Her Majesty the Queen v. James Sauvé and Richard Trudel, 2004 ONCA 9054

DATE: 20040130

DOCKET: C25967

COURT OF APPEAL FOR ONTARIO

CATZMAN, ROSENBERG and BORINS JJ.A.

B E T W E E N :

HER MAJESTY THE QUEEN

Frank Addario and P. Andras Schreck for the appellant

Respondent

• and -

Philip Campbell and Catherine Glaister for the appellant

JAMES SAUVÉ and RICHARD TRUDEL

Appellant

Scott Hutchison, John McInnes, Alexander Alvaro, Gillian Roberts and Susan Magotiaux for the respondent

Heard: February 3-7, 2003

On appeal from the convictions by Justice David L. McWilliam of the Superior Court of Justice, sitting with a jury, on May 20, 1996.

BY THE COURT:

[1] The appellants James Sauvé and Richard Trudel appeal from their convictions on May 20, 1996 by a court composed of McWilliam J. and a jury on two counts of first degree murder These convictions followed a very lengthy trial that originally included two other accused Robert Stewart and Richard Mallory. For reasons that are not germane to this appeal, Stewart and Mallory were severed from the Sauvé and Trudel trial. They were subsequently convicted and their appeal to this court is pending.

[2] The appellants have argued a large number of grounds of appeal including the adequacy of the Vetrovec caution, the charge to the jury on the co-conspirator’s exception to the hearsay rule, the admission of certain evidence said to be highly prejudicial, and the failure to sever Trudel’s trial. In addition, the appellants seek to adduce fresh evidence of post-trial recantations by two of the three principal Crown witnesses. For the following reasons we allow the appeal and order a new trial.

THE FACTS

[3] The appellants’ trial lasted fifteen months. It was marked by extensive cross-examinations of the three principal Crown witnesses. What follows is only a brief summary of the facts. We will make greater reference to the facts as necessary to explain our reasons in relation to the various grounds of appeal.

[4] Robert Stewart was a high-level drug dealer in the Ottawa area. Richard Mallory was said to be his enforcer. The appellant Richard Trudel was also a drug dealer and associated with Stewart. The appellant James Sauvé was said to be his enforcer. The two victims of the murders were Michel Giroux, a low-level drug dealer, and Manon Bourdeau his pregnant wife. They were executed in their home in Cumberland by someone using a shotgun. It was the theory of the Crown that Giroux owed money to Stewart and that Stewart ordered the killing of Giroux as an example to other drug dealers who owed him money. The Crown’s theory about why Bourdeau was killed shifted during the trial. The Crown alleged that Bourdeau was killed because she had threatened to go to the police, in response to threats from Stewart, or because she happened to be present when the killers arrived to execute Giroux.

[5] There was some evidence of a connection between the Stewart and Trudel drug ring and Giroux. Thus, a witness who purchased cocaine from Giroux testified that he had introduced Trudel to Giroux in the fall of 1989. Another witness, Denis Gaudreault, whose evidence was pivotal to the Crown’s case, testified that he had seen Giroux, Trudel and Stewart together. There was also evidence, including evidence from a defence witness, that in December 1989 and January 1990, Stewart was very agitated and concerned about money and his debts and the debts owed by others to him.

Denis Gaudreault and Related Evidence

[6] As we have said, the Crown’s case against the appellants largely depended upon the evidence of Denis Gaudreault. Gaudreault owed a substantial amount of money to Stewart as a result of his dealing and consumption of drugs. He also stored Stewart’s supply of firearms. Gaudreault testified that in December 1989 or January 1990 he was with Stewart and an associate of Stewart’s, Michel Vanasse, to view Gaudreault’s racing car, which Stewart was considering taking for the debts Gaudreault owed him. On the way, they dropped Gaudreault off in the Cumberland area because they said they had to stop somewhere to talk to somebody that owed them money. The Crown alleged that they were going to see Giroux, who, as we have said, lived in the Cumberland area of Ottawa. Another of Stewart’s drug dealers, Jamie Declare, testified that Stewart was complaining about a couple that was giving him a hard time. Stewart had tried to repossess something, and the couple told him to get out or they would call the police.

