Supreme Court of Canada **Appeal Heard:** January 16, 2019

Judgment Rendered: June 28, 2019 Docket: 38270 --- ## Parties Patrick John Goldfinch — Appellant v. Her Majesty The Queen — Respondent — and — Attorney General of Ontario and Criminal Lawyers' Association of Ontario — Interveners --- ## Coram Abella, Moldaver, Karakatsanis, Gascon, Brown, Rowe and Martin JJ. --- ## Reasons Reasons for Judgment: (paras. 1 to 76) Karakatsanis J. (Abella, Gascon and Martin JJ. concurring) Concurring Reasons: (paras. 77 to 148) Moldaver J. (Rowe J. concurring) Dissenting Reasons: (paras. 149 to 205) Brown J. --- R. v. Goldfinch, 2019 SCC 38, [2019] 3 S.C.R. 3 Patrick John Goldfinch — Appellant v. Her Majesty The Queen — Respondent and Attorney General of Ontario and Criminal Lawyers' Association of Ontario — Interveners Indexed as: R. v. Goldfinch 2019 SCC 38 File No.: 38270. 2019: January 16; 2019: June 28. Present: Abella, Moldaver, Karakatsanis, Gascon, Brown, Rowe and Martin JJ. on appeal from the court of appeal for alberta Criminal law — Evidence — Admissibility — Complainant's sexual activity — Accused charged with sexual assault — Accused seeking to introduce evidence that he and complainant were in sexual relationship at time of alleged assault — Trial judge admitting evidence and giving mid‑trial and final limiting instructions to jury on use it could make of it — Accused acquitted — Whether sexual relationship evidence admissible — Criminal Code, R.S.C. 1985, c. C‑46, s. 276. The accused was charged with sexually assaulting a woman he had dated and lived with. The two remained friends and the complainant would occasionally come to the accused's house and stay overnight. At trial, the accused requested a voir dire to determine if evidence that he and the complainant were in a sexual relationship — "friends with benefits" — at the time of the alleged assault was admissible under s. 276 of the Criminal Code. He argued that the sexual nature of the relationship provided important context without which the jury would be left with the artificial impression that he and the complainant had a platonic relationship. The accused also advanced that he did not intend to rely on this evidence to support the twin‑myth inferences that the complainant was more likely to have consented to the sexual activity or was less worthy of belief. The trial judge admitted the evidence, concluding that keeping this "relatively benign" evidence from the jury would lend an element of artificiality to the proceedings and harm the accused's right to make full answer and defence. At trial, both sides led evidence regarding the frequency of the sexual contact between the complainant and the accused. The jury found the accused not guilty. A majority of the Court of Appeal allowed the Crown's appeal and ordered a new trial, finding that the trial judge had erred in admitting the evidence. In its view, the only inferences to be drawn from the evidence were those relying on the twin myths and limiting instructions could not cure the fact that the jury had heard inadmissible evidence for which there was no permissible use. The accused appeals as of right to the Court on the issue of whether the "friends with benefits" evidence was admissible. Held (Brown J. dissenting): The appeal should be dismissed. Per Abella, Karakatsanis, Gascon and Martin JJ.: The evidence in this case did not meet the requirements of s. 276 of the Criminal Code and admitting it was a reversible error of law which might reasonably be thought to have had a material bearing on the acquittal. A new trial is required. The Canadian justice system strives to protect the ability of triers of fact to get at the truth. In cases of sexual assault, evidence of a complainant's prior sexual history — if relied upon to suggest that the complainant was more likely to have consented to the sexual activity in question or is generally less worthy of belief — undermines this truth‑seeking function and threatens the equality, privacy and security rights of complainants. Section 276 was enacted to mitigate these harms, balancing a number of trial fairness considerations and seeking to exclude evidence known to distort the fact‑finding process. It protects the integrity of the trial process by safeguarding both the dignity and privacy of complainants and the right of accused persons to make full answer and defence. It is designed to exclude irrelevant information that is more prejudicial to the administration of justice than it is probative. Sections 276(1) and (2) operate together to achieve these objectives. Section 276(1) sets out an absolute bar against introducing evidence of the complainant's prior sexual activity for the purpose of drawing twin‑myth inferences. When an accused seeks to introduce such evidence for some other purpose, that evidence is presumptively inadmissible unless the accused satisfies s. 276(2). To do so, the accused must demonstrate that the evidence is of specific instances of sexual activity, is relevant to an issue at trial, and has significant probative value that is not substantially outweighed by the danger of prejudice to the proper administration of justice. The accused must identify the evidence and its purpose with sufficient precision to allow the judge to apply s. 276(2) and weigh the factors set out in s. 276(3), which include the accused's right to make full answer and defence, the need to remove discriminatory beliefs or biases from the fact‑finding process, the potential prejudice to the complainant's dignity and privacy, and the right of every individual to the full protection and benefit of the law. Evidence of a relationship that implies sexual activity clearly engages s. 276(1), and, to be admissible, must satisfy the requirements of s. 276(2). The risk that evidence of a relationship which implies sexual activity may be used to support twin‑myth reasoning is clear. Even relatively benign relationship evidence must be scrutinized and handled with care. If the accused cannot point to a relevant use of the evidence other than the twin myths, mere assurances that the evidence will not be used for those purposes are insufficient. In this case, the evidence was barred by s. 276(1) because it served no purpose other than to support the inference that because the complainant had consented in the past, she was more likely to have consented on the night in question. Nor did the evidence satisfy the conditions of admissibility under s. 276(2). As to the first condition, the accused successfully demonstrated that the evidence was of specific instances of sexual activity. The words "specific instances of sexual activity" in s. 276(2)(a) must be read in light of the scheme and broader purpose of s. 276. Evidence of a relationship that implies sexual activity inherently encompasses specific instances of sexual activity. To satisfy s. 276(2)(a), the accused must point to identifiable activity, but the degree of specificity required in a particular case will depend on the nature of the evidence, how the accused intends to use it, and its potential to prejudice the administration of justice. Here, the accused specified the parties to the relationship, the nature of that relationship and the relevant time period. Requiring further details would have unnecessarily invaded the complainant's privacy. However, the accused failed to fulfill the second condition by establishing that the evidence was relevant to an issue at trial as required by s. 276(2)(b). The accused must identify, with precision, how the evidence is relevant to a specific issue at trial. The relevant issue cannot be one of the twin myths prohibited by s. 276(1), and generic references to credibility of the accused or the complainant, narrative or context will not suffice. While the case law provides examples of how evidence of previous sexual activity between an accused and a complainant may be relevant to an issue at trial, none of them apply in this case. There are circumstances in which evidence of a sexual relationship may be fundamental to the coherence of an accused's narrative, and by extension, credibility, but here there was nothing about the accused's testimony that cast him in an unfavourable light or rendered his narrative untenable absent the information that he and the complainant were friends with benefits. As for the third condition — which requires balancing a number of factors to determine whether the evidence has significant probative value that is not substantially outweighed by the danger of prejudice to the proper administration of justice — the accused's right to make full answer and defence would not have been compromised by excluding the sexual nature of his relationship with the complainant. Indeed, the evidence was not relevant to an issue at trial and therefore had no probative value. Admitting the evidence was a reversible error of law which might reasonably be thought to have had a material bearing on the acquittal and a new trial is therefore required. The improper admission of the evidence as "context" risked infecting the trial with the precise prejudicial assumptions s. 276 was designed to weed out. The jury should not have been privy to particulars regarding the frequency of the sexual contact or the accused's testimony characterizing the evening as "typical" or "routine". That evidence clearly engaged twin‑myth reasoning by suggesting that because the complainant had "typically" consented to sex with the accused in the past, she was more likely to have done so on that "routine" occasion. Per Moldaver and Rowe JJ.: The trial judge erred in admitting the "friends with benefits" evidence under s. 276 of the Criminal Code, having particular regard to the manifest deficiencies in the accused's application to introduce this evidence. The improper admission of the evidence for the broad purpose of providing "context" led to a significant and highly prejudicial broadening of the sexual activity evidence at trial, which might reasonably have had a material bearing on the accused's acquittal. Accordingly, a new trial is warranted, and the appeal should be dismissed. The s. 276 regime is designed to respect and preserve the rights of both complainants and accused persons by excluding evidence which would undermine the legitimacy of our criminal justice system and inhibit the search for truth, while allowing for the admission of evidence which would enhance the legitimacy of our criminal justice system and promote the search for truth. In this way, the regime seeks to promote the integrity of the trial process as a whole — a concept that is essential to the public's faith in the criminal justice system. In pursuing this objective, the s. 276 regime operates in a step‑by‑step manner. From the accused's initial application under s. 276.1 to the final limiting instruction required by s. 276.4, the s. 276 regime establishes a rigorous, multistep process through which sexual activity evidence adduced by or on behalf of the accused must be carefully vetted and winnowed down to its essentials. To make its way into evidence at trial, such evidence must withstand careful scrutiny at each stage of the process. Section 276(1) prohibits the use of sexual activity evidence to support one of the twin myths identified in R. v. Seaboyer, [1991] 2 S.C.R. 577. In doing so, it gives effect to the principle that these myths are simply not relevant at trial and can severely distort the trial process. Accordingly, if the sole purpose for which sexual activity evidence is being proffered is to support either of the twin myths, it will be ruled inadmissible under s. 276(1). But that does not mean sexual activity evidence will always be ruled inadmissible. While sexual activity evidence adduced by or on behalf of the accused is presumptively inadmissible, such evidence may be admitted where it satisfies a three‑part test under s. 276(2). Before sexual activity evidence can be admitted under this provision, the accused must file a written application under s. 276.1. If the judge is not satisfied that certain requirements have been met (e.g., the application is deficient), then he or she may dismiss the application without more. On the other hand, if the accused's written application survives scrutiny, then the process moves to the voir dire stage and the court's attention shifts to s. 276(2). The first requirement of s. 276(2) is that the evidence be of specific instances of sexual activity. As stated in R. v. L.S., 2017 ONCA 685, 40 C.R. (7th) 351, the content of the "specific instances" requirement is linked to the nature of the evidence sought to be adduced. Where the accused seeks to introduce evidence of an individual instance of sexual activity, he must identify that instance with specificity. By contrast, where the accused seeks to introduce general evidence that describes the nature of the relationship between the accused and the complainant, the specificity requirement speaks to factors relevant to identifying the relationship and its nature and not to details of specific sexual encounters. These factors will include the parties to the relationship, the relevant time period, and the nature of the relationship. The second requirement of s. 276(2) is that the evidence be "relevant to an issue at trial". To satisfy this requirement, the accused must demonstrate that the evidence goes to a legitimate aspect of his defence and is integral to his ability to make full answer and defence. This requires that the accused be able to identify specific facts or issues relating to his defence that can be properly understood and resolved by the trier of fact only if reference is made to the sexual activity evidence in question. In articulating these specific facts or issues, simply citing the need to provide greater "context" or a fuller "narrative" will not suffice. Similarly, bare invocations of credibility will not be enough. Furthermore, the requirement that the evidence be "integral" to the accused's ability to make full answer and defence means that even if the evidence can be linked to specific facts or issues relating to the accused's defence, admission is not guaranteed. There may be cases in which the evidence, while relevant to specific facts or issues relating to the accused's defence, bears only marginally on it. In such cases, the trial judge may, in his or her discretion, exclude the evidence on the basis that countervailing considerations, such as the need to protect the privacy rights and dignity of the complainant, outweigh the tenuous connection the evidence has to the accused's ability to make full answer and defence. The third requirement of s. 276(2) is that the evidence have significant probative value that is not substantially outweighed by the danger of prejudice to the proper administration of justice. As explained in R. v. Darrach, 2000 SCC 46, [2000] 2 S.C.R. 443, the requirement of significant probative value serves to exclude evidence of trifling relevance that, even though not used to support the two forbidden inferences, would still endanger the proper administration of justice. In undertaking the analysis required by s. 276(2), which is designed to ensure that any admissible sexual activity evidence is limited in scope and that its legitimate purpose is identified and weighed against countervailing considerations, the trial judge must take into account the factors listed in s. 276(3). To the extent sexual activity evidence is ultimately admitted, the trial judge must explain to the jury, in clear and precise terms, the uses for which the evidence may — and may not — be used. Finally, all trial participants — including the trial judge, Crown and defence counsel, and witnesses — must hew to the specific, legitimate purpose for which the evidence has been admitted, without expanding the scope of the ruling or using the admissible evidence for inadmissible purposes. This is essential to preserving not only the rights of the accused and the complainant, but also the integrity of the trial process as a whole. In this case, the "friends with benefits" evidence could, on its face, potentially be used to support the first of the twin myths — the myth that because the complainant consented to have sex with the accused in the past, she was more likely to have consented to the sexual activity forming the subject‑matter of the sexual assault charge. However, s. 276(1) takes this potential use off the table. As such, unless the accused could point to some legitimate use of the sexual activity evidence that would justify admission under s. 276(2), that evidence was inadmissible. Although the accused's application under s. 276.1 satisfied the specificity requirement under s. 276(2)(a), it failed to satisfy the relevance requirement under s. 276(2)(b). The accused argued that the threshold for relevance was met because the "friends with benefits" evidence: (1) was necessary to avoid an erroneous misapprehension on the part of the jury that he and the complainant were platonic friends at the time of the alleged sexual assault; and (2) provided "context" to the issues at trial. However, the accused did not explain why it was necessary to correct any potential misapprehension as to the sexual nature of his relationship with the complainant. In addition, the accused failed to identify a specific, legitimate purpose for putting the evidence before the jury — he did not link the evidence to specific facts or issues relating to his defence that could be properly understood and resolved only if reference could be made to the "friends with benefits" evidence. The evidence was also incapable of satisfying the third requirement under s. 276(2). Because the evidence was not relevant to an issue at trial based on the application presented to the trial judge, it was necessarily incapable of possessing any probative value. However, the possibility that the presiding judge at the new trial might, if presented with a properly framed s. 276.1 application, admit the evidence after applying the test and weighing the factors in s. 276(2) and (3) should not be foreclosed. Without reaching any final decision on the matter, there was at least one specific issue that the accused could have referred to in his application that might have properly supported admission of the "friends with benefits" evidence: the jury's assessment of his testimony that he mouthed the words "I'm going to fuck you" to the complainant. If the jury lacked the knowledge that the two were in a sexual relationship at the time, that statement might have seemed bizarre or even menacing. Furthermore, the accused's testimony that he made that statement to the complainant may itself have seemed implausible. In this way, withholding the sexual nature of the accused's relationship with the complainant could have had an adverse impact on the jury's assessment of his credibility, potentially infringing upon his right to make full answer and defence. Had the accused referenced this aspect of his anticipated testimony in his s. 276.1 application, the trial judge would have been better equipped to engage in the balancing exercise required by s. 276(2) and (3) and may have properly determined that the evidence was admissible for the narrow purpose of allowing the jury to assess the accused's testimony on this point. A new trial is required. A number of errors occurred at trial as a result of the trial judge's improper s. 276 ruling, which allowed for the admission of sexual activity evidence under the broad banner of "context". Grounded in this ruling, the trial judge's flawed limiting instructions failed to delineate how the sexual activity evidence was capable of assisting the jury to resolve specific facts or issues relating to the accused's defence. This flawed instruction's distorting effect was compounded when additional sexual activity evidence was admitted at trial which was not the subject of its own admissibility determination or limiting instruction. The cumulative impact of these errors can reasonably be thought to have had a material bearing on the accused's acquittal. Per Brown J. (dissenting): The evidence was admissible. The trial judge applied the correct legal principles in her evidentiary ruling and the jury rendered its verdict after being properly instructed on how to do so. The appeal should be allowed and the acquittals restored. First, the "friends with benefits" evidence did not derive its relevance solely from twin‑myth reasoning and should therefore have filtered through s. 276(1). The test for exclusion under s. 276(1) is whether the evidence derives its relevance solely from twin‑myth reasoning and not whether it merely engages that type of reasoning. Were engagement the test for categorical exclusion under s. 276(1), it would risk exclusion of all relationship evidence, or at the very least all evidence of relationships which also involve sexual activity, since that, too, would conceivably engage twin‑myth reasoning. Such an approach would resurrect the creation of pre‑determined categories of admissibility which was rejected in R. v. Seaboyer, [1991] 2 S.C.R. 577, and would downplay the text and purpose of statutory provisions like s. 276.4, which recognizes that evidence may be admissible for certain purposes yet inadmissible for others, and makes a limiting instruction mandatory where any evidence of other sexual activity is introduced, even if it only refers to other sexual activity indirectly or implicitly, to cure prejudice and to warn the jury of the impermissible uses of that evidence. Rather, relationship evidence should typically be filtered via the inquiry contemplated by s. 276(2)(b), being whether the evidence is relevant to an identifiable issue at trial. In this case, the Crown failed to explain why evidence of the "friends with benefits" relationship was objectionable, while evidence of other types of relationships which regularly pass through the filter of s. 276(1), and yet which might also suggest previous sexual activity, is not. Evidence of relationships that involve sexual activity, but which lack the expectation or desire of a more formal relationship, may also give important context to the non‑sexual interactions between the parties to the relationship, which may be, as it was in this case, necessary for the accused to make full answer and defence. Evidence of a "friends with benefits" relationship will, in certain cases, and without engaging in prohibited lines of reasoning, explain to a jury how two people know each other, consistent with how other relationships are presented to juries. The trial judge's evidentiary ruling treated the relationship between the accused and the complainant consistently with other relationships in society. Second, the "friends with benefits" evidence met the relevance test under s. 276(2)(b) because it was relevant to the accused's ability to make full answer and defence. Indeed, it was necessary for the jury to assess the credibility of the accused's evidence, which was the most relevant and material issue with which the jury would have had to grapple. To deny the accused the ability to point to his relationship would in these circumstances disable the jury from meaningfully performing its central function of finding facts and seeking out the truth, and would force the accused to tell an incomplete story — a story which includes an account of the act but no explanation for how he and the complainant "got there" and why he said what he said and did what he did. Without the evidence, the accused's actions will have appeared to have arisen out of nowhere, and the accused's right to make full answer and defence would be reduced to painting a picture of himself as (at best) crude and reckless, or (at worst) predatory. Ordering a new trial is unfair, given that the Crown's theory of the case drew directly from the sexual nature of the relationship and that it was the Crown, and not the accused, who contravened the trial judge's evidentiary ruling and explored both the details and the frequency of the sexual activity. A successful Crown appeal from acquittal in this case inevitably lowers the bar which the Crown must overcome to show that a legal error had a material bearing on the acquittal so as to secure a new trial. --- ## Cases Cited ### By Karakatsanis J. Referred to: R. v. Mills, [1999] 3 S.C.R. 668; R. v. Quesnelle, 2014 SCC 46, [2014] 2 S.C.R. 390; R. v. Darrach, 2000 SCC 46, [2000] 2 S.C.R. 443; R. v. Handy, 2002 SCC 56, [2002] 2 S.C.R. 908; R. v. Grant, 2015 SCC 9, [2015] 1 S.C.R. 475; R. v. Seaboyer, [1991] 2 S.C.R. 577; R. v. L.S., 2017 ONCA 685, 354 C.C.C. (3d) 71; R. v. Shearing, 2002 SCC 58, [2002] 3 S.C.R. 33; R. v. Ewanchuk, [1999] 1 S.C.R. 330; R. v. J.A., 2011 SCC 28, [2011] 2 S.C.R. 440; R. v. Hutchinson, 2014 SCC 19, [2014] 1 S.C.R. 346; R. v. Barton, 2019 SCC 33, [2019] 2 S.C.R. 579; R. v. Barton, 2017 ABCA 216, 354 C.C.C. (3d) 245; R. v. Rodney; R. v. A.R.C., [2002] O.J. No. 5364 (QL); R. v. Strickland (2007), 45 C.R. (6th) 183; R. v. Harris (1997), 118 C.C.C. (3d) 498; R. v. Temertzoglou (2002), 11 C.R. (6th) 179; R. v. M. (M.) (1999), 29 C.R. (5th) 85; R. v. Graveline, 2006 SCC 16, [2006] 1 S.C.R. 609. ### By Moldaver J. Referred to: R. v. Seaboyer, [1991] 2 S.C.R. 577; R. v. Darrach, 2000 SCC 46, [2000] 2 S.C.R. 443; R. v. L.S., 2017 ONCA 685, 40 C.R. (7th) 351; R. v. Graveline, 2006 SCC 16, [2006] 1 S.C.R. 609; R. v. Barton, 2019 SCC 33, [2019] 2 S.C.R. 579; R. v. Morrison, 2019 SCC 15, [2019] 2 S.C.R. 3. ### By Brown J. (dissenting) R. v. Seaboyer, [1991] 2 S.C.R. 577; R. v. L.S., 2017 ONCA 685, 40 C.R. (7th) 351; R. v. Grant, 2015 SCC 9, [2015] 1 S.C.R. 475; R. v. Harris (1997), 118 C.C.C. (3d) 498; R. v. M. (M.) (1999), 29 C.R. (5th) 85; R. v. Temertzoglou (2002), 11 C.R. (6th) 179; R. v. Blea, [2005] O.J. No. 4191 (QL); R. v. A.A., 2009 ABQB 602, 618 A.R. 137; R. v. Provo, 2018 ONCJ 474, 48 C.R. (7th) 1; R. v. W.