R. v. Pan

Reasons for Judgment — Wagner C.J. (Côté, Rowe, Kasirer, Jamal, O’Bonsawin and Moreau JJ. concurring)

I. Overview

1On the night of November 8, 2010, three armed intruders entered the Pan family home. The intruders took Bich-Ha Pan and Hann Pan into the basement and shot them each in the head and shoulders. Mrs. Pan died at the scene. Mr. Pan was seriously injured but survived. Their daughter, the respondent Jennifer Pan (“Jennifer”), was left unharmed, tied to an upstairs bannister by a shoelace.

2Jennifer and the other respondents on the main appeal, Daniel Wong, Lenford Crawford, and David Mylvaganam, were charged with the first degree murder of Mrs. Pan and the attempted murder of Mr. Pan. They were all convicted by a jury on both counts. The Court of Appeal set aside their convictions for first degree murder and ordered a new trial on that count alone.

3The Crown appeals from the Court of Appeal’s decision, seeking to restore the convictions for first degree murder. It argues that the Court of Appeal was wrong to conclude that the lesser included offences of second degree murder and manslaughter should have been left with the jury for this count. This argument invites consideration of the air of reality test as it applies to included offences. After outlining the relevant principles, I conclude that the included offences had an air of reality in this case and that the jury should have been permitted to consider them. As a result, a new trial is necessary for all of the respondents on the first degree murder count.

4The respondents each cross-appeal, asking this Court to set aside their convictions for attempted murder and order a new trial on this count. They argue that the trial judge’s error in failing to leave the included offences with the jury tainted the attempted murder convictions in addition to the first degree murder convictions, and thus a new trial is warranted on both counts. I disagree. The error affects only the first degree murder convictions and provides no basis to disturb the convictions for attempted murder.

5The respondents also raise various other decisions of the trial judge that they say were in error and provide a basis for ordering a new trial on both counts. Notably, they argue that the trial judge erred in permitting a slide show presentation prepared for the Crown to go with the jury to the jury room, and they invite this Court to consider the legal framework governing this use of jury aids. Applying the proper framework to this case, I conclude that there was no reversible error in the trial judge’s approach. I also conclude that the other grounds of cross-appeal — that the trial judge erred in dealing with allegations of juror bias and in instructing the jury on the use of propensity evidence — provide no basis for interfering with the convictions, substantially for the reasons of the court below.

6Therefore, for the reasons that follow, I would dismiss the appeal and the cross-appeals.

II. Evidence at Trial

7Jennifer and her parents, Mr. Pan and Mrs. Pan, had a difficult relationship. Jennifer regarded her parents, and especially her father, as strict and controlling. She lied to them about graduating from high school, attending what is now Toronto Metropolitan University, completing a degree in pharmacy from the University of Toronto, volunteering at The Hospital for Sick Children, and working at a Walmart pharmacy. She created fake diplomas, transcripts, graduation photos, and pay stubs to maintain the façade.

8Jennifer also lied to her parents about her boyfriend, Daniel Wong, who worked at a Boston Pizza and trafficked in marijuana. She frequently told her parents that she was staying with a friend in Toronto but instead went to Ajax to stay with Mr. Wong. Mr. Pan always disapproved of his daughter’s relationship with Mr. Wong and repeatedly forbade her from seeing him. Mrs. Pan supported her husband’s stance but, according to Jennifer, was more understanding.

9Jennifer’s relationship with her parents degenerated in 2009 and early 2010 as her lies unraveled. They discovered that she had not graduated from high school or university, that she was neither volunteering at the hospital nor working at a pharmacy, and that she was still in contact with Mr. Wong. By the spring of 2010, Mr. Pan had imposed even stricter rules on Jennifer and had issued an ultimatum: stay home and break contact with Mr. Wong, or leave and never return. Jennifer chose to stay.

10Jennifer testified that, between April and June 2010, she developed a plan with her friend. She testified that the plan was to have her friend’s roommate kill her father. Jennifer’s friend wavered in his testimony on whether the plan was to kill both parents or only Mr. Pan. This plan was ultimately aborted because the roommate stopped responding to Jennifer’s calls.

11Jennifer testified that by August 2010, she was depressed and no longer wanted to live. She claimed that she arranged for Lenford Crawford, a friend of Mr. Wong, to kill her. That way, her brother could still receive the payout from her life insurance policy. She agreed to pay Mr. Crawford $10,000. Jennifer testified that she did not want her parents to be killed but that she told Mr. Wong the plan was to kill her father so that Mr. Wong would not intervene.

12By September 2010, Jennifer testified, she was feeling more accepted in her family, in part because she had been admitted to a post-secondary program. She reached out to Mr. Crawford to cancel the plan for her death, and the two agreed to a cancellation fee of $8,500.

13On November 8, 2010, Mr. Crawford sent Jennifer a text message stating “after work ok will be game time”, which she testified was a reference to him collecting his cancellation fee (A.R., parts II-IV, vol. IV, at p. 1505; see also 2023 ONCA 362, 427 C.C.C. (3d) 4, at para. 28). Jennifer’s friend testified that he spoke to Jennifer on the phone that evening and that she told him there would be a staged home invasion in which both of her parents would be shot. Jennifer denied this.

14At 10:14 p.m. that same day, three armed intruders entered the Pan home. After demanding money, they took Mr. and Mrs. Pan to the basement and shot them each in the head and shoulders. Mrs. Pan died of her injuries. Mr. Pan miraculously survived but was severely injured. Jennifer was left unharmed, tied to an upstairs bannister by a shoelace. She was able to call 9-1-1.

15The respondents, Jennifer, Mr. Wong, Mr. Crawford, and Mr. Mylvaganam, were tried by jury for the first degree murder of Mrs. Pan and the attempted murder of Mr. Pan. Eric Carty was a co-accused during the trial, but his case was severed following the close of the Crown’s case.

16The Crown’s theory at trial was that Jennifer had arranged to have her parents killed in order to escape her dysfunctional relationship with them, resume her relationship with Mr. Wong, and collect on their estates and life insurance policies. The arrangement was said to have been made through Mr. Wong and Mr. Crawford, and carried out by Mr. Carty, Mr. Mylvaganam, and at least one other person, whom Jennifer had promised to pay.

17At trial, Jennifer denied wanting to have her parents killed. She testified that the arrangement with Mr. Crawford was to have herself killed but that she had resiled from that plan when her home life improved. The other respondents put forward various other defences but did not testify.

III. Judicial History

A. Ontario Superior Court of Justice (Boswell J.)

18Each ground of appeal before this Court relates to a decision made by the trial judge throughout the course of the long and complex trial. It is not necessary to canvass the details of this trial at length. In this section, I briefly introduce each impugned decision, which I expand upon in more detail below.

19The first is the trial judge’s decision not to charge the jury on the included offences of second degree murder and manslaughter in the death of Mrs. Pan. The theory proffered in support of second degree murder was that Jennifer and her co-conspirators only planned to have her father killed but knew that her mother’s death was a probable consequence of carrying out this plan. The manslaughter theory was that the respondents planned to kill Mr. Pan and that there was a reasonably foreseeable risk of non-trivial bodily harm to Mrs. Pan. The trial judge concluded that there was no air of reality to these theories. He also linked the two counts in his instructions and told the jury that they were “inextricably intertwined” (A.R., part V, vol. LXIV, at pp. 318-19).

20Second, the trial judge permitted a slide show prepared for the Crown, which summarized the Crown’s view of the relevant cell phone tower evidence, to go to the jury room to assist in the jury’s deliberations (2014 ONSC 6055, 318 C.C.C. (3d) 54). This ground of appeal also impugns the trial judge’s decision to dismiss the Crown’s motion to call a witness to authenticate the slide show.

21Third, the trial judge declined to discharge a juror or to declare a mistrial (2014 ONSC 4287; 2014 ONSC 4645). The juror’s spouse had been attending court and had sent him text messages that he shared with other members of the jury. The spouse had also made encouraging comments to a detective witness and had conversed with counsel for Mr. Crawford. After conducting two jury inquiries, but without requiring the production of the text messages themselves, the trial judge concluded that there was no reasonable apprehension of bias.

22Fourth, the trial judge instructed the jury regarding propensity evidence that was led against Mr. Carty, the accused whose case was subsequently severed from this trial. Mr. Crawford was concerned that the propensity evidence against Mr. Carty would also prejudice him because of his close relationship with Mr. Carty. The trial judge agreed to instruct the jury not to reason along those lines, but he did not ultimately tie that instruction to Mr. Crawford by name.

23The jury convicted the respondents of the first degree murder of Mrs. Pan and the attempted murder of Mr. Pan.

B. Court of Appeal for Ontario, 2023 ONCA 362, 427 C.C.C. (3d) 4 (van Rensburg, Miller and Nordheimer JJ.A.)

24The respondents all appealed from their convictions for first degree murder and attempted murder. The Court of Appeal agreed with them that the trial judge erred by not leaving the included offences of second degree murder and manslaughter with the jury and by withholding the option of considering each count separately. It held that this error impacted the first degree murder convictions for all of the respondents and that the curative proviso did not apply. The Court of Appeal rejected the other grounds of appeal.

25In the result, the Court of Appeal allowed the appeals from the first degree murder convictions, set aside the convictions on that count, and ordered a new trial. It dismissed the appeals from the attempted murder convictions. It also dismissed appeals by Jennifer and Mr. Wong against their sentences of life imprisonment, which are not on appeal before this Court.

IV. Grounds of Appeal

26The Crown appeals, seeking to reinstate the convictions for first degree murder. The appeal raises two main issues:

27The respondents each cross-appeal from the Court of Appeal’s decision, seeking to have their convictions for attempted murder set aside. Several of the issues raised on cross-appeal could also provide an additional basis for this Court to uphold the order for a new trial on the count of first degree murder. Their cross-appeals raise the following issues:

V. Analysis

A. The Crown’s Appeal

(1) Background and Decisions Below

28During the pre-charge conference, Jennifer’s counsel urged the trial judge to leave second degree murder and manslaughter with the jury for all of the respondents in respect of the charge of first degree murder of Mrs. Pan. The theory underlying these alternative verdicts was that Jennifer and her co-conspirators only planned to have her father killed but either they knew that her mother’s death was a probable consequence of carrying out this plan or there was a reasonably foreseeable risk of non-trivial bodily harm to Mrs. Pan. This could give rise to convictions for second degree murder or manslaughter by way of common intention party liability under s. 21(2) of the Criminal Code, R.S.C. 1985, c. C-46.

29The trial judge delivered his ruling in an email to counsel, concluding that there was no air of reality to the scenario advanced by Jennifer’s counsel:

30The trial judge therefore instructed the jury to consider only two possible scenarios. Under scenario one, there was a single planned and deliberate attack with the intention of murdering both Mr. Pan and Mrs. Pan. Under scenario two, there was a joint plan or agreement between two or more persons to commit a home invasion and robbery, in the commission of which the Pans were shot. The judge instructed the members of the jury that they were “in no way bound to find that one or the other scenario occurred” and that they “may not be satisfied that either one [had] been made out on the evidence” (A.R., part I, vol. I, at p. 409).

31The trial judge instructed the jury that it had to either find Jennifer guilty of first degree murder and of attempted murder or acquit her of both charges. Other verdicts were left open in respect of Mr. Wong, Mr. Mylvaganam and Mr. Crawford. The jury was not permitted to consider the two counts separately.

32The Court of Appeal held that the lesser included offences of second degree murder and manslaughter had an air of reality and should have been left with the jury in respect of the count relating to the death of Mrs. Pan. The court noted that the air of reality test applies to lesser included offences when a jury may not be satisfied beyond a reasonable doubt of the elements of the principal offence but may be satisfied beyond a reasonable doubt of the elements of an included offence (paras. 58-59, citing R. v. Chacon-Perez, 2022 ONCA 3, 159 O.R. (3d) 481; R. v. Ronald, 2019 ONCA 971). A trial judge should not consider how likely or unlikely he or she believes a conviction for that included offence to be, nor should the judge pay regard to concerns about the credibility or reliability of the evidence required to reach that outcome (paras. 68-70, citing R. v. Tenthorey, 2021 ONCA 324, 404 C.C.C. (3d) 457; R. v. Cinous, 2002 SCC 29, [2002] 2 S.C.R. 3).

