R. v. Hibbert, [2002] 2 S.C.R. 445, 2002 SCC 39

Kenneth Roydon Hibbert Appellant

v.

Her Majesty The Queen Respondent

Indexed as: R. v. Hibbert

Neutral citation: 2002 SCC 39.

File No.: 28021.

2001: October 10; 2002: April 25.

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 british columbia

Criminal law – Evidence – Identification – Alibi – Accused convicted of attempted murder – Trial judge erring in instructing jury that they could infer guilt from disbelieved alibi – Whether Court of Appeal erred in applying curative proviso – Whether Court of Appeal erred in finding no reversible error in trial judge’s instructions on issue of identification — Criminal Code, R.S.C. 1985, c. C‑46, s. 686(1)(b)(iii).

A real estate agent was a victim of a brutal attack while she was holding an open house. In the course of touring the house, a man provided the victim with numerous personal details. The victim took the man into the garage. When she went to plug in a light, the man struck her from behind, beat her, and strangled her until she was unconscious. The Crown’s case against the accused was based largely on circumstantial evidence and essentially rested on the victim’s knowledge of the personal details of the accused’s life; on the in‑court identification of the accused by the victim and a neighbour who saw the assailant leaving the residence; on DNA evidence that linked both the accused and the victim to a cap found hanging in a tree on the assailant’s escape route; and on various other pieces of circumstantial evidence that cumulatively point to the accused as the assailant. In his first trial, the accused was convicted of attempted murder by a jury, but the Court of Appeal set aside the conviction and ordered a new trial. In a second jury trial, the subject of this appeal, the accused was again convicted. He appealed that conviction, again alleging weaknesses in identification evidence and errors in the instructions to the jury about the evidence of identification. In addition, he alleged that the trial judge erred in telling the jury they could infer guilt from a disbelieved alibi when there was no extraneous evidence of fabrication or contrivance. The Crown conceded that the trial judge’s instructions on alibi were in error but the Court of Appeal applied the curative proviso of s. 686(1)(b)(iii) of the Criminal Code and dismissed the appeal.

Held (L’Heureux‑Dubé and Bastarache JJ. dissenting): The appeal should be allowed and a new trial ordered.

Per McLachlin C.J. and Gonthier, Iacobucci, Major, Binnie, Arbour and LeBel JJ.: The trial judge’s directions to the jury on identification evidence were not so deficient as to constitute an error of law. In the circumstances of this case, however, the trial judge should have cautioned the jury more strongly that the identification of the accused in court was highly problematic as direct reliable identification of the perpetrator of the offence. It is important to remember that the danger associated with eyewitness in‑court identification is that it is deceptively credible, largely because it is honest and sincere. The dramatic impact of the identification taking place in court, before the jury, can aggravate the distorted value that the jury may place on it. The instruction to the effect that such identification should be accorded “little weight” does not go far enough to displace the danger that the jury could still give it weight that it does not deserve. In this particular case, it would have been prudent to emphasize for the benefit of the jury the very weak link between the confidence level of a witness and the accuracy of that witness. Moreover, it should also have been stressed that the impact of the victim having seen the accused arrested by the police as her alleged assailant could not be undone. Nor could she be expected to divorce her previous recollection of her assailant from the mental image that she formed after having seen the accused on television.

The Crown properly conceded that the trial judge’s instructions on alibi were in error in the absence of evidence that the accused was implicated in putting forward a fabricated alibi. The law with respect to the rejection of a defence of alibi may be summarized as follows. In the absence of evidence of concoction (deliberate fabrication) an alibi that is disbelieved has no evidentiary value. A disbelieved alibi is insufficient to support an inference of concoction or deliberate fabrication. There must be other evidence from which a reasonable jury could conclude that the alibi was deliberately fabricated and that the accused was involved in that attempt to mislead the jury. It is the attempt to deceive, and not the failed alibi, that supports an inference of consciousness of guilt. In appropriate cases, for instance if there were multiple accused, the jury should be instructed that the fabricated alibi may be used to place the accused at the scene of the crime, but may fall short of directly implicating him in its commission. When there is evidence that an alibi was fabricated, at the instigation or with the knowledge and approval of the accused, that evidence may be used by the jury to support an inference of consciousness of guilt. In cases where such an inference is available, the jury should be instructed that it may, not must, be drawn. A fabricated alibi is not conclusive evidence of guilt.

The curative proviso cannot be used to overcome the erroneous instructions on the defence of alibi in this case. The error was not a trivial error nor one that would be unlikely to have had any effect on the verdict. The trial judge’s erroneous instructions implied that the jury could find evidence of concoction allowing them to infer that the accused was guilty. This is a serious error that provided the jury with a direct route to guilt. Despite the circumstantial evidence that points to the accused, it cannot be confidently said that a conviction is a foregone conclusion in the sense that any other reasonable jury would inevitably convict.

Per L’Heureux‑Dubé and Bastarache JJ. (dissenting): The trial judge’s directions to the jury on identification evidence did not constitute an error of law. Moreover, her warning in this regard was adequate. The trial judge’s direction to the jury pointed out the general and specific problems with the identification evidence while at the same time respecting the role of the jury to weigh the evidence that was properly before it. With regard to the in‑court identification of the accused, the trial judge clearly explained to the jury the possibility that the witnesses recognized the accused from their memory of the photographic line‑up or newscast rather than from the scene of the crime. She instructed the jury that for this reason the in‑court identification was entitled to little weight. This instruction was appropriate because it pointed out to the jury the weaknesses associated with the evidence without encroaching on the jury’s essential function as the trier of fact. Whether the victim was able to identify the accused as her assailant, or whether she was merely identifying him as the man she saw being arrested and at previous hearings, was ultimately a question for the jury to answer.

The question to be asked when considering the application of the proviso is whether there is any reasonable possibility that the verdict would have been different had the error at issue not been made. In answering this question, the strength of the Crown’s case, otherwise stated as “the legally admissible evidence untainted by the error”, is one factor to be considered, along with the seriousness of the error in question and the effect it likely had upon the jury’s inference‑drawing process. The Crown’s case was a strong one. More importantly, the trial judge mitigated the seriousness of the error contained in the then standard jury instructions by tailoring the instruction to the particular circumstances of the case. The trial judge instructed the jury that they could draw an inference of guilt if they concluded that the alibi put forward by the accused was false and was put forward to deceive them, an instruction that was erroneous in the absence of independent evidence of fabrication or concoction. She nevertheless went on to remind the jury that the Crown did not suggest concoction or fabrication. The trial judge in effect neutralized the impact of the error by immediately withdrawing it from the jury’s consideration. When the error in the trial judge’s instructions on alibi is considered in the context of the charge as a whole and against the background of the entire trial, it is evident that there was no reasonable possibility that the error would have impacted upon the reasoning process of the jury. Given the trial judge’s charge in its entirety and the circumstances of this particular case, the danger of a miscarriage of justice arising from the trial judge’s misdirection on alibi was more theoretical than real. The test for the application of the proviso is not whether it is in theory possible that the verdict would have been different had the error not be made, but rather whether there is a reasonable possibility that the verdict would have been different taking into account the background of the whole trial. Applying the proviso in the circumstances of this trial even though it was not applied in the first trial creates no inconsistency. Having determined that there was no real possibility that the error in question affected the reasoning process of the jury, it is only logical to conclude that the accused was convicted because the evidence against him was overwhelming.

Cases Cited

By Arbour J.

Referred to: R. v. Tessier (1997), 1997 CanLII 3475 (BC CA), 113 C.C.C. (3d) 538; R. v. Davison (1974), 1974 CanLII 787 (ON CA), 20 C.C.C. (2d) 424; R. v. Witter (1996), 1996 CanLII 4005 (ON CA), 105 C.C.C. (3d) 44; R. v. Coutts (1998), 1998 CanLII 4212 (ON CA), 16 C.R. (5th) 240, leave to appeal dismissed, [1999] 1 S.C.R. xii; R. v. Krishantharajah (1999), 1999 CanLII 3701 (ON CA), 133 C.C.C. (3d) 157; R. v. Carey (1996), 1996 CanLII 6039 (QC CA), 113 C.C.C. (3d) 74; R. v. Michaud, 1996 CanLII 211 (SCC), [1996] 2 S.C.R. 458; R. v. Hibbert (1996), 1996 CanLII 2174 (BC CA), 78 B.C.A.C. 277; Colpitts v. The Queen, 1965 CanLII 2 (SCC), [1965] S.C.R. 739; Wildman v. The Queen, 1984 CanLII 82 (SCC), [1984] 2 S.C.R. 311; R. v. B. (F.F.), 1993 CanLII 167 (SCC), [1993] 1 S.C.R. 697; R. v. Bevan, 1993 CanLII 101 (SCC), [1993] 2 S.C.R. 599; R. v. Charlebois, [2000] 2 S.C.R. 674, 2000 SCC 53; R. v. Khan, [2001] 3 S.C.R. 823, 2001 SCC 86.

By Bastarache J. (dissenting)

Mezzo v. The Queen, 1986 CanLII 16 (SCC), [1986] 1 S.C.R. 802; R. v. Turnbull, [1976] 3 All E.R. 549; Canning v. The Queen, 1986 CanLII 20 (SCC), [1986] 1 S.C.R. 991; R. v. Gagnon (2000), 2000 CanLII 16863 (ON CA), 136 O.A.C. 116; R. v. Charlebois, [2000] 2 S.C.R. 674, 2000 SCC 53; R. v. Edwardson (1993), 1993 CanLII 1420 (BC CA), 77 B.C.L.R. (2d) 362; R. v. Bevan, 1993 CanLII 101 (SCC), [1993] 2 S.C.R. 599; R. v. Jacquard, 1997 CanLII 374 (SCC), [1997] 1 S.C.R. 314; R. v. Jolivet, [2000] 1 S.C.R. 751, 2000 SCC 29; R. v. Khan, [2001] 3 S.C.R. 823, 2001 SCC 86; Mahoney v. The Queen, 1982 CanLII 21 (SCC), [1982] 1 S.C.R. 834; R. v. Tessier (1997), 1997 CanLII 3475 (BC CA), 113 C.C.C. (3d) 538; R. v. Krishantharajah (1999), 1999 CanLII 3701 (ON CA), 133 C.C.C. (3d) 157; R. v. Demeter (1975), 1975 CanLII 685 (ON CA), 25 C.C.C. (2d) 417; R. v. Hibbert (1996), 1996 CanLII 2174 (BC CA), 78 B.C.A.C. 277.

