Court of Appeal

Nova Scotia Court of Appeal

Citation: R. v. Barrett, 2020 NSCA 79

Date: 20201203

Docket: CAC 451442

Registry: Halifax

Between:

Thomas Barrett

Appellant

v.

Her Majesty the Queen

Respondent

 

Judge:	The Court (Wood C.J.N.S., Bourgeois and Beaton JJ.A.)
Appeal Heard:	October 26, 2020, in Halifax, Nova Scotia
Cases Considered:	R. v. K.G.K., 2020 SCC 7; R v. Toope, 2016 NSCA 32; Laframboise v. Millington, 2019 NSCA 43; R. v. Moir, 2020 BCCA 116; R. v. Youvarajah, 2013 SCC 41; R. v. Tsekouras, 2017 ONCA 290; R. v. Potter; R. v. Colpitts, 2020 NSCA 9; R. v. Lawrence, 2020 ABCA 268; R. v. Chretien, 2014 ONCA 403; R. v. Barrett, 2016 NSSC 43; R. v. Khelawon, 2006 SCC 57; R. v. Bradshaw, 2017 SCC 35; R. v. Hall, 2018 MBCA 122; R. v. Bernard, 2018 ABCA 396; D’Amico c. R., 2019 QCCA 77; R. v. Nurse, 2019 ONCA 260; R. v. Khalon, 2020 ABCA 124; R v. Herntier, 2020 MBCA 95; R. v. Taylor, 2012 ONCA 809; R. v. Naicker, 2007 BCCA 608; R. v. MacDonald, 2020 NSCA 69; R. v. Lights, 2020 ONCA 128; R. v. Roberts, 2020 NSCA 20; R. v. W.(D.), [1991] 1 S.C.R. 742
Subject:	Criminal; Criminal—murder; Evidence; Evidence—admissibility; Evidence—hearsay; Evidence—K.G.B. statement; Evidence—voir dire; Verdict; Verdict—unreasonable verdict
Summary:	Mr. Barrett was charged with murder.  The evidence included a sworn statement provided to investigators by a friend of Mr. Barrett, who subsequently died.  Following a voir dire, the trial judge admitted the statement into evidence.  Mr. Barrett was convicted of second degree murder on the circumstantial case put forward by the Crown.  Subsequent to his trial and conviction, in its decision R. v. Bradshaw, the Supreme Court of Canada refined the principles discussed in R. v. Khelawon, which had been relied on by the trial judge to admit the statement.
Issues:	(1)        Did the judge err in admitting the hearsay statement of the deceased declarant? (2)        Did the judge render an unreasonable verdict?
Result:	These grounds of appeal are dismissed.  (Additional grounds will be argued at a later date).  While R. v Bradshaw was released after Mr. Barrett’s trial, and refined the principles set out in R. v. Khelawon, the judge committed no error in her application of Khelawon to admit the hearsay statement.  Furthermore, the judge’s verdict was not unreasonable.

 

Nova Scotia Court of Appeal

Citation: R. v. Barrett, 2020 NSCA 79

Date: 20201203

Docket: CAC 451442

Registry: Halifax

Between:

Thomas Barrett

Appellant

v.

Her Majesty the Queen

Respondent

 

Judges:	The Court (Wood C.J.N.S, Bourgeois and Beaton JJ.A.)
Appeal Heard:	October 26, 2020, in Halifax, Nova Scotia
Held:	The grounds argued on October 26, 2020 are dismissed, per reasons for judgment of the Court
Counsel:	Mathieu Boutet, for the appellant Timothy O’Leary, for the respondent

 

 

 

By the Court:

[1]             On March 21, 2016 the appellant Thomas Barrett was convicted of the second degree murder of Brett Elizabeth MacKinnon.  He now appeals that conviction.  Mr. Barrett’s appeal was bifurcated to allow three of the multiple grounds of appeal to be argued first.[1]  For the reasons that follow, the grounds argued in this hearing are dismissed.  The remaining grounds of appeal will now be scheduled for hearing before the panel.

Background

[2]             Mr. Barrett was an acquaintance of the victim Ms. MacKinnon, who went missing from Glace Bay, Nova Scotia in the spring of 2006.  Mr. Barrett was involved in the local illegal drug trade.  The victim visited him at his home several times shortly before she disappeared.  On November 21, 2008 two passers-by came upon Ms. MacKinnon’s remains in a wooded area on the outskirts of Glace Bay.

[3]             Mr. Barrett was tried before Justice Robin Gogan (“the judge”) in the Supreme Court of Nova Scotia over nine days in January–February  2016.  (The judge’s decision was rendered March 21, 2016.)  As part of its case, the Crown made application to introduce the statement of the late Sheryl Flynn (“the declarant”).  Ms. Flynn had provided a warned, cautioned videotaped statement (“the statement”), commonly referred to as a K.G.B. statement, to police during their investigation, but she died before Mr. Barrett’s charges were brought to preliminary inquiry and trial.

[4]             Following a voir dire hearing to determine admissibility of the statement, the judge admitted it as evidence in the trial.  That statement, coupled with the evidence of other witnesses, led the judge to convict Mr. Barrett of murder based upon the circumstantial case put forward by the Crown. 

[5]             Mr. Barrett says the judge erred in admitting the statement as it was inadmissible hearsay.  Secondly, he maintains the judge erred in her overall assessment of the evidence led by the Crown.  We would re-frame this second ground as an analysis of whether the judge’s verdict was unreasonable.

[6]             It is not our task to, in effect, re-hear the trial or re-weigh the evidence put before the judge.  Appellate review is confined to the considerations mandated by the applicable standard of review.  The standard of review is determined by whether the assertion of error relates to matters of fact, law or a question of mixed fact and law.  (R v. Toope, 2016 NSCA 32 at para. 20; Laframboise v. Millington, 2019 NSCA 43 at para. 14).