(D.), [1991] 1 S.C.R. 742; R. v. Crosby, [1995] 2 S.C.R. 912; R. v. Nyznik, 2017 ONSC 4392, 40 C.R. (7th) 241; R. v. Darrach, 2000 SCC 46, [2000] 2 S.C.R. 443; R. v. Graveline, 2006 SCC 16, [2006] 1 S.C.R. 609; R. v. Luciano, 2011 ONCA 89, 273 O.A.C. 273; R. v. Lane and Ross (1969), 6 C.R.N.S. 273; R. v. Corbett, [1988] 1 S.C.R. 670. --- ## Statutes and Regulations Cited Bill C‑49, An Act to amend the Criminal Code (sexual assault), 3rd Sess., 34th Parl., 1992, Preamble. Bill C‑51, An Act to amend the Criminal Code and the Department of Justice Act and to make consequential amendments to another Act, 1st Sess., 42nd Parl., 2018 (assented to December 13, 2018). Canadian Charter of Rights and Freedoms, s. 11(d). Criminal Code, R.S.C. 1970, c. C‑34, s. 143. Criminal Code, R.S.C. 1985, c. C‑46, ss. 276, 276.1, 276.2, 276.4, 276.5, 676(1)(a). --- ## Authors Cited Backhouse, Constance. "Sexual Harassment: A Feminist Phrase That Transformed the Workplace" (2012), 24 C.J.W.L. 275. Boyle, Christine. "Sexual Assault as Foreplay: Does Ewanchuk Apply to Spouses?" (2004), 20 C.R. (6th) 359. Canada. Department of Justice. Research and Statistics Division. An Estimation of the Economic Impact of Violent Victimization in Canada, 2009, by Josh Hoddenbagh, Ting Zhang and Susan McDonald. Government of Canada, 2014. Canada. Department of Justice. Research and Statistics Division. Health Impacts of Violent Victimization on Women and their Children, by Nadine Wathen. Government of Canada, 2012. Canada. Statistics Canada. Canadian Centre for Justice Statistics. Police‑reported sexual assaults in Canada before and after #MeToo, 2016 and 2017, by Cristine Rotenberg and Adam Cotter. Statistics Canada, 2018. Craig, Elaine. "Capacity to Consent to Sexual Risk" (2014), 17 New Crim. L. Rev. 103. Craig, Elaine. Putting Trials on Trial: Sexual Assault and the Failure of the Legal Profession. Montréal: McGill‑Queen's University Press, 2018. Craig, Elaine. "Section 276 Misconstrued: The Failure to Properly Interpret and Apply Canada's Rape Shield Provisions" (2016), 94 Can. Bar Rev. 45. Desrosiers, Julie, et Geneviève Beausoleil‑Allard. L'agression sexuelle en droit canadien, 2e éd. Montréal: Yvon Blais, 2017. Khan, Ummni. "Hot for Kink, Bothered by the Law: BDSM and the Right to Autonomy" (Summer 2016), 41:2 Law Matters 17. Koshan, Jennifer. "Marriage and Advance Consent to Sex: A Feminist Judgment in R v JA" (2016), 6:6 Oñati Socio‑legal Series 1377 (online). Lippel, Katherine. "Conceptualising Violence at Work Through A Gender Lens: Regulation and Strategies for Prevention and Redress" (2018), 1 U of OxHRH J 142. Miller, Karen‑Lee. "You Can't Stop The Bell From Ringing." Protean, Unpredictable, And Persisting: The Victim Impact Statement In The Context Of Sexually Assaulted Women, submitted in conformity with the requirements for the degree of Doctor of Philosophy. Toronto: University of Toronto, Dalla Lana School of Public Health, 2015. Oxford English Dictionary"friend with benefits" (online). Paciocco, David M., and Lee Stuesser. The Law of Evidence, 7th ed. Toronto: Irwin Law, 2015. Randall, Melanie. "Sexual Assault in Spousal Relationships, 'Continuous Consent', and the Law: Honest But Mistaken Judicial Beliefs?" (2008), 32 Man. L.J. 144. Rosenbury, Laura A. "Friends with Benefits?" (2007), 106 Mich. L. Rev. 189. Sealy‑Harrington, Joshua. "Tied Hands? A Doctrinal and Policy Argument for the Validity of Advance Consent" (2014), 18 Can. Crim. L. Rev. 119. Silver, Lisa A. "The WD Revolution" (2018), 41 Man. L.J. 307. Stuart, Don. "Twin Myth Hypotheses in Rape Shield Laws are Too Rigid and Darrach is Unclear" (2009), 64 C.R. (6th) 74. Tanovich, David M. "'Whack' No More: Infusing Equality into the Ethics of Defence Lawyering in Sexual Assault Cases" (2015), 45 Ottawa L. Rev. 495. --- APPEAL from a judgment of the Alberta Court of Appeal (McDonald, Strekaf and Berger JJ.A.), 2018 ABCA 240, 48 C.R. (7th) 22, 72 Alta. L.R. (6th) 317, 363 C.C.C. (3d) 406, [2018] A.J. No. 830 (QL), 2018 CarswellAlta 1312 (WL Can.), setting aside the acquittals of the accused and ordering a new trial. Appeal dismissed, Brown J. dissenting. Deborah R. Hatch, for the appellant. Joanne B. Dartana and Matthew Griener, for the respondent. G. Karen Papadopoulos and Jill Witkin, for the intervener the Attorney General of Ontario. Megan Savard and Colleen McKeown, for the intervener the Criminal Lawyers' Association of Ontario. --- ## Reasons for Judgment The judgment of Abella, Karakatsanis, Gascon and Martin JJ. was delivered by Karakatsanis J. — ### I. Facts [ 1 ] Our system of justice strives to protect the ability of triers of fact to get at the truth. In cases of sexual assault, evidence of a complainant's prior sexual history — if relied upon to suggest that the complainant was more likely to have consented to the sexual activity in question or is generally less worthy of belief — undermines this truth-seeking function and threatens the equality, privacy and security rights of complainants. [ 2 ] In 1992, Parliament enacted s. 276 of the Criminal Code, R.S.C. 1985, c. C-46, to protect trials from these harms. Nearly 30 years later, the investigation and prosecution of sexual assault continues to be plagued by myths. One such myth is that sexual assault is a crime committed by persons who are strangers to their targets. In fact, in 2016-2017, Statistics Canada found that over 80 percent of reported sexual assaults occurred between people who knew one another in some way.[^1] In other words, most complainants will have some kind of relationship with the accused. This case requires the Court to review the balance between, on the one hand, admitting evidence of a sexual relationship that may be fundamental to making full answer and defence, and on the other, protecting complainants and the integrity of the trial process from prejudicial reasoning. [ 3 ] Here, the accused sought to introduce evidence that he and the complainant were "friends with benefits", a sexual relationship. He argued that the sexual nature of the relationship provided important context without which the jury would be left with the artificial impression that he and the complainant had a platonic relationship, rendering consent improbable. [ 4 ] To be admissible, relationship evidence that implies sexual activity must satisfy the requirements of s. 276 of the Criminal Code. In my view, the evidence here did not meet those requirements. Introducing evidence of the sexual nature of the relationship served no purpose other than to support the inference that because the complainant had consented in the past, she was more likely to have consented on the night in question. It was therefore barred by s. 276(1). Nor could it satisfy the conditions of admissibility under s. 276(2). While the sexual aspect of the relationship was evidence of "specific instances of sexual activity", it was not "relevant to an issue at trial". [ 5 ] A s. 276 application requires the accused to positively identify a use of the proposed evidence that does not invoke twin-myth reasoning. In other words, relevance is the key which unlocks the evidentiary bar, allowing a judge to consider the s. 276(3) factors and to decide whether to admit the evidence. Bare assertions that such evidence will be relevant to context, narrative or credibility cannot satisfy s. 276. The evidence in this case should not have been admitted and a new trial is required. I would dismiss the appeal. [ 6 ] Mr. Goldfinch and the complainant met, dated and lived together for seven or eight months, after which the complainant ended the relationship. At some point during the ensuing months, the two resumed contact. Although they each described the relationship in various ways, both ultimately agreed that their relationship could be described as "friends with benefits". [ 7 ] On the evening of May 28, 2014, the complainant called Goldfinch, who then drove to the complainant's house, picked her up, and brought her back to his place. Goldfinch testified that she had called him a few days earlier asking for "birthday sex", something the complainant couldn't remember if she had done. Goldfinch stated that he did not "100 percent" expect to have sex that evening"but that was our routine" (A.R., vol. III, at p. 228). In his view, this was a "typical evening" in that the complainant "would call in the middle of the night, want to come over, and we'd end up going to bed together" (A.R., vol. III, at p. 201). [ 8 ] Goldfinch lived in the basement of a small, older home which he shared with a roommate. After arriving at the house, Goldfinch and the complainant shared drinks and conversation with the roommate while watching television. Goldfinch testified that, during this time, he mouthed "I'm going to fuck you" to the complainant. He says she responded with a smile. The complainant couldn't remember whether this exchange had occurred, but acknowledged that it might have. [ 9 ] A few minutes later, Goldfinch invited the complainant to go downstairs. The complainant testified that she told Goldfinch "nothing was going to happen", meaning that she did not wish to have sex. Goldfinch denies ever hearing this. [ 10 ] Downstairs, the two sat on a couch together. At some point, they shared a consensual kiss. After the kiss, Goldfinch suggested that they go to bed. [ 11 ] From this point on, the two accounts of the evening diverged radically. [ 12 ] According to Goldfinch, after the consensual kiss, he followed the complainant into his bedroom where they each removed their own clothes. He and the complainant then discussed which side of the bed they wished to sleep on. Following this discussion, the two engaged in consensual foreplay and brief intercourse. He fell asleep and, hours later, she woke him up complaining that he had struck her on the head in his sleep. He was annoyed, told her to leave and called a taxi using her phone. [ 13 ] The complainant testified that she responded to Goldfinch's invitation by telling him she did not want to have sex. He then grabbed her arm and dragged her into the bedroom. She explained that Goldfinch's demeanour changed"[j]ust like something snapped" (A.R., vol. II, at p. 88), and she felt scared. She removed her clothes because he told her to. He pushed her onto the bed, struck her in the face, pushed her shoulder so hard that she believed her arm was broken, and told her "he was going to have [her], just like everyone else" (A.R., vol. II, at p. 91). Following the assault, she dressed and called a taxi from her cell phone. She called the police shortly after she arrived back at her home. Both the responding officer and a forensics officer who met the complainant at the hospital confirmed swelling on her left cheek and elbow. --- ### II. History of the Proceedings #### A. The Voir Dire, Pentelechuk J. — Court of Queen's Bench of Alberta, 140600008Q1, January 23, 2017 [ 14 ] The defence requested a voir dire to determine if evidence that the complainant and Goldfinch were "friends with benefits" was admissible under s. 276 of the Criminal Code, submitting that it was highly artificial to describe the relationship without reference to sexual activity. Counsel advanced that Goldfinch did not intend to rely on twin-myth inferences, but failed to identify any other inference or relevant use beyond "context". The Crown was willing to adduce evidence that the two knew each other for four to five years, dated and lived together for seven to eight months, and then broke up. The Crown was also prepared to adduce evidence that the two remained friends and that the complainant would occasionally come to Goldfinch's house and stay overnight. [ 15 ] The trial judge accepted that "friends with benefits" meant that "they were friends who . . . from time to time got together to have sex" (A.R., vol. I, at p. 10). She agreed that keeping this evidence from the jury would lend an element of artificiality to the proceedings and harm Goldfinch's right to make full answer and defence. She concluded this "relatively benign" evidence would not "prejudice the complainant's personal dignity, right to privacy, or personal security" if admitted in this limited form. #### B. The Trial, Pentelechuk J. — Court of Queen's Bench of Alberta, 140600008Q1, February 9, 2017 [ 16 ] The trial unfolded before a jury over four days in February 2017. [ 17 ] Before examining the complainant, Crown counsel sought clarification regarding the permissible scope of questioning with respect to previous sexual activity, noting that she would not have led this evidence had the s. 276 application not been granted. The trial judge stated that she had envisioned that the "contextual information" identified in the voir dire would be reduced to an agreed statement of facts. But, because the parties had not done so, the trial judge reiterated her expectation that any questioning would be "extremely limited", following "fairly narrow confines for the purposes of context and to simply let the jury know the nature of the relationship" (A.R., vol. II, at p. 53). [ 18 ] During direct examination, the complainant initially denied that she and Goldfinch were ever "more than just friends" following their breakup. Shortly thereafter, however, she admitted that she had been to Goldfinch's bedroom to have sex on "various" dates for "quite awhile [sic]" after the relationship ended (A.R., vol. II, at p. 81). [ 19 ] Before the cross-examination began, counsel for Goldfinch also sought clarification regarding the permissible scope of questioning. The trial judge agreed that the Crown had opened the evidentiary door and gave the defence permission to ask questions regarding "the number of times, the time frame relative to the relationship proper breaking up, and the last occasion prior to these alleged offences" (A.R., vol. II, at p. 112). During cross-examination, when defence counsel suggested the two had slept together "dozens of times", the complainant estimated they had slept together 15 times following the breakup. [ 20 ] Following the complainant's testimony, the trial judge gave this limiting instruction to the jury: > You have heard evidence that [the complainant] and Mr. Goldfinch dated and then briefly lived together. At some point after that relationship ended, [the complainant] and Mr. Goldfinch did on occasion get together and have sexual relations. This evidence provides you with some context for their relationship, but you must not use this evidence to help you decide that because [the complainant] and Mr. Goldfinch had sexual relations in the past, that [the complainant] is more likely to have consented to Mr. — consented to what Mr. Goldfinch is alleged to have done on May 29th, 2014, and you must also not use that evidence to help you decide that because [the complainant] and Mr. Goldfinch had sexual relations in the past that she is less believable or reliable as a witness in this case, all right? Thank you. (A.R., vol. II, at p. 148) [ 21 ] Goldfinch repeatedly testified — both in-chief and in cross-examination — to the frequency of his previous sexual interactions with the complainant, characterizing the evening as "typical" or "routine", indicating he had "had her many times", and stating that "when we're together, [sex] was expected, I guess, from both of us" (A.R., vol. III, at pp. 203 and 227). [ 22 ] During her final jury charge, the trial judge gave extensive instructions regarding consent, highlighting that consent must be contemporaneous and concerns only the subjective state of mind of the complainant. She reiterated the same limiting instruction she had given mid-trial, clarifying that while the jury could not use the evidence of sexual activity to infer that the complainant was less believable or reliable, it could consider any contradictions regarding the nature of the relationship in assessing the complainant's general credibility. [ 23 ] The jury found Goldfinch not guilty of sexual assault. #### C. The Alberta Court of Appeal, McDonald and Strekaf JJ.A., Berger J.A. Dissenting — 2018 ABCA 240, 48 C.R. (7th) 22 [ 24 ] The Crown appealed Goldfinch's acquittal on a question of law pursuant to s. 676(1)(a) of the Criminal Code. Section 276.5 of the Criminal Code provides that a determination respecting the admissibility of sexual activity evidence is a question of law. The majority, McDonald and Strekaf JJ.A., characterized the Crown's appeal as follows: > The Crown appellant submits that the trial judge erred in law by admitting the evidence, pursuant to section 276 of the Criminal Code, of the prior sexual relationship between [Goldfinch] and the complainant. [para. 14] [ 25 ] The majority held that finding evidence provides "context" is insufficient to demonstrate relevance for the purposes of s. 276(2). Thus, they concluded that the trial judge failed to connect the relationship evidence to any issue relevant to Goldfinch's defence. The majority also rejected the argument that the evidence was relevant to Goldfinch's credibility or as "context". In their view, permitting the defence to lead evidence of previous sexual activity to prevent the jury from concluding that consent was unlikely was no different from admitting that same evidence to establish that the complainant was more likely to consent. Given the admissions the Crown had been willing to make, there was no risk that the jury would be misled into thinking Goldfinch and the complainant were strangers. The only inferences to be drawn from the evidence of prior sexual activity could be those relying on the twin myths. In the majority's view, limiting instructions could not cure the fact that the jury had heard inadmissible evidence for which there was no permissible use. They allowed the appeal and ordered a new trial. [ 26 ] Berger J.A., writing in dissent, accepted that the evidence was not adduced to support the twin myths. He found that the trial judge had properly exercised her discretion in admitting what she characterized as "relatively benign" evidence in order to preclude misapprehensions on the part of the jury. In his view, Goldfinch's right to make full answer and defence required a "candid revelation of the true nature of the relationship" (para. 67). For Berger J.A., it was better to trust the jury to rely on the limiting instructions mandated for evidence admitted under s. 276 than to leave the jury to speculate on what lay beneath Crown admissions for which no such instructions would be required. --- ### III. Analysis [ 27 ] This case asks whether evidence of a relationship with an implicit sexual component engages s. 276 of the Criminal Code and, if so, when such evidence may be admitted. #### A. Section 276: Text, History and Objectives [ 28 ] Section 276[^2] balances a number of trial fairness considerations, seeking to exclude evidence known to distort the fact-finding process while protecting the rights of both the accused and the complainant: > 276 (1) In proceedings in respect of [various sexual offences], evidence that the complainant has engaged in sexual activity, whether with the accused or with any other person, is not admissible to support an inference that, by reason of the sexual nature of that activity, the complainant

