CITATION: R. v. Sarrazin, 2010 ONCA 577
DATE: 20100909
DOCKET: C47693/C48209
COURT OF APPEAL FOR ONTARIO
Doherty, Moldaver and Epstein JJ.A.
BETWEEN
Her Majesty the Queen
Respondent
and
Robert Sarrazin
Appellant
and
Darlind Jean
Appellant
Russell Silverstein, for the appellant, Robert Sarrazin
Philip Campbell and Howard L. Krongold, for the appellant, Darlind Jean
James K. Stewart, for the respondent
Heard: March 31 and April 1, 2010
On appeal from the convictions for second degree murder entered by Justice Michel Z. Charbonneau of the Superior Court of Justice, sitting with a jury, on November 22, 2006, and the sentences imposed on January 25, 2007.
Doherty J.A.:
I. OVERVIEW
[1] Apaid Noel was shot and seriously wounded outside an Ottawa nightclub on February 19, 1998. He died a month later on March 18, 1998. The appellants were arrested within about two weeks of the shooting. In June 2000, after a lengthy trial, both were convicted of second degree murder and sentenced to life imprisonment without eligibility for parole for 18 years. Almost five years later, in April 2005, this court ordered a new trial: R. v. Sarrazin (2005), 75 O.R. (3d) 485 (C.A.).[^1]
[2] The re-trial proceeded in the late summer of 2006. In November 2006, after a second lengthy trial, the appellants were once again convicted of second degree murder and sentenced to life imprisonment without eligibility for parole for 18 years. These appeals proceeded on March 31, 2010, almost three and a half years after the convictions on the re-trial.
[3] The appellants have been in custody for over 12 years since their arrests in late February and early March 1998.
[4] At the first trial, the appellants, who did not testify, argued that the Crown had failed to establish that they were the persons who had participated in the shooting of Mr. Noel. The jury rejected that contention.
[5] On the first appeal, the appellants raised several grounds of appeal. This court allowed their appeals primarily on the basis that the trial judge had failed to properly apply R. v. Starr, 2000 SCC 40, [2000] 2 S.C.R. 144 in holding that certain statements made by Mr. Noel were admissible under the principled exception to the hearsay rule.[^2]
[6] At the second trial, the appellants again argued that the Crown had failed to prove that they were the individuals who had participated in Mr. Noel’s shooting. In addition, the defence, through cross-examination of one of the Crown’s medical experts, elicited evidence that Mr. Noel’s death may not have been caused by the shooting, but instead caused by the consumption of cocaine shortly before his death. Causation was not an issue at the first trial. Finally, the appellants argued that even if the jury found against them on the questions of identity and causation, the appellants should only be convicted of manslaughter on the basis that the gun discharged accidentally in the course of a struggle. The jury rejected all of the defence contentions and once again convicted the appellants of second degree murder.
[7] The appellants raise two issues on their conviction appeals. First, they submit that while evidence that the appellants and Mr. Noel were members of rival street gangs was admissible, the trial judge erred in admitting extensive evidence about the manner in which those gangs operated, their criminal activities, and the culture of violence that permeated both gangs. The appellants also argue that the trial judge’s limiting instructions with respect to the “gang” evidence were inadequate. Second, the appellants submit that the trial judge failed to instruct the jury that it could convict the appellants of attempted murder if it had a reasonable doubt as to the cause of Mr. Noel’s death. The appellants submit that the trial judge’s instruction to the jury that it must acquit if it had a reasonable doubt on the issue of causation was wrong in law and was prejudicial to the appellants.
[8] In addition to the conviction appeals, the appellants appeal the parole ineligibility period of 18 years. They accept that an increase in the parole ineligibility period from the statutory ten year minimum was warranted, but submit that an ineligibility period of 18 years is manifestly excessive.
[9] The court did not call on the Crown to address the trial judge’s rulings and instructions with respect to the “gang” evidence[^3] or the fitness of the sentences imposed. On the one remaining ground of appeal, the court is unanimously of the view that the trial judge erred in law in failing to leave a conviction for attempted murder as a possible verdict. My colleague, Moldaver J.A., is satisfied that the error resulted in no substantial wrong or miscarriage of justice and would dismiss the appeal. In my view, the curative proviso cannot overcome the legal error and there must be a new trial.
II. THE EVIDENCE
[10] Mr. Noel was gunned down at point blank range with a sawed-off shotgun while standing outside of a nightclub on a public street in Ottawa. He was shot twice, once in the arm and once in the abdomen. Mr. Noel was rushed to the hospital and underwent surgery. The gunshot wound to his abdomen caused serious life-threatening injuries, particularly to his liver. Skilful surgical intervention and medical care saved Mr. Noel’s life, at least for a short time.
[11] Although Mr. Noel’s post-operative progress was somewhat uneven, by March 13, 1998, he was doing well enough to be released from the hospital. Dr. Freeman, his surgeon, expected that Mr. Noel would make a full recovery. Mr. Noel died five days later as a result of a pulmonary thromboembolism, that is, a blood clot.
[12] Moldaver J.A. has thoroughly reviewed the evidence relating to the cause of Mr. Noel’s fatal embolism. I cannot improve on that summary. It is sufficient for present purposes to indicate that the weight of the evidence strongly favoured the opinion that the fatal embolism was a product of the injuries caused by the gunshot wound to the abdomen and the subsequent surgeries necessitated by that wound. There was, however, forensic evidence that Mr. Noel had consumed cocaine shortly before his death, and expert evidence that it was possible that his consumption of cocaine, and not the gunshot wound and subsequent surgeries, had precipitated the fatal embolism.
[13] The evidence relating to the events that culminated in Mr. Noel’s death came primarily from his brother, who was with him that evening. In his testimony, Mr. Noel's brother described a verbal confrontation outside of the nightclub involving Mr. Noel on one side, and the appellants and others on the other side. Mr. Noel’s brother testified that the appellants said, “You’re going to die tonight” and then left the area outside of the nightclub. They returned about five minutes later. Unbeknownst to Mr. Noel’s brother, and presumably to Mr. Noel, the appellant, Mr. Sarrazin, had armed himself with a loaded sawed-off shotgun.
[14] Upon their return to the nightclub, the appellants renewed their verbal threats against Mr. Noel. Mr. Jean taunted Mr. Noel and attempted to distract him. Mr. Sarrazin stepped between Mr. Jean and Mr. Noel, and pointed the shotgun at Mr. Noel. When Mr. Noel reached out for the gun, Mr. Sarrazin fired, hitting Mr. Noel in the arm. Mr. Noel fell to the ground. Mr. Noel’s brother testified that as Mr. Noel lay on the ground, Mr. Jean and another person yelled at Mr. Sarrazin to kill Mr. Noel. Mr. Sarrazin pointed the shotgun at Mr. Noel and discharged the second barrel, striking Mr. Noel in the abdomen. A struggle ensued between Mr. Noel’s brother and the appellants. The appellants fled the scene and Mr. Noel’s brother was left with the sawed-off shotgun.
[15] The defence vigorously challenged the credibility of Mr. Noel's brother and the reliability of his evidence. The defence claimed that Mr. Noel and his brother had brought the shotgun to the nightclub and precipitated the altercation during which Mr. Noel was shot. Neither appellant testified.
[16] In addition to the eyewitness evidence of Mr. Noel’s brother, there was testimony from other witnesses placing the appellants at the scene. The Crown also relied on what was referred to as “gang” evidence to establish that the appellants had a motive to kill Mr. Noel.
[17] The appellants were members of one Montréal Haitian street gang, and Mr. Noel was a member of a rival Montréal Haitian street gang. There was a history of bad blood between these two gangs. The Crown led evidence as to the manner in which these gangs operated, including their strong propensity for acts of violence against each other. That violence often occurred in public places. The Crown adduced evidence that the appellants and Mr. Noel were heavily involved in gang activity, which included the use of firearms and other violent acts, and that Mr. Sarrazin was involved in the drug trade. The Crown was also allowed to lead evidence of one expert (the evidence of a second expert was excluded) outlining the manner in which the gangs operated and the significance of tattoos and other gang-related symbols.
[18] The “gang” evidence was obviously potentially prejudicial to the appellants’ right to a fair trial. The evidence was also central to the jury’s understanding and assessment of the allegations advanced by the Crown. The trial judge was obliged to address both aspects of the evidence. He was alive to the potential prejudice and excluded some of the evidence tendered by the Crown either because it could not be properly challenged on cross-examination or because it was not necessary to a proper understanding of the Crown’s allegation with respect to the motive for the murder. As indicated above, I see no error in the trial judge’s rulings on the admissibility of the evidence. The trial judge also told the jury, both when the “gang” evidence was adduced and in his closing instructions, that the “gang” evidence was admissible exclusively on the question of motive. He emphatically cautioned the jury against inferring guilt based upon the appellants’ criminal activities and gang association.
[19] The trial judge thoroughly and carefully instructed the jury on the issues relating to the identification of the appellants as the persons who participated in the shooting of Mr. Noel. He properly explained the concepts of aiding and abetting, and the mens rea for murder required by s. 229(a) of the Criminal Code. No complaint is made with respect to any of these instructions. The trial judge’s instructions on the issue of motive, specifically the use that the jury could make of the “gang” evidence, were also beyond reproach. The jury’s verdicts, placed in the context of the positions advanced by the parties, indicate that the jury was satisfied beyond a reasonable doubt that Mr. Sarrazin shot Mr. Noel with the intent necessary for the crime of murder, and that Mr. Jean aided and abetted Mr. Sarrazin with the mens rea necessary to make him a party to that murder.
III. THE FAILURE TO INSTRUCT THE JURY THAT A FINDING OF GUILT ON ATTEMPTED MURDER WAS A POSSIBLE VERDICT
(a) The trial proceedings
[20] On a murder charge, the Crown must prove beyond a reasonable doubt that the acts of the accused were a significant contributing cause of the victim’s death: see R. v. Nette, 2001 SCC 78, [2001] 3 S.C.R. 488 at para. 72. As summarized in the reasons of Moldaver J.A., there was strong evidence that the serious injuries caused by the gunshot wound to the abdomen and the subsequent surgeries necessitated by that wound were at least a significant contributing cause, if not the sole cause, of the fatal blood clot. However, both Dr. Freeman, the surgeon, and Dr. Johnston, the pathologist who performed the autopsy, agreed that Mr. Noel’s death was unexpected. Dr. Johnston also acknowledged on cross-examination that it was "a possibility" but not “likely” that the blood clot was caused by Mr. Noel’s cocaine ingestion shortly before his death, and that there was no causal connection between the gunshot wound and the subsequent surgeries, and the blood clot.
[21] Apart from the concessions made by Dr. Johnston in his cross-examination, there was no evidence to support the contention that Mr. Noel’s cocaine ingestion had anything to do with his death. Defence counsel suggested to Dr. Johnston that there would be defence expert medical evidence to support the contention that cocaine ingestion had caused Mr. Noel’s death; however, no such evidence was offered by the defence.
[22] During pre-charge discussions at the end of the evidence, counsel and the trial judge all agreed that causation was “in play”, meaning that, on the totality of the evidence, there was a realistic basis upon which a jury could have a reasonable doubt as to whether Mr. Noel’s death was caused by the gunshot wound and the subsequent surgeries. Counsel did not, however, agree as to the potential consequences of any such reasonable doubt. Defence counsel argued that the jury should be told that if it had a reasonable doubt on causation, but was satisfied beyond a reasonable doubt that Mr. Sarrazin had shot Mr. Noel with the intention of killing him, it could convict both appellants of attempted murder. Counsel submitted that an instruction on the offence of attempted murder in keeping with the principles in R. v. Ancio, [1984] 1 S.C.R. 225 should be given to the jury.