Issue 1—Did the judge err in admitting the statement?

[7]             The parties agree this ground of appeal—whether the judge erred in admitting the hearsay statement of a deceased witness—engages a standard of correctness.  We must consider whether the judge correctly applied the law.  That said, the judge’s findings of fact and determination of threshold reliability are entitled to deference absent any palpable and overriding error.  As recently stated by Newbury J.A. in R. v. Moir, 2020 BCCA 116:

[82]           I proceed, then, on the basis that the standard of correctness governs the question of what ‘test’ or standard should be applied to admissibility, and to any other issue of principle; but that in the actual assessment or ‘weighing’ of the relevant factors, the trial judge should not be ‘second guessed’ by this court. As stated in R. v. Berry 2017 ONCA 17:

The exercise of weighing the probative value of proffered evidence against its potential prejudicial effect in the course of the dynamics of a trial is a discretionary task for which trial judges are particularly suited. Their decisions in that regard are entitled to deference.… Absent an error of law or principle, a material misapprehension of the evidence, or a palpable and overriding error of fact in the exercise of that discretion, there is no basis for an appellate court to interfere. [At para. 42.]

(See also:  R. v. Youvarajah, 2013 SCC 41 at para. 31; R. v. Tsekouras, 2017 ONCA 290 at para. 146; R. v. Potter; R. v. Colpitts, 2020 NSCA 9 at para. 518; R. v. Lawrence, 2020 ABCA 268 at para. 14).

[8]             In terms of balancing the factors to reach her decision on the admissibility of the statement, the judge is entitled to deference in relation to that analysis.  As observed in R. v. Chretien, 2014 ONCA 403:

[44]      The first concerns appellate deference. The factual findings that ground a trial judge’s admissibility determination are entitled to deference from appellate courts. Trial judges are well placed to assess the hearsay dangers in individual cases and the effectiveness of any safeguards to assist in overcoming those specific dangers. Absent an error in principle, a trial judge’s determination of threshold reliability is entitled to deference on appeal: R. v. Youvarajah, 2013 SCC 41, [2013] 2 S.C.R. 720, at para. 31; and R. v. Couture, 2007 SCC 28, [2007] 2 S.C.R. 517, at para. 81.

[9]             The judge conducted a voir dire for the purpose of assessing the admissibility of a videotaped, cautioned statement Ms. Flynn provided to police under oath.  The statement was obtained when police approached her during their investigation.  In her statement Ms. Flynn related a conversation she had shared with her close friend, Mr. Barrett.  She said that during the conversation he spontaneously provided certain graphic and specific details about the death of an unidentified victim at his hand and his disposal of the body.  She reported she had not come forward sooner with her information due to fear for her safety.  Pertinent also to the judge’s considerations on the voir dire was that Ms. Flynn was a recovering drug addict, and had a pending shoplifting charge upon which she had not yet appeared in court.

[10]         The judge’s decision on the admissibility of the statement is reported as R. v. Barrett, 2016 NSSC 43.  In our view, it reflects a thorough analysis of the test for admissibility set out in R. v. Khelawon, 2006 SCC 57, which requires a statement to be both necessary and reliable.  Mr. Barrett now argues the judge erred in failing to conduct an analysis that adhered instead to the principles set out in R. v. Bradshaw, 2017 SCC 35 a decision rendered by the Supreme Court of Canada fifteen months after his conviction.

[11]         The judge did not have the benefit of the guidance provided in Bradshaw, but we are now being asked by Mr. Barrett to review the decision applying its principles.  We do so understanding that a similar sequence of events occurred in R. v. Hall, 2018 MBCA 122 at para. 122; R. v. Bernard, 2018 ABCA 396 at para. 24; D’Amico c. R., 2019 QCCA 77 at para. 200; R. v. Nurse, 2019 ONCA 260 at para. 110; R. v. Khalon, 2020 ABCA 124 at para. 39.

[12]         In her decision the judge correctly identified the burden rested with the Crown to establish on a balance of probabilities the admissibility of the presumptively inadmissible hearsay statement (Khelawon at para. 47; Bradshaw at para. 23).  As to the two prongs of analysis—necessity and reliability—she first asked herself whether the statement met the threshold requirement of necessity.  There was no dispute it was easily met as the author of the statement was then deceased.

[13]         The judge then turned to the second aspect of the analysis, whether the statement was reliable such that it could meet the threshold for admission into evidence, as distinguished from ultimate reliability.  Here, the judge considered all of the circumstances surrounding the taking of the statement, its contents, the narrative factor,[2] the motive to fabricate and the demeanour of the declarant when the statement was taken.  In addition, the judge weighed the probative value of the statement versus its prejudicial effect, and concluded the hearsay statement met the reliability threshold such that it could be admitted into evidence.  That decision did not dictate what weight the judge might ultimately accord the statement in the trial, as will be seen later in our reasons.

[14]         The judge was well-placed to make the determinations she did during the voir dire on the issue of threshold reliability.  It is our view that she properly considered what, in Bradshaw, is termed procedural reliability and substantive reliability features in reaching her conclusion on the admissibility of the statement.

[15]         Mr. Barrett maintains the judge should have examined the reliability issue in accordance with Bradshaw’s two steps:  procedural reliability and substantive reliability.  Relying extensively on Khelawon, Bradshaw maps out the analysis to be followed where there is corroborative evidence called by the party seeking to introduce the statement.

[16]         Focussing first on procedural reliability, the concept was explained in R v. Herntier, 2020 MBCA 95:

[158]    Procedural reliability requires adequate substitutes for testing the evidence which can include a video recording of the statement, the presence of an oath, and a warning about the consequences of lying.  Further, some form of cross-examination, either at a preliminary inquiry or of a recanting witness at trial, is usually required.  (See para 28.)  It is concerned with whether there is a satisfactory basis to rationally evaluate the evidence.  (See para 40.)