Statutes and Regulations Cited

Criminal Code, R.S.C. 1985, c. C‑46, s. 686(1)(b)(iii) [am. 1991, c. 43, s. 9 (Sch., item 8)].

Authors Cited

Cory, Peter deCarteret. The Inquiry Regarding Thomas Sophonow: The Investigation, Prosecution and Consideration of Entitlement to Compensation. Winnipeg: Manitoba Justice, 2001.

Ferguson, Gerry A., and John C. Bouck. Canadian Criminal Jury Instructions, vol. 2, 2nd ed. Vancouver: Continuing Legal Education Society of British Columbia, 1989 (loose-leaf updated November 1993).

APPEAL from a judgment of the British Columbia Court of Appeal (2000), 134 B.C.A.C. 281, [2000] B.C.J. No. 392 (QL), 2000 BCCA 144, dismissing the accused’s appeal from his conviction for attempted murder. Appeal allowed and new trial ordered, L’Heureux‑Dubé and Bastarache JJ. dissenting.

J. M. Peter Firestone and Catherine Tyhurst, for the appellant.

Kate Ker, for the respondent.

The judgment of McLachlin C.J. and Gonthier, Iacobucci, Major, Binnie, Arbour and LeBel JJ. was delivered by

Arbour J. —

I. Introduction

[1] The appellant appeals his conviction by a jury on a charge of attempted murder. The principal issue in this appeal is the application of s. 686(1)(b)(iii) of the Criminal Code, R.S.C. 1985, c. C-46 (the proviso). The Crown concedes that the trial judge erred in her instructions to the jury on the defence of alibi. The question is whether that error, in itself, or combined with other errors alleged by the defence, necessitates a new trial, or whether, as the Court of Appeal found, the Crown’s case was so strong that the jury would inevitably have convicted had the error not been made.

[2] In my view, s. 686(1)(b)(iii) cannot be applied in this case to preserve the conviction and the verdict of the jury must be set aside. As is common in these types of cases, the evidence must be reviewed in some detail in order to assess the nature and likely impact of the errors in the charge.

II. Overview

[3] On October 24, 1993, Janet McLeod, a real estate agent, was a victim of a brutal attack while she was holding an open house at 151 Campbell Street in Duncan, British Columbia. A man she described as wearing a felt-type English driving hat (described in other evidence as an “Andy Capp” style hat) arrived by himself to visit the house at approximately 3:00 p.m. In the course of touring the house, the man provided Mrs. McLeod with numerous personal details, including that he lived nearby on Trunk Road with his wife and two daughters aged 13 to 15, that he had moved to Duncan from Ontario, and that he was waiting for a military pension to be issued. Mrs. McLeod took the man into the garage to explain how it could be converted into an extra bedroom. When she went to plug in a light, the man struck her from behind, beat her, and strangled her until she was unconscious. Mrs. McLeod was severely beaten with a blunt object swung with tremendous force. The evidence suggested that after the beating, the assailant left the garage to clean himself in the house, and then returned and tied an electrical cord around her neck. She was found in a pool of blood by neighbours who heard the commotion. The attack was vicious and Mrs. McLeod’s injuries were life threatening.

[4] The appellant, Mr. Hibbert, was interviewed by the police on the morning of November 3, 1993. He was arrested that same day and a warrant to search his residence was executed. He was released and then arrested a second time on December 14, 1993. He has been in custody since that time.

[5] On December 20, 1994, Mr. Hibbert was convicted of attempted murder by a jury in a trial presided by Hutchison J. of the Supreme Court of British Columbia. Mr. Hibbert appealed the conviction on the grounds that the trial judge failed to charge the jury properly on weaknesses in the identification evidence and on the significance of the alibi evidence. The appeal was allowed, and the British Columbia Court of Appeal declined to apply s. 686(1)(b)(iii) of the Criminal Code to cure the error. The court set aside the conviction and ordered a new trial.

[6] In a second jury trial before Dorgan J. in March of 1997, Mr. Hibbert was again convicted. He appealed that conviction, again alleging weaknesses in identification evidence and errors in the instructions to the jury about the evidence of identification. In addition, he alleged that the trial judge erred in telling the jury they could infer guilt from a disbelieved alibi when there was no extraneous evidence of fabrication or contrivance. The Crown conceded that the trial judge’s instructions on the latter point were in error but the Court of Appeal applied the curative provision of s. 686(1)(b)(iii) and dismissed the appeal: (2000), 134 B.C.A.C. 281, 2000 BCCA 144. Mr. Hibbert appeals that decision to this Court on the grounds that the Court of Appeal erred in failing to find a reversible error in the trial judge’s charge to the jury on identification, and, having found that there was an error in the alibi jury instruction, in deciding to apply the curative proviso.

III. The Evidence

[7] The Crown’s case was largely based on circumstantial evidence. No motive was ever advanced for this unprovoked and sudden attack. In essence, the strength of the Crown’s case rested on the victim’s knowledge of the personal details of the appellant’s life; on the in-court identification of the appellant by Mrs. McLeod and a neighbour who saw the assailant leaving the residence; on DNA evidence that linked both the appellant and the victim to a cap found hanging in a tree on the assailant’s escape route; and on various other pieces of circumstantial evidence that cumulatively point to the appellant as the assailant. However, Mr. Hibbert put forward explanations, some stronger than others, for each piece of evidence presented by the Crown. I will now review this evidence in more detail.

A. Identification Evidence

(1) The Victim — Janet McLeod

[8] Mrs. McLeod could barely speak when she arrived at the hospital on the afternoon of October 24th. She related a description of the man who assaulted her and the personal details he shared with her to the attending doctor. A nurse in attendance recorded her description on the hospital record. The description was of a 55-year-old male, five feet eight inches tall, 180 pounds, with grey hair and a grey sweater who told Mrs. McLeod that he had a wife, two girls, lived on or near Trunk Road, came from Ontario, and had a military background. Mrs. McLeod also provided a physical description of her assailant to a constable who came to see her in the treatment room on the afternoon of October 24, 1993. She described her assailant as a 55-year-old Caucasian male without an accent, five feet eight inches tall, 180 pounds with grey hair, bushy eyebrows with some grey in them, a brown felt-type English driving hat, brown sweater with possibly a white shirt underneath, and dark brown pants. Mrs. McLeod also gave a physical description to another constable on October 28, 1993 describing the assailant as a male 55 years or older, five feet eight inches tall, stocky build, round face, wearing a brown “Andy Capp” style hat, large hands and fingers, brown jacket or sweater, with no recollection of whether he had an accent. On October 31, 1993, soon after the bandages were removed from her eyes, Mrs. McLeod and a police sketch artist prepared a composite sketch. She rated the composite a 6 or 7 out of 10 for accuracy.

[9] On November 4, 1993, while still in hospital, Mrs. McLeod was shown an eight-person photo line-up which included the appellant. Mr. Hibbert’s photograph was number 7. Mrs. McLeod did not positively identify anyone and she wrote nothing on her ballot sheet. While viewing the line up, she expressed that she thought that she had seen the person in picture number 7 before, and said “I feel like I’ve had a conversation with him”. She also indicated that it might not have been on the day of the attack, and could have been some time before. On the day of the photo line-up, Mrs. McLeod was tired, in a great deal of pain, and under the effects of morphine.

[10] At the time of Mr. Hibbert’s second arrest on December 14, 1993, a local news station showed footage of him being escorted by sheriffs into the courthouse in handcuffs.

[11] After being alerted by a television news station that there was an arrest in her case, Mrs. McLeod and her husband watched the early evening television news footage which showed Mr. Hibbert at the courthouse in police custody. Mrs. McLeod asked her husband to tape the nightly news segment with the same footage. The next morning they watched the tape together and freeze-framed and studied the shot of Mr. Hibbert under arrest. That same morning, one of the police constables on the case called the McLeods to update them on the status of the case and learned they had seen and taped a news film clip of the arrest. Mrs. McLeod told him they had “got the right man”. The constable came to the McLeods’ house and watched the tape with them. He then seized the tape.

[12] Mrs. McLeod identified Mr. Hibbert in court, both at the preliminary hearing and at the first trial, as her assailant. At the second trial, she was asked if she was able to recognize the man who assaulted her as being a person in the courtroom. Crown counsel added: “It’s important for the jury to know, Mrs. McLeod, whether or not you are able to identify your attacker and not someone that you have seen subsequently, do you understand?” Mrs. McLeod identified the appellant.

(2) Witness — Carol Baker

[13] Carol Baker lived immediately next door to the open house residence that was the site of the attack. At approximately 3:00 p.m., while outside, she heard a voice calling for help from the direction of the open house and what sounded like thumping on the garage door. She went inside her own house to investigate further by looking from her house window into the garage window at 151 Campbell. She made eye contact with a man in the garage. She testified that he “stared her down”. He went out of sight but she saw him again through the window shortly after. That afternoon, Mrs. Baker provided police with a description of the man she saw, describing him as between five feet eight and five feet ten inches tall, white skinned, approximately fifty-ish years old, of medium build, and wearing a tan “tam-like” hat. Mrs. Baker subsequently referred to the hat as an “Andy Capp” hat after the police officer she was talking to described it as an “Andy Capp” hat. On cross-examination, Mrs. Baker admitted that her estimate of the man’s height was a wild guess since she saw only a portion of his body through the windows.

[14] The day after the attack, Mrs. Baker worked with a police sketch artist (a different artist than the one who subsequently worked with Mrs. McLeod) on a head-on and profile composite drawing of the assailant. She assessed the composite sketches as, at best, a 5 out of 10 for accuracy of likeness to the man she saw.

