R. v. Burke, [2002] 2 S.C.R. 857, 2002 SCC 55

Howard Burke Appellant

v.

Her Majesty The Queen Respondent

Indexed as: R. v. Burke

Neutral citation: 2002 SCC 55.

File No.: 28546.

2002: March 12; 2002: June 21.

Present: McLachlin C.J. and L’Heureux‑Dubé, Gonthier, Iacobucci, Major, Bastarache, Binnie, Arbour and LeBel JJ.

on appeal from the court of appeal for ontario

Criminal law — Trial — Verdict — Validity — Foreman announcing a verdict — Court recording verdict as “not guilty” — Jury dismissed and accused discharged — Trial judge informed shortly thereafter that court may have erred in recording verdict — True verdict intended by jury may have been “guilty as charged” — Trial judge subsequently conducting limited inquiry into verdict and changing recorded verdict from “not guilty” to “guilty” — Whether trial judge had jurisdiction to conduct inquiry into verdict and alter it after discharge of jury — Whether rule in Head should be re-examined.

At the close of the accused’s trial for attempted murder, the foreman announced the verdict. The trial judge, court registrar, and both counsel heard “not guilty as charged”. A “not guilty” verdict was recorded and the jury was discharged. While escorting the jurors out of the courtroom, a court officer asked the foreman what the jury’s verdict had been. The foreman replied, “You’re kidding, guilty”. Within approximately seven to nine minutes after the announcement of the verdict in court, the trial judge was informed of the apparent error. An effort was made to locate the jurors. The foreman and another juror were found in the parking lot and brought back to the courthouse. The court officer called the remaining jurors at their homes, but was unable to reach two jurors. Approximately 25 minutes after the discharge of the appellant, court resumed in order to clarify the verdict. In the presence of both counsel and one juror, the foreman confirmed that the verdict was intended to be guilty. The trial judge reconvened court the next day. The full jury and both counsel were present, but the accused was not. Each of the jurors testified that their verdict was “guilty”. The trial judge concluded that he had jurisdiction to conduct a limited inquiry into what the actual verdict was and into whether the court had committed an error in recording the verdict. A temporary publication ban was imposed but two articles describing the incident were published in two widely read newspapers. Three days later, at the third inquiry, both the accused and the full jury were present. The court reporter testified that he originally recorded the foreman’s response as unclear but when he replayed the tape at normal speed, he heard “guilty as charged”; playing the tape on more sophisticated audio equipment, he heard “not guilty as charged” and was prepared to certify the verdict as such. The trial judge then questioned the jury. The foreman testified that he had cleared his throat in announcing the verdict, and that he had in fact said “guilty as charged”. Most of the jurors testified that they had heard something like “guilty”, but one testified that she was unable to hear the foreman as he announced the verdict. The trial judge then asked each juror two questions regarding taint or bias. Nine jurors indicated that their testimony had not been influenced by the media or anyone else. Two jurors testified that they had read or heard things in the media about the case but had not been influenced. One juror started to allude to an item he had heard in the media about the case, but the trial judge cut him off before he could give a complete answer.

The trial judge distinguished R. v. Head, 1986 8 (SCC), [1986] 2 S.C.R. 684, and held that he had the jurisdiction to change the recorded verdict from “not guilty” to “guilty as charged”. He found no air of reality to the suggestion that the jurors were or may have been tainted between the time when the verdict was announced and the time that the jurors were reconvened and testified the next morning. The majority of the Court of Appeal dismissed the accused’s appeal, finding that the error in this case was an accidental slip.

Held (McLachlin C.J. and L’Heureux-Dubé, Gonthier and Bastarache JJ. dissenting): The appeal should be allowed and a new trial ordered.

Per Iacobucci, Major, Binnie and LeBel JJ.: Head cannot be distinguished from the present case. However, the evolving jurisprudence and policy concerns warrant developing an exception to the general rule enunciated in Head that a trial judge is functus officio and lacks jurisdiction to reconvene the jury to inquire into an alleged error in the verdict once the jury in a criminal trial has been discharged. A trial judge retains the limited and exceptional jurisdiction to recall the jury and conduct a narrow inquiry into the alleged error where the error does not require the jury to reconsider the verdict or continue its deliberations with a view to handing down additional verdicts. This type of error is not properly called “clerical” or “administrative”, as those types of errors may be corrected by the judge without recalling the jury.

The first question that a trial judge must ask post-discharge is whether the error requires reconsideration of the verdict. If it does, the general rule in Head applies and there are no circumstances under which the judge will retain or otherwise possess jurisdiction to reconvene the jury and conduct an inquiry into the alleged error. If the error does not require the jury to reconsider its verdict, then the trial judge possesses jurisdiction to conduct an inquiry into whether the facts of the case disclose a reasonable apprehension of bias. In determining whether this apprehension is raised, the trial judge must consider all the relevant circumstances of the case, the most crucial circumstance usually being the dispersal of the jury and its probable effect on the minds of reasonable members of the public.

Where the trial judge concludes, post-discharge, that the facts raise a reasonable apprehension of bias, he should declare a mistrial if that is the necessary remedy to prevent a miscarriage of justice. In making that order, the trial judge must consider the rights of the accused and the public, along with the effect of not ordering a mistrial on the administration of justice. On the other hand, if a mistrial is not necessary to prevent a miscarriage of justice, then the trial judge should uphold the verdict as given at trial. Where, however, the trial judge concludes that there is no reasonable apprehension of bias, he must correct the error in the verdict; a mistrial is not available as a remedy.

The present case falls within the exception to the general rule in Head. The alleged error lay in the faulty transmission and recording of the verdict and, since this error did not involve reconsideration of the verdict, the trial judge had post-discharge jurisdiction to conduct an inquiry into the alleged error. However, the trial judge did not conduct the proper analysis. Instead of asking whether there was a reasonable apprehension in the minds of right-thinking, properly informed members of the public that the jurors might have been biased or influenced, the trial judge conducted an actual bias test. In this case, the length of time which elapsed between the delivery of the recorded verdict and the point at which the jury was reconvened was substantial; the accused was discharged from custody; the nature and the scope of the jury’s dispersal was extensive; and, lastly, the jurors were exposed to the reaction of the public to the recorded verdict during the period when they were absent from the courtroom and to the potentially prejudicial media coverage of this case which appeared before and after the temporary publication ban. When the proper test is applied and all the relevant circumstances are considered in context, there is no question that the facts, particularly the jury’s extensive and extended dispersal, establish a reasonable apprehension of bias. As a result, the trial judge did not possess the exceptional jurisdiction to correct the verdict. He retained only the remedial jurisdiction to declare a mistrial. He was therefore in error when he changed the verdict and registered a conviction.

Per Arbour J.: This is an appropriate case to re-examine the rule in Head. The trial judge’s jurisdiction to inquire as to whether the verdict was correctly recorded is based on the fact that the trial may not have been properly concluded. A verdict other than the intended unanimous verdict of the jury is a nullity. If, as here, the trial judge has a reasonable concern that the verdict might be a nullity, the trial should resume as if the verdict had not been rendered. The trial judge should determine whether the recorded verdict was in fact null and void and if not, the verdict should stand. If the recorded verdict was a nullity, it should be set aside and the trial should resume. The appropriate test is whether there is a “reasonable apprehension of taint or bias”. If there is no reasonable apprehension of bias, then there is no perceived threat to the impartiality of the jury and the jury is in the same position as it was prior to the court recording the verdict. Under this approach, the jury is free to continue deliberations if necessary. If there has been a reasonable apprehension of bias, the trial judge’s only option is to declare a mistrial and to order a new trial “on such terms as justice may require”. In this case, the extent of the jury’s dispersal established a reasonable apprehension of bias and the appropriate remedy is to order a new trial.

Per McLachlin C.J. and L’Heureux-Dubé, Gonthier and Bastarache JJ. (dissenting): There is agreement with Major J. that the strict rule in Head must be rejected in favour of a more refined and flexible analysis, and with the stated test for determining the limits of the exercise of post-discharge jurisdiction. However, whether a verdict can be corrected post-discharge is a highly fact-specific analysis which depends on the totality of the circumstances, including the probable reason for the initial mistake and the length of time which has elapsed between the original verdict and the moment the error is brought to the trial judge’s attention. While dispersal may, in some instances, provide the most compelling evidence on which the test will turn, it is not the deciding factor in a case where, as here, the evidence satisfactorily establishes error in carrying out the court’s recording of the jury’s true verdict and a failure to correct that error would not serve the administration of justice.

In this case, a reasonable and right-minded person, apprised of the totality of the circumstances, would not conclude that there was a reasonable apprehension of taint. Although the nature and scope of the jury’s dispersal and the potential exposure to media coverage raise the possibility of taint, neither of these factors is dispositive given the credible reason for the error and the manner in which it was brought to the court’s attention and confirmed by the jury. Mere exposure to media or conversations about the case does not automatically spoil the juror’s later statements, especially where, as here, it is abundantly clear that the jury had actually intended to find the accused “guilty”. On the particular facts of this case, there was no reasonable apprehension of taint, and it was proper for the trial judge to record the jury’s true verdict. To require a new trial or allow the verdict to stand in light of the totality of these circumstances would work a serious injustice to the interests of the state and the general public.

Cases Cited

By Major J.