While the judge here did not have the benefit of cross-examination of the deceased declarant, the “K.G.B.” nature of the statement meant it was taken under oath.  The declarant was also verbally warned of the consequences of providing an untruthful statement, and she then signed a document confirming her understanding of that requirement.  An additional procedural element was that the statement was videotaped, providing the judge with an opportunity to assess both the atmosphere in the room and the mood and demeanour of the declarant at the time the statement was taken.

[17]         The “K.G.B.” features of the statement cannot be ignored in considering its procedural reliability.  The statement came about when police contacted the declarant, as opposed to her having reached out to them with a “story” to tell.  As captured on video and in the written document she signed, the declarant swore an oath as to its veracity before giving her statement and was cautioned about the consequences of perjury before she provided it.  As observed in Hall, supra, at para. 66:

[66]      In assessing the inherent trustworthiness of a statement, regard should be had to the testimonial attributes of the declarant, the circumstances of the making of the statement, and whether there is corroborating or conflicting evidence (see David M Paciocco & Lee Stuesser, Essentials of Canadian Law:  The Law of Evidence, 7th ed (Toronto:  Irwin Law, 2015) at 136; and Sopinka at para 6.120).

[18]         These “testimonial attributes” were examined by the judge and permitted her to reach the conclusions she did about the safeguards present, despite the statement not being able to be tested by cross-examination.  While she may not have specifically used the distinct labels of “procedural”, and in turn, “substantive” reliability, the judge ultimately took into account the presenting factors concerning both.  Although it also predated Bradshaw, the comments of Rosenberg J.A. in R. v. Taylor, 2012 ONCA 809 resonate:

[26]           I turn then to consideration of the admissibility of the hearsay evidence in this case. No question of necessity arises; the death of the complainant fulfills the necessity criterion. This case turns on whether the complainant’s statement to the police had sufficient threshold reliability to warrant its reception. As is well known, threshold reliability may be demonstrated because of the circumstances in which it came about or because in the circumstances its truth and accuracy can nonetheless be sufficiently tested: Khelawon, at paras. 49, 62-63. However, these two different grounds are not watertight compartments: Khelawon, at para. 49.

[27]           The complainant’s statement in this case had elements of both grounds. Like testimony in court, it was taken under oath and the trier of fact could observe the declarant’s demeanour throughout because of the complete video recording. The complainant was warned of the criminal consequences of not telling the truth, which was an additional safeguard that is not explicitly found in courtroom testimony.  (Emphasis added)

[19]         In her decision the judge dealt with the factors that contributed to her conclusions regarding threshold reliability at some length:

[60]           […] It was strenuously argued that Sheryl Flynn had a motive to lie when she gave her statement to police on November 8, 2012. On this basis, the Defence says the statement is too unreliable to be admitted.

[61]           The motive to lie argument has three aspects: (1) that Sheryl Flynn bore animus toward Tom Barrett, (2) that she could have come forward with her information much sooner than she did, and (3) that she came forward with her statement at a time when she was facing her own criminal charges.

[62]      Without question, a known motive to lie is a factor to be considered on an assessment of threshold reliability. Moreover, a motive to lie in conjunction with the absence of cross-examination is at times fatal to threshold reliability. (See: R. v. Starr, [2000] 2 S.C.R. 144 at para. 215, R. v. Smith, [1992] S.C.J. No 74, paragraph 38, R. v. Blackman [2008] 2 S.C.R. 298, R. v. Scott (2005), 191 C.C.C. 183 (NSCA) and R. v. Tower, 2006 NSSC 220).

[…]

[66]           In R. v. Blackman, supra, the statement of the deceased was admitted and the accused convicted of murder. In upholding the conviction on appeal, the Supreme Court of Canada reasoned as follows respecting the role of motive to lie at para. 42:

There is no doubt that the presence or absence of a motive to lie is a relevant consideration in assessing whether the circumstances in which the statements came about provide sufficient comfort in their truth and accuracy to warrant admission. It is important to keep in mind however, that motive is but one factor to consider in the determining of threshold reliability, albeit one which may be significant depending on the circumstances. The focus of the admissibility inquiry in all cases must be, not on the presence or absence of motive, but on the particular dangers arising from the hearsay nature of the case.

[67]      The Court in Blackman then went on to consider several factors in order to determine whether a hearsay declarant may have had a motive to lie, including the nature of the relationship between the declarant and the person to whom the statement is made, the context in which the statement is made, whether the declarant had anything to gain by making false allegations and the contemporaneous nature of the statement. These are all relevant factors for consideration in the present case.

[68]      Starting with the contemporaneity of the statement, it is clear that much time passed between the purported key conversation and the eventual statement Ms. Flynn gave to police. This time gap begs the question of why it took so long for Ms. Flynn to come forward. The evidence established that she had opportunity to come forward and that she had ongoing contact with the police during the gap period. As the argument goes, if Sheryl Flynn really had such critical information, she had plenty of opportunity to disclose it and did not. I am invited to infer that she did not have the information that she later claimed to have.

[69]           However, I find that a time gap alone, even a significant one, is not evidence of motive to lie. The Crown submitted, and I agree, that there could be many reasons why Sheryl Flynn waited to disclose information to police. In her statement she said that she did not come forward because she was “terrified...like my life’s going to be at stake”. On its face that explanation seems reasonable. There are however, other factors to consider which provide context to the eventual statement to police.