[23] Crown counsel at trial initially agreed that an instruction on the offence of attempted murder was appropriate. However, after further consideration, Crown counsel took the position that the jury should be told to acquit the appellants if it had a reasonable doubt on the question of causation.[^4]
[24] During the pre-charge discussions, the trial judge expressed some doubt as to whether an accused could, as a matter of law, be convicted of attempted murder on a count alleging murder. He referred to the relevant Criminal Code provisions (ss. 660- 662) and R. v. Poole, [1997] B.C.J. No. 957 (B.C.C.A.). The trial judge told counsel that, despite his uncertainty, he was prepared to instruct the jury on the offence of attempted murder if all counsel agreed that he should do so.
[25] The trial judge ultimately determined that he would not instruct the jury that it could acquit the appellants of murder but convict them of attempted murder if it had a reasonable doubt about causation. Instead, the trial judge accepted the Crown’s position and told the jury that if it had a reasonable doubt on the question of causation, it must acquit both appellants.
(b) Analysis
[26] I would think that most Canadian lawyers and judges familiar with the criminal law would be surprised by the submission that a person facing a murder charge cannot be convicted of attempted murder if a jury was satisfied that the person made serious efforts to kill the victim with the intent of killing the victim, but had a reasonable doubt on the question of causation.[^5] Arbour J., writing for the majority in Nette, at para. 47, indicated without analysis and in obiter dicta that where “causation was not proven, a proper verdict might be attempted murder”.
[27] Mr. Stewart, Crown counsel, with his usual candour, accepts that his submission that attempted murder is not an available verdict is somewhat counterintuitive and is fairly described as a technical argument. The argument rests on a reading of ss. 660 and 662 of the Criminal Code.
[28] Section 660 of the Criminal Code provides:
Where the complete commission of an offence charged is not proved but the evidence establishes an attempt to commit the offence, the accused may be convicted of the attempt.
[29] This section, read on its own, could not be plainer. It puts an accused on notice that he or she stands to be convicted of attempting to commit the offence with which he or she is charged, if the full offence is not proved but an attempt to commit that offence is proved. Nothing in s. 660 suggests that it does not apply to a murder charge. Section 660, standing alone, leaves no room for the Crown’s argument that attempted murder is not an available verdict on a murder charge.
[30] Section 662, however, arguably muddies the interpretative waters. The relevant parts of that section read:
662.(1) A count in an indictment is divisible and where the commission of the offence charged, as described in the enactment creating it or as charged in the count, includes the commission of another offence, whether punishable by indictment or on summary conviction, the accused may be convicted
(a) of an offence so included that is proved, notwithstanding that the whole offence that is charged is not proved; or
(b) of an attempt to commit an offence so included.
(2) For greater certainty and without limiting the generality of subsection (1), where a count charges first degree murder and the evidence does not prove first degree murder but proves second degree murder or an attempt to commit second degree murder, the jury may find the accused not guilty of first degree murder but guilty of second degree murder or an attempt to commit second degree murder, as the case may be.
(3) Subject to subsection (4), where a count charges murder and the evidence proves manslaughter or infanticide but does not prove murder, the jury may find the accused not guilty of murder but guilty of manslaughter or infanticide, but shall not on that count find the accused guilty of any other offence.
(4) Where a count charges the murder of a child or infanticide and the evidence proves the commission of an offence under section 243 but does not prove murder or infanticide, the jury may find the accused not guilty of murder or infanticide, as the case may be, but guilty of an offence under section 243 [concealing the body of a child]. [Emphasis added.]
[31] Section 662(1) describes in generic terms what are commonly referred to as lesser and included offences. That section puts accused persons on notice that they could be convicted of any offences that fall within the language of s. 662(1), or an attempt to commit any of those offences, even though those offences and attempts are not expressly set out in the indictment.
[32] While s. 662(1) speaks in general terms that apply to all offences charged in an indictment, ss. 662(2)-(6) identify certain specific offences and declare that persons charged with those offences are in jeopardy of being convicted of the offences specifically identified in the subsections. For example, s. 662(5) provides that a person charged with an offence involving criminal negligence in the operation of a motor vehicle is liable to be convicted of dangerous driving, if criminal negligence is not made out but dangerous driving is proved. Accused persons are liable to be convicted of the included offence of dangerous driving by virtue of the specific language of s. 662(5) even though dangerous driving may not be a lesser and included offence within the general language of s. 662(1).
[33] Sections 662(2), (3) and (4) describe offences that are included in a murder charge. Section 662(2) declares that second degree murder and an attempt to commit second degree murder are included offences in a charge of first degree murder.[^6] Sections 662(3) and (4) identify lesser offences that are also declared to be included in a murder charge.
[34] Crown counsel places considerable reliance on s. 662(2), which specifically provides that attempted second degree murder is an included offence in a charge of first degree murder. He argues that s. 662(2) indicates that where Parliament intended that attempted murder should be a possible verdict on a murder charge, it has expressly said so.
[35] This submission reads too much into the section. The section is intended to make it clear that second degree murder is a possible verdict on a first degree murder charge. The reference in s. 662(2) to the possibility of a conviction for attempted second degree murder simply tracks the language of s. 662(1), which provides that where an accused is liable to be convicted of a lesser and included offence, he or she may also be convicted of an attempt to commit that lesser and included offence. The words of s. 662(2) that make a conviction for attempted second degree murder possible simply make the provision conform with the general provision in s. 662(1) as it relates to potential liability for an attempt to commit a lesser and included offence.
[36] I also agree with Mr. Campbell, co-counsel for Mr. Jean, that the inclusion of attempted second degree murder as a possible verdict on a first degree murder charge undermines the Crown’s argument. As Mr. Campbell observed, if the Crown’s argument is correct, s. 662(2) creates the anomalous result that an accused charged with first degree murder may be convicted of attempted murder but that an accused charged with second degree murder cannot be so convicted. Crown counsel has not suggested any justification for this distinction.[^7]
[37] Crown counsel also relies on s. 662(3). That section provides that a jury may convict a person charged with murder of manslaughter or infanticide “but shall not on [the murder count] find the accused guilty of any other offence”. Crown counsel submits that these closing words of s. 662(3) preclude a conviction for attempted murder on a murder charge.
[38] Section 662(3) is part of s. 662 and must be read in that context. I think s. 662(3) serves to limit the general language of s. 662(1) by providing that only manslaughter or infanticide are available lesser and included offences. Thus, for example, an accused cannot be convicted of assault causing bodily harm on a murder charge even though the charge of assault causing bodily harm might have been an included offence in the murder charge under the general language of s. 662(1): see R. v. Chichak (1978), 38 C.C.C. (2d) 489 (Alta. Sup. Ct. (App. Div.)).
[39] Nothing in the language of s. 662 addresses an accused’s potential liability for attempting to commit the murder with which that accused is charged. Section 660 speaks directly to that potential liability. Nothing in s. 662 limits or qualifies the broad language of s. 660. Although both sections address the same issue – possible verdicts – they speak to different aspects of that issue. Section 660 directly addresses the accused’s potential liability for an attempt to commit the offence with which the accused is charged. Section 662 speaks to liability for offences that are lesser and included in the offence with which the accused is charged either by virtue of the general language of s. 662(1) or the specific provisions of ss. 662(2)-(6).
[40] Crown counsel’s interpretation of ss. 660 and 662 is not without support. No less of an authority than Justice H. E. Taschereau, in his seminal commentary on Canada’s first Criminal Code, the Criminal Code, 1892, S.C. 55-56 Vic., c. 29, observed that s. 711, the predecessor to s. 660, “does not apply to murder”. In support of that observation, he referred to s. 713, the predecessor to s. 662: see H. E. Taschereau, The Criminal Code of Canada, (Toronto, Ont.: Carswell, 1980).[^8]
[41] Crown counsel also relies on three appellate authorities. I will deal with them in chronological order. The first, Chichak, did not involve potential liability for attempted murder. The question in Chichak was whether an accused who was charged with murder could be convicted of assault causing bodily harm. The court held that on the language of s. 589(3) [now s. 662(3)], an accused charged with murder could not be convicted of a lesser offence other than manslaughter or infanticide. I need not take issue with the ratio in Chichak, which does not assist the Crown on the issue to be decided on this appeal.
[42] R. v. Duncan (1984), 57 A.R. 362 (Alta. C.A.), the second case relied on by the Crown, is factually on point. The accused was acquitted of murder on the basis that the trial judge had a doubt as to whether the victim had died from the stab wounds inflicted by the accused or from a pre-existing heart disease. On a Crown appeal from the acquittal, the court held, relying on the language of s. 589(3) [now s. 662(3)], that a finding of guilt on the charge of attempted murder was not an available verdict on a murder charge. The court made no reference to s. 587, the predecessor to s. 660. Absent any reference to, much less analysis of, s. 660, I do not find the reasoning in Duncan to be persuasive on this point.
[43] Poole, the third case relied on by the Crown, was also a case where cause of death was in issue at trial. In Poole, the accused and a third party, acting independently of each other, fired the same handgun at the victim during an altercation and the victim’s subsequent flight from the scene. The shot fired by the third party struck and killed the victim. The trial judge instructed the jury that the accused, whose shot had missed the victim, could only be convicted of attempted murder. The jury returned a verdict of not guilty of murder but guilty of attempted murder.
[44] In the British Columbia Court of Appeal, the Crown conceded that the trial judge had misdirected the jury as to the mens rea required for attempted murder. The Crown argued, however, that attempted murder was an available verdict and that there should be a new trial on that charge.
[45] Macfarlane J.A., writing for the court, rejected the Crown’s submission for two reasons. First, he concluded that on the evidence, and on the Crown's theory at trial, the transaction referred to in the murder charge was limited to the shot that caused the victim’s death. On this view, the shot fired by the accused constituted a separate transaction. Macfarlane J.A. held that an accused could not be convicted of an offence that arose out of a different transaction than the transaction giving rise to the allegation in the indictment. This principle is not controversial, has been accepted in this court, and has no relevance to the outcome of this appeal: see R. v. Talbot (2007), 217 C.C.C. (3d) 415 (Ont. C.A.) at para. 90. In this case, the appellants’ potential liability for attempted murder clearly arises from the same transaction that gave rise to the murder charges.
[46] The second reason Macfarlane J.A. rejected the Crown’s submission is directly germane to this appeal. After setting out the relevant provisions of the Criminal Code, including ss. 660 and 662, Macfarlane J.A. referred to Ancio for the proposition that attempted murder requires an intention to kill, whereas the mens rea for murder required by s. 229(a)(ii) is less than an actual intent to kill. He said at paras. 43-44:
The offences of second degree murder and attempted murder are fundamentally different. Murder may be committed upon proof of a lesser intent than is required to support a charge of attempted murder. Conceptually, attempted murder would not appear to be a lesser included offence of second degree murder.
[47] Mr. Stewart placed heavy reliance on this aspect of the reasoning in Poole. As I understand his argument, because attempted murder requires proof of an intention to kill, whereas murder may be established either by proof of an intention to kill or proof of an intention to cause bodily harm knowing that it is likely to cause death and being reckless as to whether death ensues, attempted murder is not an included offence in the crime of murder. According to Crown counsel, the difference in the mens rea components of attempted murder and murder makes the two crimes “fundamentally different”, such that attempted murder cannot be characterized as a lesser and included offence in the charge of murder. It follows, according to this argument, that because attempted murder is not a lesser and included offence, it cannot be an available verdict under s. 660.
[48] The mens rea required for attempted murder is more limited than the mens rea that can support a conviction for murder. In that sense, the two crimes are different. This distinction is not, however, peculiar to the crime of murder. All attempts are described in s. 24 and require proof of “an intention to commit an offence”. The mens rea required for the corresponding completed offences will be found in the statutory provisions that create the offences and judicial interpretations of the relevant language. While the matter does not appear to have been decided by the Supreme Court of Canada, there is an argument that attempts require proof of an actual intention to commit the offence, even if the completed offence requires only proof of recklessness or perhaps some lower level of mens rea: see D. Stuart, Canadian Criminal Law, 5th ed. (Scarborough, Ont.: Carswell, 2007) at pp. 662-666; R. v. Janeteas (2003), 172 C.C.C. (3d) 97 (Ont. C.A.) at para. 30; E. Meehan and J. Currie, The Law of Criminal Attempt, at pp. 56 – 92.
