only rely on corroborative evidence . . . if it shows, when considered as a whole and in the circumstances of the case, that the only likely explanation for the hearsay statement is the declarant’s truthfulness about, or the accuracy of, the material aspects of the statement” (para. 44). The function of corroborative evidence is thus to “mitigate the need for cross‑examination, not generally, but on the point that the hearsay is tendered to prove” (para. 45 (emphasis in original)).

[55]                         It is true that it is the combined effect of the corroborative evidence and the circumstances of the case, and not the evidence taken in isolation, that must rule out plausible alternative explanations for the material aspects of the statement (see Bradshaw, at para. 47). However, this does not mitigate the need for a connection between the evidence and the aspect sought to be proved. In the absence of such a connection, the evidence is quite simply of no assistance in determining whether that specific aspect is true or accurate; it merely corroborates the declarant’s credibility, the accused’s guilt or one party’s theory of the case, which is not sufficient (see Bradshaw, at para. 44; see also paras. 45‑46 and 72). Evidence that is not connected with the material aspects of the statement is therefore not capable, even in combination with the circumstances of the case, of ruling out plausible alternative explanations for those aspects.

[56]                         It follows that evidence that confirms one material aspect of a statement is not necessarily admissible to establish the statement’s reliability with respect to its other material aspects. When evidence merely confirms one material aspect of a statement and no more, the support it provides for other material aspects derives entirely from the fact that it boosts the declarant’s credibility. This holds true regardless of the materiality of the aspect of the statement that is confirmed by the evidence. As established in Bradshaw, it is not sufficient for evidence to support the declarant’s credibility generally, and such evidence cannot be used to assess the admissibility of other aspects of the statement.

[57]                         On the other hand, there may conceivably be situations in which several aspects of a statement are connected, such that evidence that demonstrates the truth or accuracy of one aspect is also capable of ruling out the possible explanations for the others. In such a case, the evidence is sufficiently — albeit indirectly — connected with those other aspects. Such evidence may then be of assistance in analyzing the statement’s admissibility with respect to all of those aspects.

[58]                         Contrary to what my colleagues suggest, it is not a matter of adding a step that involves splitting up the evidence to the approach established in Bradshaw. Rather, that approach and its underlying logic are what require a connection between the evidence and each aspect of the statement that it is supposed to confirm. The Bradshaw framework serves to ensure that corroborative evidence is used only in cases where it bears on the aspect sought to be proved by adducing the statement. The requirement for a connection between the corroborative evidence and the aspect in question flows from the role that such evidence must play. Corroborative evidence must make it possible, given the circumstances of the case, to rule out plausible explanations other than the truth or accuracy of the material aspects of the statement (para. 57, point 4).

[59]                         While at first glance it may seem pedantic to deal with each material aspect separately, it is important to note that the Bradshaw criteria are designed to overcome the dangers posed by corroborative evidence. Where a statement does not have indicia of reliability of its own, “then it can add nothing to a case, but it may appear to do so if it is consistent with other evidence. Admitting a hearsay statement only because it is consistent with other evidence treats it as a makeweight: the statement would be added to the other evidence even though its own weight actually depends entirely on that other evidence” (D. M. Paciocco, P. Paciocco and L. Stuesser, The Law of Evidence (8th ed. 2020), at p. 167). Moreover, “it can be difficult to control the length and complexity of the admissibility voir dire if the reliability of the hearsay statement derives from other evidence. Where the search for consistency is taken too far, the admissibility voir dire can easily become a time‑consuming shadow trial” (p. 167).

[60]                         Karakatsanis J. addressed these issues in developing the Bradshaw analytical framework. She explained that in order to preserve “the distinction between threshold and ultimate reliability and to prevent the voir dire from overtaking the trial”, it must be possible to distinguish between evidence that is admissible to establish threshold reliability and evidence that is admissible in the main trial (para. 42). Furthermore, “[l]imiting the use of corroborative evidence as a basis for admitting hearsay also mitigates the risk that inculpatory hearsay will be admitted simply because evidence of the accused’s guilt is strong” (para. 42). Indeed, “[t]he stronger the case against the accused, the easier it would be to admit flawed and unreliable hearsay against him” (para. 42). The particular role of corroborative evidence in the analysis of threshold reliability informs the limitations on its use:

      The limited inquiry into corroborative evidence flows from the fact that, at the threshold reliability stage, corroborative evidence is used in a manner that is qualitatively distinct from the manner in which the trier of fact uses it to assess the statement’s ultimate reliability. As Lederman, Bryant and Fuerst explain, at the threshold reliability stage,

      [t]he use of corroborative evidence should be directed to the reliability of the hearsay. Certain items of evidence can take on a corroborative character and be supportive of the Crown’s theory when considered in the context of the evidence as a whole. Such evidence relates to the merits of the case rather than to the limited focus of the voir dire in assessing the trustworthiness of the statement and is properly left to the ultimate trier of fact.