(a) is more likely to have consented to the sexual activity that forms the subject-matter of the charge; or

(b) is less worthy of belief.

(2) In proceedings in respect of an offence referred to in subsection (1), no evidence shall be adduced by or on behalf of the accused that the complainant has engaged in sexual activity other than the sexual activity that forms the subject-matter of the charge, whether with the accused or with any other person, unless the judge, provincial court judge or justice determines, in accordance with the procedures set out in sections 276.1 and 276.2, that the evidence

(a) is of specific instances of sexual activity;

(b) is relevant to an issue at trial; and

(c) has significant probative value that is not substantially outweighed by the danger of prejudice to the proper administration of justice.

(3) In determining whether evidence is admissible under subsection (2), the judge, provincial court judge or justice shall take into account

(a) the interests of justice, including the right of the accused to make a full answer and defence;

(b) society's interest in encouraging the reporting of sexual assault offences;

(c) whether there is a reasonable prospect that the evidence will assist in arriving at a just determination in the case;

(d) the need to remove from the fact-finding process any discriminatory belief or bias;

(e) the risk that the evidence may unduly arouse sentiments of prejudice, sympathy or hostility in the jury;

(f) the potential prejudice to the complainant's personal dignity and right of privacy;

(g) the right of the complainant and of every individual to personal security and to the full protection and benefit of the law; and

(h) any other factor that the judge, provincial court judge or justice considers relevant. [ 29 ] Section 11(d) of the Canadian Charter of Rights and Freedoms guarantees the right "to be presumed innocent until proven guilty according to law in a fair and public hearing by an independent and impartial tribunal". This guarantee includes the accused's right to make full answer and defence, itself crucial to ensuring that the innocent are not convicted (R. v. Mills, [1999] 3 S.C.R. 668, at paras. 69 and 76). [ 30 ] For this reason, the law of criminal evidence begins with the general principle that all relevant and material evidence is admissible. The right to a fair trial does not, however, guarantee the most favourable procedures imaginable: the accused's right to make full answer and defence is not automatically breached whenever relevant evidence is excluded (R. v. Quesnelle, 2014 SCC 46, [2014] 2 S.C.R. 390, at para. 64; R. v. Darrach, 2000 SCC 46, [2000] 2 S.C.R. 443, at para. 24; Mills, at para. 75). A fair trial also requires that no party be allowed to distort the process by producing irrelevant or prejudicial evidence (Darrach, at para. 24). [ 31 ] A person's general character and past behaviour provide context for understanding specific events (R. v. Handy, 2002 SCC 56, [2002] 2 S.C.R. 908, at para. 39). However, such evidence often draws on pejorative or judgmental generalizations (D. M. Paciocco and L. Stuesser, The Law of Evidence (7th ed. 2015), at p. 54). This is problematic because "[b]ad character is not an offence known to the law" (Handy, at para. 72). Our legal system neither punishes nor protects people on the basis of lifestyle, character or reputation. To protect against propensity reasoning, trial judges must balance the probative value of such evidence against its prejudicial effects. [ 32 ] As a general rule, evidence led by the defence is excluded only where the potential prejudice substantially outweighs its probative value (R. v. Grant, 2015 SCC 9, [2015] 1 S.C.R. 475, at para. 19; R. v. Seaboyer, [1991] 2 S.C.R. 577, at p. 611). Moreover, because only the accused is being judged at trial, exclusionary rules generally do not prevent the accused from adducing evidence of another person's bad character (Paciocco and Stuesser, at p. 98). [ 33 ] Historically, no limits were placed on the defence's ability to adduce evidence of a complainant's prior sexual activities. Such evidence was routinely used to malign "the character of the complainant, distort the trial process, and undermine the ability of the criminal justice system to effectively and fairly try sexual allegations" (R. v. L.S., 2017 ONCA 685, 354 C.C.C. (3d) 71, at para. 79). Subjecting the complainant to humiliating or prolonged examination and exploiting assumptions about "communication, dress, revenge, marriage, prior sexual history, therapy, lack of resistance and delayed disclosure" was commonplace (D. M. Tanovich"'Whack' No More: Infusing Equality into the Ethics of Defence Lawyering in Sexual Assault Cases" (2015), 45 Ottawa L. Rev. 495, at pp. 498-99).[^3] These tactics shifted the focus away from the accused and essentially put the complainant on trial. [ 34 ] In 1982, Parliament enacted a blanket exclusion of all evidence of sexual activity, subject to limited exceptions. The provisions were intended to counter the twin myths that women with sexual experience are more likely to consent to sexual activity or are less worthy of belief. Because sexual assault was highly gendered and underreported, Parliament also sought to encourage the reporting of sexual crimes (Seaboyer, at p. 606, per McLachlin J., and pp. 648-50, per L'Heureux-Dubé J., dissenting in part). [ 35 ] In Seaboyer, this Court struck down that blanket exclusion, holding that Parliament had cast the net too wide, impairing the accused's right to a fair trial. Improperly excluded evidence critical to the defence included evidence going to: (i) honest but mistaken belief in consent, (ii) bias or motive to fabricate on the part of the complainant, (iii) physical conditions establishing the use of force, and (iv) evidence of a consistent modus operandi such as threatening to accuse someone of rape as a means of extortion (Seaboyer, at pp. 613-16). [ 36 ] Parliament responded to this Court's ruling in Seaboyer by essentially codifying the case's principles in s. 276 of the Criminal Code because "at trials of sexual offences, evidence of the complainant's sexual history is rarely relevant and . . . its admission should be subject to particular scrutiny" (Preamble, Bill C-49, An Act to amend the Criminal Code (sexual assault), 3rd Sess., 34th Parl., 1992). Parliament's recent amendments — reaffirming the exclusion of evidence relying on the twin myths — reinforce the provision's continued importance. [ 37 ] The mischief Parliament sought to address in enacting s. 276 remains with us today. Sexual assault is still among the most highly gendered and underreported crimes (J. Desrosiers and G. Beausoleil-Allard, L'agression sexuelle en droit canadien (2nd ed. 2017), at pp. 41-42). Even hard-fought battles to stop sexual assault in the workplace remain ongoing (compare, e.g., K. Lippel"Conceptualising Violence at Work Through A Gender Lens: Regulation and Strategies for Prevention and Redress" (2018), 1 U of OxHRH J 142, and C. Backhouse"Sexual Harassment: A Feminist Phrase That Transformed the Workplace" (2012), 24 C.J.W.L. 275). As time passes, our understanding of the profound impact sexual violence can have on a victim's physical and mental health only deepens. Parliament enacted s. 276 to address concrete social prejudices that affect trial fairness as well as the concrete harms caused to the victims of sexual assault. Throughout their lives, survivors may experience a constellation of physical and psychological symptoms including: high rates of depression; anxiety, sleep, panic and eating disorders; substance dependence; self-harm and suicidal behaviour.[^4] A recent Department of Justice study estimated the costs of sexual assault at approximately $4.8 billion in 2009, an astonishing $4.6 billion of which related to survivors' medical costs, lost productivity (due in large part to mental health disability), and costs from pain and suffering.[^5] The harm caused by sexual assault, and society's biased reactions to that harm, are not relics of a bygone Victorian era. [ 38 ] It is against this backdrop that s. 276 must be interpreted and applied. #### B. Examining Relationship Evidence Under Section 276 [ 39 ] Section 276 protects the integrity of the trial process by striking a balance between the dignity and privacy of complainants and the right of accused persons to make full answer and defence. This appeal asks us to examine that balance as it concerns evidence of a relationship from which sexual activity can reasonably be inferred. [ 40 ] As Gonthier J. explained in Darrach, s. 276 is "designed to exclude irrelevant information and only that relevant information that is more prejudicial to the administration of justice than it is probative" (para. 43). Sections 276(1) and (2) operate together to achieve this objective. First, s. 276(1) sets out an absolute bar against introducing evidence of the complainant's prior sexual activity for the purpose of drawing twin-myth inferences. Where an accused seeks to introduce such evidence for some other purpose, that evidence is presumptively inadmissible unless the accused satisfies s. 276(2). To do so, the accused must identify the evidence and its purpose with sufficient precision to allow the judge to apply s. 276(2) and weigh the factors set out in s. 276(3). [ 41 ] I proceed by first considering when such evidence engages s. 276(1). I then address two requirements of s. 276(2) which require particular attention in this case: (a) Is the evidence of "specific instances of sexual activity"? and (b) What qualifies evidence as "relevant to an issue at trial"? ##### (1) Section 276(1) [ 42 ] Turning to the first question, evidence of a relationship that implies sexual activity clearly engages s. 276(1). [ 43 ] Section 276(1) bars evidence of a complainant's previous sexual activity tendered to support the twin myths. Such evidence is "not probative of consent or credibility and can severely distort the trial process" (Darrach, at para. 33). In barring such inferences, the provision affirms the equality and dignity rights of complainants and aims to encourage reporting of sexual assault (Bill C-49). The risk that evidence of a relationship which implies sexual activity may be used to support twin-myth reasoning is clear. [ 44 ] Consider the first myth: that a complainant's prior sexual activity may support an inference of consent in a particular instance. Rejection of this myth — and its link to relationships — is intimately connected to the modern understanding of consent. Until 1983, the fact that an accused was married to a complainant was sufficient to legitimize sexual assault; indeed, rape was defined as non-consensual sexual intercourse between a man and "a female person who is not his wife" (Criminal Code, R.S.C. 1970, c. C-34, s. 143). Today, an accused may no longer argue that consent was implied by a relationship: contemporaneous, affirmatively communicated consent must be given for each and every sexual act (R. v. Ewanchuk, [1999] 1 S.C.R. 330; R. v. J.A., 2011 SCC 28, [2011] 2 S.C.R. 440, at paras. 34 and 47; R. v. Hutchinson, 2014 SCC 19, [2014] 1 S.C.R. 346, at para. 27; R. v. Barton, 2019 SCC 33, [2019] 2 S.C.R. 579, at paras. 90-94). Today, not only does no mean no, but only yes means yes. Nothing less than positive affirmation is required. [ 45 ] Consider also the second myth: that previous sexual activity renders a complainant less worthy of belief or, by extension, of full protection of the law (Barton, at para. 201, per Moldaver J. and, at paras. 222 and 231, per Abella and Karakatsanis JJ.). Before this Court, Goldfinch advanced that social mores have changed such that being "unchaste" no longer discredits a complainant. However, this Court has held that the second myth is not limited to attitudes towards "unchaste" women (Darrach, at para. 33). Moreover, while sexual activity generally carries less stigma than it once did, complainants continue to be treated as less deserving of belief based on their previous sexual conduct. The notion that some complainants "invite" assault and, by inference, do not deserve protection persists both inside and outside our courtrooms (R. v. Barton, 2017 ABCA 216, 354 C.C.C. (3d) 245, at para. 128, cited in Barton, at para. 201, per Moldaver J. and, at para. 231, per Abella and Karakatsanis JJ.; see also E. Craig, Putting Trials on Trial: Sexual Assault and the Failure of the Legal Profession (2018), at pp. 32 et seq.). This is implicit in the continued struggle to exclude inaccurate assumptions about what constitutes "typical" or "unusual" activity within a given relationship (see, e.g., E. Craig"Section 276 Misconstrued: The Failure to Properly Interpret and Apply Canada's Rape Shield Provisions" (2016), 94 Can. Bar Rev. 45, at p. 69; M. Randall"Sexual Assault in Spousal Relationships, 'Continuous Consent', and the Law: Honest But Mistaken Judicial Beliefs" (2008), 32 Man. L.J. 144; C. Boyle"Sexual Assault as Foreplay: Does Ewanchuk Apply to Spouses?" (2004), 20 C.R. (6th) 359). Finally, the suggestion that sexual assault is less harmful to those who are sexually active or in relationships is simply wrong (see, e.g., J. Koshan"Marriage and Advance Consent to Sex: A Feminist Judgment in R v JA" (2016), 6:6 Oñati Socio-legal Series 1377 (online), at pp. 1387 and 1391). [ 46 ] Even "relatively benign" relationship evidence must be scrutinized and handled with care. If the accused cannot point to a relevant use of the evidence other than the twin myths, mere assurances that evidence will not be used for those purposes are insufficient. This case highlights the dangers of accepting such assurances. [ 47 ] In this case, the obvious implication of the evidence of an ongoing sexual relationship was that because the complainant had consented to sex with Goldfinch in the past, in similar circumstances, it was more likely she had consented on the night in question. As I set out in the sections that follow, the difficulty here was not that Goldfinch and the complainant had a relationship, but that Goldfinch could point to no relevant use for evidence of the sexual nature of the relationship. Such an approach misapprehends the nature of consent and is barred by s. 276(1). ##### (2) Section 276(2) [ 48 ] Taken as a whole, s. 276 seeks to protect the privacy of complainants, encourage the reporting of sexual offences and exclude evidence which fuels propensity reasoning. In pursuit of these goals, s. 276(2) presumptively bars evidence of the complainant's previous sexual activity. [ 49 ] However, in certain circumstances, the accused's right to make full answer and defence requires that such evidence be admitted. Under s. 276(2), the accused must demonstrate that the evidence: (a) is of specific instances of sexual activity; (b) is relevant to an issue at trial; and (c) has significant probative value that is not substantially outweighed by the danger of prejudice to the proper administration of justice. [ 50 ] In determining whether these criteria are met, s. 276(3) requires judges to consider a number of factors. These include the accused's right to make full answer and defence, the need to remove discriminatory beliefs or biases from the fact-finding process, potential prejudice to the complainant's dignity and privacy, and the right of every individual to the full protection and benefit of the law. [ 51 ] Bare assertions that such evidence will be relevant to context, narrative or credibility cannot satisfy s. 276(2). A s. 276 application must provide "detailed particulars" which will allow a judge to meaningfully engage with the tests set out at s. 276(2) and (3). The accused must propose a use of the evidence that does not invoke twin-myth reasoning. These requirements are key to preserving the integrity of the trial by ensuring twin-myth reasoning masquerading as "context" or "narrative" does not ambush the proceedings. ###### (a) Section 276(2)(a): Specific Instances of Sexual Activity [ 52 ] Goldfinch suggests that relationship evidence is difficult to situate in s. 276(2) because it does not constitute "specific instances of sexual activity" (see also R. v. Rodney (Ont. S.C.J.); R. v. A.R.C., [2002] O.J. No. 5364 (QL) (S.C.J.)). This position fails to recognize the purposes of the provision. [ 53 ] The words "specific instances of sexual activity" must be read in light of the scheme and broader purposes of s. 276. The requirement that evidence be "specific" prevents aimless or sweeping inquiries into the complainant's sexual history. The accused must point to identifiable activity, but the degree of specificity required in a particular case will depend on the nature of the evidence, how the accused intends to use it, and its potential to prejudice the proper administration of justice. As Doherty J.A. noted in L.S., specificity is required so that judges may apply the scheme in a way that effectively protects the rights of the complainant and ensures trial fairness. A purposive interpretation thus calls for evidence that is sufficiently specific to support a fully informed analysis, allowing the judge to circumscribe what evidence may be adduced and how it may be used. [ 54 ] Evidence of a relationship that implies sexual activity, such as "friends with benefits", as defined by the accused here, inherently encompasses specific instances of sexual activity. Requiring further details would unnecessarily invade the complainant's privacy, defeating an important objective of the provision. I agree with the statement in L.S. that specifying the parties to the relationship, the nature of that relationship and the relevant time period satisfies the purposes of trial fairness (para. 83). Those criteria are met in this case. ###### (b) Section 276(2)(b): Relevance to an Issue at Trial [ 55 ] Turning to the second requirement, the importance of relevance to an issue at trial is highlighted by the procedural safeguards inherent in the s. 276 regime. Section 276.1(2) requires the accused to set out, in writing, the "detailed particulars" of the evidence to be adduced as well as the "relevance of th[e] evidence to an issue at trial". The application judge must be satisfied that the evidence is capable of being admitted under s. 276(2) before ordering a voir dire (s. 276.1(4)(c)). Judges who admit such evidence must also provide written reasons identifying the relevance of the evidence admitted (s. 276.2(3)(c)).[^6] These procedural requirements reflect the fact that sexual assault prosecutions require heightened attention to the general principle that no party should be allowed to distort the process by producing irrelevant evidence (Darrach, at paras. 24 and 37). [ 56 ] It goes without saying that the "relevant issue" cannot be one of the twin myths prohibited by s. 276(1).[^7] Neither will generic references to the credibility of the accused or the complainant suffice. Credibility is an issue that pervades most trials, and "[e]vidence of prior sexual activity will rarely be relevant to support a denial that sexual activity took place or to establish consent" (Darrach, at para. 58; see also Handy, at paras. 115-16). Arguments for relevance must be scrutinized to ensure "context" is not simply a disguised myth. [ 57 ] That said, a relationship may provide relevant context quite apart from any sexual activity. Where the relationship is defined as including sexual activity, as the trial judge held "friends with benefits" was here, it is critical that the relevance of the sexual nature of the relationship to an issue at trial be identified with precision. [ 58 ] At the voir dire, Goldfinch described relevance in general terms: the evidence was necessary for "context" or to prevent "faulty impressions". He was not merely concerned with dispelling the notion that he and the complainant were strangers: he specifically sought to introduce the sexual nature of the relationship. The trial judge was clearly alive to the possibility that this could be used to support the twin myths. In the end, however, the trial judge concluded that the evidence was relevant because it put the relationship "in proper context". In coming to this conclusion, she relied on R. v. Strickland (2007), 45 C.R. (6th) 183 (Ont. S.C.J.). In that case, the trial judge reasoned that the probative value of "contextual" relationship evidence did not support an inference of "an increased likelihood of consent" but rather could dispel an inference of "the unlikelihood of consent" (para. 35). [ 59 ] With respect, that is a distinction without a difference. Three paragraphs from Strickland illustrate why: > It can be said that, as a general rule, people do not have sexual intercourse with complete strangers. Generally speaking, sexual partners are involved in a relationship of some sort. . . . What does matter is that, at some point, each partner has made an assessment of the other, and decided that that person is a suitable person with whom to share this most intimate human experience.