[49] In any event, I do not see that the difference in the mens rea requirements between attempted murder and murder has any relevance to the applicability of s. 660. The Crown’s argument assumes that s. 660 applies only to attempts that also qualify as lesser and included offences within the language used in s. 662(1). Many cases have considered the meaning of the complicated language in s. 662(1) when determining whether a particular offence is a lesser and included offence of the charged offence: see, for example, R. v. G.R., 2005 SCC 45, [2005] 2 S.C.R. 371. That complicated language is not found in s. 660. Section 660 provides a free-standing statutory basis upon which attempts to commit the offence charged are available verdicts if the offence charged is not proved. Attempts are “included” offences in the sense that they are available verdicts under s. 660. They are not, however, subject to s. 662(1) and need not meet the criteria for a lesser and included offence under that section. In my view, it is irrelevant to the operation of s. 660 whether an attempt to commit an offence qualifies as a lesser and included offence under the language of s. 662(1).[^9]
[50] For the reasons set out above, I would not follow Poole. I do not think that the analysis of the mens rea requirement for attempted murder in Ancio has any bearing on whether s. 660 provides that a person charged with murder may be convicted of attempted murder if murder is not proved.
[51] Mr. Stewart also advances arguments not made in Poole or the other authorities on which he relies. He submits that although the language of s. 660 may appear to be unqualified, there are certain offences to which s. 660 clearly cannot apply. For example, he points out that a person charged with conspiracy cannot be convicted under s. 660 of attempted conspiracy. This argument is advanced to counter the suggestion that s. 660 is of general application to all charges.
[52] There is no doubt that a person charged with conspiracy cannot be convicted of attempted conspiracy: see R. v. Déry, 2006 SCC 53, [2006] 2 S.C.R. 669. The unavailability of a conviction for attempt, however, has nothing to do with s. 660 or any other pleading provision in the Criminal Code. Déry holds that an attempt to conspire is not a crime. As it is not a crime, a conviction for attempted conspiracy is not a possible verdict. To acknowledge that s. 660 does not contemplate a conviction for attempted conspiracy on an indictment charging conspiracy is simply to acknowledge that s. 660 does not create crimes.
[53] Mr. Stewart also seeks to support the submission that the general language of s. 660 does not encompass attempted murder by asserting that attempted murder is treated differently than other attempts in the Criminal Code. He refers to s. 463, which sets out the general scheme for the punishment of criminal attempts, and compares it to s. 239, which creates a separate and significantly harsher sentencing scheme for attempted murder.
[54] I do not agree that there is a meaningful difference in how the Criminal Code treats attempted murder and other attempts. Like all attempts, the requisite elements of the offence of attempted murder are found in s. 24 of the Criminal Code. Attempted murder, like other attempts, requires proof of “an intent to commit an offence” (s. 24(1)) and an act that goes beyond “mere preparation” (s. 24(2)): see R. v. Deutsch, [1986] 2 S.C.R. 2.
[55] Section 239 is a penalty provision only: see R. v. Gordon (2009), 94 O.R. (3d) 1 (C.A.) at para. 57. The penalty provided for in s. 239 reflects Parliament’s determination that attempted murder is sufficiently serious to warrant a greater penalty than is normally attached to an attempt to commit a crime. Parliament has made the same determination with respect to the crime of being an accessory after the fact to murder: s. 240 of the Criminal Code.
[56] Despite Mr. Stewart’s able submissions, I am satisfied that, as a matter of statutory interpretation, s. 660 applies to a charge of murder. Nothing in s. 662 detracts from the language of s. 660. Read together, ss. 660 and 662 (along with s. 661) put an accused on notice of the full extent of his or her potential liability on any particular charge. They also allow the trier of fact to determine that liability in a single proceeding.
[57] Although I base my conclusion that s. 660 applies to murder charges on the language of the relevant sections of the Criminal Code, my interpretation of s. 660 is further supported by practical and policy considerations. There is nothing problematic from a practical standpoint about allowing the jury to consider an accused’s possible liability on a charge of attempted murder in the same proceeding in which it considers a murder charge. The different intents required for the offence of murder and attempted murder would have to be explained to the jury. However, this added complexity pales beside the level of complexity found in some instructions that are routinely given to juries. I am confident that a jury could properly address the different intents required for the crimes of murder and attempted murder.
[58] The Crown does not suggest that there are any practical impediments to the determination of allegations of murder and attempted murder at the same trial. The Crown submits, however, that the two allegations can be considered in the same trial only where the Crown chooses to include a separate charge of attempted murder, as it is entitled to do under s. 589 as long as the two allegations arise out of the same transaction. Indeed, if the Crown had not withdrawn Count 2 in the indictment at the conclusion of the evidence, this jury would have had to consider Mr. Sarrazin’s liability for an offence that required that attempted murder be proved.
[59] The position advanced by the Crown also carries with it significant negative policy ramifications. On the Crown’s position, these appellants could not have been convicted of attempted murder at this trial. Had they been acquitted on the murder charge, the Crown could have proceeded at a subsequent trial on a charge of attempted murder. Two separate trials to decide the ultimate culpability, if any, of the appellants in the shooting of Mr. Noel would unnecessarily consume limited judicial resources, prolong the criminal process and pose the very real risk of inconsistent findings of fact in the two proceedings. The Crown might also take diametrically opposite positions on causation at the two trials, thereby potentially diminishing the repute of the administration of justice. The community as a whole and the participants in a criminal proceeding, be they accused, witness, juror, or investigator, are best served by a process that allows all issues to be resolved in a single trial.
[60] The Crown’s contention that the jury could not consider the possibility of a conviction for attempted murder also carries a real risk of potential prejudice to an accused in the position of the appellants. As indicated above, had the jury acquitted the appellants, the Crown could have proceeded at a subsequent trial on a charge of attempted murder. At that trial, the Crown would once again try to prove that the appellants were criminally culpable in the shooting of Mr. Noel and that they had intended to kill him. The Crown’s position on these issues would be the same as that advanced at the first trial. On the second trial, however, the Crown would presumably accept that it could not prove beyond a reasonable doubt that the appellants’ actions had caused Mr. Noel’s death.
[61] At the later trial, the appellants could not advance an issue estoppel claim based on their acquittals at the first trial as it would be impossible to isolate the basis upon which the jury had acquitted: see R. v. Mahalingan, 2008 SCC 63, [2008] 3 S.C.R. 316, per McLachlin C.J., at para. 54. Consequently, even if the first jury had acquitted because it had a doubt as to whether the appellants had participated in the shooting, the appellants could be retried on the attempted murder charge and the Crown would have a second opportunity to prove that the appellants had participated in the shooting.
[62] In summary, s. 660 applies to murder charges. Section 660 puts an accused charged with murder on notice that he or she is in jeopardy of being convicted of attempted murder if the Crown fails to prove the completed offence. I would think that in most murder cases there will be no air of reality to the possibility of a conviction for attempted murder as causation will not be in dispute. In those cases, it will be unnecessary to instruct the jury on the accused’s potential liability for attempted murder.
[63] This is one of those relatively unusual homicide cases in which there was a legitimate causation issue. All parties at trial agreed that causation was “in play”. There was a realistic possibility that the jury could have a doubt as to whether the appellants had caused Mr. Noel’s death. I cannot agree with the submission made by Crown counsel on appeal, but not at trial, that although causation was an issue, “[t]he defence of intervening cause of death...had no air of reality”. The defence was not required to prove an intervening cause of death. The defence was only required to point to evidence that could raise a doubt as to whether the appellants’ actions had caused Mr. Noel’s death. The evidence that Mr. Noel’s cocaine ingestion shortly before his death was a possible cause of death provided an evidentiary basis for that reasonable doubt. If the jury had a doubt based on that evidence, I do not see how a verdict of not guilty of murder but guilty of attempted murder was not a realistic possibility.
IV. THE APPLICATION OF THE CURATIVE PROVISO
[64] It is an error in law to fail to instruct the jury on the law applicable to an included offence where, on the totality of the evidence, a verdict of not guilty on the full charge but guilty of the included offence is a realistic possibility: R. v. Hughes, [1942] S.C.R. 517 at pp. 523-24; R. v. George, [1960] S.C.R. 871; R. v. Longson (1976), 31 C.C.C. (2d) 421 at p. 425 (B.C.C.A.); R. v. De Champlain (1982), 68 C.C.C. (2d) 281 at p. 283 (Que. C.A.); R. v. Jackson (1991), 68 C.C.C. (3d) 385 at p. 431 (Ont. C.A.), aff’d on this point, R. v. Davy (1993), 86 C.C.C. (3d) 385 (S.C.C.) at p. 398. In my view, the principle established in these cases has equal application where a trial judge fails to instruct a jury on potential liability for an attempt where, on the evidence, there is an air of reality to the claim that an accused is not guilty of the full offence, but is guilty of an attempt to commit the offence.
[65] If the appellant establishes an error in law, such as a failure to give a jury instruction that should have been given, the appellate court will set aside the appellant’s conviction unless the Crown can establish that the legal error occasioned no substantial wrong or miscarriage of justice: s. 683(1)(b)(iii) of the Criminal Code; R. v. Illes, 2008 SCC 57, [2008] 3 S.C.R. 134 at para. 21. The Crown bears the onus of satisfying the appellate court that there is no reasonable possibility that the verdict would have been different had the legal error not been committed. In other words, the Crown must show that the verdict would necessarily have been the same without the error. If the Crown meets this burden, the error is said to be harmless and the conviction is upheld: see R. v. Colpitts, [1965] S.C.R. 739, per Spence J., at p. 755; R. v. Jolivet, 2000 SCC 29, [2000] 1 S.C.R. 751 at para. 46; R. v. Van, 2009 SCC 22, [2009] 1 S.C.R. 716 at paras. 34-36.
[66] In deciding whether the Crown has met its burden, the appellate court does not engage in a weighing of the evidence or an independent assessment of the ultimate merits of the case. The appellate court must limit its inquiry to a determination of whether a reasonable jury, properly instructed, would necessarily have come to the same conclusion had the error not been made. Professor Mewett explained the appellate function in this way:
But the jury is not free to come to arbitrary or unreasonable conclusions, and all that a court can decide is that, within the limits of what is reasonable, no jury would have acquitted, even absent the mistake. Ruling on what is or is not reasonably supported by the evidence is not a factual determination but a legal determination. It no more usurps the function of the jury than does deciding to withdraw an issue from the jury in the absence of any evidence to support it.[^10]
[67] While the Crown clearly carries a heavy burden to demonstrate the applicability of the curative proviso, the existence of the proviso is testimony to Parliament’s determination that not all legal errors should vitiate convictions and necessitate new trials. As Binnie J. explained in Jolivet at para. 46:
Ordering a new trial raises significant issues for the administration of justice and the proper allocation of resources. Where the evidence against an accused is powerful and there is no realistic possibility that a new trial would produce a different verdict, it is manifestly in the public interest to avoid the cost and delay of further proceedings. Parliament has so provided.
[68] Recent Supreme Court of Canada jurisprudence holds that the curative proviso is properly applied either where the evidence against an accused is overwhelming or where it can be safely said that the legal error had no impact on the verdict: R. v. Khan, 2001 SCC 86, [2001] 3 S.C.R. 823 at paras. 28-31; Van at paras. 34-35. In the first category of cases, the court determines, after a review of the evidence led at trial, that the Crown’s case establishing the guilt of the accused is so overwhelming that no verdict other than a conviction was reasonably possible. No matter the seriousness of the legal error, the accused has effectively lost nothing as a conviction was the only reasonable verdict and is therefore properly viewed as inevitable: R. v. P.L.S., [1991] 1 S.C.R. 909 at p. 916; Khan at para. 31; R. v. Mahoney, [1982] 1 S.C.R. 834 at pp. 854-55, per McIntyre J.; A.W. Mewett, “No Substantial Miscarriage of Justice” at pp. 100-101. Like my colleague, Moldaver J.A., I do not think the evidence in this case is so overwhelming that it could be said that a finding of guilt on the murder charge was the only reasonably available verdict had the jury been properly instructed.