      (S. N. Lederman, A. W. Bryant and M. K. Fuerst, The Law of Evidence in Canada (4th ed. 2014), at §6.140)

      (Bradshaw, at para. 42)

[61]                         The standard articulated in Bradshaw is the product of a line of jurisprudence that fluctuated between various ways of dealing with corroborative evidence at the threshold reliability stage. At one point, this Court had even banned the use of such evidence (see R. v. Starr, 2000 SCC 40, [2000] 2 S.C.R. 144, at paras. 215 and 217), before holding in 2006 that it can be part of the analysis “in appropriate cases” (see Khelawon, at para. 4). Care must therefore be taken to preserve the balance struck by Bradshaw between the flexibility required by the principled exception and protection against the risks posed by corroborative evidence. The Crown’s position would undermine this balance; it would weaken a foundation of the Bradshaw approach, namely, the need to demonstrate a connection between the corroborative evidence and the material aspects of the statement.

[62]                         In this case, the trial judge erred in finding that the statement’s truth as regards the location of the pistols corroborated the statement in its entirety. There is no connection, even an indirect one, between that evidence and the appellant’s degree of participation. The location of the pistols therefore cannot serve to show that threshold reliability is established through that aspect of the statement alone. Aside from establishing the truth of that aspect of the statement, that evidence is not capable of ruling out plausible alternative explanations for the events. For example, it cannot confirm that it was indeed the appellant and not K.A. who uttered threats against the complainant. As will be seen in the following section, such possible explanations are quite plausible given that it was in K.A.’s interest to exaggerate the appellant’s responsibility.

[63]                         The remaining issue is whether, given the limited role played by the evidence of the location of the pistols, the threshold reliability of the statement has been established.

D.           Threshold Reliability

[64]                         The trial judge erred in principle by considering the location of the pistols to be evidence that corroborated the entire statement. As a result, there is no deference owed to his finding on threshold reliability (see Youvarajah, at para. 31; see also R. v. Mohamed, 2023 ONCA 104, 423 C.C.C. (3d) 308, at para. 37). The Court can therefore conduct its own analysis of this issue.

[65]                         The Crown argues that the trial judge properly considered K.A.’s interest in not being identified as the person who had used the weapon and uttered threats against the complainant. The combined effect of the corroborative evidence and the circumstances overcomes the specific dangers raised by the statement. It was a statement against K.A.’s interest in more than one respect, and he even consented to a search at his residence.

[66]                         I disagree. In my opinion, the majority of the Court of Appeal erred in finding that the threshold reliability of the statement had been established. Like Bachand J.A., I am of the view that the trial judge should not have admitted the statement.

[67]                         K.A.’s statement presents significant dangers. The majority of the Court of Appeal erred in ruling out the possibility that K.A. had an interest in minimizing his role in the events. At the time he made his statement, it was to K.A.’s advantage to provide an account that limited his participation to possession of the weapons. Such an account allowed him to explain the presence of the weapons at his residence while avoiding the charges that involved a greater degree of participation.

[68]                         The appellant correctly points out that because K.A. is an accomplice, his statement raises particular reliability concerns. In Youvarajah, Karakatsanis J. explained that “[c]riminal law is generally and rightfully suspicious of allegations made by a person against an accomplice. It has long been recognized that evidence of one accomplice against another may be motivated by self‑interest and that it is dangerous to rely on such evidence absent other evidence which tends to confirm it” (para. 62). Indeed, “the underlying rationale for the admissibility of admissions as against the party making them falls away when they are sought to be used against a third party” (para. 59). This is especially true when the party making the admissions assumes less responsibility than they attribute to the third party, thus potentially facing less severe consequences than the third party (see Youvarajah, at para. 60; see also Bradshaw, at para. 92).