[15] On November 4, 1993, Carol Baker was shown the same eight-person photo line-up as Janet McLeod. She noted on the ballot that the appellant’s photograph resembled the person she saw but she could not be positive. While at the photo line-up, Mrs. Baker was asked, “if you saw the person do you think you would be able to recognize him for sure, I mean if you saw a photograph similar to these of the person, would you be able to recognize him?” She replied: “I don’t know if I could or not, I might be able to.”

[16] Mrs. Baker also saw the news footage of Mr. Hibbert in handcuffs at the time of his arrest. She positively identified him at the preliminary inquiry, and at the first trial, as the man she saw on the day of the attack. As with Janet McLeod, at the second trial she was specifically asked if she recognized the person she saw that day not as someone she had seen in photographs, on television news, or at any other time, but as the man she saw on the day of the attack. Mrs. Baker stated she recognized the appellant, “I recognize him in the box”.

(3) Witness — Heather Visscher

[17] Heather Visscher was walking with her parents along a dyke close to the open house and noted a man, walking a street width away, who she thought was out of place because he seemed “less friendly” than other walkers that day. At one point, the man passed ten feet from her and turned onto Trunk Road. She described the man as being in his fifties, approximately five feet ten inches or five feet eleven inches tall with a broad face, a “good head” of black and greying hair, and bushy eyebrows. She did not recall him as having eyeglasses or any head gear but did state his clothes were wet, consistent with him sweating, and that he had a jacket tied around his waist. She was not certain of the time but set it between 2:30 and 3:30 p.m.

[18] When shown the photo line-up, she marked the appellant’s photograph on her ballot as the best likeness of the man and wrote in the comment section: “I came to this conclusion by process of elimination. I am not 100% sure that this is the man I saw, but he most closely resembles the picture I have in my mind of the man I saw that day. The complexion & eyebrows & general shape of his face was what made me think it was him.”

[19] At the trial, Ms. Visscher was asked if she thought she would be able to recognize the man she saw on the dyke that day if she saw him again. She responded that she thought so. When asked if she could identify anyone in the courtroom as the man she saw that day she responded that she could not. When defence counsel pointed to Mr. Hibbert in the prisoner’s dock and asked her if that was the person she saw when she was on the dyke, she stated, “I don’t really think so.”

(4) Witness — Bobby Johnson

[20] Bobby Johnson lived at 147 Campbell Street and was working on his truck in his driveway when he saw a man running from the vicinity of the open house holding something up against his stomach. He then saw the man jump over a fence. He described the man as six feet tall, 170 to 180 pounds, greyish hair, wearing a tan short-sleeved dress shirt and grey dress pants with reddening on the back of the calves. He put the time at 1:00 to 1:30 p.m. He was not, however, wearing a watch. He could not make an identification at trial.

(5) Position of the Defence on Identification Evidence

[21] It was the position of the defence, largely accepted by the courts below, that the identification evidence had very little probative value because it had been contaminated by the news broadcast, but also because of the unreliability of this kind of evidence in general.

[22] The defence pointed out that Mr. Hibbert had a noticeable facial scar. No witness ever described a scar on the assailant’s face. Defence counsel sought and got a re-charge by the trial judge reminding the jury that no one mentioned a scar in their descriptions. Furthermore, Mr. Hibbert testified that he regularly wore glasses. His wife testified that he wore them around 75 percent to 98 percent of the time. He was wearing glasses at the time of his arrest. The descriptions of Mrs. McLeod and Mrs. Baker did not mention glasses.

(6) Others Matching the Appellant’s Description

[23] Once the police had a composite picture, the picture was distributed widely to local newspapers and to local realtors. There were several dozen tip files created during the course of the investigation. Not one person called the police to say that the appellant matched the composite. Many of the tip files contained information from persons who identified an individual who looked like the composite, or who fit the description, and had just arrived from Ontario. In no case was the appellant identified. A defence witness testified that he had called the RCMP to say that the composite picture was a “dead ringer” for a man he had seen walking on the dyke five or six times over a period of six or seven months up to September 1993. He had conversed with the man who said he was on leave of absence from the dockyard, was with the fire department, and lived in Duncan close to the dyke. According to the witness, the RCMP constable who took his call “didn’t put too much credence in what I was telling him because he said people can look at these composite pictures and see different faces there, and so it was pretty well left at that”.

B. Circumstantial Evidence

[24] The Crown’s case rested on a combination of the identification evidence described above and various items of circumstantial evidence linking the appellant to the attack.

(1) Personal Knowledge

[25] The investigation which originally led the RCMP to the appellant was based on information that Mrs. McLeod said was communicated to her by her assailant just before the attack. She told the police that the man who attacked her had given her several specific personal details, including that he lived on Trunk Road, that he was retired from the military and waiting for a pension, that he had just moved from Ontario and that he had two daughters aged 13 to 15. At the time of the attack, all of these details were also true of the appellant.

[26] However, Mrs. McLeod was asked in examination-in-chief whether she had ever met her assailant before the day of the attack. She said:

Um, excuse me, no, I hadn’t seen him before, but he mentioned to me that he was from Ontario and waiting for his military pension to come in before he could buy something, and I kind of looked at him because I had spoken to this lady on the phone that was from Ontario with her family and they were waiting for some money to come the following year before they could purchase anything. So -- and then they came to an open house that I had. I had a little house open on Marchmount Road, and, um, I just -- when I heard this fellow was from Ontario, I just kind of looked at him again to see if it was the same person that had come to my open house and realized that it wasn’t because the man that I had met from the open house was a very, very small man and this man was quite a bit bigger.

It is therefore clear from her evidence that the personal information from her assailant reminded her of another man she had met a few weeks earlier at another open house who had a similar background. Both Mr. and Mrs. Hibbert testified that they had attended an open house held by Mrs. McLeod on Marchmount Road a few weeks before the attack, and that Mrs. Hibbert had provided the realtor with the information in question about their background.

(2) Shoes

[27] Bloody footprints were found at the site of the attack. A police footwear expert gave his opinion that the crime scene shoe impression was similar in gross characteristics, including tread pattern, to the “Venture” running shoe, a brand name of the Zellers department store. The impression at the crime scene was from a size eleven shoe. Mr. Hibbert had owned a pair of black “Venture” brand running shoes purchased at Zellers and had purchased several pairs of Zellers shoes over the years. He wore shoe sizes ranging from size eleven to twelve; his wife testified that he wore size eleven and a half to twelve. At the trial, he wore size twelve shoes. Both Mr. and Mrs. Hibbert testified that his “Venture” running shoes had been thrown away a few weeks before the attack. One of the appellant’s stepdaughters thought the shoes were thrown out shortly after they arrived in Duncan, claiming the “toe was coming off one foot and the heel was coming off the other and they were just ready to go”. The Crown was not able to link the shoe print to any shoe owned by the appellant. The actual footwear that created the specific foot impression at the scene was not discovered during the course of the police investigation.

[28] An expert testified at trial that with department store athletic footwear (that is, an economical brand and not a name brand shoe such as Nike or Reebok) foreign manufacturers, primarily in developing countries, make a large quantity of generic soles that are sold and then assembled to the upper part of the shoe by another manufacturing facility which could be in a different country. On cross-examination, the expert noted that he was unable to trace the manufacturer of the sole and could not be sure that this particular sole was not sold as an in-store brand for other companies. Further, unlike the well-known shoe brand companies whose soles are a distinctive part of their branding, with department store shoes such as the “Venture”, the same brand running shoe could change its tread pattern every year.

[29] The footprints suggested that someone tripped or lost his balance on the steps leading into the house. Mr. Hibbert stated on cross-examination that he twisted his knee the morning of October 24, 1993.

(3) Bite Marks

[30] Mrs. McLeod testified that she bit the assailant “hard” on his hand. Mr. Hibbert had small marks on his hands at the time of his arrest. A dentist called by the Crown was unable to identify those marks as consistent with a bite mark.

(4) DNA

[31] On October 24, 1993 an RCMP officer’s tracking dog found a fresh scent on a blue “Andy Capp” style hat in a tree branch, off the dyke, close to the site of the assault. The officer who seized the cap testified that the cap had two small reddish spots that looked like blood spots. On cross-examination, the officer admitted she did not make a note of the two spots of blood when she seized the hat. The hat was forwarded to a forensic lab on November 2, 1993. Forensic analysis of the cap identified two small areas of blood spots on the outside crown area of the hat. The DNA expert testified that in his expert opinion one of the blood spots matched the DNA profile of Mrs. McLeod and the other matched the DNA profile of the appellant.

[32] Mr. Hibbert testified that he had owned such a cap. Both he and his wife testified that the hat was thrown out with a box of other discarded old clothes in September 1993 when they were unpacking from the move from Ontario. The cap found in the tree branch near the dyke had a piece of masking tape inside with “25 cents” written on it. Both Mr. and Mrs. Hibbert testified that this sticker was not on the cap when it was discarded.

[33] Mr. Hibbert testified that he suffers from a form of dermatitis condition causing dry chapped skin on his hands to open and bleed. He offered that as an explanation for some of the small abrasions on his hands at the time of his arrest, as well as the presence of the small amount of his blood on the hat.

[34] In her charge to the jury, Dorgan J. stated that the exhibit officer failed to note either the blood stains or the sticker in her notebook. The defence argued that the exhibit officer was new and had failed to carefully handle the exhibits so that they were subject to contamination, and pointed out that there were mistakes in the log. The defence suggested that the victim’s blood on the hat could have been transferred from other clothing seized by the police at the hospital or from the hands of the investigating officers who handled the evidence.

(5) Hair

[35] The cap also had several hairs that a hair comparison expert examined. The hair comparison expert examined six hairs taken from the cap (four hairs and two hair fragments). She testified that one of the hair fragments found in the cap was consistent with having originated from the appellant, with a 5 percent chance of coming from another source. Two hairs were inconsistent with Mr. Hibbert’s and he was eliminated as a source for them. Four hairs were not consistent with having originated from either the appellant or Mrs. McLeod but because of some similarities to the appellant’s hair sample, he could not be eliminated as a source. The cap also contained six non-human hairs indicative of having originated from a member of the dog family and two non-human hairs indicative of having originated from the rabbit family.