33The Court of Appeal concluded that there was evidence that might have raised a reasonable doubt as to whether the murder of Mrs. Pan was planned and deliberate, and that could have supported the theory that Mrs. Pan’s death was known to be a probable consequence of carrying out a plan to kill Mr. Pan. This could give rise to a conviction for second degree murder. It was also possible for the jury to have been satisfied that there was a reasonably foreseeable risk of non-trivial bodily harm to Mrs. Pan through the implementation of the plan to kill Mr. Pan, leading to a conviction for manslaughter. The court was unmoved by the Crown’s submission that there was much stronger evidence supporting the theory that both parents were intended targets, as this did not change the fact that there was some evidence supporting the alternative theory.

34The error impacted all four respondents. The Court of Appeal rejected the Crown’s argument that the curative proviso could apply to the error, and concluded that the first degree murder convictions should therefore be set aside and a new trial ordered.

(2) Standard of Review

35A trial judge’s determination about whether there is an air of reality is a question of law reviewable for correctness (R. v. Tran, 2010 SCC 58, [2010] 3 S.C.R. 350, at para. 40; R. v. Alas, 2022 SCC 14, [2022] 1 S.C.R. 283, at para. 3).

36The Crown insists that while this question is reviewable for correctness, “in the absence of identified errors, some deference is owed to the trial judge’s decision” (A.F., at para. 73). The Crown points to some intermediate appellate court jurisprudence noting uncertainty about the nature of the deference owed to the trial judge in this context (see, e.g., R. v. Land, 2019 ONCA 39, 145 O.R. (3d) 29, at para. 71; R. v. Paul, 2020 ONCA 259, 63 C.R. (7th) 377, at paras. 26-27; R. v. Suthakaran, 2024 ONCA 50, 433 C.C.C. (3d) 175, at para. 15).

37It is true that historically there was some uncertainty about the appropriate standard of review on the question of whether there was an air of reality (see, e.g., R. v. Thibert, [1996] 1 S.C.R. 37, at para. 33). In cases like Thibert, the question was not identified as one of law reviewable on a correctness standard. However, now that the standard of review has been definitively settled as correctness, this obviates the need for any deference. Correctness means that “an appellate court is free to replace the opinion of the trial judge with its own” (Housen v. Nikolaisen, 2002 SCC 33, [2002] 2 S.C.R. 235, at para. 8). Trial judges must come to the correct answer on the air of reality question and, if they do not, they have committed a reviewable legal error (see Cinous, at para. 55). I would decline the Crown’s invitation to deviate from the understood meaning of the correctness standard in this context. To introduce some level of deference here would only invite confusion and unnecessary complication into the law on the standard of review.

38My colleague comes to a different understanding on the standard of review. She says that the trial judge is best suited to make air of reality determinations (para. 187) and that deference to the trial judge’s limited weighing of the evidence can comfortably coexist with a correctness standard on the ultimate determination (para. 184). However, as this Court held in R. v. Buzizi, 2013 SCC 27, [2013] 2 S.C.R. 248, “the trial judge is not at all in the ‘best position’ to determine whether a defence has an air of reality, since that is a question of law” (para. 15). Further, in cases like this one that turn on circumstantial evidence, a trial judge’s limited weighing of the evidence will be co-extensive with the air of reality determination. The trial judge’s decision in this case highlights the futility of trying to isolate the limited weighing process from the determination that there is an air of reality such that one standard of review can apply to the former and another to the latter. He made no such distinction.

39My colleague analogizes to cases where courts defer to the trial judge’s findings of fact but do not defer on the ultimate question of law to which those findings relate (para. 185). With respect, I fail to see the relevance of these cases to the question at hand. It is a central feature of the air of reality test and the limited weighing exercise that the trial judge is not permitted to make findings of fact (Cinous, at para. 54). Fact‑finding is a fundamentally different exercise. As the Ontario Court of Appeal recognized in Paul, there is “necessarily less” deference owed with respect to limited weighing than with respect to fact-finding, because certain reasons for deferring to trial judges, like their privileged position in assessing credibility, are not at all relevant to a limited weighing (para. 30, citing Housen, at paras. 15-18). So, while I agree with my colleague that deference is owed to the trial judge’s findings of fact, this does not support the conclusion that deference is owed here.

40What is clear from cases like Paul and Land is that confusion in this area has resulted in “complexity” for appellate courts (Land, at para. 71). This Court is able to resolve this complexity, and it is in the interests of justice to do so.

(3) Principles Governing the Air of Reality Test for Included Offences

41This Court has dealt in great depth with the question of when a defence should be left with the jury and, in that context, has formulated the air of reality test with precision and clarity (see, e.g., R. v. Osolin, [1993] 4 S.C.R. 595; R. v. Park, [1995] 2 S.C.R. 836; R. v. Davis, [1999] 3 S.C.R. 759; Cinous). The air of reality test has also been applied, both by this Court and by other appellate level courts, to determine when an included offence may be left with the jury (see, e.g., R. v. Aalders, [1993] 2 S.C.R. 482; Ronald; Chacon-Perez; Tenthorey; R. v. Nason, 2015 NBCA 34, 437 N.B.R. (2d) 259; R. v. Chalmers, 2009 ONCA 268, 243 C.C.C. (3d) 338). Despite the consistent use of the air of reality test for this purpose, little has been said about how the approach to the test may differ and what novel considerations may arise when dealing with included offences as opposed to defences. This appeal provides an opportunity to address this directly.

(a) Competing Considerations in the Air of Reality Test for Included Offences

42The air of reality test seeks to balance two competing considerations. On the one hand, far-fetched theories that have no evidentiary foundation must be excluded from the jury’s consideration. Offering these theories to the jury would serve no truth-seeking purpose and would only cause confusion, invite improper compromise, and needlessly lengthen the judge’s charge (see Park, at para. 11; Osolin, at p. 683; see also R. v. Matchett, 2018 BCCA 117, 359 C.C.C. (3d) 363, at para. 23).

43The importance of withholding unsustainable theories from the jury is amplified in the context of included offences. Whereas charging the jury on an unsustainable defence runs the risk of an acquittal that is not supported by the evidence, charging the jury on an unsustainable included offence runs the risk of an unreasonable conviction, which is “possibly the gravest error of all” (R. v. Biniaris, 2000 SCC 15, [2000] 1 S.C.R. 381, at para. 26).

44On the other hand, by setting the evidentiary bar low, the air of reality test ensures that all viable theories are submitted for the jury’s thoughtful consideration. In this way, the test facilitates the accused’s right to be tried by a jury, if he or she so chooses, instead of by a judge alone (see Osolin, at p. 690, citing P. Hogg, Constitutional Law of Canada (3rd ed. 1992), vol. 2, at p. 48-15; R. v. Fontaine, 2004 SCC 27, [2004] 1 S.C.R. 702, at paras. 58-60). Trial judges must take care not to undermine this protected choice by settling issues properly within the jury’s domain.

45Though included offences are not completely analogous to defences, the question of whether to leave such offences with the jury also engages the right to make full answer and defence insofar as leaving them gives the jury an additional route by which to conclude that the accused is not guilty of the principal offence. Indeed, this Court has recognized the risk that withholding a realistic verdict may cause a jury, faced with the choice between a conviction on the principal offence and an acquittal, to convict on the principal offence simply because an acquittal is “unpalatable” (R. v. Haughton, [1994] 3 S.C.R. 516, at p. 517).

46These competing considerations highlight a unique feature of included offences. Whereas defences are purely exculpatory, included offences have both an exculpatory dimension — in that they are exculpatory of the principal offence — and an obvious inculpatory dimension. Accordingly, while defences are advanced by the accused, the party seeking to leave an included offence with the jury will vary from case to case.

47A legal standard like the air of reality test does not change depending on who is arguing that the test is met; however, where it is the accused rather than the Crown arguing that the included offence should be left with the jury, the court must be cognizant that its ruling will bear on the accused’s right to control his or her own defence. The same consideration will not arise where it is the Crown arguing that the included offence should be left with the jury.

48Keeping these considerations in mind, I first look at what it means for an included offence to have an air of reality. Then, I explain how the approach to the evidence may differ across different types of cases, notwithstanding the fact that the test fundamentally stays the same. Finally, I elaborate upon the type of evidentiary weighing permitted in determining whether an air of reality exists on the evidence.

(b) Air of Reality Test for Included Offences

49An accused charged with an offence may be acquitted of that offence but nonetheless convicted of an included offence, even where there is no express reference to the included offence in the count. An offence is said to be “included” for this purpose where it is defined as such in the Criminal Code or where its elements form part of the offence charged, whether “as described in the enactment creating it or as charged in the count” itself (s. 662; see R. v. G.R., 2005 SCC 45, [2005] 2 S.C.R. 371, at paras. 25 and 29-33; see also M. Vauclair, T. Desjardins and P. Lachance, Traité général de preuve et de procédure pénales 2024 (31st ed. 2024), at paras. 34.51-34.53; S. Penney, V. Rondinelli and J. Stribopoulos, Criminal Procedure in Canada (3rd ed. 2022), at ¶11.21).

50Where an offence is properly an included offence in accordance with these principles, there arises the distinct question of whether that offence should be left with the jury. An included offence must be left with the jury if, and only if, it has an air of reality, meaning that there is a realistic possibility of an acquittal on the principal offence and a conviction on the included offence (see R. v. Wolfe, 2024 SCC 34, at para. 50; Joseph v. R., 2018 QCCA 1441, at para. 19; R. v. Smith, 2023 NBCA 20, 424 C.C.C. (3d) 380, at para. 33; R. v. Iyamuremye, 2017 ABCA 276, 355 C.C.C. (3d) 289, at para. 82; see also Vauclair, Desjardins and Lachance, at paras. 33.27-33.28).

51To determine whether this is a realistic possibility, the trial judge must assess whether there is a reasonable view of the evidence upon which a properly instructed jury acting judicially could be left with a reasonable doubt about elements of the principal offence that distinguish it from the included offence, while accepting beyond a reasonable doubt all of the elements of the included offence (R. v. Wong (2006), 209 C.C.C. (3d) 520 (Ont. C.A.), at para. 12; Ronald, at para. 46; Tenthorey, at para. 63; Chacon-Perez, at para. 162). The inquiry requires the judge to take a holistic view of the evidence, bearing in mind that, in accordance with the presumption of innocence, a jury may always reject evidence or refuse to draw particular inferences (Ronald, at para. 48; Joseph, at para. 25).

(c) The Application of the Test Is Contextual

52The concept of an “air of reality” arises in several distinct contexts. The question, broadly, is always whether the necessary factual inferences are available on a reasonable view of the evidence. However, the approach will naturally differ depending on the types of inferences that are in issue or, in other words, on the conclusion that is said to have an air of reality.

53Assessing whether there is an air of reality to a positive defence, for example, requires the court to examine the distinct factual inferences that together form the legal elements of the defence. The analysis involves a contextual review of the record to seek out evidence that can support these distinct inferences (see Cinous; Osolin; Park; see also Vauclair, Desjardins and Lachance, at para. 33.23).

54By contrast, in some cases, the factual inferences required to convict on an included offence will simply be a subset of the inferences needed to convict on the principal offence. The only real question in such cases is whether the distinguishing inferences are factually extricable such that a jury could be left with a reasonable doubt as to only those inferences. The question is not whether there is sufficient evidence to support them (see Ronald; Tenthorey). These are analytically distinct inquiries.