It is the fact that such a decision was made in the past by the complainant that is relevant. . . . It is this fact that makes it at least somewhat more probable that a complainant would consent to having sex with a man with whom she had an existing sexual relationship, than if no such relationship existed at all.

To restate in the language of Darrach, the inference of an increased likelihood of consent does not flow from the sexual nature of the activity, but rather from the existence of a relationship in which that activity took place. [paras. 27-28 and 30] [ 60 ] It is difficult to conceive of a more clear instance of twin-myth reasoning than the proposition that because the complainant had "at some point" consented to be intimate with the accused, it was "more probable" that she would have done so again. [ 61 ] Moreover, while the case law provides examples of how evidence of previous sexual activity between an accused and a complainant may be relevant to an issue at trial, none of them apply here. [ 62 ] Prior sexual activity may be particularly relevant to a defence of honest but mistaken belief in communicated consent (Seaboyer, at pp. 613-16; Darrach, at para. 59; Barton, at paras. 91 et seq.). However, an honest but mistaken belief cannot simply rest upon evidence that a person consented at "some point" in the past: that would be twin-myth reasoning. By definition, the defence must rely upon evidence of how the complainant previously communicated consent so that the accused can adequately support a belief that consent was expressed. Here, the trial judge properly instructed the jury not to rely on the "friends with benefits" evidence in evaluating the defence of honest but mistaken belief. [ 63 ] Evidence of a sexual relationship may also be relevant when complainants have offered inconsistent statements regarding the very existence of a sexual relationship with the accused (see, e.g., R. v. Harris (1997), 118 C.C.C. (3d) 498 (Ont. C.A.); R. v. Temertzoglou (2002), 11 C.R. (6th) 179 (Ont. S.C.J.)). There were no such contradictory statements from the complainant in the record at the time of the voir dire and Goldfinch did not proceed on this basis. [ 64 ] To the extent that Goldfinch sought to establish a pattern of behaviour, the "pattern" here was hardly distinctive; it would not be admissible as similar fact evidence (Handy, at paras. 82, 127 and 131). As I have noted, the limited admissibility of similar fact evidence protects the truth-seeking function of the trial by excluding evidence that is overly prejudicial to the accused. By imposing the same evidentiary standard under s. 276, neither the accused nor the complainant is denied equal protection of the law on the basis of lifestyle, character or reputation (Craig"Section 276 Misconstrued", at p. 71). [ 65 ] Finally, Goldfinch submits that the sexual aspect of a relationship may be relevant to the coherence of the accused's narrative, and by extension, credibility. There will, of course, be circumstances in which context will be relevant for the jury to properly understand and assess the evidence. That assessment, however, must be free of twin-myth reasoning. General arguments that the sexual nature of a relationship is relevant to context, narrative or credibility will not suffice to bring the evidence within the purview of s. 276(2). [ 66 ] That Goldfinch points to only two cases in which evidence was admitted as necessary "context" illustrates the rarity of such circumstances (*Temertzoglou*; R. v. M. (M.) (1999), 29 C.R. (5th) 85 (Ont. S.C.J.)). In both of these cases, the evidence admitted was fundamental to the coherence of the defence narrative. It was not merely helpful context. This was not the case here. [ 67 ] Before this Court, Goldfinch sought to articulate specific issues which made the sexual nature of the relationship critical to his defence. He advanced that his narrative would be inherently improbable if the jury did not know he and the complainant were "friends with benefits". Among other things, he argued that the evidence was relevant to the jury's assessment of the complainant's call suggesting Goldfinch owed her "birthday sex" as well as to his testimony that the complainant smiled when he mouthed "I'm going to fuck you". [ 68 ] In my view, there is nothing about Goldfinch's testimony that casts him in an unfavourable light or renders his narrative untenable or utterly improbable absent the information that the two were "friends with benefits". The complainant's request for "birthday sex" does not reflect on Goldfinch's character or behaviour. As well, her reaction to his comment was a smile — hardly an indication that this behaviour was beyond the pale of their relationship. Tellingly, the complainant did not deny the call, Goldfinch's comment or the smile. ###### (c) Section 276(2)(c): Balancing Probative Value and Prejudice to the Proper Administration of Justice [ 69 ] The final step in the s. 276 analysis requires judges to balance the probative value of proposed evidence against the danger of prejudice to the proper administration of justice, taking into account the factors set out in s. 276(3). Both considerations must receive heightened attention as "[the test] . . . serves to direct judges to the serious ramifications of the use of evidence of prior sexual activity for all parties" (Darrach, at para. 40). Balancing the s. 276(3) factors ultimately depends on the nature of the evidence being adduced and the factual matrix of the case. It will depend, in part, on how important the evidence is to the accused's right to make full answer and defence. For example, the relative value of sexual history evidence will be significantly reduced if the accused can advance a particular theory without referring to that history. In contrast, where that evidence directly implicates the accused's ability to raise a reasonable doubt, the evidence is obviously fundamental to full answer and defence (Mills, at paras. 71 and 94). This was not the case here: Goldfinch's right to full answer and defence would not have been compromised by excluding the sexual nature of the relationship. [ 70 ] Indeed, having found that the "friends with benefits" evidence was not relevant to an issue at trial, it follows that it has no probative value. The evidence was relevant only to suggest that the complainant was more likely to have consented because she had done so in the past. Thus the evidence went only to the twin myths which, as Gonthier J. held in Darrach, are "not probative of consent or credibility and can severely distort the trial process" (para. 33). #### C. R. v. Graveline, 2006 SCC 16, [2006] 1 S.C.R. 609 [ 71 ] The improper admission of relationship evidence from which sexual activity may be inferred risks infecting a trial with the precise prejudicial assumptions s. 276 is designed to weed out. [ 72 ] In the case at bar, the "context" laid out before the jury was clearly infected with twin-myth reasoning. The jury should not have been privy to particulars regarding the frequency of the sexual contact or Goldfinch's testimony characterizing the evening as "typical" or "routine". That evidence clearly engages twin-myth reasoning by suggesting that because the complainant had "typically" consented to sex with Goldfinch in the past, she was more likely to have done so on this "routine" occasion. Admitting that evidence was a reversible error of law which might reasonably be thought to have had a material bearing on the acquittal (Graveline, at para. 14). #### D. Final Comments [ 73 ] Evidence of sexual relationships must be handled with care in sexual assault trials. [ 74 ] Where a trial judge is concerned that the jury may improperly speculate about past sexual activity, it may be helpful to give an instruction specifying that the jury will not hear any evidence about whether the relationship included a sexual aspect. The instruction should explain that the details of previous sexual interactions are simply not relevant to the determination of whether the complainant consented to the act in question. No means no, and only yes means yes: even in the context of an established relationship, even part way through a sexual encounter, and even if the act is one the complainant has routinely consented to in the past. Giving such an instruction would both reinforce the principles which guide a proper analysis of consent and mitigate the risk that jurors will rely on their own conceptions of what sexual activity is "typical" in a given relationship. [ 75 ] How evidence is to be adduced may also impact trial fairness. Much of the evidence that ultimately came out in this case was adduced during the Crown's examination of the complainant and, to a lesser degree, its cross-examination of Goldfinch. This requires two observations. First, I note that Crown counsel would not have adduced this evidence but for the s. 276 application, which I have concluded should not have been granted. While the parties did not have the benefit of this Court's recent holding in Barton, I would reiterate that Crown-led evidence of prior sexual activity must be governed by the principles set out in s. 276(1) and Seaboyer (Barton, at paras. 68, 80 and 197). Second, proper management of evidence which falls within the scope of the s. 276 regime requires vigilance from all trial participants, but especially trial judges — the ultimate evidentiary gatekeepers. Leading evidence through an agreed statement of facts, as the trial judge suggested here, is one way to do so. [ 76 ] I would dismiss the appeal. --- ## Concurring Reasons The reasons of Moldaver and Rowe JJ. were delivered by Moldaver J. — ### I. Overview [ 77 ] This case raises the difficult yet important issue of whether — and if so, to what extent — evidence of an ongoing sexual relationship between the accused and the complainant is admissible in a sexual assault trial. [ 78 ] The complainant in this case alleges that on May 28, 2014, the appellant, Patrick John Goldfinch, sexually assaulted her in his home. The incident resulted in Mr. Goldfinch being charged with both assault and sexual assault. Mr. Goldfinch maintained his innocence, claiming that the sexual activity in question was consensual. [ 79 ] At trial, Mr. Goldfinch applied under s. 276.1 of the Criminal Code, R.S.C. 1985, c. C-46 ("Code"),[^8] to adduce evidence that at the time of the alleged sexual assault, he and the complainant were in a "friends with benefits" relationship that involved occasional sex. The Crown opposed the admission of this evidence. Following a voir dire, the trial judge ruled that the evidence was admissible for the dual purposes of providing context to the events in question and avoiding an erroneous misapprehension on the part of the jury that Mr. Goldfinch and the complainant were platonic friends. The trial ended with the jury acquitting Mr. Goldfinch of both assault and sexual assault. [ 80 ] The Crown appealed Mr. Goldfinch's acquittals to the Alberta Court of Appeal. A majority of that court overturned the acquittals, holding that the trial judge erred in admitting evidence of the nature of Mr. Goldfinch's relationship with the complainant — "friends with benefits" — at the time of the alleged assault. In the majority's view, this evidence was inadmissible because it did nothing more than advance one of the twin myths prohibited by s. 276(1) — namely, that the complainant was more likely to have consented on the night in question because she had previously engaged in sex with Mr. Goldfinch. Justice Berger dissented. In his view, the impugned evidence was admissible under s. 276(2) because it was integral to Mr. Goldfinch's ability to make full answer and defence. Mr. Goldfinch now appeals to this Court as of right. [ 81 ] The s. 276 regime is sometimes viewed as a zero-sum game pitting the rights of the complainant against those of the accused. But in my respectful view, this is a mischaracterization. The s. 276 regime is designed to respect and preserve the rights of both complainants and accused persons by excluding evidence which would undermine the legitimacy of our criminal justice system and inhibit the search for truth, while allowing for the admission of evidence which would enhance the legitimacy of our criminal justice system and promote the search for truth. In this way, the regime seeks to promote the integrity of the trial process as a whole — a concept that is essential to the public's faith in the criminal justice system. [ 82 ] In pursuing this objective, the s. 276 regime operates in a step-by-step manner. From the accused's initial application under s. 276.1 to the final limiting instruction required by s. 276.4, the s. 276 regime establishes a rigorous, multistep process through which sexual activity evidence adduced by or on behalf of the accused must be carefully vetted and winnowed down to its essentials. To make its way into evidence at trial, such evidence must withstand careful scrutiny at each stage of the process. [ 83 ] As I will explain, for sexual activity evidence to be admitted under s. 276(2), the accused must demonstrate, at a minimum, that the evidence goes to a legitimate aspect of his defence and is integral to his ability to make full answer and defence. This requires that the accused be able to identify specific facts or issues relating to his defence that can be properly understood and resolved by the trier of fact only if reference is made to the sexual activity evidence in question. Further, to the extent sexual activity evidence is ultimately admitted, the trial judge must explain to the jury, in clear and precise terms, the uses for which the evidence may — and may not — be used. Finally, all trial participants — including the trial judge, Crown and defence counsel, and witnesses — must hew to the specific, legitimate purpose for which the evidence has been admitted, without expanding the scope of the ruling or using the admissible evidence for inadmissible purposes. This is essential to preserving not only the rights of the accused and the complainant, but also the integrity of the trial process as a whole. [ 84 ] In the case at hand, for reasons that follow, I am respectfully of the view that the trial judge erred in admitting the "friends with benefits" evidence under s. 276, having particular regard to the manifest deficiencies in Mr. Goldfinch's application to introduce this evidence. Furthermore, the improper admission of this evidence for the broad purpose of providing "context" led to a significant and highly prejudicial broadening of the sexual activity evidence at trial. This, in my view, might reasonably have had a material bearing on Mr. Goldfinch's acquittal. I would therefore dismiss the appeal. ### II. Facts and Decisions Below [ 85 ] My colleague Justice Karakatsanis has summarized the facts and decisions below in her reasons, and I see no need to replicate her work. I will therefore proceed directly to the issues on appeal. ### III. Issues [ 86 ] This appeal raises two issues: (1) Did the trial judge err in admitting under s. 276 evidence of the sexual nature of Mr. Goldfinch's relationship with the complainant at the time of the alleged sexual assault? (2) If so, could this error reasonably be thought to have had a material bearing on Mr. Goldfinch's acquittal such that a new trial is warranted? ### IV. Analysis [ 87 ] Before commencing my analysis, I would point out that my colleague Karakatsanis J. and I align on many if not most of the core principles governing the s. 276 regime. The main point that separates us is that while my colleague concludes that the sexual activity evidence in question — the existence of a "friends with benefits" relationship — is categorically barred from being admitted under s. 276, I take a more nuanced approach. In particular, as I will explain at paras. 121-125 of these reasons, I would leave open the question of whether the presiding judge at the new trial might, if presented with a properly framed s. 276.1 application, admit the evidence after applying the test and weighing the factors set out in s. 276(2) and (3). Having foreshadowed these points, I turn to my analysis. #### A. Overview of the Section 276 Regime[^9] [ 88 ] Section 276 of the Code governs the admissibility of evidence of the complainant's sexual activity other than that which forms the subject-matter of the charge. In these reasons, I will refer to such evidence simply as "sexual activity evidence".[^10] As I will explain, the various components of the s. 276 regime work in harmony to root out stereotypical and discriminatory reasoning and preserve the integrity of the trial process. [ 89 ] Section 276(1) reads: > Evidence of complainant's sexual activity