C. Crime Scene Evidence

[36] No forensic evidence was found at the crime scene at 151 Campbell Street linking the appellant to the location. No forensic evidence was found at Mr. Hibbert’s residence, or on any of his clothes, linking him to Mrs. McLeod and the crime. There was a copious amount of blood at 151 Campbell Street, none of which belonged to Mr. Hibbert. The blood at the crime scene matched Mrs. McLeod’s DNA profile. Blood was on the bathroom tub faucets and spout, door lock and door frame, and there were bloody footsteps from the garage, up the steps and to the laundry. There was blood in the sink traps at 151 Campbell Street consistent with the attacker washing at the scene. The taps were running when the police arrived.

D. Evidence of the Appellant

[37] Mr. Hibbert retired as a Master Corporal in 1993, after 26 years of military service. At the time of the attack he was on his unused sick and vacation leave until his pension was to begin in March 1994. He had no criminal record and no military disciplinary record. The Hibberts moved to Duncan from Ottawa (the location of his last posting) in August 1993 and rented a house on Trunk Road. Trunk Road is about a 10-minute walk from the crime scene. The appellant had two teenage stepdaughters. Mr. and Mrs. Hibbert attended a number of open houses in September and October 1993. Both Mr. and Mrs. Hibbert testified that they attended an open house hosted by Janet McLeod in September 1993. As indicated earlier, Mrs. McLeod’s evidence confirms that a couple matching some of the Hibberts’ personal characteristics visited that open house in September.

[38] Mr. Hibbert contends that on the morning of October 24, 1993, he drove his wife to a course she was taking and came home. He testified that he was wearing grey slacks, a green golf shirt and hiking boots. In the afternoon, he drove and accompanied his stepdaughters to a cadet flag-raising ceremony. The ceremony started at 1:30 p.m. After the ceremony, Mr. Hibbert waited while the cadets mingled and talked. He accompanied his stepdaughters on a 10-minute trip to the cadet hall barracks so one of his daughters could be measured for a uniform and issued gear by the training officer. Mr. Hibbert estimated that he and his stepdaughters either returned home at 2:30 p.m. or left the cadet hall at that time. The cadet training officer estimated that Mr. Hibbert and his stepdaughters arrived at the cadet hall between 2:15 p.m. and 2:20 p.m. and left 10 to 20 minutes later. Rachel Edwards, the appellant’s stepdaughter, said that they arrived at the cadet hall at 2:15 p.m., stayed 20 minutes and arrived home at 2:45 p.m. It is a 10-minute drive from the hall to their residence on Trunk Road. The appellant testified, and his stepdaughter Rachel confirmed, that upon returning home he prepared dinner and worked on carving a walking stick. Rachel conceded that since she was babysitting a little girl that day, she was not in constant contact with the appellant throughout the afternoon. At 5:00 p.m., Mr. Hibbert picked up his wife. Mrs. Hibbert testified that his demeanor was normal, that he was wearing the same clothing as when he dropped her off in the morning, and that his clothes were orderly. Mrs. Hibbert also testified that he wore hiking boots when he dropped her off in the morning and he had the same boots on when he picked up.

[39] The appellant’s alibi was corroborated in part by his wife, his stepdaughters, and by other witnesses for the defence, such as the cadet trainer. His stepdaughter, Rachel Edwards, said that the appellant was inside the house during the period between arriving back from the cadet trip and leaving to pick up his wife, with the exception of 10 minutes when he was on the front step working on his walking stick. Ms. Edwards testified that she recalled seeing him on the step at 3:15 p.m. and him coming back into the house 10 minutes later. At trial the stepdaughter said she would have heard the appellant leave. In cross-examination, she stated that she did not wear a watch. However, in November 1993, Ms. Edwards told police that she was playing Nintendo and had music on the day of the assault and that it was possible her father could have left without her knowing. At trial she explained this inconsistency by stating that she was “tired and upset” when she spoke to the police. The other witnesses could not place Mr. Hibbert at home during the period between his arrival home from the cadet ceremony and the time he left to pick up his wife but their testimony was consistent with the defendant’s account of his day. The family acknowledged discussing the times of his whereabouts on the day of the attack after his arrest.

[40] When Mr. Hibbert was initially arrested, he cooperated with police, gave a two-and-one-half-hour tape recorded statement, was fingerprinted, photographed, and voluntarily provided hair for DNA testing.

IV. The Issues

[41] The appellant frames the issues on appeal the following way:

Did the Court of Appeal for British Columbia err in law by imposing the curative provision, s. 686(1)(b)(iii) of the Criminal Code of Canada after concluding that the trial judge had erred in law in instructing the jury that they could infer guilt from a false alibi?

Did the Court of Appeal for British Columbia err in finding no reversible error in the trial judge’s instructions to the jury on the issue of identification?

V. Relevant Statutory Provisions

[42] Criminal Code, R.S.C. 1985, c. C-46

686. (1) On the hearing of an appeal against a conviction . . . , the court of appeal

(a) may allow the appeal where it is of the opinion that

(ii) the judgment of the trial court should be set aside on the ground of a wrong decision on a question of law, or

(iii) on any ground there was a miscarriage of justice;

(b) may dismiss the appeal where

(iii) notwithstanding that the court is of the opinion that on any ground mentioned in subparagraph (a)(ii) the appeal might be decided in favour of the appellant, it is of the opinion that no substantial wrong or miscarriage of justice has occurred . . . .

VI. Analysis

[43] In order to properly address the central issue of the application of the proviso, I find it preferable first to address the adequacy of the trial judge’s charge on the issue of identification, then to examine the nature and effect of the conceded error in the judge’s charge on the defence of alibi and finally to examine the strength of the Crown’s case within the proper framework of s. 686(1)(b)(iii).

A. The Charge on Identification

[44] The appellant argues that the trial judge’s directions to the jury on identification evidence were inadequate. This was the basis on which the appeal in the first trial was successful. In this case, however, the directions on identification evidence were not so deficient as to constitute an error of law. That being said, I will take the opportunity to highlight the problems with the identification evidence in this trial and the dangers of identification evidence in general.

[45] The evidence of identification was both direct and circumstantial. The concerns expressed by the appellant are addressed to the direct evidence and, in particular, to the alleged insufficiency of the charge with respect to the in-court identifications.

[46] As pointed out by the Court of Appeal, it is true that the trial judge addressed the frailties of the identification evidence. In the middle of a passage which highlighted several potential grounds for suspicion about the reliability of eyewitness identification, the trial judge said:

However, you must consider the possibility that Mrs. McLeod identified the accused in court from her memory of either the photograph in the line up or the television newscast instead of from her memory of the person she saw at 151 Campbell Street on October 24, 1993. You must also consider the same possibility in respect of Mrs. Baker’s in court identification of the accused.

I also remind you that both Mrs. McLeod and Mrs. Baker positively identified the accused for the first time when each was asked to identify him in the courtroom at the preliminary hearing and at the first trial and at this trial and the law provides that the identification of the accused for the first time in the courtroom after a failure to positively identify him from a photo line up is to be accorded little weight.

[47] The appellant argues that this instruction was insufficient to diffuse the psychological impact on the jury of the identification of the accused in open court by the victim of a life-threatening, vicious, unprovoked beating. The defence objected to the Crown asking Mrs. McLeod whether the accused was really the perpetrator of the attack “and not someone that you have seen subsequently”. The trial judge, in reviewing the evidence, pointed out the occasions that Mrs. McLeod had had to see the appellant before the trial (on television when he was arrested, at the preliminary inquiry and at the first trial) and reminded the jury that her evidence was that the appellant was her attacker and not simply the man she had seen on these previous occasions.

[48] The appellant argues that by asking Mrs. McLeod and Mrs. Baker to make such a distinction, the Crown was asking the impossible: the witnesses simply could no longer tell where their recognition of the appellant originated from. Furthermore, the appellant argues, after so much exposure to the appellant, whom neither witness had positively identified prior to the television newscast of his arrest, their in-court identification should be accorded no weight whatsoever.

[49] One might ask, if that were the case, why the in-court identification should be permitted to occur. In this case, as in most, it of course served to confirm that the accused was, in the opinion of Mrs. McLeod and Mrs. Baker, the same man they saw throughout the chain of events (from arrest through to the second trial). In that sense, despite its almost total absence of value as reliable positive identification, the evidence of the witnesses may be given some weight at least for that purpose. In addition, generally, a jury might be concerned if a witness was not asked to identify an accused in court as the perpetrator and might draw an unjustified adverse inference against the Crown if the question was not asked. Moreover, the inability of a witness to identify the accused in court as the perpetrator is entitled to some weight. This in fact happened here in the case of Ms. Visscher who, as the trial judge reminded the jury, was unable to identify the accused in court as the man she saw on the dyke.

[50] I am of the view that, in the circumstances of this case, the trial judge should have cautioned the jury more strongly that the identification of the accused in court, by Mrs. McLeod and Mrs. Baker, was highly problematic as direct reliable identification of the perpetrator of the offence. I think it is important to remember that the danger associated with eyewitness in-court identification is that it is deceptively credible, largely because it is honest and sincere. The dramatic impact of the identification taking place in court, before the jury, can aggravate the distorted value that the jury may place on it. I am not persuaded that the instruction quoted above, to the effect that such identification should be accorded “little weight”, goes far enough to displace the danger that the jury could still give it weight that it does not deserve.

[51] The danger of wrongful conviction arising from faulty but apparently persuasive eyewitness identification has been well documented. Most recently the Honourable Peter deC. Cory, acting as Commissioner in the Inquiry regarding Thomas Sophonow, made recommendations regarding the conduct of live and photo line-ups, and called for stronger warnings to the jury than were issued in the present case (Peter deC. Cory, The Inquiry Regarding Thomas Sophonow: The Investigation, Prosecution and Consideration of Entitlement to Compensation (2001) (“Sophonow Inquiry”), at pp. 31-34).