Discussed: R. v. Head, 1986 8 (SCC), [1986] 2 S.C.R. 684; referred to: R. v. Vodden (1853), Dears. 229, 169 E.R. 706; R. v. Cefia (1979), 21 S.A.S.R. 171; R. v. Andrews (1985), 82 Cr. App. R. 148; R. v. Follen, [1994] Crim. L.R. 225; R. v. Loumoli, [1995] 2 N.Z.L.R. 656; R. v. Maloney, [1996] 2 Cr. App. R. 303; R. v. Aylott, [1996] 2 Cr. App. R. 169; R. v. Z.A., Eng. C.A., March 8, 1999, unreported; People v. Powell, 221 P.2d 117 (1950); State v. Brandenburg, 120 A.2d 59 (1956); State v. Fornea, 140 So.2d 381 (1962); Commonwealth v. Brown, 323 N.E.2d 902 (1975); State v. Edwards, 552 P.2d 1095 (1976); Webber v. State, 652 S.W.2d 781 (1983); Burchett v. Commonwealth, 734 S.W.2d 818 (1987); People v. McNeeley, 575 N.E.2d 926 (1991); State v. Myers, 459 S.E.2d 304 (1995); Montanez v. People, 966 P.2d 1035 (1998); State v. Green, 995 S.W.2d 591 (1999); Martin v. State, 732 So.2d 847 (1998); United States v. Dotson, 817 F.2d 1127 (1987), am. 821 F.2d 1034 (1987); Bricmont v. Mathieu (1987), 7 Q.A.C. 199; Chandler v. Alberta Association of Architects, 1989 41 (SCC), [1989] 2 S.C.R. 848; People v. Rushin, 194 N.W.2d 718 (1971); R. v. Budai (2001), 154 C.C.C. (3d) 289, 2001 BCCA 349; R. v. Cameron (1991), 1991 7182 (ON CA), 64 C.C.C. (3d) 96, leave to appeal refused, [1991] 3 S.C.R. x; R. v. S. (R.D.), 1997 324 (SCC), [1997] 3 S.C.R. 484; Committee for Justice and Liberty v. National Energy Board, 1976 2 (SCC), [1978] 1 S.C.R. 369; R. v. Barrow, 1987 11 (SCC), [1987] 2 S.C.R. 694; R. v. Sussex Justices, Ex parte McCarthy, [1924] 1 K.B. 256; R. v. Taillefer (1995), 1995 4592 (QC CA), 40 C.R. (4th) 287, leave to appeal refused, [1996] 1 S.C.R. x; R. v. Lessard (1992), 1992 3103 (QC CA), 74 C.C.C. (3d) 552, [1992] R.J.Q. 1205, leave to appeal refused, [1992] 3 S.C.R. vii; R. v. Woods (1989), 1989 7192 (ON CA), 49 C.C.C. (3d) 20, leave to appeal refused, [1990] 2 S.C.R. xii; R. v. Martineau (1986), 1986 4730 (QC CA), 33 C.C.C. (3d) 573; R. v. Antinello (1995), 1995 ABCA 117, 97 C.C.C. (3d) 126; R. v. T. (L.A.) (1993), 1993 3382 (ON CA), 84 C.C.C. (3d) 90; R. v. Rondeau, [1998] O.J. No. 5759 (QL).

By Arbour J.

Discussed: R. v. Head, 1986 8 (SCC), [1986] 2 S.C.R. 684.

By L’Heureux-Dubé J. (dissenting)

R. v. Head, 1986 8 (SCC), [1986] 2 S.C.R. 684; R. v. S. (R.D.), 1997 324 (SCC), [1997] 3 S.C.R. 484; Committee for Justice and Liberty v. National Energy Board, 1976 2 (SCC), [1978] 1 S.C.R. 369; R. v. Andrews (1985), 82 Cr. App. R. 148; State v. Williquette, 526 N.W.2d 144 (1995); R. v. Cinous, [2002] 2 S.C.R. 3, 2002 SCC 29.

Statutes and Regulations Cited

Criminal Code, R.S.C. 1985, c. C-46, ss. 644(1), 647(1), (2), (4).

Authors Cited

Maric, Vaso. Annotation to R. v. Burke (2001), 41 C.R. (5th) 135.

Wigmore, John Henry. Evidence in Trials at Common Law, vol. 8. Revised by John T. McNaughton. Boston: Little, Brown & Co., 1961.

APPEAL from a judgment of the Ontario Court of Appeal (2001), 2001 24043 (ON CA), 53 O.R. (3d) 600, 153 C.C.C. (3d) 97, 41 C.R. (5th) 134, 143 O.A.C. 286, [2001] O.J. No. 1119 (QL), dismissing the appellant’s appeal from a decision of the Ontario Court (General Division), [1997] O.J. No. 5568 (QL). Appeal allowed, McLachlin C.J. and L’Heureux-Dubé, Gonthier and Bastarache JJ. dissenting.

David M. Tanovich, for the appellant.

Susan G. Ficek, for the respondent.

The reasons of McLachlin C.J. and L’Heureux-Dubé, Gonthier and Bastarache JJ. were delivered by

1 L’Heureux-Dubé J. (dissenting) — The main issue in this appeal is whether a trial judge has jurisdiction, after discharge of the jury, to correct an improperly recorded verdict. More specifically, the question here is whether the trial judge had jurisdiction post-discharge to change a wrongly recorded “not guilty” verdict to the jury’s true verdict: “guilty as charged”.

2 I have had the benefit of reading the reasons of my colleague Major J., and I respectfully disagree with him on two issues, namely, the elements of the test to be applied by trial judges in these circumstances and the result he reaches.

I. Limits on Post-Discharge Jurisdiction

3 I agree with Major J., for the reasons he gives, that the strict rule announced in R. v. Head, 1986 8 (SCC), [1986] 2 S.C.R. 684, must be rejected in favour of “a more refined and flexible analysis” that confers upon trial judges the jurisdiction to correct a criminal verdict post-discharge in certain limited circumstances. I also agree with Major J.’s reasons that the test for determining the limits of the exercise of post-discharge jurisdiction is “reasonable apprehension of taint” which requires that “an informed person, viewing the matter realistically and practically — and having thought the matter through — [would] conclude” that taint likely occurred. As this Court has previously explained, the reasonable apprehension test requires a determination that an informed person would think that it is more likely than not that the jury upon recall after discharge would not decide fairly; “[t]he grounds for this apprehension must ... be substantial and I ... refus[e] to accept the suggestion that the test be related to the ‘very sensitive or scrupulous conscience’”: see R. v. S. (R.D.), 1997 324 (SCC), [1997] 3 S.C.R. 484, at para. 31 (quoting Committee for Justice and Liberty v. National Energy Board, 1976 2 (SCC), [1978] 1 S.C.R. 369, at p. 395, per de Grandpré J. (writing for the majority on this issue)).

4 I respectfully disagree, however, with the heavy emphasis Major J. places on dispersal as an element in establishing reasonable apprehension of taint. Whether or not a verdict can be corrected post-discharge is a highly fact-specific analysis and one which, in every case, will depend on the totality of the circumstances, including the probable reason for the initial mistake and the length of time which has elapsed between the original verdict and the moment the error is brought to the trial judge’s attention: see R. v. Andrews (1985), 82 Cr. App. R. 148 (C.A.), at p. 154. Often, it will not be practical to use an essentially bright-line test of dispersal to determine whether a jury verdict can be corrected; as Lamer J. pointed out in Head, supra, at p. 700 (quoting Wigmore on Evidence (McNaughton rev. 1961), vol. 8, at para. 2355):

It has occasionally been said that this correction must be claimed before the jury are discharged, but this seems unsound because such errors are seldom ascertained until after the jury have separated and conversed out of court, and if the error is satisfactorily established, there can hardly be any fixed time to limit its correction.

5 Given that the general rule prohibiting post-discharge jurisdiction serves to protect the interests of justice, it is axiomatic that an exception to that rule must serve the interests of justice in a way that adherence to the rule could not. Where it is carefully established that an error was made in recording the jury’s true verdict, the interests of justice will rarely be served by prohibiting the trial judge from altering the verdict as entered. Such a rigid adherence to procedural niceties, in my view, would undermine, rather than strengthen, the public’s confidence in our jury system by impeding the ascertainment of truth and justice. As the court in State v. Williquette, 526 N.W.2d 144 (Wis. 1995), concluded, at p. 151:

It is not difficult, however, to imagine a scenario where an incorrect verdict transcription could result in a serious injustice. The public policy considerations of promoting a jury’s freedom of deliberation, the stability and finality of judgments and protecting jurors against annoyance and embarrassment are indeed noble and must be taken into account; however, these considerations may not completely trump the ascertainment of the “truth” as actually found by the jury.

In the instance of an erroneously entered guilty verdict, it would be a serious injustice to allow an innocent person to remain in prison simply because members of the jury had separated or mingled with the public. There is no guarantee that the Attorney General would forego the holding of a new trial, given the limited inquiry concerning the verdict, and no certainty that a new trial would produce the same result. Even if bail were granted, the innocent person’s burden would be great. I believe that it is also society’s interest that those found guilty do not go free.

6 While dispersal may, in some instances, provide the most compelling evidence on which the test will turn, I do not think dispersal acts as the deciding factor in a case where, as here, the evidence satisfactorily establishes error in carrying out the court’s recording of the jury’s true verdict and a failure to correct that error would not serve the administration of justice.

II. Application of the Law to the Facts of this Case

7 In stating the verdict, the foreperson coughed or stuttered, causing listeners to hear different verdicts. The verdict as pronounced was never repeated or confirmed by the court. Although the trial judge thought he heard “not guilty as charged”, later that same afternoon, he admitted to having “a raging head cold” and acknowledged that there was something “unusual” about the verdict. The court reporter initially recorded the foreperson’s verdict as “(inaudible) guilty as charged” and, when examined the next day, testified that he “couldn’t be sure” what the foreperson said at the moment of hearing it. Even after replaying the original tape in the courtroom “eight to twelve” times, he was unable to discern what the verdict was (albeit at “normal speed” he heard “guilty as charged”). The court officer testified that she had been unable to understand the foreperson due to a “rumbling” or “gurgling” noise he made while delivering the verdict. Ten of the eleven jurors testified under oath that they heard the foreperson say “guilty”, the eleventh stated that she was not able to hear the verdict as announced. The foreperson testified that he sometimes stammered and also had a soft tone of voice, but that he had said “guilty as charged”.

8 Also significant in this case is the manner in which the error was brought to the attention of the court. After the verdict was announced on September 18, the jury was escorted by the court officer from the courtroom to the jury room. Within about 30 seconds after leaving the courtroom, the court officer asked the foreperson what the verdict was. At this point, all 12 jurors were back in the jury room and still within the care and control of the court. In the presence of all of the other members of the jury, the foreperson replied, “You’re kidding, guilty”. Although some of the jurors were in a position to overhear what was said, none of them contradicted the foreperson’s statement. The jurors were then escorted down a private elevator in order to exit the building, which is typical when there is a guilty verdict. On the way back to her office, the court officer heard that the court deputy had said that the verdict was “not guilty”. She testified that she knew immediately that an error had been made.