276 (1) In proceedings in respect of an offence under section 151, 152, 153, 153.1, 155 or 159, subsection 160(2) or (3) or section 170, 171, 172, 173, 271, 272 or 273, evidence that the complainant has engaged in sexual activity, whether with the accused or with any other person, is not admissible to support an inference that, by reason of the sexual nature of that activity, the complainant

(a) is more likely to have consented to the sexual activity that forms the subject-matter of the charge; or

(b) is less worthy of belief. [ 90 ] Stated succinctly, this provision prohibits the use of sexual activity evidence to support one of the twin myths identified by this Court in R. v. Seaboyer, [1991] 2 S.C.R. 577. In doing so, it gives effect to the principle that these myths "are simply not relevant at trial" and "can severely distort the trial process" (R. v. Darrach, 2000 SCC 46, [2000] 2 S.C.R. 443, at para. 33). Accordingly, if the sole purpose for which sexual activity evidence is being proffered is to support either of the twin myths, it will be ruled inadmissible under s. 276(1) — it is as simple as that. [ 91 ] But that does not mean sexual activity evidence will always be ruled inadmissible. Where the accused can identify a legitimate purpose for introducing the evidence — one that does not involve twin-myth reasoning — admission remains a possibility. In this regard, s. 276(2) provides that while sexual activity evidence adduced by or on behalf of the accused is presumptively inadmissible, regardless of the purpose for which it is proffered, such evidence may be admitted where it satisfies a three-part test: > Conditions for admissibility

276 (2) In proceedings in respect of an offence referred to in subsection (1), no evidence shall be adduced by or on behalf of the accused that the complainant has engaged in sexual activity other than the sexual activity that forms the subject-matter of the charge, whether with the accused or with any other person, unless the judge, provincial court judge or justice determines, in accordance with the procedures set out in sections 276.1 and 276.2, that the evidence

(a) is of specific instances of sexual activity;

(b) is relevant to an issue at trial; and

(c) has significant probative value that is not substantially outweighed by the danger of prejudice to the proper administration of justice. [ 92 ] In undertaking the analysis required by s. 276(2), the trial judge tasked with making a ruling on admissibility must take into account the factors listed in s. 276(3): > (a) the interests of justice, including the right of the accused to make a full answer and defence;

(b) society's interest in encouraging the reporting of sexual assault offences;

(c) whether there is a reasonable prospect that the evidence will assist in arriving at a just determination in the case;

(d) the need to remove from the fact-finding process any discriminatory belief or bias;

(e) the risk that the evidence may unduly arouse sentiments of prejudice, sympathy or hostility in the jury;

(f) the potential prejudice to the complainant's personal dignity and right of privacy;

(g) the right of the complainant and of every individual to personal security and to the full protection and benefit of the law; and

(h) any other factor that the judge, provincial court judge or justice considers relevant. [ 93 ] Thus, sexual activity evidence adduced by or on behalf of the accused that could potentially implicate one of the twin myths may nonetheless be admitted for a separate, legitimate purpose where it satisfies the test set out in s. 276(2) and (3). [ 94 ] But before sexual activity evidence can be admitted under s. 276(2), the accused must file an application under s. 276.1. This application must be made in writing and must set out (a) "detailed particulars of the evidence that the accused seeks to adduce" and (b) "the relevance of that evidence to an issue at trial" (s. 276.1(2)). Where the judge is satisfied that (a) the application was made in accordance with s. 276.1(2), (b) certain filing requirements have been met, and (c) the evidence sought to be adduced is capable of being admissible under s. 276(2), then he or she shall hold a voir dire — held outside the presence of the jury and the public — to determine whether the evidence is admissible under s. 276(2) (s. 276.1(4)). However, if the judge is not satisfied that each of these three requirements has been met (e.g., the written application is deficient), then he or she may dismiss the application without more. Thus, for example, where a written application under s. 276.1 does not satisfy the judge that the evidence sought to be adduced is relevant to an issue at trial, it can be dismissed without holding a voir dire. While the parties in this case have not focused their submissions on the s. 276.1 requirements, these requirements, and the consequences for failing to meet them, should be kept in mind going forward. [ 95 ] If the accused's written application survives scrutiny, then the process moves to the voir dire stage and the court's attention shifts to s. 276(2). In my view, to satisfy the relevance requirement under s. 276(2)(b), the accused must demonstrate that the evidence goes to a legitimate aspect of his defence and is integral to his ability to make full answer and defence. This requires that the accused be able to identify specific facts or issues relating to his defence that can be properly understood and resolved by the trier of fact only if reference is made to the sexual activity evidence in question. In articulating these specific facts or issues, simply citing the need to provide greater "context" or a fuller "narrative" will not suffice. Indeed, if this were enough to justify the admission of sexual activity evidence, then the s. 276 regime would be reduced to a mere statutory speed bump along the way to admission. [ 96 ] Furthermore, the requirement that the evidence be "integral" to the accused's ability to make full answer and defence means that even if the evidence can be linked to specific facts or issues relating to the accused's defence, admission is not guaranteed. There may be cases in which the evidence, while relevant to specific facts or issues relating to the accused's defence, bears only marginally on it. In such cases, the trial judge may, in his or her discretion, exclude the evidence on the basis that countervailing considerations, such as the need to protect the privacy rights and dignity of the complainant, outweigh the tenuous connection the evidence has to the accused's ability to make full answer and defence. [ 97 ] With these points in mind, when applying this regime, trial judges must take a careful, rigorous approach — one that recognizes and protects against the risks that accompany sexual activity evidence. The statutory requirements set out in s. 276(2) are designed to ensure that any admissible sexual activity evidence is limited in scope and that its legitimate purpose is identified and weighed against countervailing considerations. A careful application of these requirements is essential to the integrity of the trial process. [ 98 ] The integrity of the trial process can be further protected by adopting certain best practices for communicating general relationship evidence, such as the kind sought to be adduced in this case, after an admissibility ruling has been made. As the Attorney General of Ontario ("AGO") suggested in its intervener submissions before this Court, where general relationship evidence is ruled admissible under s. 276(2), the parties should reduce that evidence to a clear and concise agreed statement of facts prior to trial. This statement would ensure that the evidence elicited at trial does not stray beyond the parameters of the trial judge's ruling. Alternatively, the trial judge has the discretion to permit Crown and defence counsel to lead the complainant and the accused respectively through an accurate description of their relationship. These simple, prudent practices can go a long way to ensuring the scope of sexual activity evidence elicited at trial goes no further than necessary, thereby safeguarding the privacy rights and dignity of the complainant, preserving the integrity of the trial process, and promoting the search for truth. [ 99 ] Furthermore, where sexual activity evidence has been ruled admissible under s. 276(2), all trial participants — including the trial judge, Crown and defence counsel, and witnesses — must hew to the specific, legitimate purpose for which the evidence has been admitted, without expanding the scope of the ruling or using the admissible evidence for inadmissible purposes. Any new bases for admissibility that emerge during the course of the trial should be the subject of a fresh s. 276 application. And when sexual activity evidence is ultimately elicited at trial, it should be followed swiftly by a mid-trial limiting instruction to the jury that identifies both the permissible and impermissible uses of the evidence. This instruction should be reinforced by a final jury instruction. In short, everyone must get crystal clear — and remain crystal clear — about why the evidence is being admitted. [ 100 ] With these principles in mind, I will now consider whether the trial judge in this case erred in admitting the "friends with benefits" evidence under s. 276. #### B. Application of Section 276 [ 101 ] The determination of whether sexual activity evidence is admissible under s. 276 raises a question of law (see s. 276.5). A ruling on this matter is therefore reviewable on a standard of correctness. To determine whether the trial judge erred in this case, we must consider the facts as known to her at the time of the voir dire. [ 102 ] In his affidavit filed in support of his application under s. 276.1, Mr. Goldfinch stated that after their breakup, he and the complainant would "get together at [his] residence . . . to have sexual intercourse". He also noted that they "both referred to this relationship as a 'friend's [sic] with benefits' arrangement" (A.R., vol. I, at p. 5). At the s. 276.2 voir dire, Mr. Goldfinch argued that evidence of the sexual nature of their relationship was necessary to "provide proper context" and "to avoid leaving the jury with the erroneous impression that after the complainant and [he] ceased living together they maintained a purely platonic friendship" (A.R., vol. I, at p. 5). [ 103 ] In conducting the s. 276.2 voir dire,[^11] the trial judge was required to first consider s. 276(1) and then ask whether, in light of the bases for admission set out in Mr. Goldfinch's application, the proposed evidence satisfied the three requirements set out in s. 276(2), taking into account the factors listed in s. 276(3). I will consider and apply these provisions in turn, keeping in mind that the analysis must remain fixed at the time of the voir dire. [ 104 ] Before commencing this analysis, I note that the relationship evidence sought to be adduced in this case — evidence of a "friends with benefits" relationship involving occasional sex — directly implicates sexual activity and is therefore subject to the s. 276 regime. Whether general evidence of categories of relationships that may or may not involve sexual activity — such as marriage, dating, etc. — might also engage s. 276 is not before the Court. I would therefore leave that issue for another day. That said, without finally deciding the matter, I am inclined to the view that such evidence would not, without more, engage s. 276, as it does not amount to "evidence that the complainant has engaged in sexual activity" (s. 276(1) and (2)). Of course, if the accused were to seek to go into details about the relationship revealing that the complainant has engaged in sexual activity — for example, evidence that the accused and the complainant were in a spousal relationship involving regular sex (see R. v. L.S., 2017 ONCA 685, 40 C.R. (7th) 351) — then that evidence would have to pass through the s. 276 filter. ##### (1) Section 276(1) [ 105 ] As indicated, under s. 276(1), sexual activity evidence cannot be admitted to support the inference that, by reason of the sexual nature of the activity, the complainant was more likely to have consented to the sexual activity forming the subject-matter of the charge or is less worthy of belief. [ 106 ] In this case, as I have explained, Mr. Goldfinch sought to tender evidence that at the time of the alleged sexual assault, he and the complainant maintained a "friends with benefits" relationship that involved occasional sex. On its face, this evidence could potentially be used to support the first of the twin myths — the myth that because the complainant consented to have sex with Mr. Goldfinch in the past, she was more likely to have consented to the sexual activity forming the subject-matter of the sexual assault charge. However, s. 276(1) takes this potential use off the table. As such, unless Mr. Goldfinch could point to some legitimate use of the sexual activity evidence that would justify admission, that evidence was inadmissible. To determine whether a legitimate use justifying admission existed, we must turn to s. 276(2). ##### (2) Section 276(2) [ 107 ] Section 276(2) provides that in order to be admissible, sexual activity evidence must: (a) be of specific instances of sexual activity; (b) be relevant to an issue at trial; and (c) possess significant probative value that substantially outweighs the danger of prejudice to the proper administration of justice. I will consider and apply these requirements in turn. ###### (a) Specific Instances of Sexual Activity [ 108 ] The first requirement of s. 276(2) is that the evidence "be of specific instances of sexual activity". [ 109 ] In L.S., the Ontario Court of Appeal considered whether "relationship evidence" could satisfy s. 276(2)(a)'s specificity requirement. In that case, the complainant and the accused were in a spousal relationship. The complainant alleged that the accused sexually assaulted her. The accused applied to introduce evidence of the nature of his relationship with the complainant under s. 276. The affidavit filed in support of his application stated that he and the complainant "were together as a couple for the entire year of 2009" and that they had an "active sex life for that year" in which they "had sex on a regular basis" (para. 53). [ 110 ] Justice Doherty, writing for a unanimous panel, held that despite the absence of specific features of the sexual activity or individualized instances of sexual activity in the accused's affidavit, the evidence nonetheless satisfied the "specific instances" requirement under s. 276(2)(a). He reasoned that the content of the "specific instances" requirement is linked to the nature of the evidence sought to be adduced. Where the accused seeks to introduce evidence of an individual instance of sexual activity, he must identify that instance with specificity. By contrast, where the accused seeks to introduce general evidence that describes the nature of the relationship between the accused and the complainant"the specificity requirement speaks to factors relevant to identifying the relationship and its nature and not to details of specific sexual encounters" (para. 83). These factors will include "the parties to the relationship, the relevant time period, and the nature of the relationship" (para. 83). [ 111 ] I would adopt Doherty J.A.'s interpretation of s. 276(2)(a) insofar as it relates to evidence of the nature of the relationship between the complainant and the accused before, at the time of, and after the sexual activity forming the subject-matter of the charge. [ 112 ] Here, Mr. Goldfinch's affidavit specified the parties to the relationship (himself and the complainant), the relevant time period (subsequent to their breakup in 2012 until the alleged sexual assault in May 2014), and the nature of the relationship ("friends with benefits"). Therefore, the trial judge was correct in her conclusion that Mr. Goldfinch discharged the requirement under s. 276(2)(a) that the evidence be of specific instances of sexual activity. ###### (b) Relevant to an Issue at Trial [ 113 ] The second requirement of s. 276(2) is that the evidence be "relevant to an issue at trial". To fulfill this requirement, the accused must "establish a connection between the complainant's sexual history and the accused's defence" (Darrach, at para. 56). [ 114 ] Here, Mr. Goldfinch argued that the threshold for relevance was met because the "friends with benefits" evidence: (1) was necessary to avoid an erroneous misapprehension on the part of the jury that he and the complainant were platonic friends at the time of the alleged sexual assault; and (2) provided "context" to the issues at trial. I will consider each argument in turn. [ 115 ] By way of background, at the voir dire stage, Crown counsel proposed a statement of facts that did not include the existence of a "friends with benefits" relationship between Mr. Goldfinch and the complainant. Instead, the proposed statement of facts stated that Mr. Goldfinch and the complainant had previously been in a romantic relationship for seven or eight months, during which time they lived together. It further stated that they were friends at the time of the events in question and the complainant sometimes spent the night at Mr. Goldfinch's residence. [ 116 ] The problem with the first rationale advanced by Mr. Goldfinch is that it does not explain why it would be necessary to correct any potential "misapprehension" as to the sexual nature of his relationship with the complainant. A judge reviewing Mr. Goldfinch's application might legitimately conclude that the "friends with benefits" evidence was being proffered solely to inform the jury that the two were occasionally engaging in sexual intercourse during the relevant time frame, and to potentially support the impermissible inference that it was more likely that the complainant consented on the night in question. If this was the sole reason for which the defence sought to correct the supposed "misapprehension", then the evidence was inadmissible by virtue of s. 276(1). Assuming, however, that there may have been a legitimate reason to correct any potential "misapprehension" on the part of the jury, Mr. Goldfinch's application did not identify one. [ 117 ] As for the second rationale, Mr. Goldfinch argued that the sexual activity evidence was relevant to his defence because it provided "context" to the events that formed the subject-matter of the sexual assault charge. The trial judge described the existence of an ongoing sexual relationship between the complainant and Mr. Goldfinch as a "background piece of evidence" that the jurors could use when assessing the conflicting direct evidence "as to whether or not the complainant did consent to the events in question" (A.R., vol. I, at p. 9). [ 118 ] In my view, this line of reasoning paints with too broad a brushstroke. While it is true in every case that the existence of a "friends with benefits" relationship between the complainant and the accused provides "context" to the events at issue, neither Mr. Goldfinch's vaguely drafted affidavit nor his arguments at the voir dire stage specified the precise inference he wanted the jury to draw from that "context". Put differently, he failed to identify a specific, legitimate purpose for putting the "friends with benefits" evidence before the jury — he did not link the evidence to specific facts or issues relating to his defence that could be properly understood and resolved only if reference could be made to the "friends with benefits" evidence. [ 119 ] Where sexual activity evidence is concerned, the failure to identify the explicit link between the evidence and specific facts or issues relating to the accused's defence can result in twin-myth reasoning slipping into the courtroom in the guise of "context". For example, there is a risk that sexual activity evidence may be used, whether consciously or not, to "contextualize" a complainant's testimony that she did not consent to the sexual activity in question through twin-myth reasoning: because the complainant consented in the past (the "context"), it is more likely that she consented this time as well. This is, of course, precisely the sort of stereotypical reasoning s. 276(1) sought to banish from the courtroom. Yet without a clear and precise identification of the specific purpose for which sexual activity evidence is sought to be introduced, this sort of reasoning can all too easily infiltrate the courtroom through the Trojan horse of "context". [ 120 ] In sum, the "context" rationale offered by Mr. Goldfinch was insufficient to satisfy s. 276(2)(b). In my respectful view, any conclusion to the contrary would open the door to the admission of sexual activity evidence in every case where the complainant and the accused were in a previous or ongoing sexual relationship, so long as the accused cites the need for greater "context" in his application. This cannot have been Parliament's intent. The words "context" and "narrative" do not, in my view, offer the accused a means of bypassing the measured analysis required by s. 276. [ 121 ] Thus, the key flaw in Mr. Goldfinch's application was his failure to identify specific facts or issues relating to his defence that could be properly understood and resolved by the trier of fact only if reference was made to the "friends with benefits" evidence. Some discussion of what I mean by "specific facts or issues relating to [the accused's] defence" may prove helpful. As I will develop, without reaching any final decision on the matter, I note that in this case there was at least one specific issue that Mr. Goldfinch could have referred to in his application that might have properly supported admission of the "friends with benefits" evidence: the jury's assessment of Mr. Goldfinch's testimony that he mouthed the words "I'm going to fuck you" to the complainant shortly before they went downstairs. In discussing this example below, I should not be taken as foreclosing the possibility that additional facts or issues relating to Mr. Goldfinch's defence may be identified at the new trial that might support the admission of the "friends with benefits" evidence. [ 122 ] When explaining the events leading up to the sexual activity forming the basis of the sexual assault charge, Mr. Goldfinch testified as follows: > Well, we went downstairs, [the complainant] and I. I had asked her to come downstairs with me. Prior to that, we were talking and joking around and everything, getting along very well upstairs . . . and I kind of mouthed to her I'm going to fuck you. And she smiled, gave me a little cute smile back.