[52] While it is unnecessary to consider these recommendations in detail, I share the concern expressed by the Commissioner and, in this particular case, I think it would have been prudent to emphasize for the benefit of the jury the very weak link between the confidence level of a witness and the accuracy of that witness (Sophonow Inquiry, at p. 28). Moreover, here it should also have been stressed that the impact of Mrs. McLeod having seen the appellant arrested by the police as her alleged assailant could not be undone. Nor could she be expected to divorce her previous recollection of her assailant from the mental image that she formed after having seen the appellant on television.

[53] What will be required to displace the danger that the jury will give an eyewitness identification weight that it does not deserve will vary with the facts of individual cases. Here, at a second trial, and in light of the identification history, I think a stronger warning would have been appropriate.

B. The Instructions on Alibi

[54] A few days before the trial judge completed her instructions to the jury in this case, the B.C. Court of Appeal handed down its decision in R. v. Tessier (1997), 1997 CanLII 3475 (BC CA), 113 C.C.C. (3d) 538, in which it held that an instruction to the jury virtually identical to the one given in the present case was wrong. In light of that decision, the Crown concedes that the instructions here were defective.

[55] The first part of the trial judge’s instructions as to how to approach the evidence tendered in support of the defence of alibi was correct:

Keep in mind three things, that if you accept the evidence in support of the defence of alibi, you must return a verdict of not guilty if you find that these times just do not allow for this accused to have committed the acts alleged. If you do not accept the evidence in support of the defence of alibi, but you are left in a reasonable doubt about it, you must return a verdict of not guilty. Even if you are not left in a reasonable doubt by the evidence in support of the defence of alibi, you must still go on to determine whether or not on the basis of all the evidence the accused is guilty.

[56] The trial judge then proceeded to give the then standard instruction, suggested in a criminal jury instructions manual at the time (G. A. Ferguson and J. C. Bouck, Canadian Criminal Jury Instructions (2nd ed. (loose-leaf)), vol. 2, at pp. 8.04-1 to 8.04-4) as follows:

If on the other hand you determine beyond a reasonable doubt that the alibi is false and was put forward to deceive you, this allows you to draw if you wish the inference that the accused is guilty. By itself a false alibi is not conclusive evidence of guilt. You have to examine all the evidence when you are deciding whether the alibi is merely untruthful or was fabricated or concocted in a way designed to hide guilt. Keep in mind that a person may tell an untrue story for reasons other than for trying to hide guilt, that the Crown here is not suggesting to you that the times as they relate to the defence of alibi are concocted. Rather the Crown is suggesting that the times testified to have been reconstructed through conversations among and between the accused, Mrs. Hibbert and the daughters. The Crown is therefore arguing the weight and encouraging you not to give it much weight, that is the evidence of specific times as testified to by the accused and family members. [Emphasis added.]

[57] The Court of Appeal in Tessier held that in the absence of some evidence of concoction, a disbelieved alibi has no evidentiary value and cannot constitute positive incriminating evidence. This accords with the position taken also by courts of appeal in Quebec and in Ontario (R. v. Davison (1974), 1974 CanLII 787 (ON CA), 20 C.C.C. (2d) 424 (Ont. C.A.); R. v. Witter (1996), 1996 CanLII 4005 (ON CA), 105 C.C.C. (3d) 44 (Ont. C.A.); R. v. Coutts (1998), 1998 CanLII 4212 (ON CA), 16 C.R. (5th) 240 (Ont. C.A.), leave to appeal dismissed, [1999] 1 S.C.R. xii; R. v. Krishantharajah (1999), 1999 CanLII 3701 (ON CA), 133 C.C.C. (3d) 157 (Ont. C.A.); R. v. Carey (1996), 1996 CanLII 6039 (QC CA), 113 C.C.C. (3d) 74 (Que. C.A.)). As I discuss in more detail below, in this case there is no evidence of concoction. As a result, according to these authorities, the underlined passage in the above charge would constitute an error.

[58] The early leading authority on the subject was Davison, supra. Martin J.A., at p. 428, expressed the principle clearly as follows:

The inference to be drawn from the fabrication of a false alibi does not, however, arise from the mere rejection by the jury of the evidence with respect to alibi.

In my view the principle is correctly stated in Wigmore on Evidence, 3rd ed. (1940), vol. 2, at p. 126:

In particular, the failure to establish an alleged alibi is to be distinguished from the use of perjury or subornation in an unsuccessful attempt to prove the ‘alibi’; the latter admits of the usual broad inference from fraud, but the former amounts to nothing more than an inability to prove the specific fact of ‘alibi’.

Mere disbelief of the alibi advanced does not constitute proof that it was false in the sense of being fabricated. [Italics in original.]

This position was consistently followed in the cases cited above and the only doubt cast on its correctness comes from a short sentence in the oral judgment of this Court in R. v. Michaud, 1996 CanLII 211 (SCC), [1996] 2 S.C.R. 458. In the course of upholding a judgment of the New Brunswick Court of Appeal on other grounds, Sopinka J. noted a concession by the respondent that the Court of Appeal had erred on an unrelated matter, and then added: “We find no error in the charge in respect of this ground nor in the charge relating to the alibi evidence” (para. 3 (emphasis added)). It is this last passage that has created some uncertainty in that the Court of Appeal had ruled in Michaud that the trial judge’s instruction on alibi, cast in the standard terms also used in this case, was erroneous. Courts of appeal have commented on the effect of Michaud, noting that Sopinka J.’s comment was obiter (Krishantharajah, supra, per Carthy J.A., at pp. 164-65), and that the Supreme Court may have been of the view that there was sufficient evidence of concoction in that case to permit the instruction (Tessier, supra, per Rowles J.A., at p. 551, per Ryan J.A., at pp. 555-56; Krishantharajah, per Carthy J.A., at p. 165, who added that the comment might otherwise have been per incuriam).

[59] Michaud should not be taken as a departure from the well‑established position expressed and followed by the courts of appeal in the cases cited above. It is unnecessary in this case to determine the nature and scope of evidence necessary to show concoction so as to permit the instructions inviting the inference of consciousness of guilt. Suffice it to say that the cases have consistently pointed out that there must be evidence linking the accused to fabrication and that such evidence cannot emerge from a mere rejection of the alibi tendered. Many have spoken of the need for independent evidence of fabrication above and beyond a finding that the alibi is false. (See Ryan J.A. in Tessier, supra, at p. 556.)

[60] The Crown concedes that at trial it did not put forward a suggestion that the alibi had been concocted. It also concedes, as it did before the Court of Appeal, that the trial judge’s instructions were in error in the absence of evidence that the accused was implicated in putting forward a fabricated alibi.

[61] These concessions were appropriate. A defence of alibi may be disbelieved, particularly in the face of an overwhelming case for the prosecution, merely on the basis that the witnesses who testified in support of the alibi were imprecise or inconclusive, that their recollection was unreliable, or that they simply were mistaken. In such cases their evidence must be discarded, without more.

[62] Even if an alibi is advanced by the accused himself and is rejected, the finding that the alibi is untrue cannot serve to corroborate or complement the case for the prosecution, let alone permit an inference that the accused is guilty.

[63] If the alibi witnesses were found to be deliberately untruthful, their attempt at deceiving the jury could not be visited upon the accused unless he or she participated in the deceit. If, on the other hand, there was evidence that the accused attempted to put forward a fabricated defence, that effort, akin to an effort to bribe or threaten a witness or a juror, could be tendered as evidence of consciousness of guilt.

[64] In this case, the Crown concedes that there was no evidence of concoction. Apart from the fact that the appellant, his wife and his stepdaughters had discussed his whereabouts on the afternoon of the crime after his arrest, there was no evidence that the appellant was party to an effort to fabricate an alibi or that he enlisted members of his family to testify falsely about the issue.

[65] On the other hand, Crown counsel at trial suggested — to Mrs. Hibbert in particular — that she attempted to assist her daughters and her husband in reconstructing the time frame of their respective activities on the afternoon of the attack. In light of the erroneous instructions, the jury could have assumed, wrongly, that this was evidence of concoction which would allow them not only to reject the alibi evidence but to draw an inference of guilt from its rejection. The fallacy, and the danger, with that reasoning was well expressed by Ryan J.A. in Tessier, supra, at p. 556, where she said:

It seems to me that the reason we look for independent evidence that the accused fabricated his story is two-fold. In the first place as my colleague Rowles J.A. has pointed out the reasoning is circular if there is no independent proof: “The weight of the Crown’s evidence admits of no doubt therefore the accused is not telling the truth. The accused is not telling the truth therefore the Crown’s case admits of no doubt.” In the second place, because the evidence that the accused has fabricated a story can be used as part of the Crown’s case against him, care must be taken in finding that the alibi was concocted. There must be a solid evidentiary base of fabrication. It is not unreasonable to demand that this evidence be found independently of the other evidence of the proof of the crime.

If evidence of fabrication need be found only in the evidence which proves the offence then in every case where the accused testifies (alibi or not) there would be no reason not to permit a jury to use their finding that the accused has been untruthful as part of the Crown’s case against him. That does not accord with any articulation of the law that I know of.

[66] This highlights the seriousness of the erroneous instruction on the defence of alibi in the present case. It provided the jury with a direct, and incorrect, route to guilt. It implied that the jury could find evidence of concoction, which it could not on this record, and it then allowed the jury, if it rejected the alibi, to infer that the accused is guilty.

[67] Before turning to the application of the proviso in light of the above, it may be useful to summarize briefly the state of the law with respect to the rejection of a defence of alibi.

-- In the absence of evidence of concoction (deliberate fabrication) an alibi that is disbelieved has no evidentiary value.