9 Within approximately seven to nine minutes after the announcement of the verdict in court, the court officer informed the trial judge of the error. Two constables at the court found the foreperson and another juror in the parking lot. Both jurors were alone and separately confirmed to one of the constables that the verdict was guilty. Within about 25 minutes of the discharge of the appellant, the trial judge had the foreperson and the other juror back in court. Under oath, the foreperson confirmed to the trial judge that the jury’s verdict was guilty and that he had pronounced a guilty verdict in the courtroom. In my view, there is no question of the foreperson’s verdict being altered as a result of anything he heard after returning the initial verdict; any possibility that he was tainted after exiting the building into the parking lot is completely mitigated by the fact that he had already disclosed the error to the court officer while still within the care and control of the court.

10 Around the time the trial judge was questioning the foreperson, the court officer called the remaining 10 jurors at their homes and asked them to return to the courthouse. She also asked one of the jurors on the phone “off the record” what the verdict was; like the foreperson and the other juror found in the parking lot, the juror said “guilty”.

11 The trial judge reconvened court the following morning. The full jury was present, and each of the jurors unequivocally testified that the verdict was “guilty”, including the foreperson who repeated that he had said “guilty as charged”. On September 23, the jury was again reconvened and each juror was separately asked to repeat the verdict as pronounced as well as the verdict they had agreed to. Once again, all responded “guilty” to both inquiries, except for one juror who testified that she did not hear the foreperson as he announced the verdict, but that there was “not a doubt in [her] mind at all” as to what the actual verdict was. Although, as Major J. points out, two jurors testified that they had read and heard statements about the case in the media, I think it is important to stress that both unequivocally and consistently testified that the true verdict was guilty. The first indicated that he was certain what verdict the jury had reached: “I read the articles on the weekend but they had no influence. The verdict was given. The verdict is what it was. We spent almost -- close to 15 hours in deliberations. There was no question” (emphasis added). Similarly, the latter testified that there was “not a doubt in [her] mind at all” as to what the verdict was. The juror who started to allude to something he had heard in the media but was prevented by the trial judge from completing his testimony on that point was asked not only whether he heard anything that affected, but also anything that might have affected, his testimony in court. He replied: “Absolutely not”.

12 Taking into consideration the totality of the circumstances, I do not think there was a reasonable apprehension of taint in this case. Although the nature and scope of the jury’s dispersal and the potential exposure to media coverage raise the possibility of taint, neither of these factors is dispositive given the credible reason for the error and the manner in which it was brought to the court’s attention and confirmed by the jury. Even though the other jurors did not confirm the error until the next day, it is significant that the court officer was told the jury’s actual verdict within seconds after the verdict was pronounced and before the foreperson had any opportunity to mingle with the public. Also significant is the fact that the foreperson communicated the true verdict to the judge within a very short period of time. The jurors who were recalled on September 19, the morning after the verdict was returned, were simply confirming the correct verdict that had already been communicated by the foreperson to the trial judge the previous day. The evidence indicates that until they were called at home after the trial, these jurors were under the impression that the verdict announced in court was guilty.

13 In short, from the conclusion of the trial through the third and final inquiry, every single juror, whether asked off the record or under oath, consistently, repeatedly, and unequivocally confirmed that the verdict was “guilty”. A reasonable and right-minded person, apprised of these circumstances, would not conclude that there was a reasonable apprehension of taint in this case; quite simply, “[t]he clear evidence of [undue influence] required to sustain a reasonable apprehension of [taint] is nowhere to be found”: S. (R.D.), supra, at para. 58. To require a new trial or allow the incorrect verdict to stand in light of the totality of these circumstances would work a serious injustice to the interests of the state and the general public “since it forces the court to tell the jurors that despite the fact that they have dutifully carried out the oath that they had sworn to uphold by listening to [20] days of evidence and then rendering a unanimous and otherwise valid verdict, their decision must be ignored”: V. Maric, Annotation to R. v. Burke (2001), 41 C.R. (5th) 135, at p. 136. I have utmost faith in the public’s ability to understand the error that took place here and the need for flexibility on the part of judges to take appropriate steps to remedy errors of this nature. It strikes me that any confusion on the part of the public will stem from the decision to expend the Crown’s resources on another trial even though there is no reasonable apprehension of taint and no doubt that the jury’s true verdict was “guilty”.

14 In any event, I wholly disagree with Major J. that the trial judge’s conclusion was incomplete and resulted in an error of law. I believe instead that considerable deference is owed to the trial judge’s largely fact-driven conclusion that there is “absolutely no air of reality to the suggestion that there was or may have been tainting in respect of any one or more of the members of the jury between the time the verdict was announced on the afternoon of September 18, 1997 and the time the members of the jury reconvened and testified the following morning” in this case. This conclusion was based, not by “focuss[ing] solely on whether the jurors were actually influenced” (emphasis in original), as Major J. suggests (at para. 89), but rather on an objective consideration of the totality of the circumstances; indeed, the “air of reality” test applied by the trial judge, like the reasonable apprehension of taint inquiry, looks to the viewpoint of a reasonable person properly informed of the situation: see R. v. Cinous, [2002] 2 S.C.R. 3, 2002 SCC 29. The trial judge conducted an extensive three-day inquiry in which all relevant persons, including the court reporter, the court officer, the registrar, the court constables, and all 12 members of the jury, were carefully examined. As Weiler J.A. noted in her reasons in the Court of Appeal, at para. 45: “The trial judge ensured that the process in which he engaged was fair to the accused. Having regard to the fact the accused was not present [of his own accord] when some of the evidence was heard, he had the witnesses repeat all of the evidence to ensure that the appellant was not prejudiced.” In my view, the factors that the trial judge took into account, namely, the reason for the error, the manner in which it came to his attention, and the clear and unequivocal statements of the jurors that the verdict was “guilty as charged”, favour dismissal. I am not persuaded that dispersal of the jury was the overriding factor that the trial judge should have considered.

15 The trial judge was aware of the publicity surrounding the verdict as well as the possibility that some, perhaps even all, of the jurors were exposed to reactions by the public and the media. Unlike my colleague, however, I do not believe that mere exposure to media or conversations about the case automatically spoils the juror’s later statements, especially where, as here, it is abundantly clear that the jury had actually intended to find the appellant “guilty”. In another case with a different set of facts, we might find dispositive the fact that a discharged juror potentially mingled with the public or read media reports of the trial. On the particular facts of this case, however, there was no reasonable apprehension of taint, and it was proper for the trial judge to record the jury’s true verdict.

16 For these reasons, I would dismiss the appeal.

The judgment of Iacobucci, Major, Binnie and LeBel JJ. was delivered by

Major J. —

I. Introduction

17 At issue in this appeal are important considerations of our justice system, particularly that of the jury.

18 In criminal cases the jury must be unanimous to convict or acquit; a failure to reach a verdict results in a mistrial. An essential ingredient of that process is that the presiding judge is fully aware of the decision and is also aware that it is the decision of each and every juror.

19 Canadians repose confidence in our jury system, a system that has remained essentially unchanged since our country was founded. The public expects a fair and final verdict from its juries. Any departure from those standards would, to the detriment of the country, erode that confidence. We cannot permit that erosion.

20 This appeal arises from a unique set of circumstances which raises these concerns. To all public appearances, the appellant Howard Burke was acquitted by a jury on the charge of attempted murder. The foreman announced the verdict. In doing so, he either stuttered or cleared his throat such that different people in the courtroom heard different versions of the verdict. The trial judge heard “not guilty” and recorded the verdict as such. The jury had actually intended to find the appellant “guilty”.

21 The trial judge excused the jury, and shortly after their exit, discharged the accused. Soon after the jury’s discharge, a court officer discovered that the foreman believed that the jury had delivered a verdict of guilty. Upon being informed of the error, the trial judge held several inquiries into what had taken place. The trial judge concluded that he had jurisdiction to change the recorded verdict to “guilty as charged” and register a conviction.

22 The issue at the heart of this appeal is whether a trial judge has jurisdiction in a criminal trial to inquire into and possibly correct an error in the verdict after the discharge of the jury, and if so, the extent to which such jurisdiction exists. I conclude that such jurisdiction may exist in certain unusual circumstances.

23 There are four possible outcomes of a jury trial:

(1) The jury renders the verdict that it intended. The jury is discharged and the trial has concluded. This is the usual result.

(2) The jury does not render the verdict it intended. The jury is not yet discharged. The trial judge retains jurisdiction to record the intended verdict.

(3) The jury does not render the verdict it intended. The jury is discharged by the trial judge but, unlike the strict rule in R. v. Head, 1986 8 (SCC), [1986] 2 S.C.R. 684, the trial judge retains a narrow post-discharge jurisdiction to recall the jury for the purposes of an equally narrow inquiry into the alleged error, the focus of the inquiry being whether there is a reasonable apprehension of bias. This jurisdiction to recall the jury for an inquiry exists only for unintended errors; the trial judge cannot recall the jury to make any changes to the verdict that require further jury deliberation. If there is no reasonable apprehension of bias, the trial judge can and should correct the erroneous verdict.

(4) The jury does not render the verdict it intended. The jury is discharged, the trial judge recalls them for an inquiry, and the inquiry establishes a reasonable apprehension of bias. Normally, such an apprehension is only likely to arise where the jury has dispersed. Dispersal of the jury means that the jury has ceased to operate as a single unit, and has separated and mingled (or had the opportunity to mingle) with the public. Dispersal is a crucial factor in determining whether or not there is a reasonable apprehension of bias. If the trial judge concludes that there is a reasonable apprehension of bias, the trial judge cannot record the intended verdict, because the trial, in a manner of speaking, has reached the end of the road. However, in order to prevent a miscarriage of justice, the trial judge retains the ability to order a mistrial or to maintain the originally communicated verdict.

24 This appeal falls within the fourth category, above. The jury had mistakenly rendered an unintended verdict. The jury had been discharged and had dispersed into the community for a lengthy period of time. The overall circumstances establish that there was a reasonable apprehension of bias. The duration of the jury’s dispersal is an important factor in finding a reasonable apprehension of bias in this case. Given these circumstances the trial judge ought to have ordered a mistrial, as maintaining the originally recorded verdict would result in a miscarriage of justice. Therefore, I would allow the appeal and order a new trial.