[70]           The Defence pointed to the fact that Ms. Flynn’s statement contains negative references to the accused. These negative statements were characterized as animus. The court is asked to draw an inference from the negative statements that such animus exists and that it is evidence of a reason for Sheryl Flynn to come forward and present a fabricated statement to police. Having reviewed the entirety of the statement multiple times, I do not agree. While there are clearly negative characterizations of the accused in the statement, it also contains positive statements about the redeeming qualities of the accused. At times, it seems Ms. Flynn felt it necessary to point out something positive to give balance or context to the statement. Overall, I find it is not established on the evidence that Sheryl Flynn bore animus to the accused sufficient, on its own, to support a motive to lie.

[71]           There remains the consideration of the timing of Sheryl Flynn’s eventual statement. At the time she came forward, she had no criminal record but she did have charges pending against her. These charges involved thefts from local Walmart and Needs stores. The evidence on the voir dire was that these were very minor offences. Eventually, those matters were referred to Adult Restorative Justice. Sheryl Flynn died before completing the program and the charges against her were withdrawn after her death.

[72]           There is no evidence to support the conclusion that Sheryl Flynn was offered anything in return for her statement to police. There were no promises or inducements and this is confirmed in her statement. The Defence submits, however, that this does not mean that she did not hope to gain something by coming forward with information about the accused.

[73]           I agree with the Defence submission to the extent that Sheryl Flynn continued to raise the charges against her and offer to go and talk to the accused during the course of her statement. It could be that she hoped to gain something. She may have hoped that the charges would be “dealt with” by the police in return for her information. That does not mean however, that the content of her statement is not true. In other words, a motive to lie is not the only conclusion to be drawn from her hoping to have her charges dealt with at that time.

[74]           In coming to this conclusion, I am in agreement with the Crown submission that common sense does not support the view that Sheryl Flynn would fabricate the allegations against the accused in order to deal with very minor charges. The risk does not seem proportionate to the reward.

[75]           Overall then, in my view, the voir dire evidence does not support a motive to lie. It is important to note that the absence of evidence of a motive to lie does not equate to the absence of motive. It does however, somewhat neutralize this argument in the overall assessment of threshold reliably. (See Blackman, supra, at para. 40.)  (Emphasis added)

[20]         The judge also properly assessed substantive reliability.  She scrutinized all the arguments Mr. Barrett put forward during the voir dire against substantive reliability.  In doing so she considered the declarant’s motive to lie, the contemporaneity of the statement, the characterization of the evidence that could suggest animus toward Mr. Barrett and the possibility of inducements to the declarant.  The judge’s thought process also employed the drawing of permissible common-sense inferences.  There can be no question, based on the reasoning she provided that, as was her purview, the evidence presented and arguments advanced reasonably permitted the judge’s assessment of the substantive reliability of the statement.

[21]         Substantive reliability does not require the judge to be completely convinced of the trustworthiness of the statement at the voir dire stage of the inquiry.  The judge at that stage focuses on whether the statement presented is sufficiently reliable that cross-examination, had it been available, would not have enhanced the assessment (Khelawon at paras. 47–49; Bradshaw at para. 31; Herntier, supra, at para. 159).  Here, the judge concluded such was the case.

[22]         The Respondent takes the position that procedural reliability and substantial reliability are not mutually exclusive.  Relying on Bradshaw, the Respondent says they can work in tandem; the statement “must be sufficiently reliable to overcome the dangers it represents”.  The Respondent submits the judge specifically identified the core hearsay dangers in making her decision, as evidenced in this passage:

[31]      In this context, it is the Crown’s submission that the statement of Sheryl Flynn should be admitted into evidence as the core hearsay dangers can be overcome. For the sake of analysis, the core dangers may be framed in the following way:

(1) Sheryl Flynn may have misperceived the statements made by the accused during the conversation she says took place in her vehicle while parked at the Tim Hortons’ location on the Sterling Road;

(2) Even if she correctly perceived them, Ms. Flynn may have wrongly remembered them when she gave her statement to police;

(3) Sheryl Flynn may have related the facts in her statement in a misleading manner; and/or

(4) Ms. Flynn may have knowingly made a false statement.

[32]      To the foregoing list, I would add that the statement of Sheryl Flynn contains double hearsay in that Sheryl Flynn relates admissions made to her by the accused. For the Crown to be successful on this application, the “double hearsay” statements must be admissible as part of the evidence of Sheryl Flynn if she were available to testify.[3]

[23]         The overlapping nature of the procedural and substantive reliability concepts was considered in Herntier, supra.  The Manitoba Court of Appeal delved into the distinction between Khelawon and Bradshaw, commenting about the intertwined nature of procedural and substantive reliability:

[160]    Procedural and substantive reliability are not mutually exclusive.  They may work in tandem, but the standard remains high and care must be taken to ensure that the procedural safeguards and substantive guarantees of trustworthiness are sufficient to overcome the hearsay dangers.  (See para 32).

[161]    Karakatsanis J emphasised the need to preserve the distinction between threshold reliability, which concerns the admissibility of the hearsay evidence, and ultimate reliability, which concerns the degree to which the hearsay evidence should be believed and relied upon.  To do this, the corroborative evidence admitted to support threshold reliability cannot include all evidence that corroborates the declarant’s credibility, the accused’s guilt or one party’s theory of the case—those go to ultimate reliability.  Rather, it must be limited to that evidence which 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.  (See paras 42, 44.)

[162]    In assessing substantive reliability, the trial judge should identify alternative, even speculative, explanations for the hearsay statement.  Corroborative evidence is of assistance if it shows that these alternative explanations are unavailable.  If it is supportive of the truth of the statement, but is also consistent with the alternative explanations, it does not add to the statement’s inherent trustworthiness.  While the threshold reliability analysis takes place on a balance of probabilities (the declarant’s truthfulness or accuracy is more likely than any alternative explanation for the statement), the trial judge must, based on the circumstances and any evidence at the voir dire, be able to rule out any plausible alternative explanations on a balance of probabilities.  (See para 49.)  (Emphasis added)

[24]         The judge was not required to apply a standard of absolute certainty concerning the question of the reliability of the statement.  It is to be remembered the judge’s task on the voir dire was to assess threshold necessity and reliability. The ultimate task of determining whether she could rely on any part or the whole of the statement was to be conducted in the trial proper. 