[7] Gaudreault testified that before the killing he overheard Stewart yelling into the telephone, “if you think you’re going to get away with this, you and your old man, you got fucking something coming to you”. After the telephone call, Stewart said that he was talking to “that bitch from Cumberland”. It was the theory of the Crown that this referred to Ms. Bourdeau. Stewart said he was going to get Mallory to deal with the problem and a meeting was set up for later that afternoon. Gaudreault attended this meeting along with Trudel, Stewart, Mallory and Vanasse, but was not privy to any planning that may have gone on.

[8] Gaudreault testified that on the morning of January 16, 1990, Stewart and Vanasse came over to his home and complained about his dealers, including Gaudreault, not paying their bills. Stewart said that “pretty soon [Gaudreault] could read about it in the newspaper, that there was gonna be an example made out of”. Later that same day Stewart called him and said he needed a ride and for Gaudreault to bring the “tools”, meaning the firearms. At the time he received the call, Gaudreault and Declare were “free-basing” cocaine. About 20 minutes later, Trudel, Mallory, Stewart and Sauvé arrived at his home. Gaudreault brought up a bag of firearms, including a sawed off shotgun, which he gave to Mallory. Mallory gave the shotgun to Sauvé. It was the theory of the Crown that this was the murder weapon and that while other accused were armed, it was Sauvé who killed the deceased. Several associates of Gaudreault confirmed that he kept a supply of firearms and ammunition in the basement of his home.

[9] The four accused and Gaudreault drove in different vehicles to a nearby restaurant. They then all entered a white Cadillac that Gaudreault was driving. Gaudreault claimed that this was the only time he ever drove this vehicle. All four of the accused could be associated with a white Cadillac. Sauvé owned a white Cadillac, Trudel was seen driving the vehicle, and Stewart and Mallory’s fingerprints were found in that car. Declare testified that he saw Gaudreault in a white Cadillac. He began to speak to him but left when he learned Gaudreault was waiting for Stewart. Declare testified that he saw Trudel and someone who looked like Sauvé in the car. There were some frailties in this evidence. In particular, Declare did not remember until 1993 seeing Gaudreault in the vehicle. This was after he had spoken to the police on several occasions and been hypnotized in an effort to assist his memory, during which he was asked if he had seen Gaudreault in a white car. His descriptions of Sauvé and Trudel did not match how they appeared in early 1990. The police had also shown a picture of Sauvé to Declare and told him he was the “triggerman”. Garrett Nelson, Gaudreault’s girlfriend’s brother, also testified to having seen Gaudreault in a white luxury vehicle about two weeks before the end of January 1990. However, he testified that it did not appear to be the Sauvé Cadillac.

[10] According to Gaudreault, Stewart then directed Gaudreault to the home of the deceased in Cumberland where he left off Sauvé, Trudel and Mallory. Gaudreault and Stewart drove off but returned a few minutes later. When they returned, the other three were standing by the road. They picked them up and drove back to Stewart’s house. On the way, Stewart asked how it went and someone said there was no problem. Mallory mentioned that a television had been left on. When they reached Stewart’s house, Stewart directed the other accused to return the weapons to Gaudreault. According to Gaudreault, Trudel told Stewart, “No problem. He got it twice and the bitch was done in the back”. (This corresponded with the cause of death, Giroux had been shot twice and Bourdeau had been shot once in the back.) Stewart told Gaudreault to pay Sauvé and Trudel out of money Gaudreault owed Stewart. He then had his wife drive Gaudreault to his home. Stewart’s wife testified that she did not recall ever having driven Gaudreault home.

[11] After he arrived home, Gaudreault cleaned and reloaded the guns. He noticed that there were three shells missing from the shotgun. (In earlier statements to the police he said that four shells were missing.) In the following days, Gaudreault made the payments to Sauvé and Trudel. Much time was spent at trial respecting a black book that Gaudreault kept showing his various financial transactions. This was important because Gaudreault claimed to have entered the payments to Sauvé and Trudel in the book. The original book was lost prior to trial and so Gaudreault reconstructed it and presented the new book to the police as the original. In the reconstructed book, Gaudreault did not show the payment to Trudel.

[12] A few days after the killings, Stewart and Vanasse came to Gaudreault’s house. Gaudreault testified that Stewart produced a newspaper article about the Cumberland killings. The implicit message was that this was what happened to people, including Gaudreault, who owe Stewart money. Garrett Nelson, Gaudreault’s brother-in-law testified and confirmed the newspaper incident. Nelson also saw Mallory waiting outside for Stewart and Vanasse. On this same occasion, at Stewart’s direction, Gaudreault retrieved the shotgun and placed it in Vanasse’s truck. Nelson also witnessed this incident. Rhonda Nelson, Gaudreault’s girlfriend, also saw parts of the newspaper incident and a few days later Gaudreault showed her the story about the Cumberland murders and told her that Stewart had done it.