II. Facts

25 The appellant was charged with attempted murder. The charge stemmed from events that took place on October 28, 1996, in which the appellant shot Ian Francis. The dispute between the appellant and Francis concerned an unpaid debt relating to a drug trafficking enterprise, in which Francis and the appellant were joint partners. The appellant claimed the shooting was done in self-defence. The trial was presided over by Minden J. of the Ontario Court (General Division). The trial commenced on September 2, 1997. The jury retired to consider its verdict on September 17.

26 On September 18, the foreman announced the verdict. The trial judge, court registrar and both counsel heard “not guilty as charged”. The Crown asked that the jury be polled. The court registrar did not repeat the verdict to the jurors, but instead asked each juror if they agreed or disagreed with the verdict as announced. This polling method was prescribed by a court practice manual. All of the jurors stated that they agreed. The court recorded a verdict of “not guilty”, and discharged the jury. The jury exited the courtroom. Shortly thereafter the appellant was discharged and court was adjourned at 2:55 p.m.

27 While escorting the jurors out of the courtroom, a court officer asked the foreman what the jury’s verdict had been. In the presence of other jurors, the foreman replied, “You’re kidding, guilty”. The jurors were then escorted down a private elevator in order to exit the building. On the way back to her office, the court officer heard that the verdict had been recorded as “not guilty”. Within approximately seven to nine minutes after the announcement of the verdict in court and the jury’s exit from the courtroom, the court officer informed Minden J. in his chambers of the apparent error.

28 An effort was made to locate the jurors. The foreman and another juror were found in the parking lot and brought back to the courthouse. The court officer called the remaining jurors at their homes, but was unable to reach two jurors. She asked one of the jurors on the phone “off the record” what the verdict had been; the juror said guilty.

29 At 3:20 p.m., 25 minutes after the discharge of the appellant, court resumed in order to clarify the verdict. This was the first of three inquiries. The appellant was not present. Both counsel were present, as well as the foreman and the other juror who had been found in the parking lot. The foreman confirmed that the verdict was intended to be guilty. The other juror was not asked to confirm the verdict. Minden J. stated that he had a head cold and that he could have misunderstood the verdict as announced. Although he thought he had heard “not guilty”, he acknowledged that there was “something about it that was unusual”.

30 Minden J. reconvened court on September 19, 1997. The full jury was present, but not the accused. Each of the jurors testified that their verdict was “guilty”. The foreman testified that he sometimes stammered and also had a soft tone of voice, but that he had said “guilty as charged”. The trial judge concluded that he had jurisdiction to conduct a limited inquiry into what the actual verdict was, and into whether the court had committed an error in recording the verdict. He ordered the appellant to appear and issued a bench warrant for his arrest. Court was adjourned until September 22, 1997. A temporary publication ban was imposed.

31 Two articles describing the incident were published in two widely read Toronto papers, the Star and the Sun. One article appeared on September 19, 1997, and the other on September 22, 1997.

32 On September 23, court reconvened for the third and final time with both the appellant and jury present. The court reporter was called to testify. The transcript of the original proceedings reads as follows:

COURT REGISTRAR: How do you find for the accused, guilty or not guilty as charged?

JURY FOREPERSON: (inaudible) guilty as charged.

The court reporter stated that he originally recorded the foreman’s response as unclear. When he replayed the tape at normal speed, he heard “guilty as charged”. However when he played the tape again on more sophisticated audio equipment, he heard “not guilty as charged” and was prepared to certify the verdict as such.

33 The jury was questioned. The foreman testified that he had cleared his throat in announcing the verdict, and that he had in fact said “guilty as charged”. He again stated that he was soft spoken, and that he had a tendency to stammer when nervous. The trial judge asked the remaining 11 jurors whether they heard the verdict when it was announced and if so, what they heard. Most of the jurors testified that they had heard the verdict as it was announced, and that they had heard “guilty”. One juror testified that she was unable to hear the foreman as he announced the verdict. Another juror stated that “as near as I could tell”, she heard the foreman say “guilty as charged”. There was some variation among the jurors’ testimonies as to whether they heard the foreman say “guilty as charged”, “guilty” of the offence of “first degree murder”, “guilty of attempted murder”, or “guilty as accused”.

34 Counsel for the defence requested that he be permitted to cross-examine the jurors in order to test their credibility and examine the possibility of bias. The trial judge did not permit defence counsel’s cross-examination. Instead, he asked each juror two questions regarding taint or bias:

1. Between the giving of your verdict last Thursday afternoon and your testimony on Friday, was anything said to you by anyone, either by family, friends, court officials, fellow jurors or anyone else that influenced you so as to affect your testimony here in court?

2. Between the giving of your verdict last Thursday afternoon and your testimony on Friday, did you read or hear anything in the media that influenced you so as to affect your testimony here in court?

35 Two jurors testified that they had read or heard things in the media about the case but had not been influenced. One juror started to allude to an item he had heard in the media about the case, but was cut off by the trial judge before he could give a complete answer. The remaining nine jurors replied in the negative to both of the above questions. The inquiry ended on September 24, 1997.

III. Judicial History

36 Minden J. held that he had the jurisdiction to change the recorded verdict from “not guilty” to “guilty as charged”. He considered the Supreme Court decision in Head, supra, which if applied would result in him being functus officio. Minden J. found that Head was distinguishable from the present case, because “[t]he issue in Head was whether or not a jury could be reconvened to vary or reconsider its ... verdict after the jury had been discharged. . . . The issue here is: what was the jury’s verdict and was an error made . . . in recording that verdict?”

37 He found no air of reality to the suggestion that the jurors were or may have been tainted between the time when the verdict was announced and the time that the jurors were reconvened and testified the next morning. The jury’s true and unanimous verdict was “guilty”. The mandate of a trial judge is to accurately record the jury’s true verdict; the trial judge failed to follow his mandate in this case. Therefore, the inaccurate recording of the verdict and the discharge of the jury and accused constituted a “nullity”. Minden J. recorded the verdict as “guilty” and registered a conviction. He sentenced the appellant to 12½ year’s imprisonment.

38 Weiler J.A. (Simmons J.A. concurring) for the Court of Appeal for Ontario dismissed the appeal: (2001), 2001 24043 (ON CA), 53 O.R. (3d) 600. She concluded that the rule in Head makes it clear that once the jury has been discharged, it is too late to inquire into the nature of the verdict. This rule is based on policy reasons such as protecting jurors from post-trial harassment, encouraging open discussion in the jury room, and ensuring the finality of the verdict. However, Weiler J.A. held, at para. 37, that although a judicial body is normally functus officio once a jury has been discharged, sometimes a “clerical mistake or error arising from an accidental slip or omission” can be corrected. She concluded that the error in this case is an accidental slip. She also concluded that Head does not govern this case. Head dealt with an internal event, the jury’s deliberations, which are protected to encourage free and frank debate by the jurors without fear of reprisal. Unlike Head, the error in the present case was caused by events external to the jury deliberations: what the foreman said in open court and what was heard by the trial judge. As Head is distinguishable, its policy reasons were not applicable to this case.

39 Goudge J.A. dissented. He would have allowed the appeal and restored the originally recorded acquittal. Goudge J.A. disagreed with Weiler J.A.’s conclusion that Head was inapplicable to this case. Both Head and the present case were about whether, post-discharge, a jury can be reconvened to determine the true nature of the verdict in order to correct an error in the recording of the verdict. An approach which would permit an inquiry into the true verdict in some cases would “inevitably encompass the post-discharge contacts and experiences of jurors and the appearance of possible outside influence that may arise as a consequence. The inquiry will have to engage in the difficult task of evaluating that appearance” (para. 67). In contrast, the principle in Head eliminates any need for such complex inquiries. He concluded, at para. 69, that although it may indeed be that an “accidental slip” not requiring the recall of the jury can be corrected post-discharge, that was not the case here.

IV. Issues

40 (a) Does a trial judge in a criminal case have jurisdiction to conduct an inquiry into the verdict and alter the verdict after the discharge of the jury?

(b) If the answer to (a) is “yes”, what are the parameters guiding or limiting this post-discharge jurisdiction?

(c) If a trial judge has post-discharge jurisdiction, what is the scope of the trial judge’s remedial jurisdiction?

(d) Did the trial judge in this case err by correcting the verdict and entering a conviction?

V. Analysis

41 These reasons refer to a number of cases dealing with the reasonable apprehension of bias. As the test for taint is the same, and for ease in reading, the use of the term “bias” in these reasons can be interpreted as raising the same issues as “taint”. Taint is in essence a form or subset of bias. Taint and bias are intended here to be, and should be, interchangeable.

A. The Rule in Head

42 The common law rule was enunciated by this Court in Head. There, the trial judge mistakenly discharged the jury before they had delivered their complete verdict. The foreman pronounced the accused not guilty of the charge as laid. The jury was discharged. Immediately after discharge, the foreman said: “we thought we could find the Defendant not guilty of the charge as laid, but guilty of a lesser charge, is that right?” (Head, supra, at p. 696). McIntyre J. for the majority of the Court found that, upon discharge of the jury, the trial judge was functus officio and did not have the jurisdiction to correct the verdict. He stated, at pp. 688-90:

On the return of the jury if a clear and unambiguous verdict is given, it is the judge’s duty to accept the verdict and, in accordance with the practice of his court, cause it to become a part of the record of the court.

Where, on the other hand, there is ambiguity in the verdict . . ., the trial judge should inquire into the matter to ascertain the true position . . . . The judge has the discretion in such a case to accept a substituted or second verdict for the first one returned. This discretion, however, is one which is to be exercised during the course of the trial, that is, in the presence of the accused and his counsel, and prior to the dissolution of the court by the discharge of the jury. . . . It is clear, in my view, that the power or duty of the trial judge to intervene when a jury verdict is returned and to make inquiries relating to the true nature of the verdict is one to be exercised prior to the discharge of the jury . . . . [Emphasis added.]

For the majority in Head, discharge of the jury is the point of no return beyond which the trial judge in a criminal case cannot alter the recorded verdict.