[25]         Mr. Barrett contends there was no obligation on the defence to prove anything at the voir dire stage of the proceedings.  We agree the burden rested with the Crown to persuade the judge on admissibility.  Despite the absence of opportunity to cross-examine the declarant—an otherwise valuable tool to assist in establishing procedural reliability—it is not accurate to suggest the judge was entirely without ways to measure procedural reliability.  She used the tools found inside the statement itself—the cautions provided to the declarant by police regarding the consequences of perjury, and the declarant’s information given under oath. 

[26]         Mr. Barrett suggests to this Court the need for police, in securing a K.G.B. statement, to routinely incorporate the safeguard of asking a witness to explain, in their own words, what their solemn declaration means or what the implications of untruthfulness might be for them.  Mr. Barrett says these things could have been done in this case.  This cautious approach may have merit, but there was no requirement to employ it.

[27]         The judge carefully considered and analyzed four factors argued by Mr. Barrett in support of exclusion of the statement:  the witness’s motive to fabricate, her animus toward the accused, her substance abuse issues and the lack of contemporaneity in the statement.  Mr. Barrett contends that at the point the judge concluded the witness had no motive to fabricate, the burden was effectively shifted to him.  With respect, it was within the judge’s discretion to make the finding she did; that she did not adopt the position Mr. Barrett advocated did not necessarily mean the judge was in error.

[28]         It was part of the judge’s task to examine and accept or rule out alternative hypotheses.  She was required to take a functional and flexible approach in doing so (R. v. Naicker, 2007 BCCA 608 at para. 44).  The record reflects the judge’s careful consideration and her explanations of why she was satisfied the indicia of reliability present in the statement (as detailed at para. 19 herein) negated sufficiently any nefarious motive or forgetfulness by the witness.  Those were:

•                    the “K.G.B.” features of the videotaped statement—an oath and a warning against perjury

•                    the internally consistent answers in the statement

•                    the simplicity of the details provided in the statement

•                    the impact and memorability of the conversation recalled in the statement

•                    the witness’s ability to recount surrounding details

•                    the demeanour of the witness

•                    the absence of a motive to lie—animus or gain.

[29]         Mr. Barrett is critical of the absence of any indication the judge considered alternative, even speculative, explanations for the making of the statement, in the manner discussed in Bradshaw.  Mr. Barrett maintains the judge was required to look at evidence led at the voir dire in order to rule out any alternative explanations, such that the only remaining likely explanation for the statement would be the declarant’s truthfulness about, or accuracy of, material aspects of it.  Mr. Barrett again asserts because the judge did not do so, the analysis she conducted put a burden on him to demonstrate one of the other explanations was viable.  He says this constituted an error of law.

[30]         What the judge should have done, argues Mr. Barrett, was inquire whether there was sufficient corroborative evidence to rule out the explanations for the giving of the statement and/or its content put forward by him, following which the burden would return to the Crown to rebut the explanation(s). 

[31]         The role of corroborative evidence in the reliability inquiry was discussed in Bradshaw:

[47]      Second, at the threshold reliability stage, corroborative evidence must work in conjunction with the circumstances to overcome the specific hearsay dangers raised by the tendered statement. When assessing the admissibility of hearsay evidence, “the scope of the inquiry must be tailored to the particular dangers presented by the evidence and limited to determining the evidentiary question of admissibility” (Khelawon, at para. 4). Thus, to overcome the hearsay dangers and establish substantive reliability, corroborative evidence must show that the material aspects of the statement are unlikely to change under cross-examination (Khelawon, at para. 107; Smith, at p. 937). Corroborative evidence does so if its combined effect, when considered in the circumstances of the case, shows 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 (see U. (F.J.), at para. 40). Otherwise, alternative explanations for the statement that could have been elicited or probed through cross-examination, and the hearsay dangers, persist.

[48]      In assessing substantive reliability, the trial judge must therefore identify alternative, even speculative, explanations for the hearsay statement (Smith, at pp. 936-37). Corroborative evidence is of assistance in establishing substantive reliability if it shows that these alternative explanations are unavailable, if it “eliminate[s] the hypotheses that cause suspicion” (S. Akhtar, “Hearsay: The Denial of Confirmation” (2005), 26 C.R. (6th) 46, at p. 56 (emphasis deleted)). In contrast, corroborative evidence that is “equally consistent” with the truthfulness and accuracy of the statement as well as another hypothesis is of no assistance (R. v. R. (D.), [1996] 2 S.C.R. 291, at paras. 34-35). Adding evidence that is supportive of the truth of the statement, but that is also consistent with alternative explanations, does not add to the statement’s inherent trustworthiness.  (Emphasis added)

[32]         Mr. Barrett maintains the judge’s approach was flawed in that she reasoned defence evidence failed to prove the statement was false, and therefore it was admissible.  He says it was the Crown that was required to provide corroborative evidence to show the possibility the declarant lied was “substantially negated”.