(A.R., vol. III, at p. 198) [ 123 ] Evidence that the complainant and Mr. Goldfinch were in a "friends with benefits" relationship at the time of these events may have provided necessary context to aid the jury in assessing Mr. Goldfinch's testimony that he mouthed the words "I'm going to fuck you" to the complainant. If the jury lacked the knowledge that the two were in a sexual relationship at the time, this statement by Mr. Goldfinch might have seemed bizarre or even menacing. Furthermore, Mr. Goldfinch's testimony that he made this statement to the complainant may itself have seemed implausible. In this way, withholding the "friends with benefits" evidence from the jury could have led them to make an adverse credibility determination against Mr. Goldfinch that they otherwise would not have made. A jury's determination of which witnesses to find credible is a holistic exercise that involves assessing the plausibility and coherence of a given witness's testimony throughout the course of the trial. Withholding the sexual nature of Mr. Goldfinch's relationship with the complainant could have had an adverse impact on the jury's assessment of Mr. Goldfinch's credibility, potentially infringing upon his right to make full answer and defence. [ 124 ] To be clear, however, just as generic references to "context" or "narrative" will not suffice to justify the admission of sexual activity evidence under s. 276(2), bare invocations of "credibility" will not be enough. Credibility is a key issue in almost every sexual assault trial — the centrality of credibility assessments does not, however, allow the accused to bypass the rigours of s. 276. Instead, where credibility is concerned, the accused must identify specific facts or issues that require reference to the sexual activity evidence to be understood and that could have a material impact on a credibility assessment. [ 125 ] Returning to the case at hand, Mr. Goldfinch might have been able to establish a legitimate link between the "friends with benefits" evidence and his credibility by pointing to the need to enable the jury to properly assess his testimony that he mouthed the words "I'm going to fuck you" to the complainant. Had Mr. Goldfinch referenced this aspect of his anticipated testimony in his s. 276 application, the trial judge would have been better equipped to engage in the balancing exercise required by s. 276(2) and (3). Indeed, she may have properly determined that the "friends with benefits" evidence was admissible for the narrow purpose of allowing the jury to assess Mr. Goldfinch's testimony on this point. She may have further found that the impact on the complainant's privacy rights and dignity would be limited, as the bare fact that Mr. Goldfinch and the complainant were in a "friends with benefits" relationship at the time of the alleged sexual assault would be unlikely to unduly arouse the jury's sentiments and would not require a sweeping and prejudicial inquiry into the complainant's sexual history. She may have also determined that admission of the evidence was necessary to ensure Mr. Goldfinch's right to make full answer and defence. [ 126 ] Be that as it may, the fact remains that in this case the "friends with benefits" evidence was admitted for a different — and much broader — purpose. It was ushered into the courtroom under the broad banner of "context", without any identification of the specific facts or issues relating to Mr. Goldfinch's defence that required reference to this "context" to be properly understood and resolved. In short, Mr. Goldfinch's application lacked the specificity and precision required to justify admission of the evidence in question. [ 127 ] For these reasons, I conclude that Mr. Goldfinch's s. 276 application did not satisfy the "relevant to an issue at trial" requirement under s. 276(2)(b). Accordingly, it ought to have been denied. ###### (c) Probative Value Versus Potential Prejudicial Effect [ 128 ] The third requirement of s. 276(2) is that the evidence have "significant probative value that is not substantially outweighed by the danger of prejudice to the proper administration of justice". As this Court explained in Darrach"[t]he requirement of 'significant probative value' serves to exclude evidence of trifling relevance that, even though not used to support the two forbidden inferences, would still endanger the 'proper administration of justice'" (para. 41). [ 129 ] Evidence that is not relevant under s. 276(2)(b) is necessarily incapable of possessing any probative value, as the concept of probative value presupposes that the evidence bears on a relevant issue. Thus, given my conclusion that the sexual activity evidence in this case was not relevant to an issue at trial based on the application presented to the trial judge, I conclude that the evidence was incapable of satisfying the third requirement under s. 276(2). [ 130 ] That said, in the interest of providing further guidance, I will briefly explain why the lack of specificity in Mr. Goldfinch's application also adversely impacted the trial judge's ability to balance the probative value of the evidence against its potential prejudicial effect as required by s. 276(2)(c). [ 131 ] To enable the trial judge to properly assess whether sexual activity evidence sought to be adduced has significant probative value that substantially outweighs its potential prejudicial effect, the accused must situate the evidence within the particular factual matrix of the case at hand. It is only through awareness of specific facts or issues that require the admission of the evidence that the trial judge can determine whether the danger of prejudice to the proper administration of justice presented by the evidence is outweighed by its probative value. The trial judge tasked with performing this balancing exercise will not be aided by generic references to "context""narrative", or potential "misapprehensions"; more is needed. [ 132 ] This conclusion is underscored by the fact that the trial judge must consider a number of factors under s. 276(3) when making a determination under s. 276(2). These factors are specific in nature and require that the trial judge be in a position to evaluate the precise impact that the exclusion of the evidence would have on the accused's ability to make full answer and defence. The accused must provide the trial judge with the necessary tools to perform this inquiry, including a precise and clearly articulated connection between the sexual activity evidence and specific facts or issues relating to his defence. Unfortunately, that was not done here. ###### (d) Conclusion on Admissibility [ 133 ] For these reasons, although Mr. Goldfinch's application satisfied the "specific instances of sexual activity" requirement under s. 276(2)(a), it satisfied neither the relevance requirement under s. 276(2)(b) nor the balancing exercise required by s. 276(2)(c). As such, it was a legal error for the trial judge to admit the sexual activity evidence under s. 276(2). [ 134 ] In light of this conclusion, it becomes necessary to consider whether the trial judge's error warrants a new trial. #### C. Whether a New Trial Is Warranted [ 135 ] Under the test articulated in R. v. Graveline, 2006 SCC 16, [2006] 1 S.C.R. 609, an acquittal will be overturned only where the Crown can demonstrate that the trial judge made a legal error that "might reasonably be thought, in the concrete reality of the case at hand, to have had a material bearing on the acquittal" (para. 14). For reasons that follow, I am satisfied that the improper admission of the "friends with benefits" evidence meets this test. ##### (1) The Trial Judge's Limiting Instructions [ 136 ] An important feature of the s. 276 regime is its requirement of mandatory limiting instructions. Section 276.4 provides that where sexual activity evidence is admitted pursuant to s. 276(2)"the judge shall instruct the jury as to the uses that the jury may and may not make of that evidence". In this case, while the judge's limiting instructions cautioned the jury not to use the evidence of the sexual nature of Mr. Goldfinch's relationship with the complainant to support either of the twin myths, the instructions did not identify any permissible use of this evidence. The trial judge's final instruction to the jury on this point was as follows: > You heard some evidence that Mr. Goldfinch and [the complainant] had dated, briefly lived together, and continued to have sexual relations on occasion in 2014. This evidence provides you with some context for their relationship. You must not use this evidence, however, to help you decide or infer that, because of the sexual nature of their relationship, [the complainant] is more likely to have con[s]ented to sexual intercourse with Mr. Goldfinch on May 29th, 2014. . . . Further, you must not use the evidence of their relationship to help you decide or infer that [the complainant] is less believable or reliable as a witness in this case . . . .

(A.R., vol. I, at p. 37) [ 137 ] This instruction successfully identified the twin myths and cautioned the jury not to rely on them. However, the trial judge's statement that the evidence provided the jury with "some context for their relationship" was insufficient. The instruction did not identify any specific use to which the jury could properly put the evidence. As a result, the jury was left in the dark about how they could use the evidence. ##### (2) Other Sexual Activity Evidence Adduced at Trial [ 138 ] As indicated, the trial judge's instructions to the jury failed to identify specific facts or issues to which the sexual activity evidence related. Moreover, compounding the risk of forbidden reasoning based on discriminatory generalizations, the trial judge admitted evidence of the complainant's sexual activity going beyond the scope of her initial s. 276 ruling. [ 139 ] I note at the outset of this discussion that both the Crown and defence were responsible for adducing sexual activity evidence that went beyond the scope of the trial judge's ruling. In my view, the fact that both defence and Crown counsel strayed beyond the boundaries of the judge's ruling further illustrates the insufficiency of that initial ruling. Because the judge admitted the "friends with benefits" evidence under the broad banner of "context", the parties were left without any clear guidance on what other pieces of sexual activity evidence might be admissible under this broad banner. They may well have thought that the door had been swung wide open on sexual activity evidence because all such evidence would provide "context" to the events at issue in a general sense. Therefore, the overbreadth and irrelevance of the sexual activity evidence adduced at trial should serve as a cautionary tale of what can go wrong when the s. 276 regime is not properly observed. [ 140 ] At trial, both the complainant and Mr. Goldfinch were questioned about sexual activity that went beyond the scope of the s. 276 ruling. The trial judge initially determined that the only evidence which was admissible under s. 276(2) was the fact that the complainant and Mr. Goldfinch were in a "friends with benefits" relationship. However, at trial, defence counsel cross-examined the complainant about how many times she had sex with Mr. Goldfinch following their breakup. This directly contravened one of the bases for the trial judge's ruling — namely, that any prejudice to the complainant's privacy rights and dignity would be minimal because she would not be asked about the "particulars of . . . her sexual history with [Mr. Goldfinch]" (A.R., vol. I, at p. 11). That the complainant was ultimately asked about such particulars on cross-examination constituted a significant infringement on her privacy rights and dignity. [ 141 ] Furthermore, defence counsel asked the complainant about "other partners" she may have had at the time she and Mr. Goldfinch were in a "friends with benefits" relationship (A.R., vol. II, at p. 119). She testified that she had a sexual partner other than Mr. Goldfinch at the time. This evidence ran a serious risk of distorting the jury's fact-finding process by inviting stereotypical reasoning about women and consent, such as the myth that women who are "unchaste" are more likely to consent. I note that the defence did not apply to have the trial judge revisit her initial s. 276 ruling before eliciting this evidence. As such, the evidence was not the subject of a limiting instruction. [ 142 ] The Crown also strayed beyond the bounds of the trial judge's ruling. It is true that the Crown is not subject to the procedural requirements of ss. 276.1 and 276.2, which apply only where the accused seeks to adduce evidence of the complainant's other sexual activity. However, the Crown is subject to s. 276(1)'s prohibition on twin myth reasoning and must also abide by the common law principles articulated by this Court in Seaboyer. Indeed, in R. v. Barton, 2019 SCC 33, [2019] 2 S.C.R. 579, this Court stated that trial judges should determine the admissibility of Crown-led prior sexual activity evidence through a voir dire prior to trial, applying this Court's guidance in Seaboyer (Barton, at para. 80). [ 143 ] No such procedure was observed here. During direct examination of the complainant, Crown counsel elicited testimony that made reference to the complainant's other sexual partner. Further, in the Crown's cross-examination of Mr. Goldfinch, counsel led Mr. Goldfinch to provide details of his sexual "routine" with the complainant — namely, that they would generally have sex when the complainant stayed over at his house. This was prejudicial information about other instances of sexual activity by the complainant and should have been the subject of its own voir dire and ruling. [ 144 ] For reasons of procedural fairness, the fact that the Crown strayed beyond the bounds of the trial judge's ruling and elicited additional sexual activity evidence that was not the subject of a voir dire clearly cannot be used as a basis for setting aside Mr. Goldfinch's acquittal. Put simply, the Crown cannot be permitted to benefit from its own missteps. Nevertheless, much of the improper evidence was led by the defence. This resulted from the erroneous admission of the "friends with benefits" evidence under the broad banner of "context". In these circumstances, the public interest in a trial that is conducted properly according to the law and the need to protect the integrity of the justice system militate in favour of a new trial (see R. v. Morrison, 2019 SCC 15, [2019] 2 S.C.R. 3, at para. 142). [ 145 ] I would add that, as the AGO points out, vetting sexual activity evidence advanced by either Crown or defence before trial will have the salutary effect of focusing all parties on the legitimate use of such evidence. This approach will also "establish the parameters of any sexual relationship evidence that the Crown seeks to adduce" and trigger defence counsel to bring a separate application under s. 276(2) if they wish to elicit evidence going beyond the scope of that proposed by the Crown (I.F., at para. 14). The failure to do so here led to the unnecessary and inappropriate admission of sexual activity evidence that went beyond the trial judge's s. 276 ruling. This evidence came in without first having been vetted, and it was never the subject of its own limiting instruction. This was a serious error. [ 146 ] To be clear, I make no final determination as to whether any evidence going beyond the bare fact that Mr. Goldfinch and the complainant were "friends with benefits" at the time of the alleged sexual assault might ultimately be found admissible at a new trial where it is subject to proper vetting. Here, I simply point out the overbreadth of the sexual activity evidence elicited at trial as a means of illustrating the adverse impact that an improper admissibility determination under s. 276 can have on the integrity of the trial process. This case serves as a powerful illustration of how a trial can go off the rails where sexual activity evidence is admitted without being anchored to a specific, legitimate purpose. Without this anchor, the trial may drift into unsafe waters, as occurred in this case. And while the trial judge in this case made an effort to mitigate the situation, the damage was already done — Mr. Goldfinch's application should not have been allowed in the first place. [ 147 ] In conclusion, the errors at trial stemmed directly from the trial judge's improper s. 276 ruling, which allowed for the admission of sexual activity evidence under the broad banner of "context". Grounded in this ruling, the trial judge's flawed limiting instructions failed to delineate how the sexual activity evidence was capable of assisting the jury to resolve specific facts or issues relating to Mr. Goldfinch's defence. This flawed instruction's distorting effect was compounded when additional sexual activity evidence was admitted at trial which was not the subject of its own admissibility determination or limiting instruction. The cumulative impact of these errors, in my view, can reasonably be thought to have had a material bearing on Mr. Goldfinch's acquittal. Accordingly, I agree with the Court of Appeal that a new trial should be ordered. ### V. Conclusion [ 148 ] In the result, I would dismiss the appeal. --- ## Dissenting Reasons The following are the reasons delivered by Brown J. (dissenting) — ### VI. Introduction [ 149 ] Patrick Goldfinch appeals, as of right, from the decision of a majority of the Court of Appeal of Alberta setting aside his acquittals and ordering a new trial. The issue to be decided is whether the trial judge committed reversible error in admitting evidence under s. 276 of the Criminal Code, R.S.C. 1985, c. C-46, that the complainant and appellant were "friends with benefits".[^12] [ 150 ] Before the jury, the complainant testified that she was sexually assaulted after the appellant would not take "no" for an answer. Her evidence was that she called the appellant to come over to his residence. After watching television together with the appellant's roommate, she and the appellant went downstairs to his living quarters. She told him that "nothing was going to happen" (A.R., vol. II, at p. 87), meaning there would be no sexual activity. After she did not join him in his bedroom, he got angry, dragged her into the bedroom, pushed her onto the bed, hit her face, and had sex with her while she repeatedly told him to stop. [ 151 ] The appellant, on the other hand, testified that the complainant had called him to ask if she could come over because he owed her "birthday sex". He picked her up and they drove back to his residence where they watched television and drank beer with his roommate. At some point, the appellant mouthed at her "I'm going to fuck you" (A.R., vol. III, at p. 198), after which the two walked downstairs together to his living quarters. He eventually followed her into his bedroom, where she got undressed before he did. They discussed which side of the bed she would sleep on to avoid "hot flashes" that happen when she sleeps on one particular side of his bed. They had sex and fell asleep. He awoke to find her angry at him for hitting her in the face sometime while he was asleep. He became annoyed, and told her to leave. [ 152 ] In closing submissions to the jury, the appellant emphasized 11 aspects of the complainant's evidence that, in his view, exposed inconsistencies in her testimony sufficient to raise a reasonable doubt. None of these submissions referred to the relationship. Rather, it was the Crown that relied on the relationship. Its theory was that, because they were in a sexual relationship"Mr. Goldfinch was expecting sex from [the complainant] that day. That's the only reason he went to pick her up. When he found out that he was not going to get it, he snapped" (A.R., vol. III, at p. 286). [ 153 ] The trial judge would twice instruct the jury that the relationship provided context for how the appellant and complainant knew each other, but it could not be used to infer that, because they were in a relationship, the complainant was more likely to consent on this occasion, or that the complainant's account was less believable overall. [ 154 ] The jury returned a verdict of not guilty. [ 155 ] I have read carefully the reasons of the majority. I agree that this appeal turns on how to balance the admissibility of evidence of a relationship that may be a necessary link in making full answer and defence with ensuring triers of fact do not use the existence of a relationship to engage in prohibited reasoning. And while I agree with much of the majority's overview of s. 276 of the Criminal Code, we disagree on its application to this evidentiary record. The majority reasons that the relationship was simply not relevant to any issue before the jury, and therefore engaged the prohibited line of reasoning that "because the complainant had consented to sex with Goldfinch in the past, in similar circumstances, it was more likely she had consented on the night in question" (para. 47). [ 156 ] This appeal therefore presents three issues: (1) does the ordinary meaning of "friends with benefits" invariably lead the trier of fact to engage in reasoning prohibited by s. 276(1); (2) was the nature of their relationship relevant to any issue before the jury under s. 276(2)(b); and (3) should a new trial be ordered in the event the trial judge erred? [ 157 ] These are difficult questions that go to the heart of s. 276. My view is that the trial judge applied the correct legal principles in her evidentiary ruling. The majority sees it differently, describing the "clea[r] infect[ion]" of the context laid out before the jury with twin-myth reasoning, making the jury "privy to particulars" that it should not have known about (para. 72). While the majority reasons that this must have been caused by an error in the evidentiary ruling before trial, as I see it, the responsibility for any "infection" during this trial lies not in the trial judge's ruling, but rather at the feet of the Crown, who must live with its choices. [ 158 ] I would answer the legal questions as follows. First, I say that evidence of a "friends with benefits" relationship will, in certain cases, and without engaging in prohibited lines of reasoning, explain to a jury how two people know each other, consistent with how other relationships are presented to juries. I agree with the majority that relationship evidence, without proper scrutiny under s. 276 and sufficient limiting instructions, can trigger triers of fact to engage in prohibited lines of reasoning. But I do not accept the majority's solution to this problem, which would treat evidence of certain relationships (for example"married""dating", or "boyfriend-girlfriend"), each of which might also suggest previous sexual activity, differently in the s. 276 analysis based on how the particular relationship's sexual features, real or imagined, are reasonably perceived. I prefer the trial judge's evidentiary ruling, which treated the relationship between the appellant and the complainant consistently with other relationships in society. The jury could be told about the existence and nature of the relationship, but could not be told about its particularities or the frequency of sexual relations. The jury was then instructed twice on the permissible use of the evidence (to provide context for how the appellant and the complainant knew each other) and its impermissible uses (to suggest that, because they were friends with benefits, the complainant was more likely to consent on this occasion, or was less worthy of belief overall). [ 159 ] Secondly, the trial judge's evidentiary ruling was consistent with the proposition that the existence of a relationship between the complainant and the accused may be admitted as relevant, but strictly where its absence would leave the trier of fact with a distorted representation of the circumstances surrounding the incident. The trial judge concluded this was such a circumstance. While the majority disagrees, and finds nothing in the appellant's evidence that rendered his narrative "untenable" or "utterly improbable" (para. 68) absent reference to the relationship, it is hard to find fault in the trial judge's determination on this point when the Crown itself took a similar position — specifically, that admitting evidence of the relationship was necessary to prevent distortion. The Crown's objection was merely to the form, rather than the substance, of the appellant's description of the relationship. [ 160 ] I am not persuaded the trial judge committed an error in law sufficient to ground a Crown appeal from acquittal. This leads to my third point of departure from the majority. Ordering a new trial is unfair, given the following considerations: the Crown's theory of the case against the appellant drew directly from the sexual nature of this relationship; the "infection" to which the majority refers was largely as a result of a contravention by Crown counsel of the terms of the trial judge's evidentiary ruling; and, a successful Crown appeal from acquittal on this record inevitably lowers the bar which the Crown must overcome to show that a legal error had a material bearing on the acquittal so as to secure a new trial. How, after all, can the admission of this evidence be said to have had a material bearing on the acquittals, when the Crown was relying on the selfsame evidence to prove its case? The Crown was perfectly content in this case to refer, rely and ultimately weaponize the sexual past between the appellant and the complainant, until the jury returned a verdict of not guilty. Only then did the Crown on appeal take the position that the sexual nature of the relationship must have led the jury to engage in prohibited reasoning — and that, notwithstanding the Crown's reliance on the sexual nature of that relationship, the appellant should not have been able to answer it. This is fundamentally unfair. [ 161 ] These are the points that divide us. The jury rendered its verdict after being properly instructed on how to do so. I would allow the appeal and restore the acquittals. --- ### VII. The Voir Dire Decision — Court of Queen's Bench of Alberta, 140600008Q1, January 23, 2017 [ 162 ] Before the trial, the appellant applied under s. 276.1 of the Criminal Code to adduce evidence that he and the complainant were friends with benefits, in order to dispel any erroneous impression that the jury might be under that he had sexualized an otherwise platonic friendship, or to avoid misleading the jury into believing that they were complete strangers. The Crown objected, saying that friends with benefits would trigger prohibited twin myth reasoning in the jury. It did agree, however, that the jury would need some context to dispel the notion that the appellant and the complainant were strangers. To be clear, the Crown's objection was not to the injection of context, or, more broadly, to relationship evidence that insinuates past sexual activity, but to the term "friends with benefits". The Crown's solution was to tell the jury that they dated for seven to eight months, broke up, then reconnected; and that, on occasion, the complainant would visit the appellant's residence late in the evening and leave the following morning. [ 163 ] I stress that it was the Crown's position that it would not offend s. 276 or undermine the interests it is intended to protect for the jury to be told that the complainant would occasionally "visit" the appellant — a former boyfriend with whom she had reconnected — late in the evening, and that she would leave only the following morning. [ 164 ] The trial judge considered the competing positions. She directly addressed the prejudice that could flow to the complainant if the jury drew certain (prohibited) inferences from the nature of the relationship. She queried defence counsel during the voir dire: "what is the safeguard then to prevent . . . one of the twin myths that because she consented to the [friends] with benefits relationship one, two, 10 or 20 times before that she is likely to have consented on this occasion. What is the protection if this evidence is . . . permitted?" (A.R., vol. II, at p. 14). [ 165 ] In her evidentiary ruling, the trial judge agreed that without some context for how the appellant and the complainant knew each other, the appellant could not make full answer and defence. Bearing in mind that his credibility would be the central issue at trial, the jury could not properly assess (or indeed might well misconceive) the plausibility of his version of events were it prevented from understanding how the appellant and the complainant knew each other, or were it to simply assume they were strangers because no party had raised the issue of how they knew each other. [ 166 ] Admitting evidence of the relationship would be no different than admitting evidence that an accused and a complainant were "married""dating", or "boyfriend-girlfriend". The point being, the trial judge believed the jury ought to be told how they knew each other in order to avoid misinterpreting things the appellant said or did as having arisen out of the blue (A.R., vol. I, at p. 6). As she viewed the matter, keeping the existence of the relationship from the jury would leave a level of "artificiality" in the appellant's evidence, and would prevent him from being able to make full answer and defence. [ 167 ] Following a thorough examination of the relevant factors under s. 276(2) and (3), the trial judge chose to admit the friends with benefits evidence, rather than the Crown's proposed account of former lovers, who had reconnected to have adult sleepovers. As the majority notes, the trial judge remained very much alive to how a reference to the relationship could be misused to engage in prohibited lines of reasoning in relation to the complainant and her evidence. And she could not have been clearer in directing to the parties that, while evidence could be led about the relationship between the appellant and the complainant, no evidence was to be led about the relationship's particularities ("the specific sexual acts [the complainant] and the accused engaged in"), or frequency ("the number of times they had sexual relations after they ceased living together") (A.R., vol. I, at p. 10). And once again when, on the morning of trial, Crown counsel sought clarification about the scope of the evidentiary ruling, the trial judge reiterated that no evidence was to be led about particularities or frequency of the relationship.[^13] All counsel acknowledged they understood. --- ### VIII. The Trial — Court of Queen's Bench of Alberta, 140600008Q1, February 9, 2017 [ 168 ] At trial, the complainant initially testified that, at the time of the assault, she was "just friends" with the appellant, that at no time since their formal relationship ended were they anything more than "friends" (A.R., vol. II, at p. 77), and that she and the appellant "weren't speaking" in the months leading up to the alleged assault (A.R., vol. II, at pp. 77-78). Almost immediately after that, however, and still in chief, she testified that she and the appellant previously had sex in his bedroom. Crown counsel asked whether this occurred while they used to date, and the complainant answered that it occurred after they had broken up. Crown counsel then went one step further — contrary to the trial judge's clear and repeated instructions — and asked whether they had sex on a single occasion or on "various" occasions after they had broken up, and when was the last time she had sex with the appellant. The complainant confirmed that they had sex on "various" occasions after they ceased living together. [ 169 ] In light of this evidence adduced from the complainant by the Crown, the trial judge at the conclusion of examination-in-chief notified counsel that following the complainant's testimony she would provide a mid-trial limiting instruction to the jury warning against impermissible uses of evidence that refers to the complainant's sexual activity. After canvassing the text of the proposed instruction with counsel, and following the complainant's evidence, the trial judge gave the following mid-trial instruction: > THE COURT: . . .