55The facts of Ronald provide an example. There, in order to convict the prime mover of first degree murder, the jury had to be convinced beyond a reasonable doubt that she intentionally killed the victim, having planned the killing beforehand. The factual basis for a second degree murder conviction for that accused was the same, but without planning and deliberation. In that case, the question was merely whether the jury could have realistically had a reasonable doubt on the element of planning and deliberation that distinguishes first degree murder from second degree murder while accepting the other elements which make up second degree murder beyond a reasonable doubt (para. 61).

56The simplicity of the inquiry in Ronald explains Doherty J.A.’s remark, at para. 47, that the party seeking to leave an included offence with the jury does not bear an “evidentiary burden”:

57I take Doherty J.A. to be saying that the trial judge need not look for evidence contradicting the distinguishing inference — that is, there need not be evidence supporting an alternative story — but rather must look at “the totality of the evidence” to determine whether a reasonable doubt on that element alone is realistically possible. In Ronald, the distinguishing inference was planning and deliberation, and Doherty J.A. correctly instructed that the jury was entitled to have a reasonable doubt as to that inference. He observed that establishing the viability of the included offence in this circumstance did not really impose an “evidentiary burden” on the accused in the sense that it did not require the jury to accept additional factual inferences.

58However, in other cases, particularly those complicated by issues of party liability, the route to conviction on the included offence may indeed require additional factual inferences, not just fewer factual inferences. For this type of included offence to be left with the jury, there must be a sound basis for the jury to have a reasonable doubt about the distinguishing elements of the principal offence, while being without such doubt regarding all inferences, including those additional factual inferences, sustaining the included offence.

59This case provides a good example. As the Court of Appeal recognized, “in cases like the present . . ., the route to an included offence does not necessarily arise as soon as there is a reasonable doubt on planning and deliberation” (para. 64). For some of the respondents, the route to liability for the principal offence of first degree murder runs through party liability under s. 21(1) of the Criminal Code. To convict these respondents of the principal offence, the jury must accept that they aided or abetted in the planned killing of Mrs. Pan. By contrast, the route to liability for the included offences of second degree murder and manslaughter runs through common intention liability under s. 21(2). To convict on second degree murder, for example, the jury must accept that the respondents formed an intention to kill Mr. Pan, and that they knew the death of Mrs. Pan was a probable consequence of carrying out their plan. A conviction for second degree murder, in the unique circumstances of this case, rests on different, not just fewer, factual inferences.

60The trial judge in cases like this must ask whether, on the totality of the evidence, a reasonable jury could be left with a reasonable doubt about any distinguishing elements of the principal offence, and also whether the jury could realistically accept the factual inferences underlying conviction on the included offence. In this sense, cases like this resemble defence cases because it is not enough merely to conclude that the jury could have a reasonable doubt on the distinguishing inferences while not having such doubt on the remaining ones.

61It is unsurprising that the approach to the evidence differs across different types of cases. Whether there is an evidentiary foundation for a discrete defence is a different question than whether a jury could have an extricable doubt about the distinguishing element of a principal offence, which is in turn a different question than whether more complex types of included offences are sustainable on the record. At base, however, determining whether there is an air of reality will always involve a contextual inquiry into whether the record can realistically support the proposed line of reasoning (see Chacon-Perez, at para. 164; Ronald, at para. 43). Whether the analysis is described as imposing an evidentiary burden in the context of defences, or simply as an assessment of the totality of the evidence in the context of included offences, the overarching question is the same.

62In sum, an included offence will have an air of reality if there is a reasonable view of the evidence upon which a properly instructed jury could convict on the included offence and acquit on the principal offence. The trial judge must ask not only whether the reasonable view of the evidence could allow for doubt as to the distinguishing elements of the principal offence, but also whether the same reasonable view of the evidence could allow the jury to conclude that all elements of the included offence are made out. This raises the question of whether a given view of the evidence is reasonable for this purpose.

(d) Evidentiary Threshold and “Limited Weighing”

63As with all inquiries into air of reality, the evidentiary threshold set out in Cinous provides a useful starting point. The threshold is “whether there is evidence (some evidence, any evidence) on the basis of which a properly instructed jury acting reasonably could base an acquittal if it believed the evidence to be true” (para. 83). When this articulation is adapted to the context of included offences, the relevant question is whether there is some evidence upon which a properly instructed jury acting reasonably could acquit on the principal offence while convicting on the included offence. As Doherty J.A. explained in Ronald, there need not be evidence supporting doubt. Rather, on the totality of the evidence, a reasonable doubt on the distinguishing inferences must be functionally compatible with a lack of such doubt on the remaining necessary inferences.

64For cases that involve circumstantial evidence, the trial judge assessing whether there is an air of reality must conduct a “limited weighing” of the evidence (R. v. Arcuri, 2001 SCC 54, [2001] 2 S.C.R. 828, at para. 23; see also Cinous, at para. 90). This exercise is necessary because “with circumstantial evidence, there is, by definition, an inferential gap between the evidence and the matter to be established” (Arcuri, at para. 23). Through the process of limited weighing, the trial judge does not draw factual inferences, but rather comes “to a conclusion about the field of factual inferences that could reasonably be drawn” (Cinous, at para. 91). In other words:

65Factual inferences that reasonably arise on the evidence must be left available to the jury even where the trial judge believes that there are other more plausible inferences that could be drawn. In other words, the exercise of “limited weighing” is not comparative as between competing inferences. That form of comparative analysis is an example of substantive weighing, which is beyond the scope of the air of reality test (Cinous, at para. 90; R. v. Pappas, 2013 SCC 56, [2013] 3 S.C.R. 452, at para. 26).

66In conducting a limited weighing of the evidence, the trial judge is not permitted to assess credibility or reliability (Cinous, at para. 90). The narrow exception to the rule that the evidence must be assumed to be true is that a bare assertion, without more, may be insufficient to establish an air of reality (Aalders, at p. 505; Park, at para. 20).

67An example of permissible limited weighing is considering whether the proposed line of reasoning conflicts with evidence that is not materially in dispute. Per L’Heureux-Dubé J. in Park, “[w]here evidence supporting the accused which is materially in dispute is, realistically viewed, clearly logically inconsistent with evidence which is not materially in dispute, then it can be concluded as a matter of both law and logic that there can be no air of reality to the defence to which those logical inconsistencies relate” (para. 29 (emphasis deleted)). The same is true with respect to included offences.

68For example, in Aalders, the accused was charged with first degree murder after breaking into the victim’s home, lying in wait for some four hours, and shooting the victim eight times. The bullets all entered the victim’s torso and neck, except for one that entered his leg. There is a common sense inference that a person intends the consequences of his or her deliberate actions (R. v. Tatton, 2015 SCC 33, [2015] 2 S.C.R. 574, at para. 27). On the totality of the evidence in that particular case, and despite the accused’s statement that he did not intend to kill the victim but only to wound him, there was no air of reality to the included offence of manslaughter. A bare assertion that ran contrary to all other evidence would not have permitted a jury to reasonably convict on the included offence while acquitting on the principal offence (Aalders, at p. 505; Park, at para. 20).

69The trial judge’s obligation to instruct the jury on an included offence will depend not only on the evidence led, but also on the legal issues raised and the theories advanced (see R. v. Sarrazin, 2010 ONCA 577, 259 C.C.C. (3d) 293, at para. 62, aff’d 2011 SCC 54, [2011] 3 S.C.R. 505; Chalmers, at paras. 52-53). In Wong, for example, the accused argued that he had acted accidentally, or in the alternative in self-defence, when he injured his roommate’s co-worker. He was charged with aggravated assault by wounding but convicted by the jury of the included offence of assault causing bodily harm. Doherty J.A. held that the included offence should not have been left with the jury because the distinguishing element of the principal offence — the nature of the injury suffered — was not in dispute. Thus, the only verdicts that should have been open to the jury were a conviction on the principal offence or a full acquittal (paras. 12-14).

70Consider also cases in which the only live issue for the jury is the identity of the offender (see, e.g., Chacon-Perez). Views of the evidence reasonably available to the jury in such cases will generally be either that the accused committed the principal offence or that someone else did, neither scenario being compatible with the conviction of the accused on an included offence. Therefore, these cases will generally warrant “all or nothing” charges, requiring the jury to either convict on the principal offence or acquit.

(4) Application to the Facts

71I agree with the Court of Appeal that the trial judge erred in concluding that there was no air of reality to the theory of a plan to kill only Mr. Pan. On a reasonable appreciation of this record, the jury could have had a reasonable doubt that Mrs. Pan was one of the intended targets of the plan while accepting the factual inferences required to convict the respondents of second degree murder or manslaughter.

72This is not a case like Aalders, in which the entirety of the evidence belied the accused’s bare assertion that he did not intend to kill the victim. In this case, while there was strong evidence supporting the Crown’s theory, there was no undisputed evidence contradicting the idea that the plan was only to kill Mr. Pan.

73The Crown points to evidence in the form of text messages that it says disproves the included offence theories. For example, on October 31, 2010, following a phone call between Mr. Crawford and Mr. Carty, Mr. Carty and Mr. Mylvaganam exchanged the following text messages:

74The Crown submits that the reference to “5each ting” establishes that there were two murders contemplated and that Mr. Carty and Mr. Mylvaganam would be paid $5,000 for each (A.F., at para. 15). Another interpretation that was open to the jury to accept is that they would each be paid $5,000 for carrying out a single murder, unless Mr. Mylvaganam committed the murder alone, in which case he would receive the entire $10,000. To select between these interpretations, even if one is more plausible, would be to engage in substantive weighing and usurp the role of the jury.

75Another relevant text exchange was between Jennifer and Mr. Wong on November 2, 2010, in which Mr. Wong revealed that he had feelings for another woman:

76The Crown argues that Jennifer’s comment that she would “have no where to go” would make no sense unless the plan was to have both parents killed. If her mother was alive, she would have somewhere to go (A.F., at para. 15). Again, I disagree that the Crown’s interpretation is the only available interpretation of this exchange. Other interpretations are that Jennifer would no longer feel comfortable living in her house whether her mother was alive or not, or that this was simply an attempt from a heartbroken woman to stoke sympathy. The correct interpretation was for the jury to decide.

77The Crown points to other evidence probative of a plan to kill both Mr. and Mrs. Pan. This evidence includes testimony that both parents imposed strict rules and made Jennifer feel unhappy; that she was in touch with the perpetrators throughout the night and they arrived at the home only after her mother returned from line dancing; that both parents were treated the same way throughout the attack; and that the assailants did not cover their faces, suggesting that they did not intend to leave Mrs. Pan as a witness.

78Set against this evidence was the fact that Jennifer consistently distinguished between her parents, identifying her father as the one who imposed strict rules and made her feel isolated. She testified that her mother was “much more understanding” and showed empathy regarding her feelings for Mr. Wong, and that her mother would allow her to sneak time on her cell phone after it was confiscated by her father (A.R., part V, vol. L, at p. 88). Jennifer made several statements to the effect that, in a perfect world, she would be with her mother, but her father would be gone. This testimony was corroborated somewhat by the evidence of Jennifer’s brother and cousin, who both testified that Jennifer’s relationship with her mother was better than her relationship with her father.

79Most importantly, Jennifer testified that her friend told her about wanting to kill his father, and then “it clicked, something clicked between what he had said and what I was feeling. Because my dad was the one that was making me feel isolated, my mom still made me feel home, warmth” (A.R., part V, vol. XLVIII, at p. 89). Then, just months prior to the shooting, she arranged for her friend’s roommate to kill her father — only her father — and that plan failed only because the roommate never followed through. Jennifer’s distinct attitudes towards each parent and her prior plan to kill Mr. Pan were probative circumstantial evidence that Mrs. Pan may not have been a target of the November 8 plan.