43 Lamer J. (as he then was) approached the issue differently in a concurring opinion. He would have permitted a limited post-discharge jurisdiction to inquire into and correct errors made in the conveying or recording of the verdict, but no jurisdiction to correct errors that would involve the jury reconsidering its verdict or completing its deliberations. This approach had been used in civil but not criminal cases.

44 McIntyre J. disagreed with this position. He preferred to retain the bright-line rule that there is no jurisdiction to inquire into the alleged error and to alter the verdict post-discharge, regardless of the type or cause of error. He correctly stated that there are different policy concerns animating criminal law matters that are absent from civil matters, such as the interests of the state and the general public in the outcome of a criminal case (Head, supra, pp. 691-92). Although McIntyre J. did not go on to specify other relevant policy concerns, the courts below in this case articulated some of the policies animating the bright-line rule in Head, including: preserving the finality of the verdict and the certainty of the trial process; shielding jurors from post-trial harassment and promoting free and frank jury discussions by protecting the secrecy of their deliberations; and preventing the administration of justice from falling into disrepute by ensuring that the jury is not biased. These different considerations warrant a more stringent approach in the criminal law sphere.

45 The courts below held that Head was distinguishable from this case. I do not agree. In both Head and this appeal, the juries intended verdicts that were misunderstood by the trial judge. The jury in Head was not reconsidering its verdict; they had already decided to find the accused guilty on the included charge and not guilty on the charge as laid, and, similar to this case, failed to make their intentions successfully known prior to discharge. The internal-external distinction drawn by Weiler J.A. is unclear. In both cases, the courts had to determine what the intended verdict was, and in doing so, both courts had to inquire into what had been agreed upon in the jury room. Therefore, the error in this case was as “internal” as that in Head. Head cannot be distinguished from this appeal.

46 That being so, this appeal presents the Court with an opportunity to revisit the rule articulated by McIntyre J. in light of the approach now used in various common law jurisdictions. In my opinion, the common law rule in Head must be reformulated. Otherwise, apparent absurdities might result. For instance, where the error in recording the verdict is discovered immediately after the discharge of the jury but before the jurors have left the jury box, Head would say that the error could not be corrected.

47 I do not think that result can be right. Where the error is rapidly discovered after discharge but prior to the jury having separated or dispersed, many of the policy concerns supporting the rule in Head are not engaged. If the error is discovered immediately after discharge, then it would seem that the passage of time has been so slight that the finality of the verdict is not a pressing concern. An immediate discovery of the error would mean that the accused and the court could not have been under the illusion of the incorrectly recorded verdict for very long. An error made only in conveying or recording the verdict can be corrected without exposing the deliberations of the jury to undue scrutiny and subjecting them to the risk of post-trial harassment. If the jury has not dispersed beyond the jury box, then there is no realistic possibility that outside influences have tainted the jury. Thus, there is no danger to the administration of justice.

48 To the contrary, the administration of justice would be brought into disrepute by barring the court from correcting a recorded verdict where there is no perceptible injustice to the accused and no reasonable apprehension of bias. See V. Maric, Annotation to R. v. Burke (2001), 41 C.R. (5th) 135, at pp. 136-37, where it was observed:

Such rigidity jeopardizes the integrity of the jury system since it forces the court to tell the jurors that despite the fact that they have dutifully carried out the oath that they had sworn to uphold by listening to days of evidence and then rendering a unanimous and otherwise valid verdict, their decision must be ignored. . . . [T]he interests of the state and the general public would not be served . . . .

In those circumstances, not only would the policy issues used to justify the standard in Head not come into play, but the application of the rule in Head to this specific situation would run contrary to one of its own underlying policy concerns, namely the administration of justice.

49 If policy concerns animating the bright-line rule in Head are not at risk in the hypothetical facts outlined above (at para. 47), then there is no remaining rationale for prohibiting the trial judge from reconvening the jury to inquire into the alleged error, and possibly altering the verdict in those limited circumstances.

50 Head proposes a blanket approach that fails to respond to the nuances of specific contexts. Errors can vary from the jury wanting to reconsider its verdict, to a mere administrative slip, from the error being discovered immediately, to where it is not revealed until several days later, from the situation where the jurors have not yet stepped foot outside of the jury box or courtroom, to where the jurors have gone home. It appears obvious that an absolute rule with no exceptions is not attuned to modern, lengthy and expensive jury trials. Lamer J.’s separate reasons in Head are more responsive to variances in the facts noted above.

51 There is a wealth of authority supporting the adoption of a more refined and flexible analysis of this sort of problem, similar to the contextual approach suggested by Lamer J. The law in the United Kingdom has evolved to permit the clarification of a verdict in rare circumstances, after the jury has been discharged: R. v. Vodden (1853), Dears. 229, 169 E.R. 706; R. v. Cefia (1979), 21 S.A.S.R. 171 (S.C.); R. v. Andrews (1985), 82 Cr. App. R. 148 (C.A.); R. v. Follen, [1994] Crim. L.R. 225 (C.A.); R. v. Loumoli, [1995] 2 N.Z.L.R. 656 (C.A.); R. v. Maloney, [1996] 2 Cr. App. R. 303 (C.A.); R. v. Aylott, [1996] 2 Cr. App. R. 169 (C.A.); R. v. Z.A., Eng. C.A., March 8, 1999, unreported. Although American case law diverges widely on this issue, several courts have held that in certain situations verdicts can be revisited after the trial judge tells the jury that it is discharged: People v. Powell, 221 P.2d 117 (Cal. Dist. Ct. App. 1950); State v. Brandenburg, 120 A.2d 59 (N.J. County Ct. 1956); State v. Fornea, 140 So.2d 381 (La. 1962); Commonwealth v. Brown, 323 N.E.2d 902 (Mass. 1975); State v. Edwards, 552 P.2d 1095 (Wash. Ct. App. 1976); Webber v. State, 652 S.W.2d 781 (Tex. Crim. App. 1983); Burchett v. Commonwealth, 734 S.W.2d 818 (Ky. Ct. App. 1987); People v. McNeeley, 575 N.E.2d 926 (Ill. App. Ct. 1991); State v. Myers, 459 S.E.2d 304 (S.C. 1995); Montanez v. People, 966 P.2d 1035 (Colo. 1998); State v. Green, 995 S.W.2d 591 (Tenn. Crim. App. 1999).

52 These cases affirm the same general rule used by this Court in Head, i.e., that post-discharge, a trial judge is functus and has no authority to alter a recorded verdict. However, the case law in these jurisdictions also recognizes an exception to the general rule and permits a rare residual jurisdiction to inquire into the proper verdict in limited circumstances. In addition to the evolving jurisprudence, policy concerns warrant moving beyond Head to develop an exception to the general rule. A movement away from the rigidity of the Head rule is timely, given both the policy issues involved and the bulk of modern case law on this point. In my opinion, although Head forms the general rule that prohibits changes to a criminal verdict post-discharge, we should not foreclose the possibility of a limited and exceptional jurisdiction remaining with the trial judge to recall the jury for the purposes of inquiring into the alleged error, which may result in correction of the recorded verdict. The question then is under what circumstances should this jurisdiction be exercised.

B. Post-Discharge Jurisdiction to Conduct an Inquiry

53 The trial judge does not have the exceptional jurisdiction to inquire into the alleged error post-discharge when the alleged error is of the type such that its correction would involve the jury reconsidering its verdict or completing its deliberations: Head, supra, per Lamer J., at pp. 702-3. This limited residual jurisdiction may only be exercised where the errors do not “challenge the ‘validity’ of the verdict or the deliberation or mental processes of the jurors”: Martin v. State, 732 So.2d 847 (Miss. 1998), at p. 851. The rationale is that, once the jury has delivered its verdict, it should not be permitted to change its mind. If the error requires the jury post-discharge to reconsider its earlier conclusion or continue its cogitations on the matter, the error cannot be corrected, because the trial process has concluded and the judge is functus officio.

54 It is only where the error does not engage the deliberations of the jury that the exceptional jurisdiction may be exercised. It is unwise to characterize these types of errors as “clerical errors” or “accidental slips”, given the vague meaning of these terms. “Clerical errors” or “accidental slips”, as those terms are commonly understood, are administrative and may be corrected by the judge without recalling the jury: United States v. Dotson, 817 F.2d 1127 (5th Cir. 1987), modified upon rehearing 821 F.2d 1034 (5th Cir. 1987); Bricmont v. Mathieu (1987), 7 Q.A.C. 199, and Chandler v. Alberta Association of Architects, 1989 41 (SCC), [1989] 2 S.C.R. 848. These slips would be minor, such as correcting dates and duties of a similar nature. However, the error in the present appeal and similar cases necessitates a recall of the jury in order to confirm the true and unanimous verdict and the discrepancy with the announced and recorded verdict. Therefore, rather than use the label “clerical error”, it is preferable to say that the exceptional post-discharge jurisdiction can only be exercised where the correction of the error, although it requires the presence of the jury, does not require the jury to reconsider its verdict or complete its deliberations with a view to handing down additional verdicts: Head, supra, per Lamer J., at pp. 702-3.

55 If an irregularity has been detected and the jury has been discharged, the first question a trial judge should ask is whether the nature of the error is such that it would involve the jury reconsidering its verdict or completing its deliberations. If yes, then the general rule in Head applies and the trial judge is functus officio. If no, the trial judge has a narrow and exceptional jurisdiction to recall the jury for the purposes of inquiring into the alleged error.

C. The Nature of the Inquiry and its Limitations

56 If it is concluded that the error does not require reconsideration of the verdict by the jury, the trial judge has an exceptional jurisdiction to recall the jury for the purposes of conducting an inquiry. However, it remains to be determined whether the trial judge has remedial jurisdiction to change the originally recorded verdict. In some cases, it may simply be too late to change the verdict, even though the error does not involve reconsideration of the verdict. What then is the appropriate cutoff point beyond which the verdict cannot be altered? The post-discharge inquiry determines this cutoff point.

57 In the courts that permit post-discharge alteration of a verdict, the most common rule is to preclude clarification and correction of the verdict once the jurors have dispersed. The case law expresses its concern with dispersal in various ways, focussing at some times on whether the jurors have “separat[ed] from the presence and control of the trial court” and at others on the “opportunity for outside contact or influence”: Green, supra, at pp. 609 and 612.