-- A disbelieved alibi is insufficient to support an inference of concoction or deliberate fabrication. There must be other evidence from which a reasonable jury could conclude that the alibi was deliberately fabricated and that the accused was involved in that attempt to mislead the jury. It is the attempt to deceive, and not the failed alibi, that supports an inference of consciousness of guilt.

-- In appropriate cases, for instance if there were multiple accused, the jury should be instructed that the fabricated alibi may be used to place the accused at the scene of the crime, but may fall short of directly implicating him in its commission.

-- When there is evidence that an alibi was fabricated, at the instigation or with the knowledge and approval of the accused, that evidence may be used by the jury to support an inference of consciousness of guilt.

-- In cases where such an inference is available, the jury should be instructed that it may, not must, be drawn.

-- A fabricated alibi is not conclusive evidence of guilt.

C. The Application of Section 686(1)(b)(iii)

[68] This brings me to the principal ground of appeal. On appeal from the first trial in this case, the Court of Appeal held that the instructions to the jury had not adequately dealt with the weaknesses in the identification evidence and the court declined to apply the proviso. Legg J.A., for the court, noted that the error on the identification instructions was serious but the circumstantial evidence tendered by the Crown was strong. He concluded, however, that a properly instructed jury might have been left with a reasonable doubt (R. v. Hibbert (1996), 1996 CanLII 2174 (BC CA), 78 B.C.A.C. 277).

[69] In this appeal, the Court of Appeal concluded otherwise and applied the curative proviso despite the erroneous instructions on the effect of a rejected alibi.

[70] In my view, and with the greatest respect for the contrary opinion of the Court of Appeal in the present appeal, the proviso cannot be used to overcome the instructions on the inference of guilt arising from the rejected alibi which erroneously provided the jury with a direct route to conviction.

[71] The proper test for the application of the proviso has been enunciated by this Court on a number of occasions (see Colpitts v. The Queen, 1965 CanLII 2 (SCC), [1965] S.C.R. 739; Wildman v. The Queen, 1984 CanLII 82 (SCC), [1984] 2 S.C.R. 311; R. v. B. (F.F.), 1993 CanLII 167 (SCC), [1993] 1 S.C.R. 697; R. v. Bevan, 1993 CanLII 101 (SCC), [1993] 2 S.C.R. 599; R. v. Charlebois, [2000] 2 S.C.R. 674, 2000 SCC 53). It is the task of this Court to determine whether there is “any reasonable possibility that the verdict would have been different had the error at issue not been made” (Bevan, supra, at p. 617). The application of the proviso was recently reviewed in R. v. Khan, [2001] 3 S.C.R. 823, 2001 SCC 86. As I indicated in that case, the case law distinguishes between trivial errors, or errors that had no effect on the verdict, and serious errors of law which could have tainted the conviction. In the former case, the proviso can readily be applied. In the latter, it should not be applied unless the evidence of guilt is so overwhelming that any other reasonable jury would inevitably convict.

[72] The Crown concedes that the trial judge’s instructions were in error. The error in this case was not a trivial error nor one that would be unlikely to have had any effect on the verdict. As stated above, the trial judge’s erroneous instructions implied that the jury could find evidence of concoction allowing them to infer that the accused was guilty. This is a serious error that provided the jury with a direct route to guilt. Despite the circumstantial evidence that points to the accused, I do not think that one can confidently say that a conviction is a foregone conclusion in the sense that any other reasonable jury would inevitably convict. The jury deliberations were long in both trials. In the present appeal, the jury deliberated for several days.

[73] It is obvious that the jury needed proper guidance after a trial that was not an open and shut case. Not only is it possible that the erroneous alibi instructions played a part in the conviction, but without the benefit of actually hearing the evidence and assessing the credibility of the witnesses for the defence, it is not in my view possible to conclude that any other reasonable jury would also inevitably convict.

[74] In order to convict, the jury has to overcome some of the troubling features of the case: the absence of motive; the limited opportunity, even if the evidence of alibi is in part disbelieved; and the absence of any evidence linking the appellant to the scene of the crime. The jury will also have to overcome the strange confirmation by the victim, herself, that she in fact had met a man, at a previous open house, who shared many of the personal characteristics of the appellant — he and his family had recently arrived in Duncan from Ontario, he was looking for a house as he was waiting for his pension — in fact, their personal details had struck her as sufficiently similar that she wondered if it was the same man the minute her assailant conveyed that information to her. She concluded that it was not. Of course she could be mistaken, in which case such a mistake would cast serious doubt on the accuracy of all of her identification evidence. On the other hand, if she is right, this is an exculpatory coincidence that is potentially as cogent as some of the inculpatory ones. Innumerable details will have to be factored into the ultimate conclusion of the jury — the frequency of the appellant wearing glasses, the prominence of the scar on his face that was not noted by the identification witnesses, the 25-cent sticker on the “Andy Capp” hat, Mrs. McLeod’s consistent description of the hat worn by the assailant as brown while the hat found on the dyke was blue, and the hairs on that hat that did not belong to the appellant. In order to convict, a reasonable jury will have to address these matters in light of its appreciation of the credibility of the witnesses, including the appellant, who himself testified.

[75] These are all matters properly decided by the jury.

VII. Disposition

[76] I would therefore allow the appeal and order a new trial.

The reasons of L’Heureux-Dubé and Bastarache JJ. were delivered by

[77] Bastarache J. (dissenting) — Though I agree that the trial judge did not err in her instructions to the jury on identification evidence, I disagree that a stronger warning was appropriate in the circumstances of the case. In addition, I find that the error in respect of the trial judge’s instructions on alibi was not so serious as to preclude the Court of Appeal from applying s. 686(1)(b)(iii) of the Criminal Code, R.S.C. 1985, c. C-46 (the “proviso”), to cure the error. These two issues are discussed in turn below.

I. Adequacy of the Trial Judge’s Warning on Identification Evidence

[78] I agree with my colleague that the trial judge’s directions to the jury on identification evidence did not constitute an error of law. Moreover, I find that the trial judge’s warning in this regard was adequate. This Court has recognized the frailties of identification evidence. Yet it has also acknowledged that to take from the jury direct evidence of identification where the quality of that evidence is questionable risks obliterating the clear line separating the functions of the judge and jury. As stated by McIntyre J. in Mezzo v. The Queen, 1986 CanLII 16 (SCC), [1986] 1 S.C.R. 802, at p. 844, “[q]uestions of credibility and the weight that should be given to evidence are peculiarly the province of the jury.” In my view, the trial judge’s direction to the jury pointed out the general and specific problems with the identification evidence while at the same time respecting the role of the jury to weigh the evidence that was properly before it.

[79] The trial judge clearly discharged her obligation to highlight the inherent frailty of identification evidence and to explain the factors underlying its frailty: see Mezzo, supra, at p. 845, citing R. v. Turnbull, [1976] 3 All E.R. 549; see also Canning v. The Queen, 1986 CanLII 20 (SCC), [1986] 1 S.C.R. 991. In this respect, she provided the following general caution:

I am going to give you a special warning about the evidence of eye witnesses. As you know, it is such a major issue here, the issue of identity. Every once in a while in our courts a person is convicted of an offence even though she or he is innocent, and when this does happen it is often because a mistake has been made by one or more eye witnesses. It is easy to see how this can happen. You no doubt have had experience yourself where you think you have recognized someone and then found out that the person you thought it was is not the person you thought it was. An eye witness can be very convincing when that witness honestly believes that the accused person is the person she or he saw. Observation and memory can be unreliable when it comes to identification of people because people make honest mistakes. So when considering the evidence of the eye witnesses, it might be helpful for you, and I urge you to use the following guidelines paying particular attention to each eye witness’ opportunity to observe the person being described or identified.

[80] The trial judge then went on to identify with reference to the facts several factors including the length of time the witness observed the person being identified, the distance from which the observation was made, the lighting conditions and the presence of obstructions, the time that had elapsed since the time the person was identified, the eyesight and memory of the witnesses, and consistencies and inconsistencies between the testimonies given by the eye witnesses. Following this, she added an additional caution:

Bear in mind that although identification by one witness can support that of the other, even a number of honest witnesses can be mistaken.

[81] In Canning, supra, this Court confirmed that in addition to providing a general caution regarding the frailties of identification evidence, the trial judge must also relate the need for caution to the particular facts of the case. In my view, the trial judge adequately provided this aspect of the required warning. Specific vulnerabilities in the identification evidence were identified by the trial judge and discussed in considerable detail. The trial judge raised the question: “Did the eye witness see a photograph of the person described before making the identification?” In answer to this she reviewed the following evidence:

In this case you have evidence that both Mrs. McLeod and Mrs. Baker saw a photograph of the accused in the photographic line up prior to identifying the accused in court. Ms. Visscher did as well, and she was unable to identify the accused in court as the man she saw on the dike. Both Mrs. McLeod and Mrs. Baker had seen news coverage on television of the accused leaving the courthouse after he was charged. As well you have the evidence of Janet McLeod that she watched the television newscast three times, and that in so doing the television picture of the accused was stopped. The freeze frame photo has been referred to. She said she watched the news on December 14, 1993 to make sure it is the right person. Those were her words. She and her husband froze the picture of the accused in the newscast and she looked at it and she said then and she says now that the accused is the right person.

[82] The trial judge proceeded to provide the jury with a clear explanation of the significance of the above evidence:

. . . you must consider the possibility that Mrs. McLeod identified the accused in court from her memory of either the photograph in the line up or the television newscast instead of from her memory of the person she saw at 151 Campbell Street on October 24, 1993. You must also consider the same possibility in respect of Mrs. Baker’s in court identification of the accused.

I also remind you that both Mrs. McLeod and Mrs. Baker positively identified the accused for the first time when each was asked to identify him in the courtroom at the preliminary hearing and at the first trial and at this trial and the law provides that the identification of the accused for the first time in the courtroom after a failure to positively identify him from a photo line up is to be accorded little weight.

[83] After providing specific cautions regarding the possibility that the in-court identification was tainted and was in any event entitled to little weight given the inability of witnesses to pick the accused out of a photo line-up, the trial judge summed up her instruction on identification evidence as follows:

It is open to you to find that the Crown’s case on visual identification evidence is vulnerable for some of the reasons that have been outlined and applying the guidelines I have given you.