You have heard evidence that [the complainant] and Mr. Goldfinch dated and then briefly lived together. At some point after that relationship ended, [the complainant] and Mr. Goldfinch did on occasion get together and have sexual relations. This evidence provides you with some context for their relationship, but you must not use this evidence to help you decide that because [the complainant] and Mr. Goldfinch had sexual relations in the past, that [the complainant] is more likely to have consented to Mr. — consented to what Mr. Goldfinch is alleged to have done on May 29th, 2014, and you must also not use that evidence to help you decide that because [the complainant] and Mr. Goldfinch had sexual relations in the past that she is less believable or reliable as a witness in this case, all right? Thank you.

(A.R., vol. II, at p. 148) [ 170 ] I have already summarized the pertinent aspects of the appellant's evidence. In his closing submissions to the jury, the following pieces of evidence formed the central planks of his case: > [MS. HATCH] I'm going to ask you to consider 11 important items, and I'm going to list them first for you. And there may be other items that you find important as well, and obviously you will consider those. I would like you though to consider these 11 items carefully.

Number 1, the phone in the bedroom.

Secondly, wanting Mr. Goldfinch off the computer.

Thirdly, the timing, and I'm going to comment more on each of these in a few minutes.

The fourth item is the absence of any evidence of any injury to the complainant's genitals.

Number 5, the glasses being undamaged.

Number 6 is the kiss.

Number 7, phoning the cab.

Number 8, being worried that he'd lost everything.

Number 9, testimony that he may have been flaccid throughout.

And number 10, extensive lack of recall.

Lastly, number 11, all hell broke loose. There was yelling. It was loud. And I'm going to invite you to look closely at each of those items.

(A.R., vol. III, at p. 275) [ 171 ] In contrast, the Crown advanced the following theory to the jury: > [CROWN] The Crown's submission is that Mr. Goldfinch was expecting sex from [the complainant] that day. That's the only reason he went to pick her up. When he found out that he was not going to get it, he snapped.

(A.R., vol. III, at p. 286) The Crown elaborated on this theory at the outset of its closing submissions by telling the jury: > [CROWN] In this case, you heard [the complainant], a 54-year-old woman, tell you that she was sexually assaulted. She and Mr. Goldfinch dated for seven or eight months. They lived together. She ended the relationship. Her evidence on those points is uncontradicted. She told you that she called him. He picked her up. She was open to the possibility of having sex with him. When the offer was made to accompany Mr. Goldfinch downstairs, she knew what that really meant, and she made it clear that she would come downstairs but that there would be no sex.

(A.R., vol. III, at p. 283) [ 172 ] In her final charge to the jury, the trial judge instructed: > [THE COURT] Finally, I wish to talk to you about the evidence you heard of the prior sexual relationship between Mr. Goldfinch and [the complainant]. You heard some evidence that Mr. Goldfinch and [the complainant] had dated, briefly lived together, and continued to have sexual relations on occasion in 2014. This evidence provides you with some context for their relationship. You must not use this evidence, however, to help you decide or infer that, because of the sexual nature of their relationship, [the complainant] is more likely to have [consented] to sexual intercourse with Mr. Goldfinch on May 29th, 2014. The fact that they had sexual relations in the past or the extent of their sexual relations is irrelevant to the issue of consent on this particular occasion. Our law requires that consent be given to sexual activity on each and every occasion.

Further, you must not use the evidence of their relationship to help you decide or infer that [the complainant] is less believable or reliable as a witness in this case, but you may consider any contradictions in her evidence about the nature of their relationship in assessing credibility, as you would with any witness.