[69] The second category of cases where the proviso has been applied examines the effect of the legal error on the verdict by placing that error in the context of the trial. As described by Arbour J. in Khan at para. 30:
In all those cases, the appellate courts were convinced that the error could have had no effect on the verdict. Because of the nature of the errors and of the issues with respect to which they were made, it was possible to trace their effect on the verdict and ensure that they made no difference. [Emphasis added.]
[70] In tracing the effect of an error on the verdict, appellate courts have relied on different features of the trial process, including the significance of the issue in respect of which the error was made, the positions taken by counsel at trial, and any insights afforded into the jury deliberations by the trial record: see, for example, Illes at paras. 22-23; Van at paras. 38-44; R. v. Ellard, 2009 SCC 27, [2009] 2 S.C.R. 19 at paras. 44-48.
[71] The tracing exercise described in Khan, which seeks to determine the effect of the legal error, is necessarily somewhat speculative. Any doubts as to the impact of the error must be resolved against the Crown, which carries the burden of demonstrating that the verdict would necessarily have been the same: R. v. Brooks, 2000 SCC 11, [2000] 1 S.C.R. 237, per Binnie J., concurring, at para. 136; Khan at para. 30.
[72] In tracing the effect of the error on the verdict, an appellate court can consider factual findings implicit in a jury’s verdict as long as the appellate court is satisfied that those findings were not tainted by the legal error: R. v. Haughton, [1994] 3 S.C.R. 516. For example, in deciding whether the curative proviso could be applied to an error in an instruction on self-defence, an appellate court could consider that the jury, properly instructed on the elements of planning and deliberation, had returned a verdict of guilty of first degree murder. Implicit in that verdict is a finding that the accused had planned and deliberated the murder before it occurred. If the appellate court was satisfied that the finding of planning and deliberation could not have been affected by the misdirection on self-defence, that finding of planning and deliberation could negate the possibility of a successful self-defence claim and justify the application of the curative proviso to the error in the self-defence instruction: see R. v. Pilon (2009), 243 C.C.C. (3d) 109 (Ont. C.A.) at para. 76.
[73] In this case, the jury returned convictions on the murder charge, having been properly instructed on the causation issue and having been told that it could convict of murder only if it was satisfied beyond a reasonable doubt that the appellants had caused Mr. Noel’s death. It is implicit in the verdicts that the jury made a finding of causation against the appellants. Can the Crown rely on that finding to argue that the non-direction on possible liability for attempted murder occasioned no substantial wrong or miscarriage of justice? Or, to put the issue somewhat differently, can it be said that the jury, having returned murder verdicts after being properly instructed on that charge, would never have reached the question of attempt, thereby making the failure to instruct the jury on attempted murder a harmless error?
[74] As a matter of pure logic, the jury’s verdicts negate the possibility that the non-direction as to the appellants’ potential liability for attempted murder had any impact on the outcome of the trial. The jury convicted the appellants of murder. Had the jury been instructed on the offence of attempted murder, it would also have been told that it should consider liability for attempted murder only if it had a reasonable doubt on the question of causation. Given the verdicts, the jury would never have reached the question of liability for attempt. To put the same point somewhat differently, this jury was told it could only convict of murder if it was satisfied beyond a reasonable doubt that the appellants had caused Mr. Noel’s death. The murder convictions indicate that the jury was satisfied beyond a reasonable doubt on the causation issue. That finding is inconsistent with liability for attempted murder.
[75] There is authority for the proposition that an appeal court can reason backward from the verdict and rely on a conviction on the full charge to excuse any misdirection or non-direction in respect of potential liability for included offences: see R. v. Holcomb (1973), 12 C.C.C. (2d) 417 at pp. 419-20 (N.B.C.A.), aff’d without reference to this point (1973), 15 C.C.C. (2d) 239 (S.C.C.); Gilbert v. The Queen (2000), 201 C.L.R. 414 (H.C.A.), per McHugh J., dissenting, at paras. 25 and 31, and per Hayne J., dissenting, at paras. 51-52.
[76] In Jackson, this court rejected the submission that a conviction on the charge of murder, after a proper instruction on the murder charge, rendered harmless the misdirections on potential liability for the included offence of manslaughter. The two accused, Mr. Jackson and Mr. Davy, had been charged with murder. The trial judge told the jury that manslaughter was an available verdict, but misdirected the jury as to the legal principles applicable to a finding of liability for manslaughter. In refusing to apply the curative proviso, this court said at p. 431:
Furthermore, in assessing the potential impact of the failure to direct on included offences, an appellate court should not assume that a jury’s fact-finding process involves a mechanistic analysis of the evidence without regard to the verdict which flows from the jury’s findings of fact. In a case of clear culpability, a jury might be well reluctant to make findings which result in an acquittal, but may regard those same findings as more palatable if they result in some form of liability. [Emphasis added.]
[77] Only Mr. Davy appealed. In Davy, the Supreme Court of Canada also refused to apply the curative proviso to the misdirections on the law of manslaughter. McLachlin J., writing for the court, noted that the Crown argued that, as the jury was properly instructed on the elements of murder, its verdict on that charge overcame any deficiencies in the instruction on the included offence of manslaughter. She said at pp. 397-98:
It is also true that the jury found Davy guilty of murder, nevertheless, I agree with the Court of Appeal that one cannot be satisfied the verdict is just, given the failure of the trial judge to set out the basis for convicting Davy of manslaughter under s. 21(1) and (2) and the absence of any instruction that a party may be guilty of manslaughter even though the perpetrator is guilty of murder.
[78] McLachlin J., at p. 398, referred with approval to the observation of Lord Tucker in Bullard v. The Queen, [1957] A.C. 635 at 644:
Every man on trial for murder has the right to have the issue of manslaughter left to the jury if there is any evidence upon which such a verdict can be given. To deprive him of this right must of necessity constitute a grave miscarriage of justice and it is idle to speculate what verdict the jury would have reached. [Emphasis added.]
[79] McLachlin J. then put her conclusion in the terms of the curative proviso at p. 398:
I am not satisfied that it is clear that a jury, properly instructed, would necessarily have returned a verdict of second degree murder against Davy. He was entitled to have the verdict of manslaughter clearly put to the jury. We cannot be certain that if this had been done, and notwithstanding the correct instruction on murder, that the verdict might not have been different. This is consequently, not a proper case for the application of s. 686(1)(b)(iii). [Emphasis added.]
[80] Davy was applied by the majority of the High Court of Australia in Gilbert. In that case, the jury was told that it could either convict of murder or acquit. On appeal, the court held that a conviction for manslaughter was an available verdict on the evidence. The majority further held that, despite the full and accurate instruction on the charge of murder, the jury’s verdict on that charge did not render harmless the trial judge's failure to instruct on the included offence of manslaughter. Chief Justice Gleeson and Gummow J. said at para. 16:
These statements are inconsistent with the notion that an appellate court must assume, on the part of a jury, a mechanistic approach to the task of fact-finding, divorced from a consideration of the consequences. Indeed, juries are ordinarily asked to return a general verdict. They make their findings of fact in the context of instructions as to the consequences of such findings, and for the purpose of returning a verdict which expresses those consequences. [Emphasis added.]
[81] In separate concurring reasons, Callinan J. also rejected the submission that the jury’s verdict on the murder charge could be relied on to render a failure to instruct on manslaughter, harmless. He said at paras. 96 and 101:
This is to recognise the reality that a jury room might not be a place of undeviating intellectual and logical rigour. It is not to say that a jury should not perform their sworn duty to determine a case before them according to law.
The appellant was entitled to a trial at which directions according to law were given. It is contrary to human experience that in situations in which a choice of decisions may be made, what is chosen will be unaffected by the variety of choices offered, particularly when, as here, a particular choice was not the only or inevitable choice. [Emphasis added.]
[82] The approach taken in Davy and in Gilbert was followed by the House of Lords in R. v. Coutts, [2007] 1 Cr. App. R. 61 (H.L.). In Coutts, the accused was charged with a particularly sordid murder and advanced the defence of accident. The trial judge instructed the jury that it could convict of murder or acquit. Manslaughter was not left as a possible verdict. Defence counsel apparently agreed manslaughter should not be left. The Court of Appeal held that in the circumstances, it would have been unfair to the defence to instruct the jury on the included offence of manslaughter, and dismissed the accused’s appeal: R. v. Coutts, [2005] 1 Cr. App. R. 517 (C.A.).
[83] The House of Lords found that manslaughter was an available verdict on the evidence. It further held that the failure to instruct the jury on the included offence of manslaughter constituted reversible error. In coming to that conclusion, the court specifically overruled its earlier judgment in R. v. Maxwell (1991), 91 Cr. App. R. 61 (H.L.) at p. 68, which had held that an appellate court could rely on a conviction on a main charge to render non-direction on an included offence harmless, unless the appellate court was satisfied that the conviction on the main charge reflected the jury’s reluctance to see the accused “get clean away”.
[84] Following the path taken in Davy and in Gilbert, the Law Lords in Coutts declined to speculate as to why the jury had convicted on the full charge and what the impact of an instruction on the included offence of manslaughter might have been. Instead, the court focussed on the jury’s right to consider all verdicts that were reasonably available on the evidence. Lord Bingham indicated at para. 12:
The interests of justice are not served if a defendant who has committed a lesser offence is either convicted of a greater offence, exposing him to greater punishment than his crime deserves, or acquitted altogether, enabling him to escape the measure of punishment which his crime deserves. The objective must be that defendants are neither over-convicted nor under-convicted, nor acquitted when they have committed a lesser offence of the type charged. The human instrument relied on to achieve this objective in cases of serious crime is of course the jury. But to achieve it in some cases the jury must be alerted to the options open to it. [Emphasis added.]
[85] Lord Bingham went on to hold, at para. 23, that an instruction on an included offence should be given where there is clearly sufficient evidence to support that verdict. In rejecting the contention that the jury’s verdict on the full offence was a complete answer to any complaint with respect to non-direction on the included offence, Lord Bingham observed at para. 25:
The Court of Appeal rightly recognised the high sense of public duty which juries customarily bring to their task. I would not wish to belittle that in any way. But one does not belittle it to decline (as the High Court of Australia has done) to attribute to juries an adherence to principle and an obliviousness to consequences which is scarcely attainable. [Emphasis added.]
[86] Lord Rodger, in concurring reasons, echoed the language of the majority in Gilbert and the sentiment underlying Davy when he explained, at para. 89:
[Counsel for the Crown] says that the additional choice is irrelevant since the jury convicted the appellant of murder and so they would never have reached the question of manslaughter, which only arose if they were not prepared to convict of murder. But that is to make an unreal assumption that, at all stages of their deliberations, the jury would keep the various issues in separate boxes, to be considered in a prescribed order. The reality is that, in the course of their deliberations, a jury might well look at the overall picture, even if they eventually had to separate out the issues of murder, manslaughter and accident. So, introducing the possibility of convicting of manslaughter could have changed the way the jury went about considering their verdict. [Emphasis added.]
[87] Davy, and the cases from Australia and England that have followed Davy, establish that the Crown, when tracing the effect of the failure to properly instruct the jury on liability for an included offence, cannot automatically rely on factual findings implicit in the jury’s verdict on the main charge. Instead, the Crown must demonstrate, pursuant to its burden under the curative proviso, that there is no reasonable possibility that the verdict on the main charge was influenced by the misdirection or non-direction with respect to potential liability for an included offence. On this analysis, the failure to afford a jury an opportunity to consider returning a verdict on an included offence, where that verdict is reasonably available, will in most circumstances constitute reversible error unless the Crown can show that it had an overwhelming case on the main charge.[^11] While Davy and the other authorities all involve errors in respect of the included offence of manslaughter, I think the same analysis applies in this case to the appellants’ potential liability on the charge of attempted murder.