[33]         In Bradshaw, the Court expressly identified the role of any available corroborative evidence, finding that such evidence:

[71]           […] will only assist in establishing the substantive reliability of the re-enactment statement if it shows, when considered in the circumstances of the case, that the only likely explanation is that Thielen was truthful about Bradshaw’s involvement in the murders. When the hearsay danger is sincerity, substantive reliability is only established when the circumstances and corroborative evidence show that the possibility that the declarant lied is substantially negated, that “even a sceptical caution would look upon [the statement] as trustworthy” (Wigmore, at p. 154; Khelawon, at para. 62; Couture, at para. 101). Corroborative evidence or circumstances showing that the statement is inherently trustworthy are required to rebut the presumption of inadmissibility.  (Emphasis added)

[34]         As will be discussed later, in this case the role of corroborative evidence became significant in the trial, but was not an element for consideration in the voir dire.

[35]         We are not persuaded the judge reversed the onus upon Mr. Barrett requiring him to prove the statement was unreliable.  Rather, the record tells us the judge reviewed and properly considered the alternate explanations put forward by him.  Bradshaw reminds judges:

[49]      While the declarant’s truthfulness or accuracy must be more likely than any of the alternative explanations, this is not sufficient. Rather, the fact that the threshold reliability analysis takes place on a balance of probabilities means that, based on the circumstances and any evidence led on voir dire, the trial judge must be able to rule out any plausible alternative explanations on a balance of probabilities.

Having considered all the arguments, the judge was able to rule out alternatives and was satisfied any hearsay dangers were sufficiently alleviated such that the statement had threshold reliability.

[36]         As adverted to earlier, it was not only the evidence found in the statement that contributed to the conviction.  In the trial proper, the judge noted the threshold indicia of reliability (which had earlier allowed her to admit the hearsay statement into evidence at the voir dire) could not go toward her assessment of the whole of the evidence and the question of the burden of proof, had she not had the benefit of other corroborating evidence.

[37]         The judge found corroboration in the evidence of no less than four other witnesses.  While Bradshaw addresses the use of corroborative evidence in establishing threshold reliability, here it was not until the trial proper that the corroborative evidence of others was available and went to enhancing the overall reliability of the statement.

[38]         The Crown’s case was circumstantial.  Ms. MacKinnon died 9.5 years prior to the trial, and the Crown had no eyewitness evidence to tender.  However, there was a body of evidence, drawn from various witnesses, about what Mr. Barrett had said to them and/or done with them in the weeks and months after Ms. MacKinnon went missing.  Notably, each of those witnesses testified to independent conversations and/or events, but all described the common thread of acts of violence, or death, and in some instances, identification of the victim by Mr. Barrett, and concealment of the body. 

[39]         The analysis circumstantial cases command was recently examined by this Court in R. v. MacDonald, 2020 NSCA 69:

[37]      Where proof beyond a reasonable doubt is based on circumstantial evidence, a trial judge must guard against “too readily drawing inferences of guilt”. An inference of guilt “drawn from circumstantial evidence should be the only reasonable inference that such evidence permits.” (R. v. Villaroman, at para. 30) Reasonable, alternative inferences other than guilt must not be overlooked. As established by the Supreme Court of Canada in Villaroman, “[i]f there are reasonable inferences other than guilt, the Crown’s evidence does not meet the standard of proof beyond a reasonable doubt”. (at para. 35)

[38]      In assessing circumstantial evidence, a trial judge is to consider alternative plausible theories and reasonable alternative inferences inconsistent with guilt. Evidence or the lack of evidence may support a reasonable, alternative inference. Villaroman requires trial judges to conduct their analysis in accordance with the “basic question”:

38...whether the circumstantial evidence, viewed logically and in light of human experience, is reasonably capable of supporting an inference other than that the accused is guilty.

[40]         Is there merit in Mr. Barrett’s argument that Bradshaw has re-written the rules set out in Khelawon?  In our view, Bradshaw refines Khelawon in those circumstances where corroborative evidence is called to support the substantive reliability analysis.  That said, in this case there was no corroborating evidence available at the threshold reliability stage.  Given that corroborative evidence may not always be proffered at the threshold stage, it can only be that Bradshaw does not wholly replace Khelawon, but instead augments it.  We adopt the reasoning of Mainella J.A. in Hall, supra:

[68]         The Bradshaw rules as to corroborative evidence are more complex to apply than the single rule in Starr which prohibited altogether considering extrinsic evidence as corroborative of the hearsay evidence for the purpose of determining admissibility (see Starr at para 217).  The Bradshaw rules focus on the relevancy, sufficiency and reliability of the proposed corroborative evidence.  According to the majority in Bradshaw, the purpose of these three rules is to preserve the distinction between the trial judge deciding threshold reliability and the trier of fact deciding ultimate reliability (at para 42).

[69]         As Newbury JA explained in R v Poony, 2018 BCCA 356, the effect of Bradshaw is to create a “high bar” (at para 27) before evidence can be considered to be corroborative of hearsay in the analysis of threshold reliability.  […]

[…]

[74]         This exercise is a cumulative assessment.  As Karakatsanis J explained, “substantive reliability is concerned with whether the circumstances, and any corroborative evidence, provide a rational basis to reject alternative explanations for the statement, other than the declarant’s truthfulness or accuracy” (at para 40; see also para 48; R v Thyagarajah, 2017 ONCA 825 at para 11; R v Johnston, 2018 MBCA 8 at paras 115-16; R v Larue, 2018 YKCA 9 at para 93; and R v Klimitz, 2018 ONCA 553 at para 8).

[75]     Karakatsanis J summarised the framework for a trial judge to determine whether corroborative evidence is of assistance in the substantive reliability inquiry as follows (at para 57):

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.

[…]

[77]     Given some of the arguments advanced on this appeal, in my view, it is important not to stray too far afield from exactly what Bradshaw decided.  The only point Bradshaw decides is clearly identified by Karakatsanis J as being, “When can a trial judge rely on corroborative evidence to conclude that the threshold reliability of a hearsay statement is established?” (at para 3).  Khelawon remains the leading decision on determining threshold reliability (see Johnston at para 98; Larue at para 98; and Brousseau c R, 2018 QCCA 1140 at paras 21-22).