58 Some courts have taken a rigid stand on dispersal, and concluded that even the briefest exit from the courtroom eliminates any possibility of changing the verdict: Brandenburg, supra; People v. Rushin, 194 N.W.2d 718 (Mich. Ct. App. 1971). For instance, in Green, supra, the jury was reassembled 58 seconds after their discharge; it was held to be too late for the trial judge to correct the verdict. This appears to me to be too narrow and rigid an approach to defining the circumstances under which the remedial jurisdiction to correct the verdict exists. If the “discharge” rule in Head is to be varied, it would make no sense to replace that inflexible test with one of equal rigidity.

59 In addition, there may be circumstances where it would be inappropriate for the trial judge to correct the verdict, even though the jury has not yet dispersed or separated. For instance, a discharged jury which has not yet left the jury box could be influenced or tainted by the strong reactions of the accused or the public gallery to the initially announced verdict (see Follen, supra). Therefore, as other factors impinge on whether the recorded verdict should be corrected, it seems that dispersal cannot be the sole determinative test for the trial judge’s remedial jurisdiction to correct the verdict.

60 The appropriate cutoff point beyond which the verdict cannot be corrected is found in the rationale underlying the “dispersal” case law. The reason for using dispersal as the outside limit is that once the jurors have dispersed, they are no longer functioning as a single, cohesive unit, and are exposed to outside contacts which may reasonably raise the apprehension of bias. The jury’s dispersal is relevant only insofar as it establishes the presence or absence of the reasonable apprehension of bias. Logically, if the jurors have left the controlled courtroom environment, then the more time that has passed between the announcement of the verdict and the reconvening of the jury, the more convincing the argument becomes for establishing a reasonable apprehension of bias. Similarly, “reasonable apprehension of bias” is the test used by trial judges in deciding whether to discharge a juror (s. 644(1) of the Criminal Code, R.S.C. 1985, c. C-46; R. v. Budai (2001), 154 C.C.C. (3d) 289, 2001 BCCA 349), and in determining whether there has been improper contact between a sequestered jury and the outside world (s. 647(1) and (2) of the Criminal Code; R. v. Cameron (1991), 1991 7182 (ON CA), 64 C.C.C. (3d) 96 (Ont. C.A.), leave to appeal refused, [1991] 3 S.C.R. x). In my opinion, if the dominant consideration driving the “dispersal” case law, the law on sequestering and the law on the discharge of jurors is the fear of a tainted or biased jury, or the appearance of unfairness, then this driving consideration should form the relevant test for the trial judge’s post-discharge jurisdiction to reconvene the jury and inquire into the alleged error.

61 The appropriate post-discharge inquiry into an alleged error is “reasonable apprehension of bias”. In determining whether this apprehension exists, the trial judge should examine the circumstances of the individual case. This contextual approach has a basis in the English case of Andrews, supra, at p. 154:

[T]he judge has a discretion whether to allow the alteration [to the verdict] to be made. In exercising that discretion he will, it goes without saying, take into account all the circumstances of the case; in particular the important considerations will be the length of time which has elapsed between the original verdict and the moment when the jury express their wish to alter it, the probable reason for the initial mistake, the necessity to ensure that justice is done not only to the defendant but also to the prosecution. The fact that the defendant has been discharged from custody is one of the factors but is not necessarily fatal to the judge’s discretion to alter the verdict to one of guilty.

As stated in R. v. S. (R.D.), 1997 324 (SCC), [1997] 3 S.C.R. 484, at para. 11, the precise phrasing of the test is not crucial, if the substance is plain. It is interchangeably expressed as a “reasonable apprehension”, “real likelihood” or “real danger” of bias, a “reasonable suspicion” of prejudice or taint, and so forth. Whatever the exact formulation of the test, the essence of the inquiry is the same; namely, the test is “what would an informed person, viewing the matter realistically and practically — and having thought the matter through — conclude”: Committee for Justice and Liberty v. National Energy Board, 1976 2 (SCC), [1978] 1 S.C.R. 369, at p. 394, per de Grandpré J.

62 The list of factors in Andrews does not purport to be exhaustive. All circumstances relevant to determining reasonable apprehension of bias in an individual case should be considered. By way of illustration, a trial judge may wish to consider whether the jury was exposed to an immediate and strong reaction to the recorded verdict by the public (as in Follen, supra) and/or the media.

63 Although the above-listed factors may be useful to take into account, I think that in many cases a significant element in establishing reasonable apprehension is dispersal. It is very important for the trial judge to examine the nature and extent of the jury’s dispersal. How much dispersal has occurred? Did the jury merely step back into the jury room, which is not accessible to the public? Or did the jury leave the courthouse and enter public spaces? The trial judge should take care to examine the length of time during which the jury was separated, because obviously the jury’s absence from the courtroom for a few seconds or minutes is often less indicative of a reasonable apprehension of bias, as compared to a more extended absence. If a jury has dispersed for a relatively lengthy period of time, such as several days, I would think that a trial judge would almost always conclude that this extensive degree of dispersal would raise a reasonable apprehension of bias. In my opinion, the evidence on dispersal will in many cases provide the most compelling evidence on which the reasonable apprehension of bias determination will turn. Dispersal may often be the linchpin of the reasonable apprehension of bias test.

64 Dispersal may in some cases be reparable but in the circumstances of this case, the wide media coverage and the public statements that it was “scary” that the judge could acquit the accused and then “change his mind”, and that the process was “a joke” (infra, at para. 84) is some evidence that the legal process was unsatisfactory. The public confidence in the jury system would be diminished and undermined by these circumstances.

65 As the test is the reasonable apprehension of bias, the concern in determining whether there has been dispersal is not whether there has been the opportunity for momentary outside influence, but whether there is a reasonable apprehension of bias. Similar to judges, juries should be presumed to be impartial. There are numerous procedural safeguards designed to ensure the impartiality of jurors (R. v. Barrow, 1987 11 (SCC), [1987] 2 S.C.R. 694, at p. 714). Section 647(1) of the Criminal Code ensures that the jury remains sequestered, stating that the trial judge can permit the jury to separate only prior to entering deliberations and rendering the verdict. Section 647(4) contemplates a mistrial if this rule is violated. In addition, there are other safeguards, such as jury selection, the process that requires jurors to swear an oath to give a true verdict according to the evidence, and the judge’s instructions in the form of his or her charge to the jury.

66 As a result, the presumption of impartiality should not be displaced by the slightest suspicion of outside influence. As an allegation of bias or outside contact calls into question the entire integrity of the jury and the overall administration of justice, the impartiality of the jury should only be impugned where on the evidence there is a reasonable apprehension of this allegation. The same standard also applies to judges or administrative tribunals when there are allegations of bias: S. (R.D.), supra, at paras. 109-15. Courts should be reluctant when presented with a whisper of bias to immediately conclude that the system ensuring impartial decisions has failed; consequently, we should not be quick to find a biased jury unless there is evidence to support a reasonable apprehension of the allegation.

67 Along those lines, the nature of the post-discharge inquiry into the alleged error does not require proof of actual bias of the jury. The test is “reasonable apprehension of bias”, not a “certainty of bias”. Not only are there practical concerns about the difficulty of ever being able to prove actual bias, but it is also unnecessary to say that it is important to maintain a high degree of public confidence in the integrity of the criminal justice system. Actual bias is not the appropriate test. Instead we should look again, albeit unimaginatively, to Lord Hewart’s famous admonition that it “is of fundamental importance that justice should not only be done, but should manifestly and undoubtedly be seen to be done”: R. v. Sussex Justices, Ex parte McCarthy, [1924] 1 K.B. 256, at p. 259.

68 In summary, the jurisprudence has shifted since Head. The general rule has been that after the jury in a criminal trial has been discharged, the trial judge is functus officio and lacks jurisdiction to reconvene the jury and inquire into the alleged error in the verdict.

69 However, an exception to the general rule exists where the error is one that does not require the jury to reconsider the verdict or continue its deliberations with a view to handing down additional verdicts; there the trial judge retains the limited jurisdiction to recall the jury and conduct a narrow inquiry into the alleged error. This type of error has sometimes been called “clerical”, but given the mixed meanings of such phrases as “clerical error” or “accidental slip”, the description is inadequate.

70 The first question that a trial judge must ask post-discharge is whether the error is one that requires reconsideration of the verdict. If it requires redeliberation of the verdict, there are no circumstances under which the judge will retain or otherwise possess jurisdiction to reconvene the jury and conduct an inquiry into the alleged error. The trial has concluded and the jury’s function is finished. The jury is not then permitted to change its mind. If the error does not require the jury to reconsider its verdict, then the trial judge possesses jurisdiction to conduct an inquiry, the nature of the inquiry being whether the facts of the case disclose a reasonable apprehension of bias. In determining whether this apprehension is raised, the trial judge must consider all of the relevant circumstances of the case, an important circumstance usually being the dispersal of the jury and its probable effect on the minds of reasonable members of the public. If the facts disclose a reasonable apprehension of bias, the trial judge’s remedial jurisdiction is necessarily limited, such that the trial judge cannot alter the originally recorded verdict. The scope of the trial judge’s remedial jurisdiction is discussed in detail in the next section.

D. Remedial Jurisdiction of the Trial Judge

71 As in some cases a trial judge may possess a residual jurisdiction to conduct an inquiry into the verdict post-discharge, it is logical that there exists a concomitant remedial jurisdiction. The question is the content of this remedial jurisdiction.

72 In my opinion, the trial judge is, as is frequently the case, in the best position to select the appropriate remedy based on the particular circumstances of the case. Having seen the trial through to the discharge of the jury, the trial judge is in a superior position vis-à-vis appellate courts to determine the probable reason for the error and to assess the resulting damage, if any. The trial judge should have the discretion to settle upon the proper remedy for the irregularity that occurred under his or her supervision.

73 The trial judge must of course be aware of the policies animating the general rule of “discharge is too late” and the rule permitting exceptional jurisdiction to inquire into certain errors. If the trial judge concludes that, in the circumstances, there exists a reasonable apprehension of bias, then he or she cannot correct the verdict because in order to reach that conclusion the trial judge has found that the jury has ceased to function as a single, cohesive unit, resulting in potential exposure to outside influence. To attempt to correct the verdict in the face of a reasonable apprehension or perception by the public that the system is biased or prejudiced is not allowed.