80The question, looking at all of this evidence, is not whether one theory emerges stronger. My colleague recognizes this (at para. 199) but, in my respectful view, engages in this kind of substantive weighing in her application of the law to this case. The proper question, having regard to the presumption of innocence and the onus on the Crown to prove each element of an offence beyond a reasonable doubt, is whether the record realistically left room for the jury to have a reasonable doubt that the respondents planned for Jennifer’s mother to be killed, while accepting beyond a reasonable doubt that they planned for her father to be killed. In my view, it was realistically possible for the jury both to accept beyond a reasonable doubt that Mr. Pan was the target of the respondents’ plan, and to have a reasonable doubt that Mrs. Pan was also a target.

81With regard now to the additional factual inferences required to convict on the included offence, the additional mens rea element for a conviction on second degree murder is that the respondents knew that the death of Mrs. Pan was a probable consequence of carrying out a common intention to kill Mr. Pan. The additional mens rea element for manslaughter is that non-trivial bodily harm was an objectively foreseeable consequence of the unlawful plan to kill Mr. Pan.

82There was evidence upon which a properly instructed jury could realistically conclude that the elements of the included offences had been established and also conclude that there was a reasonable doubt about the existence of a plan to kill Mrs. Pan. It was open to the jury to accept that the respondents had planned the death of Mr. Pan, had recruited the aid of several men whom they did not know well to execute that plan, and had contemplated that the men would be armed. It was also open to them to accept that Jennifer, with or without the knowledge of the other respondents, had unlocked the door to the house and had contemplated the men’s arrival while Mrs. Pan was home. On the basis of that evidence, the jury could realistically conclude that the respondents subjectively knew Mrs. Pan’s death was a probable consequence of carrying out their plan or that bodily harm to Mrs. Pan was objectively foreseeable.

83To summarize, I agree with the Court of Appeal that there was an air of reality to the included offences of second degree murder and manslaughter and that they should have been left with the jury.

(5) Curative Proviso

84The Court of Appeal held that the curative proviso set out in s. 686(1)(b)(iii) of the Criminal Code did not apply to the trial judge’s failure to leave the included offences with the jury. Relying on Ronald, at para. 66, the court concluded that “the fact that an included offence was not left with the jury, when it ought to have been, means that the error was not harmless” (para. 143).

85The Crown argues that the Court of Appeal erred because the evidence of a plan to kill both parents was overwhelming, while the evidence of a plan to kill only Mr. Pan was marginal at best. The respondents say that the Court of Appeal was correct to conclude that the proviso has no application to these circumstances.

86Failure to instruct the jury on a potential verdict is an error of law within the meaning of s. 686(1)(a)(ii) of the Criminal Code (Sarrazin, at para. 64). An error of law will warrant setting aside the conviction under s. 686(2). However, under s. 686(1)(b)(iii), a court of appeal may dismiss the appeal and uphold the conviction where the Crown can establish that no substantial wrong or miscarriage of justice flowed from the error (para. 65; R. v. Illes, 2008 SCC 57, [2008] 3 S.C.R. 134, at para. 21).

87Specifically, the curative proviso set out in s. 686(1)(b)(iii) is available where: (1) the error or irregularity in question is minor or harmless, such that it had no impact on the verdict; or (2) the error or irregularity, despite being serious enough to warrant a new trial, caused no substantial wrong or miscarriage of justice because the evidence against the accused is so overwhelming that a trier of fact would inevitably convict (see R. v. Tayo Tompouba, 2024 SCC 16, at para. 76).

88In simple terms, the question is whether there is no reasonable possibility that the verdict would have been different had the legal error not been committed (Sarrazin, at para. 65). I am not satisfied that the Crown has met its burden in this regard.

89This Court has established that a conviction on a more serious charge cannot be taken to mean that a jury would not have convicted on a lesser charge had it been available (Haughton). In such circumstances, as mentioned above, there is always the risk that “the jury convicted because they had no other alternative than acquittal and acquittal was unpalatable” (p. 517).

90Only in rare cases will it be possible to conclude that an error in failing to leave an available verdict with the jury did not cause a substantial wrong. As Doherty J.A. suggested in Ronald, the proviso may apply where “the court can take into account findings of fact implicit in the verdict or verdicts returned by the jury as long as those verdicts are not tainted by the legal error, and those findings are unambiguously revealed by the verdict” (paras. 68-69). With that said, the court must remain mindful of the risk that the jury convicted on the principal offence solely because an acquittal was the only other option. While I would not foreclose the possibility that, in rare instances, the air of reality to an included offence will be so marginal as to permit the application of the proviso, this is not one those cases.

91Had the jury been properly instructed on the included offence theories, it would have had to grapple with the evidence suggesting that Jennifer did not have the same animus towards her mother as she did towards her father. I would note that much of this evidence came from Jennifer herself, whose theory at trial was that the only plan in November 2010 was that she herself would be killed, and who thus had no interest in admitting any animus towards either parent. That she did admit her animus towards her father while maintaining that her mother was “the perfect mother” is not insignificant (A.R., part V, vol. L, at p. 310). Reasonable doubt about planning and deliberation, as I have stressed, is a low bar.

92I do not dispute that a conviction on one of the included offences, particularly second degree murder, which has a subjective mens rea element, would require the jury to draw important inferences based solely on circumstantial evidence. However, I would not consider a conviction on either one of those offences to be outside the realm of possibility. I therefore conclude that the curative proviso does not apply.

93I would accordingly dismiss the Crown’s appeal. I move now to the cross-appeals by the respondents.

B. Cross-Appeals

(1) Should a New Trial Be Ordered for the Attempted Murder Count Given the Order for a New Trial for First Degree Murder?

(a) Background and Decisions Below

94After concluding that a new trial was required on the count of first degree murder in the death of Mrs. Pan, the Court of Appeal turned its attention to the count of attempted murder in the shooting of Mr. Pan. It concluded that the respondents’ convictions for the attempted murder of Mr. Pan could stand notwithstanding the trial judge’s error in failing to leave the included offences with the jury in respect of the first degree murder count regarding Mrs. Pan. This was because the included offence theories that should have been put to the jury also posited the attack on Mr. Pan, meaning that they would not have raised alternative lines of reasoning or additional bases for doubt on the attempted murder count.

95The respondents argue that the Court of Appeal erred in refusing to order a new trial on the count of attempted murder of Mr. Pan. Without a new trial on that count, they argue, there is a risk of inconsistent verdicts. That is, in a new trial, the jury may reject entirely the theory that there was a plan to murder either of Jennifer’s parents and acquit on that basis. Jennifer, and potentially others of the respondents, would then be condemned to serve their sentences with respect to the attempted murder of Mr. Pan even though an equally qualified jury found, by necessary implication, that they were not responsible for it.

96The Crown’s position is that the error impacted only the first degree murder count and not the attempted murder count. There is no statutory basis for an appellate court to quash a conviction and order a new trial on a count that was not itself tainted by a legal error. Even if a court could order a new trial on that count to avoid future inconsistent verdicts, the Crown argues that there is no real risk of inconsistency in this case. Its reasoning is that even a full acquittal at a new trial on the shooting of Mrs. Pan could be understood in a way that is consistent with the convictions for attempted murder.

97The parties’ submissions thus raise two questions. The first is whether the prospect of inconsistency alone can warrant ordering a new trial on a count untainted by legal error. The second is whether a prospect of true inconsistency arises in this case. As I will explain, the answer to the first question is “no”, so it is not necessary to address the second question.

(b) Principles Governing New Trials for Associated Counts

98An appellate court’s power to order a new trial originates in statute. The relevant statutory provision is s. 686(1) and (2) of the Criminal Code, which reads as follows:

99Section 686(1)(a) authorizes an appellate court to allow an appeal and set aside a verdict if the appellant can show that it was unreasonable or cannot be supported by the evidence, that an error of law was made, or that a miscarriage of justice occurred. In Tayo Tompouba, this Court identified the principle underlying each of these three grounds of appeal, at para. 54:

100Section 686(1)(a) is set out in broad terms to capture any verdict rendered in a manner that was truly unfair to the accused. However, appellate intervention is not unbounded. There is no authority for appellate courts to interfere with a verdict that is unaffected by error and untarnished by a miscarriage of justice (Tayo Tompouba, at para. 56). It is a basic rule, fundamental to our system of adjudication, that appellate courts do not disturb jury verdicts that are both reasonable and fair.

101As discussed above, the respondents correctly point out that failure to leave an included offence with the jury is an error of law that triggers appellate intervention under s. 686(1)(a)(ii) (Sarrazin, at para. 65). A conviction on any count that is materially impacted by that error should be set aside. Indeed, there will be situations where a failure to leave an included offence in respect of one count materially impacts not only that count but also an associated one. In such cases, it will be appropriate to set aside both verdicts and order new a trial on each count.

102In Ronald, for example, two co-accused, Ms. Ronald and Mr. Gill, were charged with the first degree murder of the victim, Mr. Gill’s wife. The Crown’s theory was that Mr. Gill was involved in the planning but that Ms. Ronald alone committed the killing. The Court of Appeal held that the trial judge erred by failing to leave second degree murder open as a possible verdict for Ms. Ronald because there was an air of reality to the theory that she committed the murder without having planned to do so with Mr. Gill.

103The Court of Appeal correctly ordered a new trial both for Ms. Ronald and for Mr. Gill, because on the alternative theory where the killing was unplanned, Mr. Gill was innocent. Therefore, in effect, the jury had been deprived not only of a valid pathway to conviction on a lesser included offence for Ms. Ronald but also of a valid pathway to an acquittal for Mr. Gill. This was a case in which both convictions were materially impacted by a legal error made in respect of instruction on only one of them.

104However, in cases where there are convictions that are left unaffected by a legal error, the risk of inconsistency as a result of ordering a new trial on the affected count is not in itself a valid basis for ordering a new trial on the unaffected count.

105The issue of factual inconsistency across separate trials has arisen several times in cases where co-accused are tried separately with different results. In Rémillard v. The King (1921), 62 S.C.R. 21, this Court addressed whether it is possible for a principal to be convicted of a lesser offence than a party at two separate trials. The Court unanimously held that there was no inconsistency in those results. As Anglin J. explained:

106Jennifer argues that Rémillard is distinguishable and instead relies on this Court’s decision in R. v. Nygaard, [1989] 2 S.C.R. 1074. In that case, two co-accused, Mr. Nygaard and Mr. Schimmens, were charged with first degree murder. Having found a legal error relating exclusively to Mr. Schimmens’ alibi defence, the Court concluded that both accused had to have a new trial:

107A similar issue was taken up two years later in R. v. Hick, [1991] 3 S.C.R. 383. There, two co-accused, Mr. Hick and Mr. Marshall, were charged with unlawful confinement and sexual assault. Mr. Hick entered a guilty plea upon an indictment alleging commitment of the offence of unlawful confinement, while Mr. Marshall went to trial, where he was acquitted of both counts. At issue was whether Mr. Hick should have been entitled to withdraw his plea in light of Mr. Marshall’s acquittal. Citing Rémillard, this Court held that he could not, as “the jury verdict is only conclusive as between the Crown and the accused at that trial” (p. 386).

108In my view, Rémillard and Hick articulate the correct legal principles. To the extent that it is inconsistent, Nygaard should not be followed because it was decided per incuriam. A decision is made per incuriam, or by inadvertence, when it fails to consider a relevant statute or binding authority (see R. v. Sullivan, 2022 SCC 19, [2022] 1 S.C.R. 460, at para. 77). For a decision to be overturned on the basis that it was made per incuriam, it is not enough to point to an authority that is missing from the reasons. The inadvertence must “have struck at the essence of the decision” (para. 77). In other words, “it must be shown that the missing authority affected the judgment” (ibid., citing M. Rowe and L. Katz, “A Practical Guide to Stare Decisis” (2020), 41 Windsor Rev. Legal Soc. Issues 1, at p. 19).