[78]     Practically, the relevance of Bradshaw in a given case will depend primarily on how the moving party seeks to establish threshold reliability of the evidence in question; particularly if there is attempted reliance on corroborative evidence.  If corroborative evidence of the statement is important to establishing threshold reliability, so, too, will be the Bradshaw rules regarding corroboration.  If, however, the case is like this one, where corroborative evidence plays little, if any, role on the question of threshold reliability, Bradshaw will be of less significance.  (Emphasis added)

[41]         As in Hall, supra, here the judge did not have corroborative evidence to consider in conducting the voir dire on admissibility of the statement.  Therefore, we can be satisfied the specific application of Bradshaw beyond its general discussion of procedural and substantive reliability as established in Khelawon would not have assisted the judge in her analysis.

[42]         While Bradshaw has served to refine the law on admissibility of hearsay evidence, we do not accept Mr. Barrett’s argument that it supplants Khelawon, which in our view remains the standard for the necessity–reliability analysis.  Bradshaw steers the analysis when there is corroborative evidence available to assist the trier of fact in assessing reliability at the threshold stage.  Again, corroborative evidence was not available to the judge conducting the voir dire in this case.

[43]         In response to a question posed during oral argument, Mr. Barrett presented the alternative argument that regardless of Bradshaw the judge had also failed to properly apply the principles set out in Khelawon.  For the same reasons already discussed, we do not discern any such difficulty and are satisfied Khelawon was properly considered and applied by the judge at the voir dire stage of the trial.

Issue No. 2—The reasonableness of the verdict

[44]         Mr. Barrett asserts the judge committed “an error of mixed law and fact when reviewing the evidence she found supported a conviction” and the judge’s recollection of the evidence was “erroneous, improperly summarized, and unsupported by the evidence at trial”.

[45]         In R. v. Lights, 2020 ONCA 128, Watt J.A. described our task on the assertion of an unreasonable verdict as this:

[30]      A verdict is unreasonable if it is one that no properly instructed jury, acting judicially, could reasonably have rendered. This test requires not only an objective assessment of the evidence adduced at trial, but also, to some extent at least, a subjective evaluation of that evidence. To discharge this responsibility, we are required to review, analyse, and, within the limits of appellate disadvantage, weigh the evidence. This weighing is only to determine whether that evidence, considered as a whole, is reasonably capable of supporting the verdict rendered: R. v. R.P., 2012 SCC 22, [2012] 1 S.C.R. 746, at para. 9; R. v. Yebes, [1987] 2 S.C.R. 168, at p. 186; R. v. Biniaris, 2000 SCC 15, [2000] 1 S.C.R. 381, at para. 36; R. v. Burns, [1994] 1 S.C.R. 656, at p. 663.

[31]      A verdict may also be unreasonable where a judge has drawn an inference or made a finding of fact that is plainly contradicted by the evidence or is incompatible with evidence that is not otherwise contradicted or rejected: R.P., at para. 9, citing R. v. Sinclair, 2011 SCC 40, [2011] 3 S.C.R. 3, at paras. 4, 16, 19-21.

[46]         Potter, supra, also discussed the analysis required to conclude unreasonableness of a verdict in a judge-alone trial:

[579]       Determining whether a verdict is reasonable involves asking this question: on the basis of the evidence presented at trial, could a properly instructed jury, acting judicially, have returned it? We must also assess “whether it was based on an inference or finding of fact that, (1) is plainly contradicted by the evidence relied on by the trial judge in support of that inference; or (2) is shown to be incompatible with evidence that has not otherwise been contradicted or rejected by the trial judge”.  (Footnotes omitted)

[47]         Was the evidence before the judge reasonably capable of supporting her decision?  Mr. Barrett argues because there was no physical evidence to link him to the offence, the judge overemphasized the importance of the hearsay statement, without which it would have been unreasonable to convict him. 

[48]         Without casting aside either the presumption of innocence or the burden of proof resting with the Crown, an observation by the court in Lights, supra, echoes in this case.  Like Mr. Barrett, the accused in that case did not testify at trial.  The court noted:

[33]      When the claim of an unreasonable verdict rests on the assertion that, based on the evidence, the trier of fact could not have reasonably rendered the guilty verdict, an appellate court is entitled to consider that the accused did not testify at trial or adduce other evidence to support any other reasonable inference consistent with innocence: Corbett v. The Queen, [1975] 2 S.C.R. 275, at pp. 280-81; R. v. Wu, 2017 ONCA 620, at para. 16.

[49]         All his criticisms of the perceived flaws of the quality of the corroborative evidence are without the benefit of any alternative evidence from Mr. Barrett.  This observation is not intended to suggest any burden on him or any other accused at trial, but it does now permit us to regard the trial evidence through that lens.  (See also R. v. Roberts, 2020 NSCA 20 at para. 52.)

[50]         In Lights, supra, the court went on to address the specific analysis required to assess, as here, the reasonableness of the verdict in a circumstantial case:

[36]      When the Crown’s case consists wholly or substantially of circumstantial evidence, the standard of proof requires the trier of fact be satisfied beyond a reasonable doubt that the accused’s guilt is the only reasonable inference to be drawn from the evidence as a whole: R. v. Villaroman, 2016 SCC 33, [2016] 1 S.C.R. 1000, at para. 20.

[37]      To determine if the circumstantial evidence meets the required standard of proof, the trier of fact must keep in mind that it is the evidence, assessed as a whole, that must meet this standard of proof, not each individual piece of evidence that is but a link in the chain of proof: R. v. Smith, 2016 ONCA 25, 333 C.C.C. (3d) 534, at paras. 81-82; R. v. Morin, [1988] 2 S.C.R. 345, at pp. 360-61; Côté v. The King (1941), 77 C.C.C. 75 (S.C.C.), at p. 76.