[84] The appellant argues that the trial judge should have instructed the jury that Mrs. McLeod’s and Mrs. Baker’s in-court identification should be accorded no weight whatsoever given the possibility that these witnesses were basing their identification on subsequent observations of the appellant (on television, at the preliminary hearing at the first trial) rather than on their recollection of the appellant at the scene of the crime. He further asserts that the failure of the witnesses to positively identify him prior to seeing him on television and in court strengthens the argument that the in-court identification should be given no weight.

[85] My colleague rejects the appellant’s argument that the trial judge should have instructed the jury to accord no weight to the in-court identification. She is of the opinion that the in-court identification was entitled to some weight. She notes in this respect that the in-court identification served to confirm that the accused was, in the opinion of Mrs. McLeod and Mrs. Baker, the same man they saw through the chain of events (from arrest through to the second trial). In addition, she notes that a jury may be concerned if a witness was not asked to identify an accused in court as the perpetrator and might draw an adverse inference against the Crown if the question was not asked. Moreover, in-court identification serves some purpose since the inability of a witness to identify the accused in court as the perpetrator is entitled to some weight.

[86] While she would not require the trial judge to instruct the jury that the in-court identification was entitled to no weight, she would have preferred the trial judge to instruct the jury that they could not rely on the in-court identification as direct reliable identification of the perpetrator of the offence. For the reasons discussed below, it is my opinion that a warning of that nature was neither necessary nor desirable.

[87] The first problem I have with the instruction suggested by my colleague is that it appears inconsistent and therefore risks confusing the jury. On the one hand, the jury is told that in-court identification has some purpose and in particular can be used to refute the Crown’s theory that the accused is the perpetrator if the witness is unable to identify the accused as the perpetrator in court. On the other hand, they are told that a positive in-court identification of the accused as the perpetrator should not be relied upon as direct reliable evidence of the identification of the perpetrator. In my view, a jury would find these two views confusing and inconsistent and would tend to disregard in-court identification for any purpose. The approach of the trial judge in this case was preferable. Rather than instructing the jury to disregard the identification of the accused in court, she warned the jury that this evidence was entitled to little weight and accompanied this warning with a clear explanation as to why the evidence was entitled to little weight given the particular circumstances of the case.

[88] Perhaps a more significant problem with the instruction suggested by my colleague is that it encroaches upon the jury’s essential function as the trier of fact. Once the decision has been made to allow evidence to be put before the jury, it is ultimately up to the jury to determine what weight to attribute to that evidence. Though this Court recognizes that there are certain frailties associated with identification evidence that must be brought to the attention of the jury, properly admitted identification evidence should not be effectively removed from the jury’s consideration by means of the trial judge’s instruction. As was stated by Weiler J.A. in R. v. Gagnon (2000), 2000 CanLII 16863 (ON CA), 136 O.A.C. 116, at para. 91:

The trial judge was correct that the generally accepted state of the law is that, where evidence is tainted, either because identification was suggested by the accused’s presence in the prisoner’s box or as a result of inappropriate police procedures, the evidence is not thereby rendered inadmissible. Rather, the evidence of tainting is a factor going to the weight of the evidence which is exclusively the province of the jury. See R. v. Mezzo, 1986 CanLII 16 (SCC), [1986] 1 S.C.R. 802; 68 N.R. 1; 43 Man. R. (2d) 161; 27 C.C.C. (3d) 97; R. v. Miaponoose (A.) (1996), 1996 CanLII 1268 (ON CA), 93 O.A.C. 115; 110 C.C.C. (3d) 445, at p. 458 (C.A.); R. v. Buric (G.J.) et al. (1996), 1996 CanLII 1525 (ON CA), 90 O.A.C. 321; 106 C.C.C. (3d) 97, at p. 112 (C.A.), per Labrosse, J.A, whose reasons were affirmed at (1997), 1997 CanLII 380 (SCC), 209 N.R. 241; 98 O.A.C. 398; 114 C.C.C. (3d) 95 (S.C.C.). [Emphasis added.]

In this case, it was open to the jury to determine that Mrs. McLeod and Mrs. Baker were able to identify the accused in court because they recognized the accused as the perpetrator of the crime. The trial judge cautioned the jury regarding the dangers associated with drawing this conclusion without unduly impinging upon the jury’s function to weigh the evidence.

[89] The final problem that I have with the instruction suggested by my colleague is that it seems to create a spectrum of warnings from which the trial judge must choose when instructing a jury regarding in-court identification. While in some cases it may be appropriate for the trial judge to instruct the jury to accord "little weight" to the in-court identification, in other cases the trial judge must instruct the jury that in-court identification should not be used as evidence of "direct reliable identification", a standard which presumably lies somewhere in between “little weight” and “no weight”. In my view, creating a myriad of appropriate cautions ranging from “little weight” to “no weight” is unnecessary and confusing and will have the undesirable effect of spawning a new ground for appeal.

[90] Furthermore, where the Court concludes that the warning provided by the trial judge was adequate taking into account the circumstances of the case, it is unnecessary to suggest an alternative preferable method of cautioning the jury. As I stated in R. v. Charlebois, [2000] 2 S.C.R. 674, 2000 SCC 53, at para. 24:

Even if the trial judge could have expressed himself more clearly, “a standard of perfection is not the test when an appellate court reviews a jury charge” (R. v. Malott, 1998 CanLII 845 (SCC), [1998] 1 S.C.R. 123, at para. 15). In R. v. Jacquard, 1997 CanLII 374 (SCC), [1997] 1 S.C.R. 314, our Court recognized at para. 2 that while there is an entitlement to a properly instructed jury, “(t)here is . . . no requirement for perfectly instructed juries” (emphasis deleted). One must accept that it is not incumbent on the trial judge to make an argument for the defence or to fashion the most favourable charge to the defence; see R. v. Dickhoff (1998), 1998 CanLII 12386 (SK CA), 130 C.C.C. (3d) 494 (Sask. C.A.).

[91] The above principle, which eschews a formulaic approach to the trial judge’s charge to the jury, was echoed in the specific context of the trial judge’s instructions on identification evidence in R. v. Edwardson (1993), 1993 CanLII 1420 (BC CA), 77 B.C.L.R. (2d) 362, a decision of the British Columbia Court of Appeal, at paras. 39 to 41:

As I have noted, there has been a reluctance on the part of this and other appellate courts to adopt a general rule such as that found in the Turnbull case. The reason for that is undoubtedly the fear that in some cases a special instruction on the inherent frailties of eyewitness identification will induce juries to reject accurate identifications and acquit where they should convict.

But that concern can be met by casting the special instruction in language appropriate to the case. The precise point which has been made time and again is that there is no rigid formulation for such instruction. Each should be designed with specific reference to the evidence which either supports or casts doubt on the eyewitness identification. . . .

In all cases the jury should be told that they ought not to resile from acting upon an eyewitness identification if, after taking into account the various weaknesses which have been pointed out to them by both the judge and counsel, and exercising the required caution, they are nonetheless satisfied beyond a reasonable doubt that it is an accurate identification. [Emphasis added.]

[92] For the reasons expressed above, I also think that the suggestions put forward by my colleague at para. 52 are unnecessary, and in some instances, undesirable. My colleague suggests that the trial judge should have emphasized “the very weak link between the confidence level of a witness and the accuracy of that witness”. In my view, the trial judge communicated that idea to the jury when she explained that an eyewitness can be very convincing when he or she honestly believes that the accused person is the person she or he saw, but that “[o]bservation and memory can be unreliable when it comes to identification of people because people make honest mistakes.”

[93] In the same paragraph (para. 52), my colleague says that the trial judge should have stressed that Mrs. McLeod could not have divorced her previous recollection of the assailant from the mental images she formed after having seen the appellant arrested by the police on television. I disagree that a caution of this nature should have been given by the trial judge. As I discussed above, whether Mrs. McLeod was able to identify the accused as her assailant, or whether she was merely identifying him as the man she saw being arrested and at previous hearings, was ultimately a question for the jury to answer.

II. Application of the Proviso to the Trial Judge’s Erroneous Instruction on Alibi

[94] This Court has established that the question to be asked when considering the application of s. 686(1)(b)(iii) of the Criminal Code is whether there is any reasonable possibility that the verdict would have been different had the error at issue not been made: R. v. Bevan, 1993 CanLII 101 (SCC), [1993] 2 S.C.R. 599, at p. 616; R. v. Jacquard, 1997 CanLII 374 (SCC), [1997] 1 S.C.R. 314, at para. 58; Charlebois, supra, at para. 11; R. v. Jolivet, [2000] 1 S.C.R. 751, 2000 SCC 29, at para. 48; R. v. Khan, [2001] 3 S.C.R. 823, 2001 SCC 86, at para. 28.

[95] In answering the above question, the strength of the Crown’s case, otherwise stated as “the legally admissible evidence untainted by the error”, is but one factor to be considered along with the seriousness of the error in question and the effect it likely had upon the jury’s inference-drawing process (Jolivet, supra, at para. 54). I agree with the Court of Appeal that the Crown’s case was a strong one. However, in my view, it is the latter two factors that are dispositive of this case. The trial judge mitigated the seriousness of the error contained in the then standard jury instructions by tailoring the instruction to the particular circumstances of the case. When the error in the trial judge’s instructions on alibi is considered in the context of the charge as a whole and against the background of the entire trial, it is evident that there was no reasonable possibility that the error would have impacted upon the reasoning process of the jury.

[96] I agree with my colleague that case law applying the proviso distinguishes between trivial errors, or errors that had no effect on the verdict, and serious errors of law which could taint the conviction. Where the error is determined to be “serious”, the proviso should not be applied unless the evidence of guilt untainted by the error is so overwhelming that any other reasonable jury would inevitably convict (Khan, supra, at para. 26). My colleague suggests that the error made in this case was, by its very nature, serious. In my view, there is no independent yardstick by which to gauge the seriousness of the error. As McIntyre J. remarked in Mahoney v. The Queen, 1982 CanLII 21 (SCC), [1982] 1 S.C.R. 834, at p. 852, “[t]he Court of Appeal must consider the errors against the background of the whole trial.” Whether an error is considered “trivial” or whether it is “serious” depends on the extent to which it can be said to have affected the jury’s determination.