[69]                         The risk referred to by Karakatsanis J. in Youvarajah is clearly present in this case. Although K.A. incriminated himself to some extent, he portrayed his own conduct as being much less serious than that of the appellant. As the Crown points out, it is true that K.A. described the incident as an immature joke, which tended to minimize the extent of the appellant’s responsibility. However, in doing so, K.A. did not assume any greater responsibility himself; instead, he played down the seriousness of the incident as such. The specific risk in this case is not that K.A. wanted to exaggerate the appellant’s involvement in the abstract, but rather that he did so with the aim of evading his own responsibility. In short, there is a very real danger that K.A. tried to shift his responsibility onto the appellant in his statement. Indeed, as the appellant notes, K.A.’s statement places much of the responsibility on Fares and the appellant.

[70]                         In the absence of external evidence confirming that the appellant played the primary role in the washroom, the circumstantial guarantees cannot overcome the dangers presented by K.A.’s statement.

[71]                         As is the case for the absence of leading questions (see R. v. Couture, 2007 SCC 28, [2007] 2 S.C.R. 517, at paras. 100‑101), the absence of inconsistent statements, promises of benefit or a criminal lifestyle simply points to an absence of factors that, if present, would have detracted from an otherwise trustworthy statement. Nothing in the factual context of this case suggests that these circumstances have a higher degree of trustworthiness in relation to the specific dangers arising from K.A.’s statement.

[72]                         Moreover, some clarifications are required regarding the usefulness of the factors considered by the trial judge in the factual context of the case. The absence of a criminal lifestyle is not at all clear. On the contrary, K.A. indicated that he intended to sell the pistols on the black market with Fares. As for the presence of K.A.’s mother, that circumstance is not actually an indicium of reliability, because it is possible that K.A. did not want his mother to know his degree of involvement, which could have motivated him to lie. Accordingly, it is instead a neutral factor. The temporal proximity between the statement and the events serves to mitigate the risk that K.A. did not have an accurate recollection of the events. However, this factor is not useful in assessing the specific danger raised by the statement, namely, that K.A. had incentive to lie.

[73]                         My colleagues in dissent, citing para. 38 of R. v. L.T.H., 2008 SCC 49, [2008] 2 S.C.R. 739, refer to the fact that K.A. consulted counsel before making his statement. However, since one cannot speculate on the content of the discussions between K.A. and counsel, that consultation does not make it possible to exclude the risk that K.A. tried to minimize his responsibility. This risk is quite different from the reliability issues discussed by this Court at para. 38 of L.T.H., namely, “false confessions by young people inclined to make a statement in order to end the pressure of interrogation or to please an authority figure” and the need to “ensur[e] that any statement given manifests the exercise of free will”. Moreover, in L.T.H., the Court was dealing with s. 146(2)(b) of the Youth Criminal Justice Act, S.C. 2002, c. 1, a provision whose scope extends beyond simply consulting with counsel.

[74]                         Nor are the indicia of procedural reliability reassuring. The usual substitutes for the traditional safeguards are absent. There was no recording of the statement or the interview that preceded it (of which the police investigator had only a limited recollection at the voir dire). K.A. was not under oath, and although he was read his rights, he was not given a warning by the investigators concerning the need to tell them the truth and the consequences associated with lying (see B. (K.G.), at pp. 795‑96). In addition, since K.A. claimed to have no recollection of the events, the defence was deprived of any opportunity to cross‑examine him. Yet some form of cross‑examination is usually required to establish procedural reliability (Bradshaw, at para. 28).

[75]                         In short, the indicia of reliability — whether substantive reliability, procedural reliability or both — do not support the admissibility of K.A.’s out‑of‑court statement. It cannot be said that cross‑examining K.A. at the time he made his statement to the police would have added little if anything to the process. There are many aspects of the statement that, without cross‑examination, remain impossible to verify. For example, there is no way to verify that K.A. was not aware of the issues between the appellant and the complainant or that the pistols belonged to Fares. There remains a real concern regarding the truthfulness of the statement given the opportunity that K.A. had to minimize his responsibility and to exaggerate the appellant’s. The indicia of reliability do not rule out this possibility. Thus, the combined effect of the corroborative evidence and the circumstances does not overcome “the specific hearsay dangers raised by the . . . statement” given by K.A., such that its “only likely explanation . . . is [K.A.’s] truthfulness about, or the accuracy of, the material aspects of the statement” (Bradshaw, at para. 47 (emphasis in original)).