[88] I come now to Haughton, the authority primarily relied on by my colleague, Moldaver J.A., in support of his conclusion that the curative proviso can be applied in this case. Haughton has much in common with Davy. In both cases, the trial judge properly instructed the jury on the main charge of murder, but misdirected the jury on the law applicable to the included offence of manslaughter. In both cases, the possibility of a conviction on the lesser and included offence of manslaughter was left with the jury and, in both cases, the jury convicted of murder. In Davy, as noted above, the court rejected the Crown’s contention that the jury’s conviction on the murder charge rendered any misdirection on manslaughter harmless. In Haughton, the court looked to the conviction on the main charge to support the application of the curative proviso to a misdirection on manslaughter.
[89] I do not regard Haughton as a jurisprudentially significant case. It is a one-paragraph oral judgment in which the legal error to which the proviso is applied is not even identified.[^12] I read Haughton as an application of the curative proviso that is consistent with the jurisprudence both before and after Haughton.
[90] As I understand the reasons in Haughton, Sopinka J., for the court, was satisfied for two reasons that the factual findings implicit in the murder conviction were not undermined by the misdirection on manslaughter. First, because the jury was left with the possible verdict of guilty of manslaughter, it could not be said that the conviction on the murder charge was potentially the product of the jury’s reluctance to see the accused escape liability entirely. Second, the finding implicit in the murder conviction that the appellant had the requisite subjective foresight of death required for murder could not have been affected by an erroneous instruction with respect to the lesser mens rea required for manslaughter. Haughton does not turn exclusively on the fact that manslaughter was left as a possible verdict, but also looks at the potential impact of the specific legal error made by the trial judge on the mens rea finding implicit in the jury’s verdict of guilty of murder.
[91] I do not think that Haughton pronounces that, in every case where an included offence is left with a jury and the jury convicts on the main charge, the appellate court may take the findings of fact implicit in that verdict into account in deciding whether a legal error in the instruction on potential liability for an included offence resulted in no substantial wrong or miscarriage of justice. To read Haughton that broadly would be to render legal errors in respect of instructions on included offences irrelevant if the jury convicted on the main offence. It would also mean that Davy was implicitly overruled by Haughton. Davy was decided only 10 months before Haughton. All seven members of the unanimous court in Haughton were part of the unanimous nine-person court in Davy. There is no mention of Davy in Haughton. I conclude that Haughton was not intended to in any way detract from the authority of Davy. I see no inconsistency in the two judgments. Both address the curative proviso by asking whether the Crown can demonstrate that the legal error had no effect on the verdict. In the circumstances of Davy, the court was not prepared to accept the findings of fact implicit in the jury’s verdict in deciding whether to apply the curative proviso. In the different circumstances of Haughton, the court was prepared to accept those findings for the purposes of the curative proviso analysis.
[92] In holding that the failure to instruct the jury on potential liability for the offence of attempted murder occasioned no substantial wrong, Moldaver J.A. stresses that the jury was instructed that the appellants could be convicted of the included offence of manslaughter. The jury was told that manslaughter was a possible verdict if the appellants caused Mr. Noel’s death, but the Crown failed to prove the mens rea required for murder. Moldaver J.A. observes that the instruction on the included offence of manslaughter meant that this jury was not faced with the stark choice of convicting the appellants of murder or acquitting them outright. Consequently, the finding of fact in respect of causation implicit in the jury’s verdicts cannot have been the product of the jury’s determination to avoid setting the appellants free entirely, but must have been a true reflection of their assessment of the relevant evidence.
[93] I cannot agree that, since the jury was instructed to convict the appellants of manslaughter if it had a doubt regarding intent, the Crown has demonstrated that this court can safely rely on the causation finding implicit in the jury’s verdicts of guilty of murder. The manslaughter instruction was only relevant to the jury’s deliberations if the jury was satisfied beyond a reasonable doubt that the appellants had caused Mr. Noel’s death. The manslaughter instruction was directed at the appellants’ mental state and left it open to the jury to convict the appellants of manslaughter if the Crown had proved causation beyond a reasonable doubt but had not proved the mens rea required for murder. I do not see how an instruction as to manslaughter that becomes relevant only after a finding of causation and which focuses on the appellants’ mental state can cure a misdirection on the consequences of the jury’s finding with respect to causation. Nothing in the instruction linked the possibility of manslaughter to the jury’s causation inquiry.
[94] My concern, and I think the primary concern of the courts in Gilbert and Coutts, is not that the jury would deliberately disregard the trial judge’s instructions and search out some middle ground verdict to ensure that an unsympathetic accused was punished for something even though he may not have been guilty of anything in the eyes of the law. My concern is that a jury’s fact-finding on a specific issue such as causation could be subconsciously influenced by what it is told about the legal consequences of that fact-finding.
[95] Manslaughter instructions directed at a possible doubt regarding the mental state required for murder do not allay my concern that the jury’s finding of fact on the distinct question of causation may have been influenced by an instruction that a reasonable doubt regarding causation should lead to the appellants’ outright acquittal. This is not a case like Haughton where the possibility of a manslaughter conviction arose out of the same factual issue - the accused's mental state - as did the legal error. In that situation, I find it much easier to infer that the jury's finding against an accused was not the product of the jury's unwillingness to let the accused escape entirely.
[96] In coming to my conclusion, I do not suggest that the jury was in fact influenced by what it was told about the consequences of its fact-finding on the causation issue. I mean only to say that there is sufficient uncertainty as to whether the jury’s fact-finding on the question of causation was tainted by the non-direction on attempted murder to cause me to conclude that the Crown cannot rely on the implicit finding of causation in the jury’s verdict in seeking to demonstrate that the non-direction caused no substantial wrong or miscarriage of justice.
[97] For the reasons set out above, I do not regard this case as one in which the Crown can rely on findings of fact implicit in the verdict to demonstrate that a legal error occasioned no substantial wrong or miscarriage of justice. The verdict does not assist the Crown in tracing the effect of the error. The curative proviso analysis should be undertaken without regard to any findings of fact implicit in the verdicts returned by the jury.
[98] I agree with Moldaver J.A. that causation was a secondary issue at trial. I also agree that the evidence on the causation question strongly favoured the Crown’s position. That said, I again agree with Moldaver J.A. that this is not a case where the curative proviso can be applied on the basis that the Crown’s case against the appellants was overwhelming. I note that this is the sole argument advanced by Mr. Stewart in support of his contention that the proviso should be applied.
[99] My analysis of the possible application of the curative proviso comes down to this. On all of the medical evidence, Mr. Noel’s death was entirely unexpected. From a medical standpoint, the jury could conclude that his death was a puzzle and its cause far from obvious. It was open to the jury to prefer the opinion of the pathologist, Dr. Johnston, who had the benefit of his observations during the post-mortem and the subsequent pathology reports, over the opinion of the surgeon, Dr. Freeman, who was not familiar with the evidence gathered post-mortem. On Dr. Johnston’s evidence, Mr. Noel’s cocaine ingestion was a possible, although not a probable, cause of death. The trial judge correctly told the jury that it could find the appellants liable for Mr. Noel’s death only if the Crown proved beyond a reasonable doubt that the appellants had caused Mr. Noel’s death. Dr. Johnston’s opinion arguably stood directly in the way of the Crown’s meeting that burden.
[100] Given Dr. Johnston’s evidence, I cannot be satisfied that the failure to properly instruct the jury on the legal consequences of its giving effect to Dr. Johnston’s opinion caused no substantial wrong or miscarriage of justice. During the pre-charge discussions as to whether the offence of attempted murder should be left with the jury, the trial judge observed that an instruction on potential liability for attempted murder would inure to the benefit of the appellants and that the appellants would be prejudiced should he decide not to give that direction. Crown counsel at trial agreed with the trial judge’s observation. I think they were right in their assessment of the impact of the failure to instruct on potential liability for attempted murder. There was a realistic possibility that prejudice would flow to the appellants from that non-direction.
[101] I also cannot agree that the Crown’s position on the application of the proviso is improved because the causation issue was thoroughly explored by counsel at trial and was, except for the error discussed above, the subject of a careful and fair instruction by the trial judge. The treatment of the issue of causation by counsel and the trial judge demonstrates that, while it was a secondary issue, it was still very much a live issue at trial. That same treatment makes the failure to properly instruct the jury as to the legal consequences of its findings on causation more significant.
[102] I would not apply the curative proviso.
V. CONCLUSION
[103] I think the appeal must be allowed and the verdicts set aside. Given that the jury was fully and properly instructed on all issues other than causation, one might think that the re-trial could be limited to the question of causation. Unfortunately, the relevant provisions of the Criminal Code do not permit this court to place any restrictions on the new trial: see R. v. Thomas, [1998] 3 S.C.R. 535 at para. 28. Consequently, there must be a third trial at which all of the relevant legal and factual issues may be re-litigated. I would order a new trial for both appellants on the charge of second degree murder.
“Doherty J.A.”
“I agree Epstein J.A.”
Moldaver J.A. (Dissenting):
THE CURATIVE PROVISO
[104] I have read the detailed and scholarly reasons of my colleague Doherty J.A. and I agree with his analysis and conclusions on the substantive issues raised by the appellants. Respectfully, however, I am unable to agree with my colleague on the application of the curative proviso.
[105] In view of our mutual conclusion that the trial judge erred in law in failing to instruct the jury on the offence of attempted murder, the application of the curative proviso in s. 686(1)(b)(iii) of the Criminal Code “requires the court to consider whether a jury properly instructed could, acting reasonably, have come to a different conclusion absent the error.” See R. v. Haughton, [1994] 3 S.C.R. 516 per Sopinka J.
[106] I mention Haughton because, as will be seen, Haughton speaks to the application of the proviso in cases where an included offence (or in this case a lesser offence) is not left with the jury and the jury convicts of a more serious offence.
[107] More recent cases have expanded on the general principle enunciated in Haughton. Two types of situations have now been identified in which the proviso will generally be applied: (1) where the evidence pointing to the guilt of the accused is so overwhelming that a trier of fact would inevitably convict; and (2) where the error of law is trivial or, if more serious, can be said to have had no effect on the verdict and caused no prejudice to the accused. See for example, R. v. Khan, [2001] 3 S.C. 823 at paras. 29‑31 and R. v. Trochym, 2007 SCC 6, [2007] 1 S.C.R. 239 at para. 81.[^13]
[108] In his able argument for the Crown, Mr. Stewart submitted that the proviso could safely be applied here because the evidence of causation – that the gunshot injuries sustained by Apaid Noel caused or substantially contributed to the formation of the blood clot that resulted in his death – was so overwhelming that no properly instructed jury, acting reasonably, could have come to any different conclusion.
[109] For reasons that follow, I am unable to accept that submission. The evidence called at trial – and particularly that of Dr. Brian Johnston, the pathologist who performed the autopsy on Apaid Noel – does not permit that conclusion. As will be discussed further, Dr. Johnston’s evidence was that he could not discount the possibility that the blood clot that caused Apaid’s death was unrelated to the gunshot injuries he suffered, but was instead the result of an unknown quantity of cocaine he ingested a short time before his death.
[110] In the face of that evidence, much as I think that Dr. Johnston was being ultra-cautious in his reluctance to completely rule out the ingestion of cocaine as being the sole cause of the clot, I cannot say that the Crown has met its onerous burden of showing that the evidence of causation is so overwhelming that a new trial would be an exercise in futility. However, I do take Dr. Johnston’s evidence of the remoteness of the “cocaine-alone” possibility into account in concluding that the harmless error branch of the proviso can be applied to sustain the convictions. To that end, I propose to review the evidence pertinent to the causation issue.