109The Court in Nygaard did not reconcile its decision with Rémillard, which was binding authority. Had the Court in Nygaard considered the legal constraints imposed on it by Rémillard, it would have reached a different result; it could not have concluded that Mr. Nygaard had to receive a new trial solely because his co-accused Mr. Schimmens was to receive one. To the extent that Nygaard stands for the proposition that a party must receive a new trial solely because the principal offender has received a new trial, it should not be followed.

110This conclusion brings the common law into conformity with s. 23.1 of the Criminal Code, which was enacted prior to the Court’s decision in Nygaard but after the offence in that case was committed. Section 23.1 states that ss. 21 to 23, which are provisions on party liability, apply to an accused “notwithstanding the fact that the person whom the accused aids or abets, counsels or procures or receives, comforts or assists cannot be convicted of the offence”. The enactment of s. 23.1 put to rest any argument that a party must receive a new trial solely because the principal will receive a new trial.

111A jury’s factual findings at a trial are confined to that trial because they are wholly dependent on, and inextricable from, the evidence and argument at that trial (Rémillard, at pp. 23-24; Hick, at p. 386; R. v. Chol, 2021 BCCA 279, 73 C.R. (7th) 78, at para. 31). Parties make tactical decisions about what evidence to adduce, how to frame it, and what theories to argue. It is a necessary feature of our system of justice that different trials may yield results that imply different factual findings. This follows directly from the fact that innocence is presumed and that this presumption can be rebutted only through evidence properly adduced at the trial in which a person’s guilt is considered.

112In sum, there is no statutory authority for an appellate court to order a new trial solely to pre-empt the risk of factually inconsistent verdicts sometime in the future, nor would such an order be appropriate in any event. As a basic principle, appellate courts do not disturb validly rendered, untainted jury verdicts.

(c) Application to the Facts

113The only reason to set aside the attempted murder convictions in respect of Mr. Pan and order them to be considered at a new trial would be if they were materially impacted by the trial judge’s failure to leave second degree murder and manslaughter with the jury in respect of the death of Mrs. Pan. The court below was correct that the attempted murder verdict was not materially impacted by that error.

114Unlike in Ronald, the trial judge’s failure to leave second degree murder and manslaughter with the jury in the killing of Mrs. Pan in no way impeaches the jury’s verdict on the attempted murder of Mr. Pan. The entire basis for the argument that those included offences should have been put to the jury was counsel for Jennifer’s insistence that the respondents’ plan may plausibly have been to kill only Mr. Pan. Far from being exculpatory with respect to the attempted murder of Mr. Pan, the theories of liability underlying those included offences both imply the respondents’ guilt on that count. Allowing the jury to consider the alternative theories would thus not have raised any basis for doubt on that count.

115In sum, the error in failing to leave the included offences with the jury on the first degree murder count did not taint the jury’s verdict on the attempted murder count. There is thus no statutory authority for an appellate court to order a new trial on the attempted murder count on the basis of that error. I turn now to the remaining grounds of cross-appeal, which the respondents argue provide additional bases for ordering a new trial on both the first degree murder count and the attempted murder count.

(2) Did the Trial Judge Err in Permitting the Crown Slide Show to Go to the Jury Room?

116This next ground of cross-appeal raises an issue of broader significance about the use of jury aids. Exponential advances in technology have rendered both criminal activity and police investigations increasingly complex. This growing complexity means that the evidence presented at modern criminal trials is often vast and thorny. Jury aids, which synthesize and explain evidence, play an important role in the pursuit of truth in these cases. At the same time, jury aids pose risks that require the careful attention of trial judges. Here, I consider when they ought to be permitted to go to the jury room to assist in deliberations.

(a) Background and Decisions Below

117Shortly before the close of its case, the Crown disclosed a 231-slide PowerPoint presentation. That slide show summarized evidence from disparate sources, including call detail records, cell tower evidence, data recovered from cell phones, bank withdrawal records, the time of the 9-1-1 call, and video surveillance footage records. Slides detailed how over 400 calls, almost 100 text exchanges, and the contents of over 250 individual text messages implicated all of the co-accused.

118The Crown sought to have the slide show admitted as an exhibit and to call as a witness the Ontario Provincial Police analyst who created it. The respondents agreed that the Crown could use the slide show in its closing submissions, but Jennifer objected to the late disclosure, Jennifer and Mr. Crawford objected to its admissibility as an exhibit and to calling the analyst as a witness, and Mr. Mylvaganam and Mr. Wong sought to cross-examine the analyst and opposed the slide show going to the jury room as a numbered exhibit if changes could not be made to it as a result of cross-examination.

119The trial judge ruled that the Crown could not call the analyst but could use the presentation in closing submissions (2014 CanLII 74050). After the defences had closed, the trial judge ruled that he would permit the slide show and related charts to go to the jury room but would not enter the slide show as an exhibit (2014 ONSC 6055). According to the trial judge, the slide show would greatly assist the jury in making sense of a massive volume of evidence, and although it brought the Crown’s case into focus, it also revealed its limitations. The slide show also served as a starting point for the defence to demonstrate shortcomings in the Crown’s theory of the case. The trial judge instructed the jury on the limitations of relying on the slide show in its deliberations.

120The Court of Appeal held that the trial judge did not commit a reviewable error in his decisions regarding the slide show. It noted in particular that the evidence in this case was dense, complicated and lengthy and that defence counsel had a copy of the slide show before they made their closing addresses and were able to address any shortcomings in it. It was also significant that defence counsel were allowed to give the jury other charts that reflected evidence they planned to emphasize and that the trial judge gave careful and explicit instructions to the jury on limitations inherent in the slide show.

(b) Defining Jury Aids

121Jury aids are tools to elucidate the significance of a piece or collection of testimonial, documentary, or real evidence. They include maps, timelines, casts of characters, charts, and slide shows. Whereas testimonial, documentary, and real evidence “provides the trier of fact with an opportunity to draw a relevant first hand impression”, jury aids perform a secondary function, namely “to assist the jury [to] interpret, understand or analyze” this evidence (S. C. Hill, D. M. Tanovich and L. P. Strezos, McWilliams’ Canadian Criminal Evidence (5th ed. (loose-leaf)), at § 23.1).

122Unlike evidence, jury aids do not provide a standalone basis for factual findings or legal conclusions. Their utility is wholly dependent on whether the trier of fact independently accepts the evidence on which they are based (R. v. Scheel (1978), 42 C.C.C. (2d) 31 (Ont. C.A.), at p. 34; see also R. v. Kanagasivam, 2016 ONSC 2250, 29 C.R. (7th) 201, at para. 41; R. v. Shaw, 2004 NBQB 260, 277 N.B.R. (2d) 306, at para. 8).

123Jury aids come in all forms and vary in their sophistication, and the distinction between jury aids and evidence may sometimes be difficult to draw (see Hill, Tanovich and Strezos, at § 23.1). In some instances, tools that may accurately be classified as jury aids will nonetheless be admitted as evidence through a witness and made exhibits in accordance with the ordinary rules of evidence. This case does not concern the practice of admitting jury aids as evidence in accordance with those rules.

124Rather, the issue raised is when jury aids may be permitted to go to the jury room outside of the normal rules of evidence for use by the jury in the course of its deliberations. In the Court of Appeal, Nordheimer J.A. observed that, in the usual course, only two categories of material will go to the jury room: trial exhibits and jury aids proffered on consent (para. 114). As he rightly pointed out, notwithstanding the consent of the parties, the trial judge retains discretion not to send an aid to the jury room, and conversely, the trial judge has discretion to permit a contested aid to go to the jury room in some circumstances. I turn to those circumstances now.

(c) Framework for Permitting the Use of Jury Aids in the Jury Room

125As a conceptual matter, the discretion to permit jury aids to go to the jury room falls within the domain of the trial judge’s trial management powers, which flow from the court’s inherent or implied jurisdiction to control its own processes and to “ensure that trials proceed in an effective and orderly fashion” (R. v. Samaniego, 2022 SCC 9, [2022] 1 S.C.R. 71, at para. 20; see also R. v. Anderson, 2014 SCC 41, [2014] 2 S.C.R. 167, at para. 58).

126The party seeking to have a jury aid go to the jury room should disclose it to the opposing party as soon as reasonably possible after it is prepared and make an application to the court to tender the aid. Where the opposing party contests the aid going to the jury room, the trial judge should solicit submissions. Aids that satisfy the criteria laid out below should be permitted to go into the jury room and should be formally marked in a manner that distinguishes them from the evidence.

127If an aid is permitted for use by the jury, it is incumbent on the trial judge to instruct the jury as to the proper and improper uses of the aid (see, e.g., R. v. Hovila, 2013 CarswellAlta 2965 (Q.B.), at para. 20). For an example of a well-formulated caution, I would point to the trial judge’s instruction in this case, reproduced below. Proper instruction is vital to ensure that the jury does not rely on an aid as “a convenient shortcut” in lieu of examining the evidence (R. v. Belcourt, 2012 BCSC 2128, at para. 10).

128Determining what procedural mechanisms and instructions are necessary in the circumstances will depend on the nature of the aid, the purpose to which it is being put, and the positions of the parties. For instance, a straightforward jury aid such as a map or cast of characters may call for a streamlined or informal approach. By contrast, a complex or obscure aid may require the proffering party to call a witness to explain and authenticate the aid. The trial judge should make decisions with a view to preventing unfairness, preserving trial efficiency, and enhancing the jury’s truth-seeking capacity.

129With respect to substantive criteria, I agree with Nordheimer J.A. that the trial judge erred in relying on the statement from R. v. Bengert (1980), 15 C.R. (3d) 114 (B.C.C.A.), that “the members of the jury [are] entitled to anything which would assist them in dealing with the evidence reasonably, intelligently and expeditiously” (p. 160). This test is too broad and fails to recognize the potential prejudice flowing from overreliance on aids.

130Instead, I would adopt the test laid out by Nordheimer J.A. in the Court of Appeal, with some slight modification. Before permitting a jury aid to go to the jury room during deliberations, the trial judge must be satisfied that the aid is reasonably necessary, accurate, and fair. These criteria serve to balance the value that these aids can provide in elucidating the evidence against their potential to distract or mislead the jury. Jury aids should be permitted to go to the jury room if the former outweighs the latter.

131The first criterion is that the aid must be reasonably necessary for the jury to understand the evidence. An aid will be reasonably necessary where the evidence that it incorporates is so vast, complex, or technical in nature that a jury would struggle to make sense of it without assistance or without expending an unreasonable amount of effort and time. The trial judge need not be satisfied that it would be impossible for the jury to perform its task without the aid; it is enough to show that it would be unreasonably onerous, or unreasonably time-consuming, for the jury to sift through the relevant data points required to understand the evidence without the aid. As Fairburn J. (as she then was) explained in Kanagasivam, at para. 42:

132Juries are not expected to “locate needles in haystacks” (Kanagasivam, at para. 48). However, where the aid merely repackages already accessible evidence in a more attractive or convenient way, the necessity criterion will not be met.

133Second, the aid must summarize the evidence accurately. It cannot distort, misstate, or obscure any evidence, whether intentionally or unintentionally (Kanagasivam, at para. 52). In R. v. Poitras (2002), 57 O.R. (3d) 538 (C.A.), Doherty J.A. described the imperative of accuracy as follows, at para. 48:

134Additionally, an aid may be deemed inaccurate if it fails to capture the complexity of the evidence. As Conlan J. warned in Woods v. Jackiewicz, 2019 ONSC 2069, “we must be cautious to avoid oversimplification of technical evidence through the use of a demonstrative aid” (para. 13(vi)). Abbreviation is not always possible without sacrificing accuracy.

135Third and finally, the trial judge must be satisfied that permitting the aid to go to the jury room would be fair. Assessing fairness involves an overarching inquiry into the aid’s explanatory value and prejudicial effect (see Jackiewicz, at para. 13(iii)). A key consideration is the extent to which the aid reflects one party’s theory of the case, though there is no strict prerequisite that an aid be completely free of the proffering party’s perspective. If the aid does reflect one party’s theory of the case, the opposing party should have an opportunity to submit its own aid, or to submit alterations or additions to the aid. Neutral aids that do not reflect either party’s perspective will more easily satisfy the dictates of fairness.