74 However, a conclusion that there is a reasonable apprehension of bias does not mean that the verdict as originally announced must always stand. Where the trial judge has concluded that the apprehension is raised and, on these circumstances, the verdict cannot be corrected, the trial judge retains the remedial jurisdiction and has the discretion to declare a mistrial. There are broad common law powers to declare a mistrial. Mistrials have been ordered or considered as a potential solution in a range of situations: where a jury member is discharged (R. v. Taillefer (1995), 1995 4592 (QC CA), 40 C.R. (4th) 287 (Que. C.A.), leave to appeal refused, [1996] 1 S.C.R. x; and R. v. Lessard (1992), 1992 3103 (QC CA), 74 C.C.C. (3d) 552, [1992] R.J.Q. 1205 (C.A.), leave to appeal refused, [1992] 3 S.C.R. vii); where inadmissible evidence is adduced during trial which might influence the jury (R. v. Woods (1989), 1989 7192 (ON CA), 49 C.C.C. (3d) 20 (Ont. C.A.), leave to appeal refused, [1990] 2 S.C.R. xii); where there is inadmissible communication between a witness and a juror causing prejudice (R. v. Martineau (1986), 1986 4730 (QC CA), 33 C.C.C. (3d) 573 (Que. C.A.)); where disclosure is made immediately prior to or during the trial (R. v. Antinello (1995), 1995 ABCA 117, 97 C.C.C. (3d) 126 (Alta. C.A.); R. v. T. (L.A.) (1993), 1993 3382 (ON CA), 84 C.C.C. (3d) 90 (Ont. C.A.)); and where the jury had already rendered a verdict but had not decided on the issue of mental disorder, making it impossible for the judge to enter the intended conviction without “taint” (R. v. Rondeau, [1998] O.J. No. 5759 (QL) (Gen. Div.)). The common theme running through this case law is the test of whether there is a [translation] “real danger” of prejudice to the accused or danger of a miscarriage of justice: Lessard, supra, at p. 562 C.C.C.

75 In declaring a mistrial, the trial judge therefore turns his or her mind to the question of whether a mistrial is needed to prevent a miscarriage of justice. This determination will necessarily involve an examination of the surrounding circumstances. Injustice to the accused is of particular concern, given that the state with all its resources acts as the singular antagonist of the individual accused in a criminal case. This factor should be balanced against other relevant factors, such as the seriousness of the offence, protection of the public and bringing the guilty to justice. It may be fitting to allow the announced verdict to stand where the period the accused has been at liberty and under the mistaken impression that he or she had been acquitted has been lengthy, and where the charge is not so egregious as to bring the administration of justice into disrepute. As has already been stated, the trial judge is in the best position to assess the circumstances of each individual case and select the most appropriate remedy.

76 Where, however, the trial judge concludes that, on the facts, a reasonable apprehension of bias is not raised, then the trial judge can and should correct the error and register the proper verdict. It is difficult to think that a mistrial would be available in those circumstances. There is no need to declare a mistrial where the trial judge has retained jurisdiction, if the fear of the reasonable apprehension is not engaged. Declaring a mistrial may lead to a new trial. Given the time, cost and resources expended on new trials, where a trial judge has found no reasonable apprehension of bias, simply correcting the error is a more efficient and fair result, without any prejudice to the accused or the state.

77 In summary, where the trial judge concludes, post-discharge, that the facts raise a reasonable apprehension of bias, then the trial judge should declare a mistrial, if that is the necessary remedy to prevent a miscarriage of justice. In making that order, the trial judge considers the rights of the accused, the public, and the effect of not ordering a mistrial on the administration of justice. On the other hand, if a mistrial is not necessary to prevent a miscarriage of justice, then the trial judge should uphold the verdict as given at trial. Where, however, the trial judge concludes that there is no reasonable apprehension of bias, the trial judge must correct the error in the verdict; a mistrial is not available as a remedy.

E. Application to this Appeal

78 Before commencing an analysis of the substance of the appeal, I observe, as an aside, that the method of polling prescribed in the court practice manual and used in this case was flawed. Polling the jury is an optional safeguard designed to ensure the true unanimity of the verdict. Merely asking the jurors to assent or dissent with “the verdict as announced” does not ensure that each juror actively confirms the verdict and the unanimity of the verdict. Where polling is requested, the procedure for polling should be for the court clerk to ask each juror to repeat the verdict that the foreman has announced, or for the court clerk to repeat the verdict to each juror and ask if he or she agrees.

79 Applying the principles outlined above to the facts of this case, the appeal should be allowed, and a new trial ordered. The general rule is that, post-discharge, a trial judge in a criminal matter is functus officio and does not have jurisdiction to conduct an inquiry into the alleged irregularity. However, the present case falls within the exception to the general rule. Namely, the error at issue in this case is one that, while necessitating the recall of the jury, does not involve the jury reconsidering its verdict. The alleged error in this case lay in the faulty transmission and recording of the verdict, rather than in the deliberations leading to the verdict. A residual jurisdiction to inquire into the irregularity exists where the error in question did not involve recalling the jury to continue or complete its deliberations.

80 As the error in this case is one that does not involve reconsideration or redeliberation of the verdict, the trial judge had post-discharge jurisdiction to conduct an inquiry into the alleged error. However, the trial judge did not conduct the proper inquiry. As I concluded above, the nature of the inquiry is directed to whether there is a reasonable apprehension of bias. The trial judge instead conducted an actual bias test. Furthermore, an assessment of the relevant circumstances by applying the contextual analysis outlined above (at paras. 60-66) leads me to conclude that the apprehension of bias is raised in this case. In particular, the extent of the jury’s dispersal provides the most compelling evidence establishing a reasonable apprehension of bias.

81 A careful analysis of the relevant circumstances of this case supports the conclusion that a reasonable apprehension of bias was raised. First, the length of time which elapsed between the delivery of the recorded verdict and the point at which the jury was reconvened was substantial. Although a court officer discovered the potential error soon after the delivery of the original verdict, the jury was not reassembled as a single entity until the next day. Second, the accused Burke was discharged from custody.

82 Third, and perhaps most telling, the nature and the scope of the jury’s dispersal in this case was extensive. This was not a case where the jurors had momentarily separated, or where the jury had not strayed outside of or far from the courtroom, as in Aylott, supra, Brandenburg, supra, Edwards, supra, and Webber, supra. Instead, by the time the court had discovered the error, the entire jury had left the courtroom, and some had left the courthouse. The jurors were free until the next day to mingle with members of the public. Two jurors were found in the courthouse parking lot. The other ten had left the vicinity of the courthouse, and gone home or elsewhere. Only eight of these jurors were contacted directly, when the court officer telephoned them at home later that day. The whereabouts of two of the jurors that day remains unknown, because the court officer making the phone calls could not reach them. The jury was not together as a cohesive unit until the next day when, in the absence of the accused, they confirmed the guilty verdict. The combined circumstances indicate that the nature of the dispersal in this case was wide and uncontrolled. The relatively extensive degree of separation of the jurors in this case is evidence capable of establishing a reasonable apprehension of bias.

83 Other circumstances also shed light on the nature and extent of the jury’s dispersal. Although the foreman confirmed the verdict at the first inquiry, the other juror found in the parking lot was not then questioned as to the intended verdict. He was simply told to return the next day. In addition, when the court officer called the jurors at their homes, she did not have a prepared response to the jurors’ questions about why they were being asked to return to the courtroom the next day. She asked one of the jurors on the phone “off the record” what the verdict had been. During polling, one juror assented to the verdict although she later testified that she had not heard the verdict as announced. There is evidence that defence counsel had spoken with two of the jurors in the hallway after the jury was discharged. Two constables at the court had discussed the verdict with the foreman and another juror before escorting them back to the courthouse for the first inquiry.

84 Fourth, not only were the jurors exposed to the reaction of the public to the recorded verdict during the period when they were absent from the courtroom, but they would also have been exposed to the potentially prejudicial media coverage of this case which appeared before and after the temporary publication ban. The police issued a press release explaining what had occurred and citing the reason for the error. Two newspaper articles were also published, one dated the day of the second inquiry. This article was quite likely available that morning, before the jurors confirmed their intended verdict. Indeed there was a possibility that there were media reports of the incident the night prior to the September 19 inquiry. For instance, juror number one testified on September 19, that he had “information” that the incident “has already been leaked to the press this morning or last night”. The second article was dated the day that the third and final inquiry began. The report described the joyous reaction of the accused to the verdict as announced, which he thought was “not guilty”, and also referred to “news reports” that the accused had “bolt[ed]” from the courtroom. The articles called the inquiries conducted by the trial judge “unprecedented” and quoted the defence counsel as saying that it is “scary” that the judge could acquit the accused and then “change his mind”. One article also quoted a police source calling the whole incident “a joke”.

85 While the trial judge concluded that there was no “air of reality” to the proposition that there was or may have been bias or taint, his conclusion was, with respect, incomplete and resulted in an error of law. The inquiry he conducted did not constitute a full and proper investigation of the reasonable apprehension of bias. It is true that all of the jurors testified that no one had said anything to them during their dispersal that influenced their testimony on the verdict. Two jurors testified that they had read and heard things in the media but that they were not influenced by them. One juror started to allude to something he had heard in the media but was prevented by the trial judge from completing his testimony on that point.

86 Although the jurors all testified that they had not been influenced by anything in the media or by any conversations with the public, their testimony, as elicited, is of little use. As previously stated, the relevant inquiry is not whether the jurors were actually biased, but whether there was a reasonable apprehension in the minds of right-thinking, properly informed members of the public that the jurors might have been biased or influenced. As has already been stated, in applying the reasonable apprehension of bias test, the trial judge should consider all factors that might lead to the apprehension of bias. Of necessity, the most significant consideration will be the length of time that the jury has been dispersed. In order to maintain the confidence of the public in the important function of the jury, its verdict must be unanimous and bring finality to the trial.

87 While the trial judge is in most cases in the best position to assess the consequences of the length of dispersal and other relevant factors, he or she must be careful not to apply a subjective test. An example of the potential mischief created by the subjective test is where the recorded verdict is corrected weeks or months after the jury has been allowed to separate. In my opinion, the period of separation does not have to be very long before the public would grow uneasy about the efficacy of the jury system, and the apprehension of bias then arises.