E.            Use of the Complainant’s Testimony in the Threshold Reliability Analysis

[76]                         The Crown also argues that the complainant’s testimony corroborates K.A.’s statement insofar as the appellant’s role is concerned, with the result that threshold reliability is established for that aspect of the statement. As discussed above, despite the fact that the complainant testified before the trial judge rendered his decision on the voir dire, his testimony was not part of the voir dire. In fact, the trial judge specifically asked the parties if all of the trial evidence would be tendered in the voir dire, and the defence refused to consent. The parties agreed that only K.A.’s demeanour during his testimony would be considered. It was therefore clear that the complainant’s testimony would not form part of the threshold reliability analysis. Nevertheless, at the end of its analysis, the majority of the Court of Appeal wrote that the [translation] “striking similarity” between the complainant’s testimony and K.A.’s statement tended to confirm that the statement was sufficiently reliable to be admissible (para. 71). According to the Crown, this Court should also consider that evidence in analyzing threshold reliability, despite the fact that it was not formally tendered in the voir dire.

[77]                         This position must be rejected.

[78]                         On appeal, the only possible avenue for considering the complainant’s testimony is the curative proviso. However, the Crown chose not to invoke that provision before this Court. In any event, as the dissenting judge noted, the trial judge expressly referred to K.A.’s statement in his findings concerning the appellant’s guilt. It therefore cannot be concluded that the trial judge’s admission of the statement was a harmless or trivial error. Nor is it “clear that the evidence pointing to the guilt of the accused is so overwhelming that any other verdict but a conviction would be impossible” (see R. v. Khan, 2001 SCC 86, [2001] 3 S.C.R. 823, at para. 31).

[79]                         As a result, the Court of Appeal could not consider the complainant’s testimony in deciding the appeal, except in the context of the curative proviso. Furthermore, it is not necessary to decide the issue raised by the Crown regarding the self‑contained nature of the voir dire at the trial stage. This issue raises concerns related to the particular function of the voir dire, notably the principle that the voir dire must not “overtak[e] the trial” (see Bradshaw, at para. 42) as well as procedural fairness. Given that the Court of Appeal dealt only in passing with the complainant’s testimony and that the appellant has not really addressed the self‑contained nature of the voir dire in his factum, I will leave consideration of this issue for another day.

VI.         Conclusion

[80]                         For these reasons, I would allow the appeal, quash the convictions and order a new trial.

                   English version of the reasons of Côté, Rowe and Kasirer JJ. delivered by

                   Côté and Kasirer JJ. —

I.               Overview

[81]                         We have had the benefit of reading the reasons of our colleague Moreau J. With respect for our colleague’s views, and substantially for the reasons of the majority of the Court of Appeal, Doyon and Cournoyer JJ.A., we are of the opinion that the appeal should be dismissed (2022 QCCA 1013, 82 C.R. (7th) 373). We agree with the majority of the Court of Appeal that there is no reviewable error in the trial judge’s decision to admit K.A.’s out‑of‑court statement into evidence under the principled exception to the rule against hearsay.

[82]                         In support of his appeal to this Court, the appellant invites us to adopt the position of the dissenting judge at the Court of Appeal on this issue. The dissenting judge found that the trial judge erred in relying on evidence that only partially corroborated the material aspects of K.A.’s statement to assess the admissibility of the statement as a whole. The dissenting judge also held that the trial judge committed an error in principle in considering circumstances which had only very relative weight.

[83]                         The respondent adopts the explanations provided by the majority of the Court of Appeal on these two points. The respondent also notes the following comment made by the majority in the second last paragraph of its reasons: [translation] “Finally, it is difficult to ignore the striking similarity between the victim’s testimony and the statement by K.A., who obviously did not discuss the content of his statement with the victim, which tends to confirm that the hearsay evidence was sufficiently reliable to be admissible” (para. 71). According to the respondent, the Court of Appeal could validly consider the complainant’s testimony at trial, which was given before the closing of the voir dire but was not produced on it.

[84]                         We agree with the respondent that the two grounds raised by the appellant are without merit. However, on the question of the self‑contained nature of the voir dire, we agree with the conclusion of our colleague Moreau J. First, the curative proviso is the appropriate mechanism for considering, on appeal, evidence admitted at trial that was not produced on the voir dire, like the complainant’s testimony in this case (para. 4). Second, it is not necessary for this Court to address, in the context of this appeal, the distinct question of the self‑contained nature of the voir dire at the trial stage (at para. 79), especially since the appellant himself rightly recognizes that the majority’s comment was an obiter dictum, made as a passing remark, regarding a question that was not raised at first instance.