136In exercising their discretion with respect to jury aids, trial judges should aim to preserve the proper and fair operation of the adversarial system. Trial judges must be alive to resource imbalances and should not permit a one-sided aid to go to the jury room where it would be unduly burdensome for the opposing party to provide competing material. An important consideration is how quickly an aid can reasonably be disclosed after it is prepared. Some aids may need to be updated throughout the trial to satisfy the accuracy criterion or to fairly represent all the evidence. Naturally, it will be more burdensome for a party to provide competing material if the original aid is disclosed for the first time shortly before it is to be presented to the jury.

137The ultimate aims of these criteria are to prevent improper reasoning on the part of the jury and to avoid the appearance of unfairness. The decision to send a jury aid to the jury room or not to do so is, in the end, a discretionary decision for the trial judge to make. As a trial management decision, it is owed deference absent an error in principle or an unreasonable exercise of discretion (Samaniego, at para. 26).

(d) Application to the Facts

138There was no reviewable error in the trial judge’s handling of the Crown slide show. Regarding the procedure by which the jury aids made it into the jury room, the trial judge reasonably balanced the values of fairness and efficiency. He ordered the Crown to provide an unlocked version of the slide show to the defence two and a half months before the closing addresses so that defence counsel could submit modifications, create their own presentations, or address the slide show’s shortcomings in their closing submissions.

139Moreover, his dismissal of the Crown’s motion to call a witness to authenticate the slide show itself was justifiable because the jury had already heard lengthy testimony from three telecommunications company representatives explaining the evidence summarized in the slide show. Additionally, counsel for Jennifer, Mr. Carty, and Mr. Crawford each argued against the motion, protesting that the witness would needlessly rehash the Crown’s case.

140Finally, the trial judge carefully instructed the jury on the slide show’s limitations and on impermissible lines of reasoning:

141With respect to the criteria of necessity, accuracy, and fairness, while the trial judge did not have the benefit of the guidance set out in this decision, I am satisfied that there was no error in permitting the slide show to go to the jury room.

142The slide show was reasonably necessary, because the cell phone evidence that it summarized was voluminous and fragmented across thousands of data points from three separate service providers. Slide shows have been used in courtrooms to synthesize and summarize cell phone tower data (see, e.g., Kanagasivam; R. v. Hamilton, 2011 ONCA 399, 271 C.C.C. (3d) 208). It would have been unreasonable to expect the jury to piece together a timeline on its own when this slide show already existed. The cell phone evidence was a significant portion of the evidence in this case, and it was reasonably necessary to provide the jury with a tool that would assist in interpreting it.

143The respondent Mr. Mylvaganam argues that the slide show was inaccurate because it attributed certain calls to him when the identity of the caller was a live issue. He notes that the Crown conceded that some of the calls attributed to Mr. Mylvaganam in the slide show must have been made by Mr. Carty even though they came from Mr. Mylvaganam’s cell phone. He further complains that the slide show omitted evidence of calls to an alternative suspect.

144In my view, these are not inaccuracies. The slide show labelled the calls with Mr. Mylvaganam’s name because his cell phone was used, but the jury was well aware of the theory that Mr. Carty made some of those calls. This was the subject of a large portion of the closing address by Mr. Mylvaganam’s counsel. The trial judge instructed the jury that part of its task was to determine who was using Mr. Mylvaganam’s phone at any given time. Moreover, the Crown slide show was sent to the jury room alongside an additional aid prepared by Mr. Mylvaganam’s counsel called the “Weird Calls List”. That list fully outlined Mr. Mylvaganam’s claim that he did not make the impugned calls and eliminated any concern that the Crown slide show may have misled the jury on this point. These safeguards reasonably assuaged the concern that the jury might interpret the slide show as settling an issue that was expressly put to it to decide.

145The same is true with respect to the alleged omission. Counsel for Mr. Mylvaganam provided the jury with a chronological list of phone calls involving the alternative suspect, which was allowed to go to the jury room alongside the slide show. The aids together painted an accurate picture of the evidence.

146Overall, it was fair to admit the Crown slide show and to send it to the jury room. Although the slide show reflected the Crown’s theory, in the unique circumstances of this case, its explanatory value outweighed any potential prejudice. The prejudice was greatly reduced by the presence of defence counsel’s competing aids, which rounded out the evidentiary picture, as well as the trial judge’s forceful caution about the dangers of relying on the slide show. The trial judge’s decision to permit the slide show to go to the jury room is owed deference.

(3) Did the Trial Judge Err in His Approach to the Juror Bias Issues?

147About four months into the trial, the jury sent the trial judge a note raising concerns about text message exchanges between juror 4 and his spouse who was attending the trial as a member of the public:

148The trial judge conducted an inquiry of each juror, starting with juror 4, whose spouse was referred to in the note. Juror 4 said that the texts did not reveal any information about what had transpired in court in the absence of the jury but rather dealt with some hallway discussions. The trial judge found as fact that the information passed along by the spouse of the juror consisted of the following: delays because a witness or counsel was late or because Mr. Carty did not receive his sandwich, a comment on the physical appearance of an upcoming witness, and the name of an upcoming witness. All of the jurors confirmed that they remained impartial.

149The trial judge dismissed the application for a mistrial. He found that although the texts were improperly sent, they were not so serious that they destroyed or undermined the appearance of justice and fairness. Juror 4’s spouse did not advise him of discussions that took place inside of the courtroom in the absence of the jury. It was naive to believe that non-sequestered jurors in the circumstances would be completely insulated.

150A subsequent issue arose when the same spouse expressed strong opinions on the case to a detective who was a key Crown witness. He testified that juror 4’s spouse made a clapping motion to him, which he understood to mean she was applauding his work, that she said he “needed to fight back more”, and that she “didn’t see any rubber hoses” (2014 ONSC 4645, at para. 7), which he understood to mean that he had not forced anyone to say anything against their will. Additionally, counsel for Mr. Carty informed the court that he too had interacted with juror 4’s spouse, asking for her impressions of the case and treating her “sort of like a 13th juror” (A.R., part V, vol. XLIV, at p. 136).

151The trial judge conducted a second inquiry during which juror 4 confirmed that he was not aware of any comments that his spouse made about any of the participants in the proceeding. He also confirmed that nothing he heard about the case from outside of the courtroom affected his view of the process or any of the participants. Defence counsel jointly argued that juror 4 should be discharged, and one defence counsel requested a mistrial.

152The trial judge denied the request for a mistrial and declined to discharge juror 4. He held that there was no evidence of actual bias and that the spouse’s conduct did not give rise to a reasonable apprehension of bias or undermine the appearance of justice and fairness. In coming to this conclusion, the trial judge accepted juror 4’s statement that he was not aware his spouse made the comments in question and that she had not expressed any opinions to him. Although the conduct was “inappropriate” and the “optics aren’t great”, a reasonable and informed observer would not consider the conduct sufficient to raise a reasonable apprehension of bias (2014 ONSC 4645, at paras. 41 and 48). This ruling dealt with the detective witness’s testimony but did not expressly address defence counsel’s interactions with the spouse.

153The Court of Appeal held that, although it would have been preferable in the first inquiry to ask to see the text messages, it was open to the trial judge to accept juror 4’s statement that he was not aware of his spouse’s comments and that her views had not been shared with him. A trial judge’s decision addressing a jury issue is entitled to significant deference (R. v. Durant, 2019 ONCA 74, 144 O.R. (3d) 465). The trial judge held that the events had not undermined the appearance of fairness and, ultimately, that was his call to make (R. v. Kossyrine, 2017 ONCA 388, 138 O.R. (3d) 91).

154I agree with the Court of Appeal that decisions addressing juror issues are entitled to significant deference (Durant, at para. 152; see also Vauclair, Desjardins and Lachance, at para. 27.54). An inquiry is sufficient if it allows the trial judge to “determine the true basis of the claim for discharge and to resolve it” (Durant, at para. 141).

155The Court of Appeal properly dismissed this ground of appeal. The trial judge gathered the information necessary to determine whether there was actual bias or a reasonable apprehension of bias. Whether to go further and call juror 4’s spouse or require the production of the text messages was a discretionary decision that depended on the trial judge’s assessment of the entire situation, from the perspective of a reasonable and informed observer. His decision not to do so did not amount to a reviewable error.

(4) Did the Trial Judge Err in His Instruction on the Propensity Evidence?

156Mr. Mylvaganam’s defence was that he never left the car on the night of the shootings and believed that the others were there only to carry out a home robbery. Central to this defence was Mr. Pan’s initial description of the assailants as two black men and a white man. Counsel for Mr. Mylvaganam argued that the two black men that Mr. Pan described were an alternative suspect and Mr. Carty. As Mr. Mylvaganam is not white, he argued he could not have been the third assailant.

157The trial judge permitted Mr. Mylvaganam to adduce propensity evidence against Mr. Carty to support the proposition that Mr. Carty was one of the shooters. The evidence included Mr. Carty’s nickname, “Sniper”, two tattoos, including “SNYPA” with an AK-47 on Mr. Carty’s calf and a neck tattoo, and three of Mr. Carty’s convictions for violent offences, including a 2013 conviction for first degree murder.

158During the pre-charge conference, counsel for Mr. Crawford argued that the propensity evidence against Mr. Carty could prejudice his own client because of Mr. Crawford’s close relationship with Mr. Carty. The trial judge agreed. On the second day of his charge to the jury, he warned against improper use of the propensity evidence:

159On the third day of the charge, the trial judge repeated that instruction:

160The Court of Appeal held that the instructions on how to use the propensity evidence were sufficient and that the trial judge did not have a duty to tailor this caution to Mr. Crawford specifically. The Court of Appeal noted that “[i]n a trial with multiple accused persons, that complaint could be made with respect to almost any instruction that the trial judge gives” (para. 133).

161I agree with the Court of Appeal that the trial judge’s approach to the propensity evidence provides no basis for interfering with the convictions. The jury instructions were adequate. It was clear that the trial judge’s reference to “any” of the accused included Mr. Crawford, particularly given that Mr. Crawford’s counsel raised this concern and issued his own warning about impermissible reasoning in his closing statement.

162There is no duty on a trial judge to relate every instruction to each accused by name. Doing so in this case would have needlessly added further complexity to already long and complex jury instructions. As this Court stated in R. v. Abdullahi, 2023 SCC 19, at para. 35, “[t]he accused is entitled to a jury that is properly, not perfectly, instructed”. Mr. Crawford’s argument comes across as a demand for perfection. It cannot succeed.

VI. Disposition

163I would dismiss the appeal and all of the cross-appeals.

Reasons Dissenting in Part — Karakatsanis J. (Martin J. concurring)

I. Overview

164I join with the Chief Justice in dismissing the cross-appeals. However, respectfully, I find no grounds to interfere with the trial judge’s decision not to instruct the jury on lesser included offences regarding the killing of Jennifer Pan’s mother, Bich-Ha Pan.

165I agree with the Chief Justice that a trial judge must instruct a jury on a lesser included offence if and only if there is an air of reality on the totality of the evidence that the accused is guilty of that offence but not guilty of the charged offence. I also agree that the air of reality standard is the same in this context as for an affirmative defence. However, I would emphasize key differences in the application of the test in the context of a lesser included offence.

166An air of reality analysis for a lesser included offence is necessarily a broader and more nuanced analysis than an air of reality analysis for an affirmative defence. The question is not simply whether the jury could conclude that the elements of the lesser included offence have been proven beyond a reasonable doubt. Rather, the question is whether the jury could realistically reach that conclusion, while also concluding that the Crown has not proven beyond a reasonable doubt any distinguishing element of the charged offence. Thus, the trial judge must assess the totality of the record to determine whether there is direct evidence or reasonably available inferences that could support a plausible theory of the case leading both to an acquittal on the main charge, and a conviction on the included offence. In this sense, the trial judge’s task does not relate to one discrete issue, but to whether a route to both verdicts is plausible given all the evidence. This added breadth and nuance entails a qualitatively different exercise.