I. The Causation Issue
(i) Background
[111] Causation played no part in the appellants’ first trial or in their first appeal to this court.[^14] In R. v. Sarrazin (2005), 75 O.R. (3d) 485 (C.A.), the only mention of causation appears at para. 2, where Blair J.A. simply notes that after being shot in the early morning hours of February 19, 1998, Apaid Noel “died 26 days later of a pulmonary embolism suffered after treatment and release from hospital but nonetheless caused by the shooting” (emphasis added). It seems that in the timeframe between the first appeal in 2004 and the second trial in 2006, the medical profession gained greater insight into the risks of blood clotting associated with the use of cocaine. Dr. Johnston acknowledged this at the re-trial in cross-examination by Ms. Mulligan, counsel for Mr. Sarrazin:
Q. And just to be fair to you, the last time we discussed this back in 2000, the reason I started with the suggestion to you that medical knowledge continues to advance, the last time we were here, you were unaware of the fact that cocaine could cause the kind of hypercoagulability [propensity for blood clotting], right?
A. That’s correct.
Q. But certainly you know a lot more about it now.
A. Correct.
This new-found knowledge opened the door for the defence to explore the issue of causation with Dr. Johnston at the second trial.
(ii) Non-Contentious Issues Related to Causation
[112] The causation issue is not open-ended. It begins with certain undisputed facts.
[113] In the early morning hours of February 19, 1998, Apaid Noel, age 19, was shot twice at close range with a .12 gauge sawed-off shotgun: once in the arm and once in the abdomen. The gunshot wound to Apaid’s arm was serious but non-life-threatening. The wound to his abdomen was both serious and life-threatening.
[114] Apaid was taken to the Ottawa General Hospital where Dr. Joel Freeman, a distinguished surgeon, operated on his abdominal wounds. Apaid’s liver had been softened and separated by the shotgun blast and Dr. Freeman removed as much foreign material – pellets and wadding – as he safely could before closing the wound site. In his view, the remaining foreign material could be left inside the liver without undue risk.
[115] Prior to the shooting, Apaid was fit and otherwise healthy. In the first few days after his surgery, he did well. However, his progress stalled and on March 5, Dr. Freeman performed a second abdominal surgery in which he removed scar tissue blocking Apaid’s intestine. While doing so, Dr. Freeman examined the liver repair and found it to be “perfect”. On March 13, Apaid was released from the hospital. Prior to his release, he had been taking a blood thinner to prevent the formation of clots. Dr. Freeman stopped this medication on March 10 and on March 13, the date of Apaid’s discharge, he regarded Apaid as having no risk of clotting.
[116] A mere five days later, however, Apaid died from a massive blood clot that blocked the flow of blood to his lungs. According to Dr. Johnston, the fatal clot formed in Apaid’s veins, travelled to his heart, entered his pulmonary artery and completely blocked the two branches of that artery that supply fresh blood to the lungs. According to members of Apaid’s family who were at his side when he died, death came quickly. Apaid clutched his chest and lost consciousness almost immediately. Dr. Johnston confirmed that when blood is unable to reach the lungs, death will ensue rapidly, sometimes preceded by an inability to breathe and lapses into unconsciousness.
[117] Blood samples taken from Apaid at the time of the autopsy revealed trace amounts of cocaine in his cardiac blood. According to Dr. Johnston, such traces will disappear from the cardiac blood within 30 to 45 minutes of consumption. Hence, Apaid would have ingested some amount of cocaine – how much Dr. Johnston could not say – very shortly before he died.
[118] Dr. Johnston detected no signs that Apaid was a regular cocaine user. In particular, he saw no needle marks on Apaid’s body, his arteries showed no indication of hardening, and there were no cocaine metabolites in his blood.
[119] Dr. Freeman, who testified at trial, regarded death from a blood clot in a patient such as Apaid to be “totally unexpected” and a “freak” occurrence. Apaid was young, he had been healing well and he was ambulatory. The case was sufficiently puzzling to Dr. Freeman that he presented it to the hospital’s morbidity conference. None of the physicians in attendance had heard of anything comparable.
(iii) Contentious Issues Related to Causation
[120] The debate between the Crown and defence at trial centred primarily on the reason or reasons why the clot had formed. Sub-issues included when the clot had formed and where in Apaid’s body it had originated.
[121] Dr. Freeman was adamant that the blood clot that killed Apaid would not have occurred had Apaid not been shot and required surgery:
Q. Right. In your opinion, what led to the formation of the clot, the pulmonary embolism, that killed Apaid Noel?
A. What led – his initial – if he hadn’t have had an operation, he wouldn’t have formed it. Is that ….
Q. Thank you. And you’ve told us why operations and multiple operations lead to forming clots, if he had not been shot and then had to have the operation and so forth, if he’d not been shot would he have died of his pulmonary embolism?
A. No.
[122] Dr. Freeman was certain that the clot was linked to the surgeries and the gunshot wounds. In his opinion, the clot likely formed in Apaid’s pelvis or femoral veins while he was hospitalized, probably between his first and second operations.
[123] Dr. Johnston did not share Dr. Freeman’s opinion about the timing or location of the clot formation. His post-mortem examination revealed that the clot was fresh. In his view, it had formed no later than 48 hours before Apaid’s death and probably sooner, perhaps within 24 hours. Moreover, Dr. Johnston saw no evidence of clotting in the pelvis or femoral veins and nothing from which he could infer that a clot may have passed through those veins.
[124] According to Dr. Johnston, wherever the clot formed in Apaid’s body, it did so in his veins, not in his arteries. Upon sectioning the liver (a procedure Dr. Freeman would not have been privy to), Dr. Johnston found what appeared to be relatively fresh clotting around foreign material that was lodged in Apaid’s liver, specifically wadding and pellets that Dr. Freeman had decided not to remove in the first operation. Dr. Johnston offered the following explanation for the clotting around the foreign material:
[O]n the inside of the liver, there’s still some red stuff around this foreign material, and that stands to reason because the body has to somehow wall off the foreign material, and as part of the healing process, the body will start developing small blood vessels, and these are very fragile thin wall or blood vessels which might continue to leak and bleed. So, there is to be some bleeding expected as part of a long-term healing process.
Eventually, if things go properly, all of the foreign material should be walled off by scar tissue, but that may take months or even a year or so.
Q. And is clotting, then, part of that ongoing healing process?
A. That’s correct.
[125] In the end, Dr. Johnston formed the opinion that the clot that killed Apaid must have formed in the small branches of the hepatic vein located in the liver. His reasons for coming to this conclusion were three-fold:
(1) He saw “no external manifestations” of blood clots in the veins of Apaid’s legs or arms.
(2) The area of injury in Apaid’s liver was the only place where he saw blood clots.
(3) On microscopic examination, the smaller branches of the hepatic vein appeared to be dilated, without any inflammation, and the hepatic vein itself was “very dilated … which might indicate that a blood clot passed through it.”
[126] In sum, Dr. Johnston deduced that the site of the blood clot had to be the liver:
[B]ecause we don’t have a demonstrable source of the origin of the blood clot except in the area of injury in the liver, I concluded that the blood clot undoubtedly came from the liver and not from somewhere else.
[127] In cross-examination by Ms. Mulligan, Dr. Johnston was asked about the role of platelets in the healing process. He responded that platelets are tiny elements in the blood that “migrate to areas of injury to try and prevent bleeding.” They “stick together” and “promote clotting”, which prevents bleeding, and thus loss of blood, in areas that are injured and that require blood to repair damaged tissues.
[128] Dr. Johnston agreed in cross-examination that blood clots can form in people who are not injured or healing, for example, in individuals who “produce too many platelets.” He further acknowledged that cocaine ingestion can cause “hypercoagulability”, a term used to describe a propensity in someone to develop blood clots. Cocaine, he agreed, even in small doses, can cause “blood platelets to become excited or the blood to hypercoagulate” and “hypercoagulability causes platelets to stick together, which causes clots to form.” While Dr. Johnston did not know how common it was for blood clots to form as a result of cocaine ingestion, he thought it would not be “very common, but it certainly can happen.” He then conceded that, although he considered it unlikely, it was possible that the clot that killed Apaid may have had no relationship to his injuries or surgeries:
Q. Okay. So, given that we had a young man here with a clot, the source of which you could not determine definitively, you’re not really able to rule out cocaine as having caused the clot to form, right?
A. I guess that is true, yes.
Q. Because, although – it’s a little difficult in Mr. Noel’s case because he’s had these injuries, he’s had surgery, and then a month later he’s deceased. It would be clearer if he’d had these injuries, had surgery, and a month later he walked out and was hit by a bus. We’d be able to say, you know, there’s no relationship.
A. Correct.
Q. Here it is possible that his ingestion of cocaine is just like getting hit by that bus. The clot may have no relationship to any of the surgery or any of the injuries.
A. It’s possible. I wouldn’t think it’d be likely, but it’s possible. [Emphasis added.]
[129] At a later point in her cross-examination, Ms. Mulligan asked Dr. Johnston if he knew a Dr. Markestein. Dr. Johnston acknowledged that he did and that Dr. Markestein was a respected pathologist. She then put a question to Dr. Johnston in which she purported to summarize what Dr. Markestein would say about the clot if asked, namely: “that death was most likely caused by the ingestion of cocaine resulting in hypercoagulability, which caused the clot”.[^15] Crown counsel objected to this question. Discussion occurred in the absence of the jury and upon the jury’s return, Ms. Mulligan reframed the question and engaged in the following exchange with Dr. Johnston:
Q. Okay, I’m going to try it a different way. I’m going to put it sort of in a hypothetical way. Hypothetically, looking at the evidence that you had at the time of the autopsy, the medical reports, everything that you had in front of you, can you categorically disagree with my suggestion that the clot did not emanate from the legs, it’s very unlikely that it emanated from the liver, and that the most likely cause was the ingestion of cocaine causing hypercoagulability of the blood?
A. Okay. Cocaine was always a possibility. The reason why – and I still think that the injury to the liver, which caused a blood clot to be propagated, is the most likely source of the thromboembolism – is that, yes, cocaine causes a hypercoagulable state, or any problem with that, the effects of cocaine tend to be mostly on the smooth muscle of arteries, which causes spasm and damage to the inner lining, and a lot of blood clots are formed as the result of cocaine mostly in arteries. That’s not to say they can’t arise in veins; they certainly can.
So, even though I agree with the possibility, I still think the liver injury is the most likely source of the clot rather than it being an effect of cocaine. For other reasons, in that I did not see evidence of blood clots in other organs, in other small blood vessels throughout all the organs of the body that I sampled during the time of autopsy.
Can you have it in one particular area and cause a vein to thrombose and the blood clot to go up into the lungs and cause pulmonary embolism as a result of cocaine? Yes. But if we’re talking about a generalized hypercoagulable state, I would expect to see manifestations in other organs, and not to say they have to be. I would expect to see that.
So, although cocaine remains in the realm of possibility – and I don’t discount that – I think, again, the most likely source is the injured liver, and that’s what I will maintain and continue to express as my opinion.
Q. All right. Even though, doctor, you’ve agreed that you can’t say with any degree of certainty that, in fact, the thrombosis – sorry, thromboembolism came from the liver.
A. That’s right.
Q. So, you would agree that it is an issue in this particular case as to whether cocaine or the injury to the liver was the, I suppose the cause of the clot, you’d agree it’s an issue about which reasonable pathologists could disagree.
A. Correct.
[Emphasis added.]
[130] In sum, Dr. Johnston remained of the view that the clot that killed Apaid formed in his liver and resulted from trauma associated with the gunshot injuries. Although he could not discount the possibility that the clot may have formed independently by reason of cocaine ingestion, he considered this unlikely because clotting from cocaine typically occurs in the arteries, not the veins, and because hypercoagulability caused by cocaine is usually generalized, such that if it existed, clotting elsewhere in Apaid’s body should have been seen, but was not.
(iv) Options Available to the Jury on Causation
[131] The jury had available to it the following options on the issue of causation. On the defence side of the ledger, the appellants had to be acquitted if the jury believed or had a reasonable doubt that within 30 to 45 minutes of ingesting an unknown quantity of cocaine, Apaid died from a massive veinous clot that formed in an area of his body unrelated to the areas where he had been shot and independent of the gunshot injuries he had sustained.
[132] On the Crown’s side of the ledger, in order to find causation, the jury had to be satisfied beyond a reasonable doubt that the gunshot injuries caused or substantially contributed to the formation of the clot that killed Apaid. There were two ways in which the jury could arrive at that conclusion.