II.            Analysis

[85]                         Before we begin the analysis, it is appropriate to identify the applicable standard of review, a point on which the parties agree. In itself, the question of whether a statement may be introduced into evidence under the principled exception to the rule against hearsay is a question of law, which is reviewable on a standard of correctness on appeal (see R. v. Le, 2019 SCC 34, [2019] 2 S.C.R. 692, at para. 23; R. v. Youvarajah, 2013 SCC 41, [2013] 2 S.C.R. 720, at para. 31). However, the findings of fact made by the trial judge are entitled to deference. Similarly, “absent an error in principle, the trial judge’s determination of threshold reliability is entitled to deference” (Youvarajah, at para. 31; see also R. v. Couture, 2007 SCC 28, [2007] 2 S.C.R. 517, at para. 81).

[86]                         Having articulated the applicable standard of review, we must now examine the parties’ respective arguments.

A.           Consideration of the Search Results in the Assessment of Threshold Reliability

[87]                         Relying in particular on R. v. Bradshaw, 2017 SCC 35, [2017] 1 S.C.R. 865, the appellant argues that the search results cannot, in this case, constitute evidence which is relevant to the substantive reliability inquiry. He agrees with the dissenting judge that the trial judge erred in considering the search results in the assessment of threshold reliability because that evidence [translation] “was irrelevant to the threshold reliability analysis with respect to K.A.’s assertion as to the appellant’s conduct and comments during the incident in the washroom” (C.A. reasons, at para. 35).

[88]                         We disagree.

[89]                         The appellant was charged with assault with a weapon, using an imitation firearm in the commission of an assault and uttering threats. Doyon and Cournoyer JJ.A. rightly pointed out that at the time of the voir dire on the admissibility of K.A.’s statement, the appellant had not yet testified on the merits and nothing had been admitted. The Crown therefore had to prove all of the essential elements of each offence beyond a reasonable doubt.

[90]                         The majority of the Court of Appeal was thus correct to point out that it [translation] “was . . . necessary to demonstrate both the appellant’s participation and the use of a weapon” and that, “[b]y demonstrating the presence of a weapon at the witness’s residence, the evidence confirmed its existence and corroborated the witness’s account of the appellant’s use of a weapon, a weapon that he himself had given him” (para. 58). In other words, since the handling of a weapon necessarily implies the presence of a weapon, it follows that proof of the handling of the weapon by the appellant requires proof of the presence of the weapon. Given the charges laid and the burden they entailed for the prosecution, there is indeed a logical connection between the aspect of the statement pertaining to the presence of the weapon in the washroom, corroborated by the discovery of the weapon at K.A.’s residence, and the aspect of the statement pertaining to the handling of that same weapon by the appellant at the same time and in the same location. While K.A.’s statement does evince the aforementioned material yet distinct aspects, they are logically intertwined.

[91]                         Such a logical connection makes it possible, in the assessment of threshold reliability, to consider corroborative evidence that does not relate to all of the material aspects of a statement. Evidence of this nature generally cannot suffice, on its own, to establish threshold reliability. Our Court has previously held that this threshold can then be met through the combined effect of the corroborative evidence and the circumstances that constitute indicia of reliability (see Bradshaw, at paras. 44, 47 and 57).

[92]                         The Manitoba Court of Appeal was therefore correct in concluding in R. v. Hall, 2018 MBCA 122, [2019] 1 W.W.R. 612, that “the . . . submission that substantive reliability requires the existence of corroborative evidence on all material aspects of the hearsay statement must be rejected” (para. 85). In such a case, threshold substantive reliability may be established by the combined effect of the corroborative evidence and the circumstances that constitute indicia of reliability, as those circumstances make remedying the insufficiency of the corroborative evidence possible (see Hall, at paras. 82‑85; R. v. Burns, 2016 SKCA 67, 337 C.C.C. (3d) 523, at para. 30; R. v. Allary, 2021 SKCA 110). As the majority of the Court of Appeal in our case pointed out (at para. 53), this approach reflects the inherent flexibility of the principled exception (see, in particular, R. v. U. (F.J.), [1995] 3 S.C.R. 764, at para. 35; R. v. Khelawon, 2006 SCC 57, [2006] 2 S.C.R. 787, at para. 45).