[38]      Inferences consistent with innocence need not arise from proven facts. Rather, they may arise from a lack of evidence: Villaroman, at para. 35. Accordingly, a trier of fact must consider other plausible theories and other reasonable possibilities inconsistent with guilt so long as these theories and possibilities are grounded on logic and experience. They must not amount to fevered imaginings or speculation. While the Crown must negate these reasonable possibilities, it need not negate every possible conjecture, no matter how irrational or fanciful, which might be consistent with an accused's innocence: Villaroman, at paras. 37-38. See also R. v. Bagshaw, [1972] S.C.R. 2, at p. 8.

[39]      When a verdict that rests wholly or substantially on circumstantial evidence is challenged as unreasonable, the question appellate courts must ask is whether the trier of fact, acting judicially, could reasonably be satisfied that the guilt of the accused was the only reasonable conclusion available on the evidence taken as a whole: Villaroman, at para. 55. Fundamentally, it is for the trier of fact to determine whether any proposed alternative way of looking at the case as a whole is reasonable enough to raise a doubt about the guilt of the accused: Villaroman, at para. 56.  (Emphasis added)

[51]         We are unable to agree with Mr. Barrett’s characterization of failures on the part of the judge in reaching her ultimate conclusion that the Crown had proved its case beyond a reasonable doubt.  The judge did not rely strictly on the hearsay statement of Ms. Flynn; indeed, the judge stated that standing alone the statement would not have been sufficient to ground a conviction.  It was the availability of corroborative evidence, from a variety of witnesses in relation to a variety of different points, that ultimately tipped the scales for the judge. 

[52]         We conclude the judge could be reasonably satisfied of Mr. Barrett’s guilt based on the Villaroman test.  It was the judge’s task to assess and weigh the evidence, and there is no support found in the record for the suggestion the evidence was not reasonably capable of supporting the verdict.  As stated in Herntier, supra:

[365]    While the accused now seeks to review, individually, pieces of evidence and suggest that there is a reasonable alternative theory for each piece, these points were all argued by defence counsel in his address to the jury and were, therefore, clearly before the jury.  The test for an unreasonable verdict is not whether there is an alternative explanation for each piece of evidence, but whether the evidence as a whole could reasonably satisfy a trier of fact of the guilt of the accused beyond a reasonable doubt.  […]

[…]

[372]    While other interpretations of the evidence, particularly when looked at individually, were possible, it was for the jury to determine whether, when taken as a whole, the evidence led to a reasonable alternative explanation to that of guilt.  […]

[53]         In addition, Mr. Barrett asserts the judge wrongly concluded that after Ms. MacKinnon’s death someone had moved her body a second time, without proof of that having been offered by the Crown.  Even if it could be said the judge made improper inferences or findings of fact in relation to whether the body was moved after it was hidden, of which we are not persuaded, ultimately that post-offence related evidence did not impact upon all of the other elements of murder, upon which the judge was ultimately satisfied beyond a reasonable doubt.

[54]         It is also argued the judge wrongly opined Mr. Barrett had experienced an “adrenaline rush” in the course of committing the murder.  The submission before us is there was no evidence before the judge concerning what was in Mr. Barrett’s mind at any time.  With respect, we disagree.  The evidence about Mr. Barrett’s state of mind was contained in his discussion with Ms. Flynn, as found in her statement.  When that statement was admitted into evidence at trial, the information concerning Mr. Barrett’s state of mind was then before the judge for consideration.

[55]         Furthermore, there was no error committed by the judge in drawing the inference that Mr. Barrett’s admission to Ms. Flynn that he had killed “someone” was a reference to the victim.  The judge also heard from three other Crown witnesses (Andrews, Newell and MacDonald) to whom Mr. Barrett had made various admissions at various times.  The nature and details of all these admissions were such that the judge’s inferences were rational and logical and flowed from the evidence.  The whole of the evidence, and the putting of the hearsay statement in context, permitted both the inferences made and conclusions reached by the judge.

[56]         For the reasons set out earlier, we are satisfied there was no error with respect to the admission of the statement, nor the ultimate weight attached to it by the judge.  In addition, we consider there was overwhelming evidence of Mr. Barrett’s intention found in the various admissions he made to other witnesses who testified in the trial.

[57]         We are unable to conclude the judge’s decision to convict was erroneous, improperly summarized or unsupported by the evidence.  The judge conducted a thorough analysis of the evidence.  She identified the presumption of innocence, the burden on the Crown and the elements of the alleged offences, and followed with the appropriate W.D. (R. v. W.(D.), [1991] 1 S.C.R. 742 at para. 28) analysis to make her credibility findings.  Throughout, the judge explained why she was or was not prepared to accept or rely on the evidence of each witness.  The evidence before her permitted the judge to make the findings and inferences she used to reach her conclusion about the guilt of Mr. Barrett.  We are satisfied her decision, easily understood, and the reasoning pathway to it were supported by the body of evidence.

Conclusion

[58]         To summarize, the judge did not err in her application of Khelawon, nor in failing to apply Bradshaw, to determine the threshold necessity and reliability of the hearsay statement.  She committed no palpable and overriding error in reaching her factual conclusions.  Further, we are not persuaded the judge’s decision the Crown had met its burden on the charges was unreasonable.  Based on both the evidentiary record and the standard of review that guides us, we find no basis to interfere with the judge’s conclusions.

[59]         For the foregoing reasons, these grounds of appeal are dismissed.  We direct the matter return to Chambers to be scheduled for a hearing before this panel on the remaining grounds of appeal, which include an allegation of ineffective assistance of trial counsel.

Wood C.J.N.S.

Bourgeois J.A.

Beaton J.A.