(A.R., vol. III, at p. 362) [ 173 ] The jury returned a verdict of not guilty. --- ### IX. The Appeal — 2018 ABCA 240, 48 C.R. (7th) 22 [ 174 ] The Crown appealed, arguing the trial judge erred by allowing the jury to hear that the appellant and complainant were friends with benefits, because it was used to advance inferences that were prohibited under s. 276(1) of the Criminal Code — specifically, that the complainant was more likely to consent to the sexual activity with the appellant, or that she was less worthy of belief overall. A majority at the Court of Appeal of Alberta agreed, and ordered a new trial (at paras. 1-55), with Berger J.A. dissenting. The relationship evidence had no relevant purpose other than to engage in prohibited reasoning, and its admission ran afoul of s. 276(1). --- ### X. Analysis [ 175 ] In order to appeal from an acquittal, the Crown must demonstrate that the trial judge erred on a point of law. The legal question before us is, therefore, whether the trial judge erred in law because some legal principle precludes relationship evidence, namely "friends with benefits" evidence, from being admitted in these circumstances. Any alleged errors reduce to this: either the trial judge erred, as the ordinary meaning of "friends with benefits" invariably leads the jury to engage in prohibited reasoning under s. 276(1), or she erred because the evidence that they were "friends with benefits" was simply not relevant to any issue at trial under s. 276(2)(b). [ 176 ] With respect for those holding contrary views, no such errors are shown here. #### A. The Trial Judge Did Not Err in Permitting the Evidence to Filter Through Section 276(1) [ 177 ] Evidence that an accused and a complainant are in a friends with benefits relationship, without more, does not solely derive its relevance in a criminal trial from lines of reasoning prohibited by s. 276(1). But this is exactly what the majority finds: the obvious implication of friends with benefits is that the parties are in a relationship understandable only by reference to its sexual features. The evidence, then, is seen as denoting to the jury nothing more than a relationship characterized by (and only by) a repeated pattern of sexual activity; the allure of twin myth reasoning becomes irresistible. [ 178 ] Respectfully, I do not share this starting premise. While I agree that the jury could draw from prohibited lines of reasoning in these circumstances, the test for exclusion under s. 276(1) is not satisfied merely where a jury could draw such inferences. The risk that the jury might erroneously infer consent from evidence of the nature of a relationship is a realistic danger in all relationship evidence — which, I note, the majority does not categorically exclude under s. 276(1). Rather, the test for exclusion under s. 276(1) is whether the evidence solely derives its relevance from prohibited lines of reasoning. [ 179 ] In other words, s. 276(1) categorically excludes evidence of other sexual activity where its relevance is premised on one or both of the twin myths. But that is only part of the inquiry. The analyses under ss. 276(1) and 276(2) are not isolated from each other. Rather, they apply together since s. 276(2) creates a presumption of inadmissibility, making all evidence of the complainant's sexual history inadmissible unless it is (a) of a sufficient specificity for the interests underlying s. 276 to be properly weighed, (b) relevant to an issue at trial that does not draw its relevance from prohibited lines of reasoning under s. 276(1),[^14] and (c) of significant probative value that is not substantially outweighed by its prejudicial effect (R. v. L.S., 2017 ONCA 685, 40 C.R. (7th) 351, at paras. 82-83; E. Craig"Section 276 Misconstrued: The Failure to Properly Interpret and Apply Canada's Rape Shield Provisions" (2016), 94 Can. Bar Rev. 45, at p. 53). If the evidence cannot meet each of these statutory criteria under s. 276(2), it is inadmissible. [ 180 ] In practice, therefore, the voir dire judge should determine whether the evidence is of specific instances of sexual activity, is relevant to an identifiable issue at trial that does not draw from prohibited lines of reasoning set out in s. 276(1), and has significant probative value that is not outweighed by substantial prejudice. [ 181 ] To be clear, however, and to repeat: the test for exclusion under s. 276(1) is whether the evidence derives its relevance solely from twin myth reasoning. Section 276(1) does not operate to exclude evidence which merely "engages" (majority reasons, at para. 42) twin myth reasoning. Were "engagement" the test for categorical exclusion under s. 276(1), it would risk exclusion of all relationship evidence, or at least all evidence of relationships which also involve sexual activity, since that, too, would conceivably "engage" twin myth reasoning. Rather, relationship evidence will typically be filtered via the inquiry contemplated by s. 276(2)(b), being whether, in this case, the relationship evidence is relevant to an identifiable issue at trial. [ 182 ] The heart of the Crown's appeal is that "friends with benefits" is too "sexual" in its ordinary meaning, and therefore solely derives its relevance from prohibited lines of reasoning such that it is categorically excluded under s. 276(1). But the Crown does not explain why evidence of the relationship which this term denotes is objectionable, while evidence of other relationships which regularly pass through the filter of s. 276(1), and yet which might also suggest previous sexual activity — for example"married""dating", or "boyfriend-girlfriend", or, for that matter, the Crown's preferred relationship description in this case of former lovers who enjoy adult sleepovers — is not. [ 183 ] The trial judge reasoned, entirely correctly in my view, that Canadian jurors understand that relationships come in many forms. And, in any event, she made the jury aware of the permissible purpose for which the evidence of this relationship was admitted (to explain how the appellant and the complainant knew each other), and of the impermissible inferences they could not draw from its use in popular imagination. [ 184 ] It follows that the foundation of the majority's test for admitting relationship evidence under s. 276(1) is that some relationships have aspects which transcend sexual intimacy, while other relationships (which also involve sexual activity) can be understood only through the lens of that sexual activity. The majority reduces the relationship in this case to its most carnal features, and leaves it to trial judges, moving forward, to determine whether the sexual features, real or imagined, of the particular relationship term before them would likely be used to engage in prohibited lines of reasoning. But, and I say this respectfully, this posture is unprincipled and rests on assumptions about relationships that I do not share. At the risk of stating the obvious, even relationships that involve sexual activity, but which lack the expectation or even desire of a more formal relationship such as marriage or common law partnership, can be about much more than the sexual activity. At the very least, evidence of such relationships will give important context to the non-sexual interactions between the parties to the relationship, which may be (as it is here) necessary to making full answer and defence. [ 185 ] The majority sweeps these concerns aside. Instead, we are left only with the undesirable effect of the majority's reasons, being their creation of legally significant distinctions in this area of law between the "right" kind of relationship with sexual aspects (i.e., monogamous and stable) and the "wrong" kind of relationship with sexual aspects (friends with benefits, i.e., casual or occasional) (see U. Khan"Hot for Kink, Bothered by the Law: BDSM and the Right to Autonomy" (Summer 2006), 41:2 Law Matters 17; L. A. Rosenbury"Friends with Benefits?" (2007), 106 Mich. L. Rev. 189, at pp. 207-8). [ 186 ] In the result, those who live outside the scope of relationships privileged by the majority (i.e., those who live in relationships which are neither monogamous nor quasi-marital), are sacrificed, so to speak, to "majoritarian sexual morality" (E. Craig"Capacity to Consent to Sexual Risk" (2014), 17 New Crim. L. Rev. 103; J. Sealy-Harrington"Tied Hands? A Doctrinal and Policy Argument for the Validity of Advance Consent" (2014), 18 Can. Crim. L. Rev. 119, at p. 145). The majority may not find it troubling to privilege, in a way that goes to the heart of the ability to make full answer and defence, certain kinds of relationships over others, but I do, and I respectfully decline to follow my colleagues through this moral thicket. [ 187 ] Further, the majority effectively reads out crucial parts of the common law from Seaboyer governing how trial judges should approach defence evidence (see also R. v. Grant, 2015 SCC 9, [2015] 1 S.C.R. 475, at para. 19) and resurrects the approach to evidence — specifically, of creating pre-determined categories of admissibility — which this Court rejected in Seaboyer. And finally, this approach downplays the text and purpose of statutory provisions like s. 276.4, which recognizes that evidence may be admissible for certain purposes yet inadmissible for other purposes, and makes a limiting instruction mandatory where any evidence of other sexual activity is introduced, even if it only refers to other sexual activity indirectly or implicitly, to cure prejudice and to warn the jury of the impermissible uses of that evidence. [ 188 ] Here, the evidentiary ruling was crystal clear: the jury could be told the basic fact that the appellant and the complainant were "friends with benefits", but could not be told about any details, including the frequency of sexual activity. The jury was instructed twice on the permissible use of the evidence (to provide context for how the appellant and the complainant knew each other) and its impermissible uses (to suggest that, because they were friends with benefits, the complainant was more likely to consent on this occasion, or was less worthy of belief overall). I therefore cannot agree with the majority that the evidence should not have filtered through s. 276(1). #### B. The Trial Judge Did Not Err in Her Relevance Determination Under Section 276(2)(b) [ 189 ] Turning to s. 276(2)(b), the majority finds nothing in the appellant's evidence which, absent the information that he and the complainant were friends with benefits, casts him in a particularly unfavourable light, or which would render his narrative "untenable" or "utterly improbable". On this basis, the majority finds the trial judge's determination that the nature of the relationship was relevant to the appellant's ability to make full answer and defence to constitute reversible error. [ 190 ] On this point, I cannot find fault with the principles relied on by the trial judge which led her to that conclusion. Section 276(2)'s evidentiary filter recognizes the relevance of evidence of a relationship between the complainant and the accused and therefore allows it to be admitted, but only where its absence would prevent the accused from making full answer and defence, by leaving the trier of fact with a distorted representation of the circumstances surrounding the incident (see R. v. Harris (1997), 118 C.C.C. (3d) 498 (Ont. C.A.); R. v. M. (M.) (1999), 29 C.R. (5th) 85 (Ont. S.C.J.); R. v. Temertzoglou (2002), 11 C.R. (6th) 179 (Ont. S.C.J.); R. v. Blea, [2005] O.J. No. 4191 (QL) (S.C.J.); R. v. A.A., 2009 ABQB 602, 618 A.R. 137; R. v. Provo, 2018 ONCJ 474, 48 C.R. (7th) 1; see also Craig"Section 276 Misconstrued", at p. 75). The trial judge concluded this was such a circumstance. And, I repeat: the Crown was substantially of the same view. Its objection was not to the admissibility of evidence of the relationship, but to the label. Indeed, the Crown's position has remained consistent that the nature of the relationship is not in dispute (A.R., vol. II, at p. 40; R.F.C.A., at para. 6; R.F., at para. 8): a former relationship — now a friendship — that, from time to time, involves sexual activity. [ 191 ] Preventing distortion was necessary for the jury to assess the credibility of the appellant's evidence, which was the most relevant and material issue with which the jury would have had to grapple. Indeed, it went to what Sopinka J. (albeit in dissent) once described as "the most fundamental rule of the game" (R. v. W.(D.), [1991] 1 S.C.R. 742, at p. 750; see also L. A. Silver"The WD Revolution" (2018), 41 Man. L.J. 307), since it went to this accused's capacity to raise a reasonable doubt on the strength of his own evidence. It is difficult to conceive of evidence of greater probative value than evidence that is essential to an accused's ability to make full answer and defence. This is particularly so where, as here, the evidence of the complainant and the accused are diametrically opposed in every material respect, leaving credibility the most important issue at trial (R. v. Crosby, [1995] 2 S.C.R. 912, at para. 12, per L'Heureux‑Dubé J.; see also R. v. Nyznik, 2017 ONSC 4392, 40 C.R. (7th) 241, at paras. 15-16). [ 192 ] The majority gestures to these principles, but ultimately finds that this jury gained nothing of value by knowing the appellant and the complainant were in a relationship. It in no way assisted them as it weighed the plausibility of the appellant's account, as it could have only been used to advance impermissible lines of reasoning. [ 193 ] I am afraid I see the matter quite differently. To deny the appellant the ability to point to his relationship would in these circumstances disable the jury from meaningfully performing its central function of finding facts and seeking out the truth. How else, for example, could the jury properly assess his explanation for why he went over to the complainant's residence to pick her up that evening (he owed her "birthday sex")? Or his mouthing, unprompted, at her "I'm going to fuck you" in the presence of his roommate? Or their bickering about which side of the bed she would sleep on to avoid "hot flashes"? Or his knowledge of her hot flashes? Or that the hot flashes occurred when she was on a certain side of his bed? [ 194 ] To prohibit the appellant from explaining these statements, or this knowledge, by testifying that he was in a relationship with the complainant, is to force him to tell an incomplete story — a story which includes an account of the act but no explanation for how he and the complainant "got there", and why he said what he said, and why he did what he did. More particularly, without this evidence, his actions (including his words and gestures) will have appeared to have arisen out of nowhere, creating a completely misleading impression on the jury. His right to make full answer and defence would be reduced to painting a picture of himself as (at best) crude and reckless, or (at worst) predatory. [ 195 ] Ultimately, the majority suggests that the trial judge would not have erred had she simply accepted the Crown's proposed language of former lovers, who, from time to time, had adult sleepovers. But it is far from clear that this is an improvement over "friends with benefits", and indeed in my view it would appear more likely to invite prurient and speculative inferences. And, as Berger J.A. noted in his dissent in the court below, the Crown's language would not have been necessarily accompanied by either of the two limiting instructions the trial judge gave at this trial. The jury's collective mind would truly be left to wander. Instead, the trial judge concluded that allowing the jury to hear about the existence of a relationship, without hearing evidence about its particularities or frequency, prevented the appellant's evidence from being considered through a distorted lens and ensured the appellant's right to make full answer and defence was preserved. I find nothing in the trial judge's relevance assessment that constitutes reversible error. #### C. A New Trial Is Procedurally Unfair and Lowers the Graveline Standard [ 196 ] My third point of departure relates to fairness. I observe that it was the Crown's cross-examination of the appellant which directly put the sexual nature of his relationship with the complainant at issue. And it was the Crown who suggested that he had sexually assaulted her because they hadn't had sex recently. And it was the Crown who suggested that they had a "tumultuous relationship" marked with many arguments and her being upset with him but returning to him eventually. The difference, however, is that the Crown leaned into the existence of the relationship in order to advance a theory that, because they were in a sexual relationship, the jury should find that he expected sex and was angry when she refused. At the same time, however, the Crown insisted that the appellant be prohibited from rebutting that theory, if his evidence would in any way indirectly reveal to the jury that the complainant had engaged in the same sexual activity that the Crown pointed to in support of its own theory. This is, as I have already observed, fundamentally unfair. [ 197 ] What also strikes me as fundamentally unfair is the majority's characterization of the effects of the evidentiary ruling, in particular because the evidentiary ruling would not have revealed any direct evidence of other sexual activity (which is a critical consideration in the balancing mandated by s. 276), had not the Crown chosen to pursue a line of questioning that directly contravened it. The ruling, it must be remembered, should have had the effect of only indirectly or incidentally revealing the complainant's other sexual activity, because no evidence could be elicited about "the specific sexual acts [the complainant] and the accused engaged in", or "the number of times they had sexual relations after they ceased living together" (A.R., vol. I, at p. 10). The s. 276 regime recognizes that differing degrees of prejudice will flow from certain types of sexual history evidence over others. A trial judge assesses the extent to which sexual history evidence will risk leading the jury in this case to draw prohibited inferences. This is a core feature of the discretion afforded to trial judges by s. 276 (R. v. Darrach, 2000 SCC 46, [2000] 2 S.C.R. 443, at paras. 19, 30, 36 and 71). [ 198 ] The majority provides a helpful guide for trial judges on the exercise of discretion (paras. 74-75): where a trial judge is concerned about the jury improperly speculating on past sexual activity, the trial judge should clearly explain to the jury that they will not hear any evidence about whether the relationship included a sexual aspect; the jury must be instructed that no means no and that consent in the past in no way implies or assumes consent in the future; and if relationship evidence is admissible, it should be adduced through an agreed statement of facts, to minimize the need for examination and cross-examination. [ 199 ] And yet, each of these principles was observed by the trial judge in her evidentiary ruling, her conversations with counsel, her cautions to the jury, and her final instructions. It was the Crown's decision not to adduce this evidence through an agreed statement of facts (as suggested by the trial judge) and the Crown's later contravention of the evidentiary ruling (which had the effect of introducing otherwise inadmissible evidence before the jury), that caused the problems at trial which the majority now, remarkably, lays at the feet of the trial judge. This strikes me as a wholly unacceptable basis for a successful Crown appeal from an acquittal. [ 200 ] That Crown counsel's caution in seeking clarification of the ruling (which is to be commended) did not prevent a contravention of that same ruling is demonstrative of a problem of application arising from the delicate balancing mandated by s. 276 (see D. Stuart"Twin Myth Hypotheses in Rape Shield Laws are Too Rigid and Darrach is Unclear" (2009), 64 C.R. (6th) 74, at pp. 75-76). What is agreed to in advance can go off the rails when the witness takes the stand. The majority identifies this as such a case, but then resolves the problem of application in favour of the Crown. This is unfortunate, because the trial judge's evidentiary ruling, as well as her numerous interventions in the trial, were faithful to (and not a departure from) the principles articulated in Seaboyer and Darrach. She applied the correct legal principles: such relationship evidence will be relevant where the accused would otherwise be prevented from making full answer and defence, and the features of the evidence only indirectly or incidentally reveal the complainant's other sexual activity. In such a circumstance, full answer and defence cannot be achieved by confining accused persons to saying what happened and not allowing them to explain how, or why, or to produce any corroborating evidence. The significant probative value of such evidence would outweigh any prejudicial effect (particularly when accompanied, as it was here, by an appropriate limiting instruction), and the proper balance under s. 276 would be struck. [ 201 ] Finally, the majority's approach relieves the Crown of its high burden of demonstrating that a legal error had a material bearing on the jury's verdict (see R. v. Graveline, 2006 SCC 16, [2006] 1 S.C.R. 609). While the majority reasons that the relationship evidence tendered in this case would have had a material bearing on the acquittals, this does not account for the fact that the relationship evidence was relied on by the Crown against the appellant. And again, it was the Crown, and not the appellant, who contravened the evidentiary ruling and explored both the details (the location of previous sexual activity), and the frequency of the sexual activity. [ 202 ] As Watt J.A. observed in R. v. Luciano, 2011 ONCA 89, 273 O.A.C. 273: > It is worth reminder that the appeal of the Attorney General is from the acquittal, not the admissibility ruling that is said to constitute an error in law. What is required is a demonstration of a legal error (in the admissibility ruling) and a nexus between the legal error and the verdict rendered (an acquittal). The authorities teach that acquittals are not to be overturned lightly. [Emphasis in original; para. 260.] Here, the Crown has not established a sufficient nexus between the legal error and the verdict rendered. Again, I ask, how can the admission of this evidence be said to have had a material bearing on the acquittals, when the Crown was relying on the selfsame evidence to prove its case? The Graveline standard is not met on this record. I say respectfully that the majority's position undesirably waters down the Graveline standard in remitting this matter for a new trial. ### XI. Conclusion [ 203 ] Ultimately, I agree with the trial judge's chain of reasoning under s. 276. Making full answer and defence includes being allowed to tell a story that is complete and coherent by drawing from relevant evidence whose probative value outweighs its prejudicial effects. This cannot be achieved by confining accused persons to saying what happened and not allowing them to explain how, or why, or to produce any corroborating evidence. [ 204 ] Recognizing this, the trial judge in my view deftly managed a circumstance made difficult not only by the subject matter of the evidence, but by the Crown having exceeded the scope of the evidentiary ruling in her examination of the complainant. Her decision to admit this highly probative evidence for the limited purpose of allowing jurors to assess the plausibility of the appellant's account of what happened was faithful to Seaboyer (at p. 609) and to s. 276. And, her careful instruction to the jury regarding that limited purpose was sufficient to address any prejudicial effect arising. Jurors are "[not] morons, completely devoid of intelligence", but are "conscientious""anxious to perform their duties", and would not "be forgetful of instructions" (R. v. W.(D.), at p. 761, per Cory J., citing R. v. Lane and Ross (1969), 6 C.R.N.S. 273 (Ont. S.C.) (emphasis deleted); see also R. v. Corbett, [1988] 1 S.C.R. 670, at pp. 695-96, per Dickson C.J.). [ 205 ] This jury rendered its verdict, having been properly instructed on how to do so. I would allow the appeal and restore the acquittals. --- Appeal dismissed, Brown J. dissenting. Solicitor for the appellant: Deborah Hatch Law, Edmonton. Solicitor for the respondent: Attorney General of Alberta, Edmonton. Solicitor for the intervener the Attorney General of Ontario: Attorney General of Ontario, Toronto. Solicitors for the intervener the Criminal Lawyers' Association of Ontario: Addario Law Group, Toronto; Daniel Brown Law, Toronto. --- [^1]: Statistics Canada, Police-reported sexual assaults in Canada before and after #MeToo, 2016 and 2017 (2018). [^2]: On December 13, 2018, Bill C-51, An Act to amend the Criminal Code and the Department of Justice Act and to make consequential amendments to another Act, 1st Sess., 42nd Parl., 2018, received royal assent. This Act makes minor amendments to "clarify" the application of s. 276. They do not apply to the case at bar. Nor do they affect the analysis in this case. [^3]: R. v. Shearing, 2002 SCC 58, [2002] 3 S.C.R. 33, at para. 173, and Mills, at para. 90, take up this phrase coined by Cristin Schmitz in "'Whack' sex assault complainant at preliminary inquiry", published in The Lawyers Weekly, vol. 8, No. 5, in May 1988. [^4]: See, e.g., Department of Justice, Research and Statistics Division, Health Impacts of Violent Victimization on Women and their Children (2012); K.-L. Miller"You Can't Stop The Bell From Ringing." Protean, Unpredictable, And Persisting: The Victim Impact Statement In The Context Of Sexually Assaulted Women (2015). [^5]: Department of Justice, Research and Statistics Division, An Estimation of the Economic Impact of Violent Victimization in Canada, 2009 (2014). [^6]: Now s. 278.94(4)(c). [^7]: This has been made explicit in the December 2018 amendments. [^8]: All section number references are to the Code. [^9]: On December 13, 2018, various amendments to the s. 276 regime came into force. These reasons deal with the s. 276 regime as it existed before that date. [^10]: This type of evidence is often referred to as "prior sexual history evidence". However, I will use the phrase "sexual activity evidence" in these reasons to reflect the fact that such evidence can also relate to sexual activity that is ongoing or takes place after the alleged offence. [^11]: As will become apparent, I consider it doubtful that Mr. Goldfinch's application would have met the relevance test under s. 276.1 (see para. 94 above). But that issue was not raised on appeal. [^12]: The term is defined by the Oxford English Dictionary as "a friend with whom one has an occasional and casual sexual relationship" (online). The origin of the concept is unclear, but the term appears to originate in a lyric by Alanis Morissette in "Head Over Feet" (1995): "You're the best list'ner that I've ever met. You're my best friend, best friend with benefits." [^13]: All counsel acknowledged they understood. [^14]: Parliament has clarified this by amending s. 276(2)(a) of the Criminal Code to state that evidence of other sexual activity cannot be adduced for the purpose of supporting an inference described in s. 276(1).