[93]                         In this context, and with respect for the opinion of the dissenting judge, we agree with the majority that the dissenting judge erred in limiting the value of the corroborative evidence to only one of the two material aspects of the statement, despite the fact that they were nevertheless logically intertwined. The majority stated the following on this issue:

     [translation] We disagree with our colleague, who is of the opinion that this corroboration can confirm only the second aspect of the statement, and not the first aspect, which concerns the appellant’s participation. Evidence cannot be split up in this manner to limit its corroboration to one material aspect at the expense of the other. At that stage of the trial, the scope of the statement was much broader than that described by our colleague, and the discovery of a weapon, according to the judge, corroborated the entire statement. We see no error in that reasoning. [para. 59]

[94]                         Since the corroborative evidence at issue is logically connected to the two aspects of the statement, it is not necessary to address the distinct question of whether, when a statement has several material aspects that are not interconnected, the trial judge must carry out a global analysis of the statement or examine each of its material aspects separately (see C.A. reasons, at paras. 30‑34, 56‑57 and 59).

[95]                         As we need not decide this question, we will only comment briefly on this point.

[96]                         First, even though this Court, when it rendered its decision in Bradshaw, was mindful of the fact that a statement may have several material aspects, it did not see fit to include the splitting up of the statement in the steps of the approach that the trial judge must take when assessing the admissibility of an out‑of‑court statement under the principled exception to the rule against hearsay (see para. 57). This strongly suggests that it is not for the trial judge to split up the evidence.

[97]                         Second, we are of the view that it is for the trier of fact to give credit to certain aspects of a statement and not to others. This seems to be the best way to respect the separation between the role of trial judge — responsible for questions of law — and the role of trier of fact — responsible for factual determinations — which rests on the essential distinction between threshold reliability and ultimate reliability (see Bradshaw, at para. 39; Khelawon, at para. 50). It is solely for the trial judge to decide whether a hearsay statement is sufficiently reliable to justify its admissibility. If so, the trial judge must leave it to the trier of fact to assess its probative value in light of the evidence as a whole (S. N. Lederman, M. K. Fuerst and H. C. Stewart, Sopinka, Lederman & Bryant: The Law of Evidence in Canada (6th ed. 2022), at ¶6.147).

[98]                         In view of the foregoing, we conclude that the trial judge made no reviewable error in taking the discovery of the weapons into account as corroborative evidence. Respectfully, it is rather the analysis of the dissenting judge of the Court of Appeal that is tainted by the error that led him to exclude the search results.

B.            Consideration of Certain Circumstances in the Assessment of Threshold Reliability

[99]                         The appellant argues that the trial judge committed an error in principle which tainted his finding that threshold reliability was established in this case. The appellant claims that this error was committed in weighing the circumstances relating to threshold reliability. In that weighing exercise, the trial judge allegedly failed to distinguish the circumstances that provided a true circumstantial guarantee of trustworthiness from those that had only relative weight, contrary to the teachings of Bradshaw and Couture. In the appellant’s opinion, the dissenting judge correctly stated that the trial judge [translation] “relied mainly on circumstances identical or similar to those that the Supreme Court considered as having very relative weight in [Couture and Bradshaw], without ever mentioning that they could play at the most a secondary role in the threshold reliability analysis” (C.A. reasons, at para. 42 (footnote omitted)). The dissenting judge found that [translation] “[t]he possibility that [the trial judge] gave them too much weight is therefore real” (para. 42).

[100]                     The appellant is mistaken.

[101]                     It appears from the passages of the Court of Appeal’s decision quoted above that the dissenting judge did not identify any error in principle, but instead used the presence of factors of relative weight as a basis for conducting his own examination of the circumstances. With respect, this is not permitted by the standard of review applicable to the assessment of threshold reliability, which, as noted, is entitled to deference on appeal (Youvarajah, at para. 31; Couture, at para. 81).

[102]                     It is well established that the relevance of the circumstances depends on the specific dangers associated with the hearsay in question and thus on the facts of the case (see Khelawon, at paras. 45 and 55; Youvarajah, at para. 21). That being so, it would be wrong to categorize, objectively and independently of the facts of the case, the circumstances that are neutral or secondary and those that are more important. This is therefore not a process that the trial judge is required to undertake.