[133] First, the jury could accept the opinions of Dr. Johnston and Dr. Freeman that the gunshot injuries had a direct bearing on the formation of the fatal clot.
[134] Second, the jury could find that the cocaine which Apaid ingested increased his platelet count and in turn, promoted clotting in those areas of the liver where Dr. Johnston believed the clot had formed. The fact that the cocaine may have exacerbated the formation of clotting in the liver was, in my view, an inference that the jury could legitimately draw from the evidence, even though Dr. Johnston was not specifically asked to address the matter.
II. The Error was Harmless
[135] In my view, the failure to instruct the jury on the offence of attempted murder was harmless in the circumstances. In so concluding, I should not be taken as suggesting that a failure to leave a lesser available offence with the jury is a trivial error. On the contrary, in many instances, such an error will necessitate a new trial. Here, however, I believe that the failure to leave attempted murder with the jury was harmless in its effect because of five factors that are particular to this case. Taken together, they satisfy me that the proviso can safely be applied. They are as follows:
(1) The jury was not faced with the stark choice of convicting the appellants of murder or letting them go free; manslaughter was left to the jury as a viable alternative.
(2) The causation issue was canvassed with the jury in depth by Crown and defence counsel. Both sides treated the issue as one of substance and the jury had before it their competing positions.
(3) The trial judge’s charge to the jury on causation was exemplary. He treated the issue seriously and was not dismissive of it in the least. His instructions were thorough, scrupulously fair, and designed to impress upon the jury the importance of the issue and the need to consider it with care.
(4) The jury gave the case its most careful consideration. Over the course of five days of deliberation, the jury asked two questions, both of which related to the primary issue of identification. The questions were insightful and indicative of the care with which the jury was pursuing its task and the attention it was paying to the smallest detail.
(5) Identity was the primary issue at both the first and second trials. It accounted for the great bulk of the evidence and counsel devoted the lion’s share of their closing addresses to it. Causation, while admittedly a live issue, was at best secondary, and while the case for the Crown may not have been overwhelming, the evidence favouring the Crown’s position on causation readily surpassed the evidence favouring the defence position.
[136] Of these five factors, two stand out as particularly important and worthy of elaboration. They are:
• the jury was not faced with a stark choice of murder or acquittal; and
• the charge on causation was exemplary.
(i) The jury was not faced with a stark choice
[137] The concern about the jury being faced with a stark choice – convicting the appellants of murder or letting them go free – is central to understanding the potential harm created by the trial judge’s failure to instruct the jury on the offence of attempted murder. The governing principle is that the curative proviso will generally not be available in cases where an included offence (or in this case, a lesser offence) is not left with the jury and the jury convicts of a more serious offence. Ironically, this principle comes from R. v. Haughton, a case in which the issue did not arise.
[138] In that case, Haughton was charged with second degree murder in connection with the death of a store owner who was bludgeoned to death in the course of a robbery. Counsel at trial agreed that whoever killed the store owner was guilty of murder and the Crown maintained that Haughton acted alone. Haughton testified and acknowledged his participation in the robbery, but claimed that an associate, whom he identified as “Freddy”, was responsible for killing the store owner. Haughton admittedly knew that Freddy had armed himself with a steel bar for purposes of overcoming any resistance they might encounter in the robbery. However, he denied knowing that Freddy would commit murder as a possible consequence of the robbery. The trial judge instructed the jury on murder, as well as the included offence of manslaughter, but the jury convicted Haughton of murder.
[139] On appeal to this court [reported at (1992), 11 O.R. (3d) 621], Doherty J.A. accepted Haughton’s submission that the trial judge had misdirected the jury on Haughton’s potential liability for manslaughter as a party under s. 21(2) of the Criminal Code. Nonetheless, he was satisfied that the curative proviso could safely be applied to uphold Haughton’s conviction for murder. In his view, the evidence that Haughton had the requisite state of mind for murder under s. 21(2) was overwhelming. At p. 628, after reviewing the whole of the evidence, including Haughton’s evidence, Doherty J.A. observed: “the conclusion that the appellant knew that it was probable that Freddy would use this lethal weapon against the victim is an overwhelming one.” In light of that finding, Doherty J.A. applied the proviso and dismissed the appeal.
[140] On further appeal to the Supreme Court of Canada, Sopinka J. provided brief reasons on behalf of the court for dismissing Haughton’s appeal. In the end, like Doherty J.A., he applied the curative proviso – but not because the evidence against Haughton was overwhelming; rather, because it was clear in the circumstances that the trial judge’s error was harmless. Given the brevity of Sopinka J.’s reasons and their importance to the issue at hand, I have reproduced them in full below:
The application of s. 686(1)(b)(iii) of the Criminal Code, R.S.C., 1985, c. C-46, requires the Court to consider whether a jury properly instructed could, acting reasonably, have come to a different conclusion absent the error. In applying this test the findings of the jury in the case under appeal may be a factor in determining what the hypothetical reasonable jury would have done, provided those findings are not tainted by the error. In cases in which an included offence is not left with the jury, a conviction by the jury of the more serious offence cannot generally be relied on by reason of the fact that it may very well be a reaction against a complete acquittal. There is an apprehension that the jury convicted because they had no other alternative than acquittal and acquittal was unpalatable. In this case, the jury had an alternative: they could have convicted of manslaughter. It cannot be said that it did not do so by reason of the failure to charge them by reference to the objective standard of liability with respect to manslaughter. In convicting of murder the jury must have found that the appellant had subjective foresight of death. It is impossible to hold that they came to this conclusion because they were unable to conclude that the appellant had subjective foresight of bodily harm. [Emphasis added.]
[141] As I understand Sopinka J.’s reasons, the trial judge’s error in failing to leave objective foresight of bodily harm as a basis upon which Haughton could be found guilty of manslaughter, was harmless in light of the jury’s verdict of murder. As he explained, it was apparent from the verdict that the jury did not reject manslaughter because it was unable to conclude that Haughton “had subjective foresight of bodily harm.” Rather, by its verdict, it was evident that the jury was satisfied that Haughton “had subjective foresight of death.” Hence, the failure to properly instruct the jury on the objective basis by which Haughton could be found guilty of manslaughter, while erroneous, was harmless in the circumstances.
[142] Haughton is important to the issue at hand in two respects. First, it defines the approach that must be taken in deciding whether the “harmless error” branch of the curative proviso can be invoked in any given case. In short, it eschews a “cookie-cutter” approach and calls for a contextual and purposive analysis. Second, in line with that approach, it draws a distinction between cases where an included offence is not left with the jury and the jury convicts of a more serious offence (not Haughton) and cases where the jury is instructed on an included offence but the instruction is tainted by legal error (Haughton).
[143] The distinction that Haughton draws between those two scenarios and the different considerations that apply in deciding whether the curative proviso can be invoked is central to the problem at hand – can the harmless error branch of the proviso be applied here, where the jury did not receive instruction on the offence of attempted murder? In deciding that issue, the focus must be on the concern identified by Sopinka J. that:
[A] conviction by the jury of the more serious offence … may very well be a reaction against a complete acquittal. There is an apprehension that the jury convicted because they had no other alternative than acquittal and acquittal was unpalatable. [Emphasis added.]
[144] The notion that a jury might arrive at a verdict not because it considers it to be a “true verdict based on the evidence” but because the only other available verdict is one that it finds “unpalatable”, initially strikes me as being at odds with established jurisprudence from the Supreme Court in which the court has consistently expressed its faith and trust in the jury system and its firm belief that juries can be trusted to perform their duties in accordance with the law. See for example, R. v. Corbett, [1988] 1 S.C.R. 670 and R. v. Pan; R. v. Sawyer, 2001 SCC 42, [2001] 2 S.C.R. 344.
[145] But on closer analysis, I do not read Haughton as suggesting that jurors cannot be trusted to obey the law in difficult cases involving hard choices. On the contrary, Haughton takes a practical look at how people reason through problems and recognizes that the reasoning process can be influenced, consciously or subconsciously, by the end result.
[146] Accepting that to be the case and acknowledging that the jury in the instant case may have balked at the prospect of letting the appellants go free, the fact remains that the jury was not faced with a stark choice. Manslaughter was left to the jury as a viable “third” option – and it was “viable” in every sense of the word.
[147] First, the choice of manslaughter addressed the Haughton concern. The jury had a “palatable” option that relieved against the “unpalatable” prospect of letting the appellants go free. Again, strictly from the standpoint of Haughton, it can hardly be suggested (and no one does) that the jury would have viewed a verdict of manslaughter as a less than palatable substitute for a verdict of attempted murder. While there might be cases where the “third option” could itself be viewed as unpalatable, particularly when measured against the offence for which the jury did not receive an instruction, this is not one. Manslaughter and attempted murder are close cousins. They are both serious and they both attract possible life sentences.
[148] Second, the parties and the trial judge accepted that intent was a live issue and manslaughter was accordingly left to the jury as a viable verdict. No complaint is taken with the trial judge’s instructions on manslaughter, which were both thorough and objective.
[149] Third, manslaughter was not just a viable verdict in theory. It was a viable verdict in fact, depending on the jury’s view of the evidence. The appellants take issue with the notion that leaving manslaughter with the jury effectively alleviated the Haughton problem. In his able argument, Mr. Campbell submitted that in light of the strength of the Crown’s case on the issue of intent, which he characterized as “overwhelming”, it would have been very difficult for a principled jury to return a verdict of manslaughter.
[150] While Mr. Campbell’s concern may be valid, it only holds sway if the evidence is viewed in its most favourable light for the Crown, a matter to which I will return shortly. However, I should not be taken as accepting that the Haughton problem can never be overcome in cases where the more palatable middle-of-the-road verdict may be difficult to justify on the facts. To be consistent, once it is accepted that the reasoning process may be influenced by the end result – in this case, a jury troubled by the prospect of letting the appellants go free – it would be illogical to suggest that the same result‑oriented reasoning would not occur with respect to conviction – in this case, a jury troubled by the prospect that causation may not have been proven, but nonetheless convicting the appellants of the more serious crime of murder because manslaughter may not have fit well on the facts.
[151] But I need not finally resolve that debate in this case. As indicated, Mr. Campbell’s argument that the Crown’s case on intent was overwhelming only holds true if the evidence is viewed in its most favourable light for the Crown. But that does a major disservice to the approach taken by Ms. Mulligan at trial, in which she marshalled the evidence of a number of witnesses, accepted by the Crown to be credible and reliable, in an effort to show, or at least raise a reasonable doubt, that the shotgun blasts that injured Apaid occurred in the midst of a physical struggle between Apaid and his brother, Aschley, on the one hand, and their assailants on the other.
[152] Without going into detail, suffice it to say that Ms. Mulligan spent the better part of 25 pages of her closing address (just 5 short of the number she spent on causation) weaving together a credible scenario in which the gun was either produced by Aschley and went off accidently in the course of a struggle, or it was produced by one of the assailants and went off accidently in a struggle.
[153] That being so, it hardly lies with the appellants to suggest at this stage that the scenario presented by Ms. Mulligan was fanciful and not worthy of the jury’s consideration – or more pointedly, that no self-respecting jury could have returned a verdict of manslaughter.
(ii) The Charge to the Jury
[154] The charge to the jury on causation was anything but perfunctory. It was detailed and objective, and it treated the issue of causation as a serious matter worthy of the jury’s careful attention.
[155] The trial judge devoted the better part of 30 pages of his charge to the issue of causation. He provided the jury with clear instructions on the governing legal principles, he reviewed the pertinent evidence in considerable detail, and he provided the jury with a succinct overview of the parties’ position. He reminded the jury on several occasions that the onus rested on the Crown to prove beyond a reasonable doubt that “the shooting of Mr. Noel was a significant contributing cause of the death” and that if the jury entertained a reasonable doubt on this issue, they must acquit.