[97] Whether or not the error was in this case “serious” necessitates an examination of the nature of the error and the likely effect it had on this jury’s reasoning process. The trial judge’s error consisted of giving the standard instruction concerning false alibi as contained in G. A. Ferguson and J. C. Bouck, Canadian Criminal Jury Instructions (2nd ed. (loose-leaf)), vol. 2, at pp. 8.04-1 to 8.04-4. Those instructions were found to be in error in R. v. Tessier (1997), 1997 CanLII 3475 (BC CA), 113 C.C.C. (3d) 538 (B.C.C.A.), a decision which the trial judge did not have the benefit of when charging the jury. Tessier holds that in the absence of some evidence of actual concoction, a disbelieved alibi has no evidentiary value and cannot be placed on the scales to contribute to a finding of guilt. It is therefore an error to instruct the jury that they can draw an inference of guilt from a disbelieved alibi in the absence of independent evidence of fabrication. The error is described in the following terms in R. v. Krishantharajah (1999), 1999 CanLII 3701 (ON CA), 133 C.C.C. (3d) 157 (Ont. C.A.), at pp. 163-64:

This court has consistently held that in the absence of evidence of concoction, (there is no suggestion by the Crown that such exists in this case), an alibi that is disbelieved has no evidentiary value. It should not be added to other evidence to contribute to a finding of guilt. . . . As explained by Doherty J.A. in Coutts, this cannot be justified by pure logic because in many cases the inference of concoction, and thus consciousness of guilt, may flow naturally from disbelief. However, the safeguard is necessary to prevent the jury from short-circuiting the Crown’s obligation to prove guilt beyond a reasonable doubt by moving directly from disbelief of an alibi to a finding of guilt.

[98] The trial judge in this case instructed the jury: “If . . . you determine beyond a reasonable doubt that the alibi is false and was put forward to deceive you, this allows you to draw if you wish the inference that the accused is guilty.” To have acted impermissibly on this instruction, the jury would have had to have made an initial determination that the accused’s alibi was concocted. In addition, the jury would have had to have applied the inference to go directly to a finding of guilt without examining the whole of the evidence. For the reasons discussed below, I think that it is highly improbable that the jury reasoned in this way.

[99] If the jury heeded the trial judge’s instructions they would have been unlikely to reach the conclusion that the accused’s alibi was concocted. In the portion of the instructions immediately following the erroneous instruction on alibi, the trial judge gave the following caution:

You have to examine all the evidence when you are deciding whether the alibi is merely untruthful or was fabricated or concocted in a way designed to hide guilt. Keep in mind that a person may tell an untrue story for reasons other than for trying to hide guilt, that the Crown here is not suggesting to you that the times as they relate to the defence of alibi are concocted. Rather the Crown is suggesting that the times testified to have been reconstructed through conversations among and between the accused, Mrs. Hibbert and the daughters. The Crown is therefore arguing the weight and encouraging you not to give it much weight, that is the evidence of specific times as testified to by the accused and family members. [Emphasis added.]

I agree with the respondent that this passage “could only have served to neutralize the effect of the erroneous passage of the instruction because it immediately withdrew it from the jury’s consideration”.

[100] My colleague states that despite the trial judge’s clear instruction that the Crown was not suggesting that the alibi was concocted, the jury might nonetheless have reached that conclusion from the Crown’s cross-examination of Mrs. Hibbert. I find no suggestion in the Crown’s cross-examination of Mrs. Hibbert that the witnesses were lying. In cross-examination of Mrs. Hibbert, the Crown merely suggests that Mrs. Hibbert assisted the family members to remember the times of things that occurred on the day of the crime. In doing so the Crown puts forward the suggestion that there was “some confusion” regarding the times. Nor did the Crown’s closing argument suggest a concocted or false alibi.

[101] Even in the unlikely event that the jury did conclude that the alibi was concocted, it seems improbable that the jury would have applied the inference to proceed directly to a finding of guilt without examining the whole of the evidence. Immediately preceding the erroneous instruction, the trial judge provided the jury with the following comprehensive instructions:

It is open to you to consider what effect if any the conversations among and between the various family and the accused have on the various times testified to, and it is for you to decide what weight you attribute to the evidence of the timing of the various events. Clearly the defence of alibi and issues of credibility are intertwined, and you must know that the Crown must prove beyond a reasonable doubt that the defence of alibi cannot succeed. The accused does not have to prove anything.

Keep in mind three things, that if you accept the evidence in support of the defence of alibi, you must return a verdict of not guilty if you find that these times just do not allow for this accused to have committed the acts alleged. If you do not accept the evidence in support of the defence of alibi, but you are left in a reasonable doubt about it, you must return a verdict of not guilty. Even if you are not left in a reasonable doubt by the evidence in support of the defence of alibi, you must still go on to determine whether or not on the basis of all the evidence the accused is guilty. [Emphasis added.]

In my view, the jury would likely take these comprehensive instructions as their guide to assessing the alibi evidence as opposed to drawing an inference of guilt based on a theory of concoction that was, as the trial judge carefully reminded, not before them.

[102] My colleague submits that the case was by no means “open and shut” and notes that the jury deliberated for several days (paras. 72-73). In my view, the fact that the jury deliberated for several days indicates that they did not proceed on the erroneous basis that they could move directly from a rejection of the accused’s alibi to guilt. Perhaps more significant is that after three days of deliberations the jury asked the trial judge to read back the charge on circumstantial evidence. The jury’s request to have the direction on circumstantial evidence repeated suggests that the jury, having rejected the defence of alibi, went on to determine whether the accused was guilty on the basis of the circumstantial evidence adduced by the Crown. Had the jury found the alibi to be concocted and followed the erroneous instruction that they could infer guilt on this basis alone, they would have required no further clarification on the distinction between direct and circumstantial evidence.

[103] Given the trial judge’s charge in its entirety and the circumstances of this particular case, I conclude that the danger of a miscarriage of justice arising from the trial judge’s misdirection on alibi was “more theoretical than real” (Tessier, supra, at p. 562). The test for the application of the proviso is not whether it is in theory possible that the verdict would have been different had the error not been made, but rather whether there is a reasonable possibility that the verdict would have been different taking into account the background of the whole trial. As the Ontario Court of Appeal once observed, “[t]here will probably never be a perfect charge or one that cloistered appellate counsel cannot find objectionable after minute scrutiny” (R. v. Demeter (1975), 1975 CanLII 685 (ON CA), 25 C.C.C. (2d) 417, at p. 436). Though the test for the application of the proviso is strict, it nonetheless recognizes that while the accused is entitled to a properly instructed jury he or she is not entitled to a perfectly instructed one (Jacquard, supra, at para. 2).

[104] My colleague notes that on appeal from the first trial in this case, the Court of Appeal held that the instructions to the jury had not adequately dealt with the weaknesses in the identification evidence and the court declined to apply the proviso to cure this error (see para. 68, citing R. v. Hibbert (1996), 1996 CanLII 2174 (BC CA), 78 B.C.A.C. 277). In my view, applying the proviso in the circumstances of this trial even though it was not applied in the first trial creates no inconsistency. In the first trial the Court of Appeal commented at length upon the seriousness of the trial judge’s failure to give an adequate warning on identification evidence at para. 63:

The case for the Crown against the appellant was indeed a strong circumstantial case and the address of counsel for the appellant at trial had stressed the weakness of the identification evidence shortly before the judge gave his charge. However, in my view, it was important that the judge deal with the weaknesses in the identification evidence and avoid giving the jury the impression that Mrs. McLeod and Mrs. Baker had identified the appellant as the man who committed the assault. This was particularly important in the case under appeal because the appellant had testified, supported by his wife and two stepdaughters, that he was not the person who had entered the house and assaulted Mrs. McLeod. There was the possibility that Mrs. McLeod was confused in her evidence over when she had learned of the military background of the appellant. Further, there was no evidence that the appellant had any motive for committing the assault on Mrs. McLeod. The appellant’s credibility and the credibility of Mrs. Hibbert and the two daughters were significant matters for the jury to consider and weigh with the weaknesses in the evidence of identification in deciding whether it was convinced beyond a reasonable doubt that the appellant was the person who committed the crime. If the learned judge had warned the jury of the weaknesses in the identification evidence, the jury might have been left with a reasonable doubt that the Crown had proved that it was the appellant who committed the assault. [Emphasis added.]

[105] Whereas the Court of Appeal concluded that the error made by the trial judge in the first trial was a serious one, I conclude that the error that was made by the trial judge in the second trial was not, in all the circumstances, serious. In my view, there is a compelling rationale underlying the rule that when applying the proviso, the court looks to the error and whether that error might have affected the jury. If the court were to look only to the case as a whole and whether it was overwhelming for any jury, the result would be that trials would be held over and over again as long as the trial judge’s charge contained some error, no matter how minute.

[106] As I stated at the outset, I am of the view that the proviso should be applied in this case for the reason that the trial judge’s error was not, in the circumstances, serious and was unlikely to have affected the jury. This conclusion is nonetheless inextricably linked to a finding that the weight of the Crown’s evidence was overwhelming and I do not reject the Court of Appeal’s conclusion in this regard. Having determined that there was no real possibility that the error in question affected the reasoning process of the jury, it is only logical to conclude that the accused was convicted because the evidence against him was overwhelming.

[107] I would therefore apply the curative provision of s. 686(1)(b)(iii) and dismiss the appeal.

Appeal allowed and new trial ordered, L’Heureux‑Dubé and Bastarache JJ. dissenting.

Solicitors for the appellant: Firestone & Tyhurst, Victoria.

Solicitor for the respondent: The Ministry of Attorney General, Vancouver.