[103]                     It is true that, in Bradshaw, this Court stated that circumstances that “‘. . . in essence simply point to an absence of factors that, if present, would detract from an otherwise trustworthy statement’ . . . do not provide a circumstantial guarantee of trustworthiness” (Bradshaw, at para. 92, quoting Couture, at para. 101). It must not be forgotten, however, that, in the same paragraph, this Court acknowledged that those circumstances are “relevant”. In the words of the majority of the Court of Appeal in the present case, while such circumstances are not sufficient on their own to establish threshold reliability, [translation] “the fact remains that such circumstances, considered in conjunction with others (for example, corroboration even insufficient on its own), may lead to the conclusion that the statement has all the attributes required for acceptable threshold reliability” (para. 62). This reading of Bradshaw is also consistent with R. v. Khan, [1990] 2 S.C.R. 531, where this Court found that “the absence of any reason to expect fabrication in the statement” was relevant to the analysis (p. 547).

[104]                     In the absence of an error in principle that tainted the analysis at first instance, the majority was correct in finding that the trial judge’s decision was owed deference by the Court of Appeal. It is thus neither necessary nor appropriate for us to repeat the analysis of the circumstances.

[105]                     On this point, and with all due respect, we note only that the dissenting judge seems to have committed two errors in his analysis of the circumstances. First, contrary to what his reasons suggested at para. 38, Bradshaw does not impose a rigid rule that substantive reliability is always subject to a high threshold; it merely recognizes that the greater the risk posed by the statement in light of the facts of the case, the greater the guarantees of trustworthiness required to rule out that risk (see R. v. Larue, 2018 YKCA 9, 434 D.L.R. (4th) 155, at paras. 99‑101, aff’d 2019 SCC 25, [2019] 2 S.C.R. 398; R. v. Bernard, 2018 ABCA 396, 80 Alta. L.R. (6th) 258, at para. 24). Second, [translation] “the fact that the declarant received advice from a relative or counsel before making their statement” is not a circumstance that “simply point[s] to an absence of factors that, if present, would detract from an otherwise trustworthy statement” (C.A. reasons, at para. 39, citing Bradshaw, at para. 92). In R. v. L.T.H., 2008 SCC 49, [2008] 2 S.C.R. 739, at para. 38, this Court expressly recognized that s. 146(2)(b) of the Youth Criminal Justice Act, S.C. 2002, c. 1, which deals with, among other things, the right of a young person to consult counsel and certain adults whom the young person trusts, “[is] indeed meant to ensure the reliability of a statement”. In our opinion, the involvement of the relative or counsel achieves this objective in particular by ensuring that the declarant understood the gravity of the situation and their obligation to tell the truth (see C.A. reasons, at para. 64, citing L.T.H., at para. 38). The fact that K.A. spoke with counsel, the fact that his mother accompanied him when he made his statement and the fact that his mother was made aware of the tenor of K.A.’s rights are therefore indicia of reliability that the trial judge could validly consider.

C.            Consideration of the Complainant’s Testimony at Trial, Which Was Not Produced on the Voir Dire

[106]                     The respondent argues that the majority of the Court of Appeal could validly consider the complainant’s testimony at trial, which was given before the closing of the voir dire but which was not produced on it (see C.A. reasons, at para. 71). Our colleague Moreau J. is of the view that “[o]n appeal, the appropriate mechanism for considering the complainant’s testimony is the curative proviso in s. 686(1)(b)(iii) of the Criminal Code, R.S.C. 1985, c. C‑46” (para. 4). We agree.

[107]                     Given the state of this Court’s jurisprudence, we recognize that the distinct question of the self‑contained nature of the voir dire at the trial stage is an open one. We also take due note, however, that this question was not raised by the appellant in support of his appeal and that, according to him, it was not raised before the trial judge. The appellant is of the view — and rightly so — that the majority’s comment is an obiter. The question is moot in the present appeal given that neither party challenges before us the trial judge’s decision to not consider the complainant’s testimony, a decision that was, we might add, made in accordance with the parties’ wishes.

[108]                     In the circumstances, we agree with our colleague Moreau J. that caution must be exercised: this appeal is not an appropriate one for dealing with the issue of the limits of the self‑contained nature of the voir dire at the trial stage.

III.         Conclusion

[109]                     For these reasons, we would dismiss the appeal.

                    Appeal allowed, Côté, Rowe and Kasirer JJ. dissenting.

                    Solicitor for the appellant: Emmanuelle Rheault, Montréal.

                    Solicitor for the respondent: Director of Criminal and Penal Prosecutions, Montréal.