[156] In short, the trial judge’s instructions on causation were a model of clarity and fairness and there was no attempt on his part to influence the jury, subtly or otherwise, to resolve this issue in favour of the Crown. The jury was equipped with all of the tools it needed to give the matter the careful consideration it deserved. I have every reason to believe that the jury did just that in the course of its extensive deliberations, which resulted in the conclusion that the Crown had met its onus on the charge of murder. Were it to be assumed that any of the jurors were troubled by the issue of causation, but considered it unpalatable to let the appellants go free, I would have expected a compromise verdict, i.e., manslaughter, or alternatively, a hung jury. I find it unreasonable, if not unfathomable, to think that a jury in those circumstances would render a verdict of murder, which would not only have been unresponsive to the concern, but a dramatic betrayal of it.
[157] And that, in essence, is where my colleague and I part company. At the end of the day, the issue that divides us is the integrity of the murder convictions.
[158] As I understand my colleague, he takes the position that the convictions for murder tell us nothing about whether the jury might have reached a different conclusion on causation had the jury been instructed on the offence of attempted murder. I, on the other hand, believe that the murder convictions tell us all we need to know to conclude that the appellants were not prejudiced by the failure to leave the offence of attempted murder with the jury and the verdicts would necessarily have been the same had the jury been so instructed.
[159] Ironically, insofar as the integrity of the verdicts is concerned, I believe that the appellants’ position would be much stronger if the jury had convicted them of manslaughter. At least then, they could argue that manslaughter was a compromise verdict reached to avoid an unpalatable result and that they should rightfully have been convicted of attempted murder instead – a hollow victory to be sure, but one which might realistically (or more realistically) be put forward as a basis for challenging the integrity of the manslaughter verdicts had they been so convicted.
[160] When the focus is placed on the integrity of the verdicts, as I believe it should be, the case on which my colleague places considerable stock, R. v. Davy (1993), 86 C.C.C. (3d) 385 (S.C.C.), is readily understandable and one with which I wholeheartedly agree.
[161] In its essence, Davy is a case where the integrity of the murder conviction could not be guaranteed because the jury had not been instructed on the way or ways by which Davy might be found guilty of manslaughter and his co-accused Jackson guilty of murder.
[162] Unlike the case at hand, Davy is not a case where the palatability of a potential verdict may have played a role in the jury’s decision. The issue of concern on appeal, both in this court and the Supreme Court, was the degree of Davy’s culpability. Nowhere in his instructions did the trial judge expressly tell the jury that it could find Jackson guilty of murder and Davy guilty of manslaughter. Indeed, in the various scenarios which the trial judge outlined to the jury, none of the possible verdicts for Davy was manslaughter. On the contrary, the trial judge stated: “It may be unlikely in this case that Davy would be guilty of manslaughter”.
[163] To repeat, in those circumstances, where the offence of manslaughter was not properly left for the jury’s consideration insofar as Davy was concerned, one could hardly look to the murder conviction as a basis for “overcoming” the error. The error is what put the integrity of that verdict in doubt.
[164] But that is not this case. And while my colleague is correct to note my focus on Haughton, it is not so much because Haughton is the case that I rely on in support of my conclusion that the proviso can be applied, as my colleague suggests, but rather that Haughton explains the rationale/practical considerations underlying the principle that a new trial will generally be required when an included offence has not been left to the jury and the jury is left with a stark choice.
[165] To be clear, in response to my colleague’s comments at para. 91, I do not for a moment suggest that the court in Haughton intended to implicitly overrule Davy. Nor do I suggest that Haughton stands for the broad proposition that in every case where an included offence is left with the jury and the jury convicts on the main charge, the appellate court may take the findings of fact implicit in that verdict into account in deciding whether a legal error in the instruction on potential liability for an included offence resulted in no substantial wrong or miscarriage of justice. Rather, I see the decision in Haughton as supporting the more narrow proposition that the fact-finding on the issue of causation implicit in the jury’s verdict of murder can be relied on in a case like the present one, where the jury was properly instructed on the issue of causation and also properly instructed on the included offence of manslaughter. As I said at para. 156 above:
Were it to be assumed that any of the jurors were troubled by the issue of causation, but considered it unpalatable to let the appellants go free, I would have expected a compromise verdict, i.e. manslaughter, or alternatively, a hung jury. I find it unreasonable, if not unfathomable, to think that a jury in those circumstances would render a verdict of murder, which would not only have been unresponsive to the concern, but a dramatic betrayal of it.
[166] There are two other brief matters from my colleague’s reasons that I feel obliged to address.
[167] First, my colleague refers on two occasions to the exchanges between the trial judge and counsel and the trial judge’s comment that a failure to leave attempted murder would be “fatal” to any conviction should it turn out that such an instruction was required. To the extent that my colleague relies on those exchanges to support his conclusion, I would respectfully respond by observing that while the views of counsel and the trial judge are certainly worthy of consideration, they do not answer the question that we must answer in our appellate role, namely, whether the proviso can safely be applied. And if this is a case where it can be safely applied, it matters not what the trial judge or trial counsel may have thought. (See R. v. Prokofiew (2010), 100 O.R. (3d) 401 at para. 41 (Ont. C.A.)).
[168] So too with the position taken by the Crown in its factum before us concerning the applicability of the proviso i.e., that the case for the Crown on causation was overwhelming – a submission that my colleague and I cannot accept. The basis upon which I respectfully believe the proviso can and should be applied was put to counsel in argument and the matter was fully canvassed. If the proviso was available on that basis, then it is up to the court to apply it, irrespective of the position taken by the Crown. (See R. v. Pickton, [2010] S.C.C. 32 per Charron J. at para. 27).
CONCLUSION
[169] I do not pretend that manslaughter was a perfect substitute for attempted murder. To be precise, I cannot gainsay the possibility that the jury may have looked at the causation issue through a slightly different lens had it been instructed on the offence of attempted murder. To that extent, this was not a perfect trial. For the reasons stated however, I am satisfied that there was no miscarriage of justice here and no wrong of sufficient import to warrant yet a third trial, assuming the case can still be put together, some twelve years after the event.
[170] Identity was the primary issue at both the first and second trials. Causation was a secondary issue at the second trial. It took shaky hold during Dr. Johnston’s cross-examination. And while Dr. Johnston’s evidence passed the “anything is possible” mark, it did so by very little. No defence evidence was called on the subject and the Crown was not permitted to call two further experts, including Dr. Michael Pollanen, Chief Pathologist of Ontario. In the end, a carefully instructed, conscientious jury found that the Crown had met its burden of proving causation. Taking that into account, along with the fact that the jury had – but obviously rejected – the option of finding the appellants guilty of manslaughter if outright acquittals were considered “unpalatable”, I am satisfied that the failure to leave attempted murder as a possible verdict amounted to harmless error. In short, I am satisfied that the verdicts would inevitably have been the same had the jury been instructed on the offence of attempted murder.
[171] Accordingly, I would dismiss the appeals from conviction.
SENTENCE APPEALS
[172] In view of the position of the majority that a new trial must be ordered, I need not address the sentence appeals.
RELEASED: “DD” “SEP 09 2010”
“M.J. Moldaver J.A.”
[^1]: A third accused, Wolfson Cetoute, was convicted of manslaughter at the first trial. He too was granted a new trial. The Crown did not proceed against him on the re-trial.
[^2]: The parts of Starr relied on by this court in ordering a new trial were subsequently overruled by the Supreme Court in R. v. Khelawon, 2006 SCC 57, 215 C.C.C. (3d) 161.
[^3]: The appellants challenged the admissibility of the “gang” evidence on the first appeal, but abandoned that ground of appeal in oral argument: see Sarrazin at paras. 7-9.
[^4]: At the same time that Crown counsel took the position that a finding of guilt on the charge of attempted murder should not be left as a possible verdict on the murder charge, he also withdrew Count 2 in the indictment. Count 2 in the indictment alleged that Mr. Sarrazin had used a firearm during the commission of the indictable offence of attempted murder.
[^5]: In I. Grant, D. Chunn and C. Boyle, The Law of Homicide, looseleaf (Scarborough, Ont.: Carswell, 1994), at s. 5.2, the authors indicate, without reference to any authority, that “[a]ttempted murder is always an included offence to the crime of murder”. See also E. Meehan and J. Currie, The Law of Criminal Attempt, 2nd ed., (Scarborough, Ont.: Carswell, 2000) at p. 300.
[^6]: The section was introduced when murder was divided into capital and non-capital murder, the former being punishable by death and the latter by life imprisonment. The explanatory notes state that the provision was intended to make it clear that non-capital murder was an included offence in the obviously more serious charge of capital murder: Bill C-92, An Act to amend the Criminal Code (Capital Murder), 4th Sess., 24th Parl., 1960-61, cl. 7.
[^7]: While s. 662(2) speaks of a conviction for “attempted second degree murder”, there is no such offence. The offence is attempted murder simpliciter: see s. 239(1).
[^8]: Final 235 in Watt’s Manual of Criminal Jury Instructions (Toronto, Ont.: Carswell, 2005) also implies that attempted murder is not a possible verdict if there is a reasonable doubt on causation. The model charge instructs the jury to acquit if it has a reasonable doubt on the question of causation.
[^9]: The appellants make an argument that under the test in Luckett v. The Queen, [1980] 1 S.C.R. 1140, attempted murder is a lesser and included offence under the criteria set out in s. 662(1). I need not address that argument.
[^10]: A.W. Mewett, “No Substantial Miscarriage of Justice” in A. Doob and E. Greenspan, eds., Perspectives in criminal law: essays in honour of John Ll.J. Edwards (Aurora, Ont.: Canada Law Book, 1985) 81 at p. 101.
[^11]: In the recent decision of the High Court of Scotland in Johnston v. H.M. Advocate, 2009 SCCR 518 (Scot. H.C.), the court applied the Coutts analysis but held that the failure to instruct on the included offence of attempted murder was a harmless error. In that case, the accused was charged with murder. Causation was in issue. The jury was not told about potential liability for attempted murder, but was instructed that it could return a verdict on certain included assault charges. The High Court reasoned that any doubt on causation would have resulted in a verdict on one of the lesser assault-related offences on which the jury was properly instructed, and that the failure to instruct on attempted murder did not, therefore, affect the validity of the murder conviction.
[^12]: This court found errors in the manslaughter instruction pertaining to liability under s. 21(2), as well as liability as an aider and abetter: (1992), 11 O.R. (2d) 621 (C.A.). Given the subsequent pronouncement of the Supreme Court of Canada in Davy with respect to potential liability for manslaughter under s. 21(2), it is doubtful that the trial judge in Haughton made any error in respect of s. 21(2). This would leave only the error arising out of the aiding and abetting provisions.
[^13]: With respect, this is perhaps an area that the Supreme Court of Canada may wish to re-visit in view of the present-day complexity of the criminal law, the ever-increasing length of criminal trials, the practical difficulties involved in re-trying cases many years after the event, the added burden that lengthy re-trials place on an already overburdened justice system, the right of victims to closure, the interests of finality and the public’s overall confidence in the administration of justice. Perhaps in light of these modern-day realities, a more holistic approach to the application of the proviso would be preferable to the existing “pigeon-hole” model so that the proviso could be applied where an appellate court is satisfied that the evidence of guilt is very strong, although not quite overwhelming, and the legal error or errors, though not insignificant, are highly unlikely to have affected the result.
[^14]: The first appeal was allowed primarily because of a Starr error [R. v. Starr (2000), 147 C.C.C. (3d) 499 (S.C.C.)] which may have been viewed differently following the Supreme Court’s subsequent decision in R. v. Khelawon (2006), 215 C.C.C. (3d) 161.
[^15]: Following Dr. Johnston’s testimony, Crown counsel advised the trial judge that he wished to call Dr. Pollanen, Chief Pathologist of Ontario, and Dr. Rodger, a haematologist, to confirm that trauma due to gunshots and subsequent surgeries were at the origin of the fatal pulmonary embolism, not cocaine ingestion. The trial judge denied the Crown’s request but stated that he was prepared to revisit the matter after hearing from Dr. Markestein. Nothing further came of this as Dr. Markestein was not called to testify.