[13] On January 31, 1990, Gaudreault left town with Garrett Nelson. He took with him a large quantity of hashish that belonged to Stewart. At the time he also still owed Stewart $13,000. He went to British Columbia. Before leaving town he told his sister Sylvie Gravelle that Stewart was involved in the murder of Paulo Trudel (the appellant Trudel’s brother) and Denis Roy. It was agreed at trial that these two deaths were suicides, not murders. When Sylvie Gravelle asked about the Cumberland killings, Gaudreault said that the four murders were all connected. Soon after Gaudreault left town, Stewart began making threatening telephone calls to Ms. Gravelle and to Rhonda Nelson about the money that Gaudreault owed him. Because of threats from Stewart, Ms. Gravelle and her husband sought police protection and Ms. Gravelle pressed Gaudreault for information about the Cumberland killings and also urged him to go to the police.

[14] In February and March 1990, Gaudreault met with police officers and gave them some information about the Cumberland murders. He did not, however, tell the police about his role in driving the car nor did he implicate Sauvé. Over time, and once Gaudreault had been given assurances about witness protection, Gaudreault gradually disclosed his involvement and the involvement of the four accused. In June 1990, Gaudreault returned to Ottawa and retraced the route he took on January 16th. The videotape of this drive was played at the trial. The Crown relied on the fact that Gaudreault was able to identify the place where he dropped off the appellants and Mallory, near the home of the deceased. It was the theory of the defence that many of the details about the killing, such as the number of shots fired and the manner of death, were available through the media and other sources. The deceased’s address, however, was not accurately identified in the media.

[15] Gaudreault was cross-examined for over 30 days. That cross-examination disclosed that he had lied to the police, fabricated evidence and lied at the preliminary inquiry. He had a lengthy criminal record and essentially had been living a life of crime for most of his life. He had continued to commit criminal offences after he went to the police and intended to do so in the future. Further, in January 1990, Gaudreault was using large amounts of drugs and he had been using cocaine on the day of the killing. On at least two occasions in the course of the court proceedings he asked Rhonda Nelson whether he had been hallucinating about the murders.

Jack Trudel

[16] Jack (or Jacques) Trudel is the appellant Trudel’s brother. He was involved with Trudel in the drug business. He was in jail at the time of the Cumberland murders and Trudel was handling his share of the business. Jack Trudel left jail in the spring of 1990. He became involved in a dispute with his brother over the profits from the drug business and clearly believed that his brother was stealing from him. In May 1990 he was staying with his brother and Sauvé at a residence in Constance Bay. The three of them were in the living room and his brother seemed not to be himself. Jack asked him what was wrong. His brother looked over at Sauvé and Sauvé indicated, “It’s okay”. His brother told him about his and Sauvé’s involvement in the Cumberland murders. He told him that Stewart had said that he was having a problem with a guy who owed him $2,000 and he asked Rick Trudel to help him out. Trudel agreed and called Sauvé and Mallory. The four of them went to the guy’s house. A girl opened the door. Stewart was supposed to intimidate the guy but the next thing they knew, Stewart had shot him. Sauvé then spoke up and told Jack Trudel that he had to shoot the girl because they did not want to go down for murder. So Sauvé went into the bedroom and shot the girl. He said, “I have no choice you know.” After the shooting they went to a strip bar and later disposed of the car, which was a white Cadillac. Rick Trudel and Sauvé then disposed of the guns in the river. It is apparent that the version as related by Jack Trudel is inconsistent with Gaudreault in many details such as where the group went after the killings and the disposal of the firearms and, most importantly, that Stewart shot Mr. Giroux.