167In a jury trial, the jury is the sole trier of fact, and is responsible for returning a verdict (R. v. Gunning, 2005 SCC 27, [2005] 1 S.C.R. 627, at paras. 27-28). At the same time, the trial judge plays a “fundamentally important role” as a gatekeeper (R. v. Grant, 2015 SCC 9, [2015] 1 S.C.R. 475, at para. 44). The trial judge must screen what evidence is put before the jury, and what legal instructions to give the jury as to the available verdicts (R. v. Arp, [1998] 3 S.C.R. 339, at paras. 47-48; R. v. Tran, 2010 SCC 58, [2010] 3 S.C.R. 350, at para. 41; R. v. Khill, 2021 SCC 37, [2021] 2 S.C.R. 948, at para. 114). This duty exists to make sure “the jury understands its task and is properly equipped to make its decision” (R. v. Abdullahi, 2023 SCC 19, at para. 32).

168While the air of reality threshold is low, it is a meaningful and important gatekeeping exercise. Failing to properly instruct a jury on a lesser included offence could lead the jury to improperly find the accused not guilty of any charge. But, in the words of Doherty J.A., instructing the jury on implausible theories of lesser included offences invites only “confusion and compromise. Neither are conducive to a true verdict” (R. v. Wong (2006), 209 C.C.C. (3d) 520 (Ont. C.A.), at para. 12).

169Appellate courts should not presume to be on equal footing to trial judges in their assessment of what plausible theories of the case are available on the totality of the record. This case amply shows the advantage a trial judge has in seeing a case play out live, in all its manifold complexity.

170The trial judge properly exercised his important gatekeeping function in refusing to leave lesser included offences with this jury. There was no direct evidence for the theory that on the night in question, the respondents acted on a plan to kill only Hann Pan, Ms. Pan’s father. And the trial judge — who observed the voluminous evidence firsthand — also concluded that the evidence could not realistically support circumstantial inferences in favour of that theory. I find no error in his firsthand assessment of the evidence. It is not the role of an appellate court to draw inferences from isolated pieces of evidence that are speculative and contrary to an overwhelming body of evidence. This is especially so given much of the evidence relating to what happened the night of the killing was undisputed. I would not interfere with the trial judge’s conclusion that there was no air of reality to a verdict of second degree murder or manslaughter for the killing of Ms. Pan’s mother.

171Accordingly, I would allow the appeal, dismiss the cross-appeals, and restore the convictions for first degree murder.

II. Instructing a Jury on Lesser Included Offences

A. The “Air of Reality” Test in the Context of Lesser Included Offences

172A trial judge must instruct jurors that they may find the accused not guilty of a charged offence but guilty of a lesser included offence, where there is an air of reality on the totality of the evidence for that verdict (Smith v. The Queen, [1979] 1 S.C.R. 215; R. v. Aalders, [1993] 2 S.C.R. 482, at pp. 504-5). An offence is a “lesser included offence” of a charged offence when it is stated to be so in the Criminal Code, R.S.C. 1985, c. C-46, or when its elements are included within those of the charged offence (R. v. G.R., 2005 SCC 45, [2005] 2 S.C.R. 371, at para. 25; Fergusson v. The Queen, [1962] S.C.R. 229, at p. 233).

173This Court has applied the air of reality standard in several contexts, most notably when assessing whether a jury ought to be instructed on an affirmative defence (see, generally, R. v. Cinous, 2002 SCC 29, [2002] 2 S.C.R. 3). The air of reality test exists to respect the “basic division of tasks between judge and jury” (R. v. Osolin, [1993] 4 S.C.R. 595, at p. 682). It is the jury’s responsibility as the trier of fact to reach a verdict (R. v. Krieger, 2006 SCC 47, [2006] 2 S.C.R. 501, at para. 18). However, the trial judge plays an important role as gatekeeper in screening what verdicts are plausibly available to the jury. I agree with the Chief Justice that “[t]he trial judge must assess whether there is a reasonable view of the evidence upon which a properly instructed jury acting judicially could be left with a reasonable doubt about elements of the principal offence that distinguish it from the included offence, while accepting beyond a reasonable doubt all of the elements of the included offence” (para. 51).

174When a party requests that the trial judge instruct the jury on specific lesser included offences, on the basis of a specific theory of the case, the trial judge’s analysis consists of two elements. The trial judge must assess all the evidence and determine whether the range of reasonable factual inferences necessary for the theory is available on the record. And they must determine if the theory is plausible, meaning that a reasonable, correctly instructed jury could adopt it and return a guilty verdict for a lesser included offence. This theory must enable the jury to have a reasonable doubt on the charged offence, while at the same time satisfying the jury that the lesser included offence is established beyond a reasonable doubt (see, generally, Cinous, at para. 82).

175Determining the range of reasonable available inferences requires a different assessment for direct and circumstantial evidence. Direct evidence is evidence which, if believed, is enough to establish a fact in issue. The trial judge must accept as true all facts for which there is some direct evidence, except for bare assertions that are inconsistent with an overwhelming body of evidence or a fact not in dispute (R. v. Park, [1995] 2 S.C.R. 836, at para. 30; R. v. Gauthier, 2013 SCC 32, [2013] 2 S.C.R. 403, at paras. 60-61; Aalders).

176Circumstantial evidence is evidence which, if believed, can give rise to inferences in support of a fact (R. v. Villaroman, 2016 SCC 33, [2016] 1 S.C.R. 1000, at para. 23). In assessing circumstantial evidence, the trial judge must engage in a “limited weighing” exercise, whereby they determine and accept all inferences that can reasonably be drawn from the evidence as a whole, without regard to the inherent reliability of any individual piece of evidence (Cinous, at paras. 89-91; R. v. Arcuri, 2001 SCC 54, [2001] 2 S.C.R. 828, at paras. 23-24).

177The limited weighing exercise is tightly constrained. The trial judge’s task at this stage is not to make findings of fact, or determine whether any given evidence is credible or reliable. At the same time, the trial judge’s role is not to consider what inferences can be drawn from each piece of evidence in isolation. The assessment is conducted on the totality of the evidentiary record. So, where an isolated piece or body of evidence could reasonably give rise to an inference, but that inference is unreasonable when the evidence is viewed as a whole, that inference cannot form part of the range of reasonable inferences (Arcuri, at para. 34; M.M. v. United States of America, 2015 SCC 62, [2015] 3 S.C.R. 973, at para. 47). Given the low threshold, the range of reasonable inferences may include contradictory inferences, and inferences that would require the jury to reject pieces or bodies of evidence.

178In assessing the theory of the case put forward by the party asking for the jury to be instructed on the lesser included offence, the trial judge should ask two questions: Given the range of reasonable inferences, could a reasonable, correctly instructed jury find that the proposed theory is plausible? And does that theory mean that the lesser included offence is established beyond a reasonable doubt, while allowing for reasonable doubt on the additional elements of the charged offence?

179A theory of the case consistent with a lesser included offence must be more than “tenuous and speculative” (R. v. Kelsie, 2019 SCC 17, [2019] 2 S.C.R. 101, at para. 2). Rather, it must be a “realistic possibility” (R. v. Wolfe, 2024 SCC 34, at para. 50). The theory can also rely on different modes of liability than the Crown’s theory on the charged offence. For example, in this case, the Crown’s theory of the charged offence relied on aiding and abetting liability (Criminal Code, s. 21(1)) for Ms. Pan, while the theory supporting the lesser included offences relied on common purpose party liability (s. 21(2)). It follows that that the theory supporting the lesser included offence may rely on different inferences from the evidence than the Crown’s theory supporting the charged offence, and not necessarily fewer inferences.

180The trial judge must not instruct the jury on a theory that relies on inferences that are each reasonably possible in isolation, but contradict each other such that they cannot plausibly coexist. Theories constructed on contradictory inferences are unreasonable. Instructing a jury on such a theory would be to invite the jury to potentially return an unreasonable verdict (R. v. R.P., 2012 SCC 22, [2012] 1 S.C.R. 746, at para. 9). This would violate the judge’s role as gatekeeper. At the same time, a failure to instruct the jury if the trial judge is satisfied that an air of reality exists for a lesser included offence would undermine the jury’s role as fact-finder.

181The air of reality test is the same in all contexts, in that the trial judge must determine whether the range of reasonable inferences can support a realistic theory of the case. However, applying the air of reality test in the lesser included offence context will require a more nuanced analysis than in an affirmative defence context. When considering an affirmative defence, the trial judge need only consider whether there is an air of reality to each element of the defence. But for lesser included offences, the trial judge must find that there is an air of reality to a theory of the evidence where the Crown has proven each element of the lesser included offence, but did not prove the additional elements of the charged offence. For example, in a murder trial, where the evidence is overwhelming that the victim was intentionally killed and the only contested fact is the identity of the killer, there may be an air of reality to each element of the lesser included offence of manslaughter. But since there would be no air of reality to the notion that the Crown could have proven the accused guilty of manslaughter but not murder, it would be an error to instruct the jury on manslaughter.

182The analysis also differs in terms of the scope of the trial judge’s inquiry. In the affirmative defence context, the trial judge’s task is simplified because they need only consider direct evidence and circumstantial inferences that favour the accused, excluding bare, self-serving assertions (Cinous, at para. 98; Gauthier, at para. 25; Osolin, at p. 687). But in the lesser included offence context, the trial judge must take a broader view, and consider the full range of all reasonable inferences, regardless of whether they seem to favour one party or another. Further, while trial judges must always consider all the evidence when applying the air of reality test, an affirmative defence is a more discrete issue, for which large bodies of evidence may give rise to no relevant inferences. Finally, an air of reality analysis for a lesser included offence must consider not just the range of reasonable inferences in isolation, but also whether those reasonable inferences can support a plausible theory of the case that would permit the jury to find the accused not guilty of the charged offence but guilty of a lesser included offence. In this sense the trial judge’s exercise does not relate to one discrete issue, but must engage with all the evidence to determine whether a given theory of lesser liability is plausible.

B. The Standard of Review for Assessing an Air of Reality

183The standard of review applicable to a trial judge’s air of reality determination has been called an area of “some complexity” (R. v. Land, 2019 ONCA 39, 145 O.R. (3d) 29, at para. 71). On the one hand, the Crown points to cases where this Court has recognized the trial judge’s advantage in seeing the evidence firsthand, and called for appellate courts to show deference in reviewing whether an air of reality existed (see A.F., at para. 6, citing R. v. Thibert, [1996] 1 S.C.R. 37, at para. 33; R. v. Cairney, 2013 SCC 55, [2013] 3 S.C.R. 420, at para. 63). But Ms. Pan points to cases where this Court has stated that the standard of review for whether an air of reality existed is correctness (see R.F., at para. 124, citing R. v. Buzizi, 2013 SCC 27, [2013] 2 S.C.R. 248, at para. 15; Tran, at para. 40; R. v. Alas, 2022 SCC 14, [2022] 1 S.C.R. 283, at para. 3).

184These lines of jurisprudence are reconcilable. In my view, a correctness standard on the ultimate question of whether an air of reality exists can coexist with a degree of deference to a trial judge’s “limited weighing” based on the assessment of the evidence.