88 I think the present appeal amply illustrates this point. The trial judge concluded that there was no apprehension of taint or bias. He did not consider the nature of the dispersal in this case. The jury had gone their separate ways in the large metropolitan areas of Toronto; they were potentially exposed to a great amount of media coverage. While it is true that each juror may correctly have believed they had not been influenced or tainted, the inquiry should not end there. The question the trial judge should have asked himself is not whether any juror felt tainted, but rather would a reasonably informed person conclude that in the period of dispersal spanning two days, the apprehension of bias could occur. Instead, the trial judge’s question of “Did you read or hear anything that influenced you?” was essentially asking the jury member to come to their own conclusion about whether he or she was biased. A jury member faced with that question could have thought that although he had read the news reports and spoken with his neighbours about the case, he did not think that they had influenced him. Therefore, he would answer “no” to the question.

89 In short, we do not know how many of the jurors actually spoke with people about the case, or how many of the jurors read or heard reactions or items in the news about the case, because the judge’s style of questioning permitted the jurors to draw the conclusions themselves about whether they were influenced. The trial judge should not have left the jurors to perform internal evaluations on bias. Jurors do not know the legal test of reasonable apprehension of taint. A juror is not responsible for canvassing what he has heard or read in the media, and then conducting the legal inquiry of whether a reasonable person would conclude that there was bias. A juror is not in a position where he can detach himself and objectively conclude that he was not biased, even though he, the juror, adamantly and honestly believed that he himself was not biased. Rather, this legal test is something that the trial judge is responsible for applying, and failed to do in this case, because he focussed solely on whether the jurors were actually influenced, rather than on the correct question, which is whether a reasonable person would apprehend influence. The distinction is significant, for it demarcates the difference between an actual bias inquiry and a reasonable apprehension of bias inquiry. The reasonable apprehension of bias inquiry is much less forgiving, for justice must manifestly be perceived to have been done.

90 A preferable way to conduct the inquiry about reasonable apprehension of bias would have been for the trial judge to ask whether the juror read or heard anything in the media or spoke to anyone about the case, rather than whether they had read anything that had influenced them. This is desirable because it establishes the existence of any media coverage, and places the trial judge in a position to assess the credibility of the jurors regarding potential bias, rather than requiring the juror to do it for him- or herself. Indeed, in some cases the trial judge may not even need to ask the jurors themselves if they have read or heard anything in the media or spoken to anyone, if the press coverage of the case has been particularly prominent. As it stands in this case, although we know that at least two and possibly three jurors heard things in the media about the case, we do not know if any other jurors were exposed to reactions by the public and the media. Given the inflammatory nature of the news reporting, which at least a few of the jurors admitted to having read, and the fact that the jurors were separated for almost 24 hours and had the opportunity to mingle with the public, in my opinion, the evidence supports a finding of a reasonable apprehension of bias.

91 The error in this case is such that it was not a matter of the jury reconsidering its verdict or completing their deliberations. Therefore, the trial judge had post-discharge jurisdiction to reconvene the jury and conduct a limited inquiry into the alleged error. However, the trial judge in this case did not conduct the proper and relevant inquiry, namely whether there was a reasonable apprehension of bias. Given the combination of circumstances in this case, there is in my opinion no question that the facts, particularly the jury’s extensive and extended dispersal, establish a reasonable apprehension of bias. As a result, the trial judge did not have remedial jurisdiction to correct the verdict. He was, therefore, in error when he changed the verdict and registered a conviction. The only two remedies available to the trial judge, given that the facts of this case disclosed a reasonable apprehension of bias, were to let the original acquittal stand, or to declare a mistrial.

92 The appropriate remedy in these circumstances would have been for the trial judge to use the broad common law power to declare a mistrial. A mistrial is necessary to prevent a miscarriage of justice. It is not an injustice to the accused in these circumstances to declare a mistrial and order a new trial. While the accused was discharged, he was notified of the error by his defence counsel the same day that he was mistakenly acquitted. The time of the notification will have some effect in the period he was at liberty. It is obvious that being notified immediately imposes minimal harm as opposed to being notified days or weeks later. The assessment of the effect of the time the accused was at liberty is one more factor that should be left to the trial judge.

93 There is no doubt that the public would wonder what had happened to the administration of justice in this case. A verdict of not guilty had been announced, the accused was released, the jury discharged, and left to mingle with the community and wander the streets of one of Canada's largest metropolitan areas, where they were subject to media comments and questions from the curious. This uncertainty as to what had transpired continued on and off for two days. Then, inexplicably, to the public at least, there were three additional hearings, one of which was held in the absence of the accused. The most natural impression would be that something had gone wrong. This is an impression that should not be left with the public, as it would erode public faith in the jury and in the general propriety of criminal verdicts. The clumsy sequence of events in this case cannot help but result in confusion in the community, as well as skepticism about the efficacy of the jury system. The importance of the public's confidence in the Canadian jury system and in the administration of justice cannot be overstated. This public trust, respect and acceptance if eroded will be at great cost to the effective operation of the criminal justice system. Therefore, the acknowledgment of the errors in this case and the order for a new trial will hopefully assuage these concerns and go towards restoring public confidence in our jurisprudence and in the function of the jury in particular. The memory of recent miscarriages of justice suffered by David Milgaard, Donald Marshall and Guy Paul Morin, to name only three, have raised concerns about the efficacy of our system, where years after conviction and imprisonment of an accused, the convicted person has been cleared of the crime. A truncated proceeding such as that which occurred here will only heighten that concern.

94 The appeal is allowed and a new trial is ordered.

The following are the reasons delivered by

95 Arbour J. — The central issue in this appeal is whether a trial judge has jurisdiction to correct an error in the recording of a jury’s verdict after the jury has been discharged. This involves, in turn, the preliminary question of whether the trial judge even has jurisdiction to inquire into the verdict to determine if it was correctly recorded. I agree with my colleague Major J. that jurisdiction to inquire into and correct an error exists. However, I find that this jurisdiction is based on the fact that in such a case the trial has not, contrary to appearances, been properly concluded.

96 The common law rule that after the discharge of the jury the trial judge lacks jurisdiction to alter a recorded verdict was enunciated by this Court in R. v. Head, 1986 8 (SCC), [1986] 2 S.C.R. 684. McIntyre J., for the majority of the Court, held that the trial judge only had the power to make inquiries relating to the true nature of the verdict prior to the discharge of the jury. Once the jury was discharged, the court lacked jurisdiction to make any correction of the recorded verdict. The Court concluded that once the jury has been discharged, the trial court is functus (p. 694).

97 I agree with Major J. that Head cannot be distinguished from this appeal. I also agree that this is an appropriate case in which to re-examine the rule in Head and formulate a new rule that better accommodates the various interests and policy concerns at stake.

98 In Head, discharge was the point after which it was “too late” to alter or amend a recorded verdict because the Court was focussed on jurisdiction. Major J. does not address directly this jurisdictional concern that plagued the Court in Head. Rather, he identifies a rare residual jurisdiction remaining with a trial judge to conduct an inquiry into the jury’s intended verdict and possibly correct an error post-discharge, provided it is not necessary for the jury to reconsider its verdict or deliberate further. This exceptional residual jurisdiction is to be exercised only when there is no reasonable apprehension of taint or bias of the jury.

99 In the normal course of a trial, jurors are sequestered once they begin their deliberations until they return a verdict: s. 647 of the Criminal Code, R.S.C. 1985, c. C-46. I believe that this case is best characterized within this context. The jury in this appeal communicated its intended verdict to the court. The court recorded a different verdict. It was brought to the trial judge’s attention that an incorrect verdict might have been recorded. It is the duty of the trial judge to ensure that the court records the intended verdict of the jury upon its return from deliberations (Head, supra, at p. 688). A verdict other than the intended unanimous verdict of the jury is a nullity. If the trial judge has a reasonable concern that the verdict might be a nullity, the trial should resume as if the verdict had not been rendered. The trial judge should determine whether the recorded verdict was in fact null and void and if not, the verdict should stand. If the trial judge concludes that the recorded verdict was a nullity, he or she should set it aside and the trial should resume. In such a case, the provisions of s. 647 of the Code come into play and the question is whether a jury which should have remained sequestered has dispersed, and if so, what the consequences of dispersal ought to be. In these circumstances, I agree that the appropriate test is, as enunciated by Major J., “reasonable apprehension” of taint or bias. If the trial judge determines that there is no reasonable apprehension of bias, then there is no perceived threat to the impartiality of the jury and the jury is in the same position as it was in prior to the court recording the verdict.

100 Under this approach, I would conclude, contrary to Major J., that, if there is no reasonable apprehension of bias, the jury is free to continue deliberations if necessary. Assume, for example, that there is a trial on an indictment containing two counts of murder. The jury returns a verdict in count one and finds the accused not guilty. However, nothing is said by anyone about count two involving a different victim. Before the jury is discharged, it is clear that the trial judge may ask the jury whether they have a verdict on count two. If the jury indicates that they simply forgot to deliberate on count two, there is no impediment to sending them back to continue their deliberations on that count. I see no difference if the same scenario were to occur after the jury has been discharged but prior to the point in time at which there is a reasonable apprehension of bias or taint. In such a case, I would see no impediment, in principle, to asking the jury to resume deliberations on count two. In other words, the critical moment should be the point at which there is a reasonable apprehension of bias, as identified by my colleague, and not the jury’s discharge.

101 When the trial judge finds that there has been a reasonable apprehension of bias, the situation is the same as if the jury had not remained properly sequestered pending the completion of its deliberations and the rendering of its verdict. In such a case, the trial judge’s only option is to declare a mistrial and to order a new trial “on such terms as justice may require” (s. 647(4)(b) of the Code).

102 I concur with the conclusion of Major J. that, in this case, the extent of the jury’s dispersal established a reasonable apprehension of bias and that the appropriate remedy is to order a new trial.

Appeal allowed, McLachlin C.J. and L’Heureux-Dubé, Gonthier and Bastarache JJ. dissenting.

Solicitors for the appellant: Pinkofsky Lockyer, Toronto.

Solicitor for the respondent: The Attorney General for Ontario, Toronto.