[17] Jack Trudel also had a lengthy criminal record and was extremely violent. He had a history of making deals with the police in the sense of exchanging firearms for leniency. He only agreed to provide information to the police in January 1991, after the arrest of the appellants, after the police contacted him and asked him for information about the Cumberland murders. At the time, Jack Trudel was in custody awaiting trial on serious drug charges. After lengthy negotiations about the witness protection programme, Jack Trudel gave a statement to the police concerning the conversation with his brother and Sauvé. In February 1991, he pleaded guilty to a drug charge and was sentenced to ten years imprisonment. The Crown promised to reduce the sentence to seven years on appeal if Trudel cooperated. On appeal in 1994, that sentence was reduced to seven years on the recommendation of the Crown. In October 1992, Trudel was released on parole after a senior police officer, Detective Inspector MacCharles, appeared at the parole hearing. He was released to the supervision of the OPP witness protection officers. Trudel was unhappy with his accommodation. Apparently to cheer him up, one of the officers put him in touch with Gaudreault, who was also in witness protection. Trudel later visited Gaudreault and they eventually began to talk about their evidence. In April 1993, Trudel was charged with aggravated assault and rearrested. He was in custody at the time of the trial and no longer in the witness protection programme.

[18] Jack Trudel falsely implicated his brother and Sauvé in another murder and attempted to obstruct the course of the preliminary inquiry. He was unhappy with his treatment by the authorities, especially the witness protection programme, and was concerned that a videotape statement he had given to the police was circulating in the Ottawa underworld. He demanded a new identity and $300,000 from the witness protection programme. These demands were not met. The appellants seek to introduce as fresh evidence statements made by Jack Trudel in which he was said to have recanted his trial testimony.

Scott Emmerson

[19] Scott Emmerson is a life-long drug user and known to be a jailhouse informer. He was 26 years old when he testified at the trial. His family were friends with the Mallory family. He testified that while in custody in July 1991 he ended up on the same range as Sauvé and Mallory. He testified that Sauvé told him that he and Mallory had been paid to collect a debt but since there was no way that the guy could pay the debt they knew it was going to be a “hit”. Sauvé mentioned the names of the victims but Emmerson could not remember their names. Sauvé killed the man. The woman was not supposed to have been home. She ran into the kitchen and Sauvé chased her while Trudel laughed. Sauvé grabbed the woman, at which point Mallory put a hand on her shoulder. Sauvé told Mallory something to the effect that he had learned his lesson and was not going to make more mistakes; he was not going to leave any witnesses behind. He put a pillow over her head and shot her.

[20] Emmerson testified that the next day he overheard Sauvé telling another prisoner that he had “learned his lesson” before by leaving a witness and that he had done 4 years out of 7 years for killing a taxi driver in Montreal. One of the appellants’ grounds of appeal concerns the evidence the Crown led with respect to Sauvé’s Montreal manslaughter conviction.

[21] There were frailties in Emmerson’s account of the conversation with Sauvé. He too had a lengthy criminal record. Emmerson lied to the police initially about Mallory’s role in the killings and it is apparent that he received leniency from the courts on the recommendation of the Crown for his cooperation with the police in the Cumberland murder investigation. He also told the police that he overheard Trudel shout out an inculpatory statement while on the range with him in July of 1991. Emmerson later retracted that allegation and independent evidence showed that Trudel was not on the range with Emmerson in July of 1991. Emmerson testified that someone else must have made the statement but he did not know who it was. Emmerson continued to commit crimes after providing information to the police and was in and out of jail and the witness protection programme up to the time of the trial.

[22] The defence alleged that there were sources other than Sauvé for Emmerson’s knowledge of the killings. Emmerson’s father attended parts of the preliminary inquiry, that began in 1991, and there was a video tape circulating in the Ottawa underworld that shows the crime scene, including a very brief shot of the pillow on Manon Bourdeau, and police interviews with Gaudreault and Jack Trudel.

[23] Emmerson first came forward with his information in 1993. He had just been released from hospital after being severely injured by his uncle. He was on a cocaine binge and getting paranoid and called the police. The police put him in a cell until the drugs wore off and called Sergeant Davidson. Davidson was aware of the connection between the Mallory and Emmerson families and had asked to be notified if any of the Emmersons were arrested. Davidson asked Emmerson if he knew anything about the Cumberland killings. Emmerson responded yes and gave a version of the conversation he had with Sauvé in the jail. In this initial version he said that Mallory was not involved and was at Emmerson’s cottage water-skiing at the time of the killings. It was in this first statement that he said that Trudel had yelled information about the killings. In later statements and prior to testifying, Emmerson said that it was not Trudel, after the police challenged him about this part of his story.

[24] Sergeant Davidson raised the possibility of witness protection with Emmerson and asked him to think about it. About a month later, Emmerson was charged with fraud and related offences. He then telephoned Davidson and asked to make an official statement because he claimed that he now wanted to go straight and enter a strict drug treatment programme. He asked Davidson for help in getting bail. Davidson said he would look into it but later told him he could not intervene. Emmerson was released on bail on condition that he attend a residential drug treatment programme at Harvest House.