185Appellate courts routinely defer to a trial judge’s assessment of the evidence, while reviewing their ultimate legal determination for correctness. Thus, this Court stated in R. v. Shepherd, 2009 SCC 35, [2009] 2 S.C.R. 527, while reviewing whether reasonable and probable grounds exist to make a breath demand, that “[a]lthough the trial judge’s factual findings are entitled to deference, the trial judge’s ultimate ruling is subject to review for correctness” (para. 20). More generally, the admissibility of evidence is a question of law always reviewed on a standard of correctness, but a trial judge is owed deference on their fact-finding feeding into admissibility determinations (R. v. Samaniego, 2022 SCC 9, [2022] 1 S.C.R. 71, at para. 25; R. v. Couture, 2007 SCC 28, [2007] 2 S.C.R. 517, at para. 81). Appellate courts apply these dual standards of review every day. Trial judges do not make findings of fact in an air of reality analysis, but the limited weighing of the evidence similarly engages a trial judge’s assessment of all of the evidence. In my view, deference to the trial judge’s assessment of the evidence when reviewing an air of reality analysis does not confuse or complicate the law.

186I agree with the Chief Justice that whether an air of reality exists is a question of law. The ultimate decisions on whether a theory of the case is realistic and whether it is capable of sustaining a conviction for a lesser included offence are questions of law that must be reviewed on a standard of correctness. These questions are binaries that must give rise to a single, correct answer.

187However, correctness is an inappropriate standard in reviewing a trial judge’s limited weighing exercise in determining the range of reasonable inferences on the totality of the evidence. As this Court clearly held in Thibert, at para. 33, the trial judge is best suited to make these determinations in an air of reality analysis, since, unlike an appellate court, they have the advantage of experiencing the entire trial record firsthand. In recognition of this, an appellate court should be reluctant to overturn a trial judge’s determination of what inferences are realistically available from the totality of the evidence.

188This appeal amply shows the justification for this posture of appellate deference. The trial judge had a front-row understanding of this massive trial record, with proceedings spanning over 100 court days, and giving rise to over 300 exhibits. Against this backdrop, absent an extricable error in principle, a measure of appellate deference is grounded in judicial humility and respect.

III. Application

189Before the charge to the jury, counsel for Ms. Pan asked the trial judge to instruct the jury on the lesser included offences of second degree murder and manslaughter in relation to the killing of her mother, Bich-Ha Pan. The trial judge declined, stating in an email sent to counsel that

190Ms. Pan and the other respondents to the Crown’s appeal contend that the trial judge erred by reducing liability to an “all-or-nothing” proposition, and identifying only two plausible theories: either that Ms. Pan and the other respondents were involved in a plan to kill both her parents, or that the shootings occurred as the result of a home invasion “gone wrong”. The trial judge instructed the jury that Ms. Pan must be acquitted if they concluded the shooting occurred because of a home invasion. According to the respondents, an alternative theory should have been left to the jury — that the murder of the father had been planned and deliberated upon by Ms. Pan and the co-accused, but that they had not intended for her mother to be targeted. If so, Ms. Pan and the other respondents could be guilty of the second degree murder or manslaughter of her mother, rather than first degree murder.

191Under this theory, liability would be grounded under s. 21(2) of the Criminal Code. As the Court of Appeal noted, under this provision, Ms. Pan and the other respondents could be found guilty of second degree murder or manslaughter based on the common intention of the respondents to carry out the unlawful purpose of murdering her father (2023 ONCA 362, 427 C.C.C. (3d) 4, at para. 72). Ms. Pan and the other respondents would be guilty of second degree murder if they knew that the murder of the mother was a probable consequence of carrying out that plan. They would be guilty of manslaughter if they knew or ought to have known that the infliction of non-trivial bodily harm to the mother was a probable consequence of the plan. In both scenarios, the jury would first need to have a reasonable doubt that the death of the mother was planned and deliberate while still finding that the respondents had participated in the unlawful common purpose to murder the father. I agree that this theory, if plausible and available on the range of reasonable inferences, could have supported a conviction for a lesser included offence. However, as I will demonstrate, the contrary is true — there was no air of reality for this theory, as the inferences required could not be reasonably drawn from the evidence.

192In support of the alternate theory of the evidence, Ms. Pan points to evidence regarding her lack of animus toward her mother, as opposed to her father, and to evidence of an earlier plan she had formed in the spring of 2010 to have only her father killed, which was abandoned well before the events of November 8, 2010. According to her testimony, she had paid the roommate of her friend to shoot her father as he left his place of work. However, Ms. Pan stated that the plan fell through when the roommate took the money and never carried out the shooting. She testified that she had never planned for the death of her mother. In addition, Ms. Pan argues that her more positive relationship with her mother, as corroborated by testimony from her brother and cousin, could ground a reasonable doubt that the killing of the mother was planned and deliberate. Ms. Pan testified at trial that most of the strict rules in her household were imposed by her father. Finally, other relevant evidence also includes Ms. Pan’s testimony regarding her conduct during the home invasion, where she said she had pleaded with the men to be allowed to go to her mother as her parents were led down into the basement.

193However, the evidence of Ms. Pan’s better relationship with her mother and her previous plan to kill her father must be assessed against the totality of evidence of what actually occurred on November 8, 2010 — it cannot be considered in isolation to determine only whether it could raise a reasonable doubt about whether the killing of her mother was planned and deliberate. This is precisely what the Court of Appeal did when it concluded that it would be possible for a jury to have a reasonable doubt as to planning and deliberation while also finding that the route to liability under s. 21(2) had been satisfied. It did not elaborate how, on the totality of the evidence, the jury could plausibly arrive at that conclusion of liability.

194As discussed above, it is not enough in an air of reality analysis to ask the isolated question of whether there could be a reasonable doubt as to planning and deliberation regarding the killing of the mother. Rather, the question is whether a reasonable doubt on that element could realistically coexist with a theory where there was no reasonable doubt on the remaining elements of second degree murder or manslaughter.

195To accept the alternative theory, the jury would need to reject all of Ms. Pan’s trial testimony that no plan for the death of her parents existed on November 8, 2010, and that she had merely cancelled an earlier plan to have herself killed. The jury would need to conclude beyond a reasonable doubt that the respondents had planned and deliberated the death of the father, and find that the events of November 8, 2010 had been a staged home invasion orchestrated by Ms. Pan with the other respondents during which the mother died, despite not being a target of the plan.

196However, there was no evidence based in the actual events surrounding the shootings that can support the reasonable inference that the father was the only target. Much of the evidence about the events of November 8, 2010 was undisputed, and overwhelmingly indicated that both of Ms. Pan’s parents were equally targeted. The mother and the father were treated in the same way by the men who carried out the home invasion, aside from the fact that one of the men struck the father on the head with his gun. The men who entered the home did so only after the mother arrived home from her scheduled line dancing as expected, even though the father had been home in the two and a half hours before the mother returned from her class. Then they took both parents downstairs, covered both their heads with blankets and shot them both in the head, in the same way, execution style. At no time did Ms. Pan direct the men not to harm her mother, or try to stop the apparent home invasion, even though under the proposed alternate theory she was the directing mind of the plan to kill only her father. That at least one of the intruders did not cover their face during the home invasion also suggests that they did not intend to leave any witnesses who could identify them.

197The evidence of what occurred in the hours before the home invasion further dispels any air of reality to the proposed alternate theory of the evidence. Ms. Pan was in continuous contact with the other respondents throughout the day of November 8, 2010. Her call records indicate calls with Lenford Crawford, Daniel Wong, and David Mylvaganam between 6:12 p.m. and 10:05 p.m., along with many texts with Mr. Wong during that same time frame. She spoke on the phone for nearly one minute with Mr. Wong at 9:11 p.m., and her call records indicate two phone calls with Mr. Mylvaganam’s phone in the hour before the home invasion occurred, with the latter of those calls ending mere minutes before the home invasion began around 10:15 p.m. To convict on second degree murder or manslaughter, the jury would have to conclude these calls related to the plan to kill the father. However, the mother was absent from 7:00 p.m. to 9:30 p.m. for her weekly line dancing, which Ms. Pan testified she knew was a regular weekly occurrence. It is significant that despite the continuous communication between Ms. Pan — the directing mind of the plan — and the co-accused on the night in question, the home invasion occurred only after the mother returned home around 9:30 p.m., rather than in the two-and-a-half-hour period before when only the father was home.

198The Crown invites us to consider additional incriminating evidence which it argues counters the theory that the father was the sole target. It points to: Ms. Pan’s statements to police regarding her “parents” and “their” rules and restrictions, rather than singling out her father as the object of her animosity; testimony from Ms. Pan’s friend and his roommate that Ms. Pan had stated that she wanted both her parents dead; testimony from her friend that minutes before the home invasion occurred, Ms. Pan informed him on the phone that a staged home invasion was going to happen during which both her parents were to be killed; and Ms. Pan’s financial motive for eliminating both of her parents — an inheritance which she would not receive if only her father was killed.

199However, it is not the trial judge’s role to evaluate the evidence to determine which theory was more plausible. Although there was evidence based on her statements to her friends that she had planned to kill both parents, and would inherit and benefit financially only if they both died, the jury could have a reasonable doubt about that evidence. A jury can always accept all, some, or none of a witness’s testimony.

200Still, the totality of the undisputed evidence of the events of November 8, 2010, dispels any air of reality to the proposed alternate theory that only Ms. Pan’s father was intended to die that night. As stated by L’Heureux-Dubé J. in Park, “[w]here evidence supporting the accused which is materially in dispute is, realistically viewed, clearly logically inconsistent with evidence which is not materially in dispute, then it can be concluded as a matter of both law and logic that there can be no air of reality” (para. 29 (emphasis deleted)). The test is not to point to “some” evidence which may support an alternate theory. The evidence must be such that the jury could reasonably make the inferences required to convict on the lesser included offences while having a reasonable doubt as to whether the murder of the mother was planned and deliberate. Here, even when Ms. Pan’s evidence is taken at its most favourable, given the totality of the undisputed evidence, it is speculative to find that because Ms. Pan had less animus toward her mother and had months earlier planned the death of only her father, that the plan on November 8, 2010 was also to kill only her father.

201In this sense, the facts in this appeal are similar to those in Aalders. There, this Court dismissed the accused’s appeal from his conviction of first degree murder, finding that there was no air of reality to the included offence of manslaughter. This conclusion was based on the undisputed events of the murder — this Court held that the accused’s statement that he had not intended to kill the victim but only to shoot him in the legs was irreconcilable with the fact that the accused had concealed himself in the victim’s home and waited for the victim to return for several hours, before ambushing the victim and shooting him eight times with all but one shot striking the victim in his upper body (pp. 504-5). In Aalders, the evidence in favour of the accused’s theory was his own testimony, direct evidence, but here the respondents can point only to speculative inferences.

202If the jury accepted that Ms. Pan had planned, executed, and orchestrated the home invasion on November 8, 2010 to kill her father, there can be no reasonable inference on the totality of the evidence that the plan did not include the murder of her mother. There is no evidence on the record to indicate that Ms. Pan, the directing mind of the scheme under this theory, tried to prevent the men from targeting her mother, nor is there any evidence that the mother was treated substantially differently from the apparent target of that plan.

203I find no error in the approach taken by the trial judge, who concluded that only “two plausible scenarios arise from the evidence”. Although the reasons of the trial judge are brief, he applied the correct test. He was alive to the importance of plausibility in the air of reality analysis, stating that there were only two plausible scenarios, or theories of liability. This focus on the plausibility of the possible scenarios, regarding the totality of the evidence, was correct. Moreover, the trial judge was best placed to make this determination, having presided over a long and complex trial for over nine months and having seen the testimony of the witnesses firsthand. To instruct the jury on the lesser included offences, in a trial which was already exceptionally long and complex, would have added unnecessary complexity and potential confusion, particularly when one considers that the charge to the jury took three and a half days to deliver and spanned 573 pages. Such instructions could also have led the jury to reach an unreasonable verdict. On the totality of the evidence here, instructing the jury on the lesser included offences would run contrary to the trial judge’s important gatekeeping role.

IV. Conclusion

204I would allow the Crown’s appeal, dismiss the respondents’ cross-appeals, and restore the convictions for first degree murder entered at trial.