[25] Emmerson suffered a further relapse and found himself back in hospital after injecting drugs using dirty water. The police visited him and promised to take care of his security and help him with his drug problem if he testified. He decided to cooperate because he thought if he did not deal with his problems, including what he knew about the Cumberland killings, he would die. He gave a videotape statement under oath after being released from hospital. He was then placed in witness protection and placed in another drug rehabilitation clinic. He received a suspended sentence for the fraud charges. He subsequently re-offended and was incarcerated and removed from witness protection.

[26] The appellants seek to adduce fresh evidence of Emmerson’s post-trial recantation of his trial testimony.

The Defence Case

[27] The appellants did not testify but did introduce alibi evidence. There were some frailties with the alibis. It was part of the theory of the defence that the killings actually occurred on January 17th not the 16th. (The bodies were not discovered until January 18th.) Sauvé had a very strong alibi for the 17th. It is fair to say, however, that the weight of the evidence tended to support the Crown theory that the killings were on the 16th.

[28] The defence also adduced some evidence to contradict parts of the testimony of Gaudreault, Jack Trudel and Emmerson. This evidence also had some frailties.

[29] In our view, in the end, the case depended upon Gaudreault and the extent to which the jury found support for his evidence in the testimony of Jack Trudel and Emmerson and to a lesser extent in the testimony of the Gravelles, the Nelsons and Jamie Declare. We now turn to the grounds of appeal, beginning with the fresh evidence.

ANALYSIS

The Fresh Evidence

Scott Emmerson

[30] Scott Emmerson was released on mandatory supervision from a British Columbia penitentiary on May 15, 1998 and began using a variety of street drugs. On May 22, 1998, he took an overdose of Elavil tablets and became unconscious; in fact he thought he had died and was brought back to life by police and health care workers. On May 26th, he was transferred back to the Matsqui Institution and his mandatory release was revoked.

[31] Gary Barnes, an Ottawa lawyer, represented the appellant Trudel at the trial. On May 29, 1998, some two years after the trial, Scott Emmerson unexpectedly called Mr. Barnes. Mr. Barnes was not in the office the first two times Emmerson called. The third time, Mr. Barnes took the call and took notes of the conversation. Emmerson said that Sauvé had never told him the things that he had testified about and he thought the “driver” did the killings, although he had no evidence to support that theory. Emmerson said that he wanted to recant his evidence on television. He said that he had obtained the information about the killings from someone whom he did not identify and that “they” had “manufactured the whole thing”. He also said that he had been threatened by Detective Gary Dougherty, one of his police handlers. Emmerson said that he was scared, that everyone around the case was dying and he mentioned someone named John Last. He wanted to “get his soul back” and was willing to plead guilty to perjury.

[32] On August 17, 1998, Emmerson again called Mr. Barnes from Matsqui Institution. On this occasion he mentioned the Milgaard case and said it was “the same thing”.

[33] Emmerson spoke to Mr. Barnes a third time, sometime in the late summer or fall of 1998. By this time he was out of jail and confirmed that he was “still sticking to what I told you”.

[34] Mr. Barnes notified Trudel’s appellate counsel about the calls six months later. The respondent Crown was informed of the proposed fresh evidence in July 2001. In November 2001, at the request of the Crown, Toronto police officers conducted an investigation of the proposed fresh evidence. The Toronto officers asked [now] Superintendent Davidson, the officer who first approached Emmerson in 1993, to approach him on their behalf. Emmerson admitted talking to Mr. Barnes but told the police that he had not lied at trial and that he was “all fucked up and things were not going the way I expected”. He agreed to meet with the Toronto officers.

[35] On November 8, 2001, this court made an order authorizing Emmerson’s release into the custody of the Toronto officers. Those officers attempted to interview him. He denied that Detective Dougherty had threatened him. He denied making the calls to Mr. Barnes and suggested that Mr. Barnes fabricated the information. We understand the Crown on this appeal to accept that Mr. Barnes was truthful about the calls from Emmerson. In any event, there is abundant independent evidence to support the fact that Emmerson did make the calls to Mr. Barnes. Emmerson went on to say that he wished he had never become involved and that Sauvé had never spoken to him. He said that if he had made the calls to Mr. Barnes, it would have been out of sheer anger and rebellion about what had happened to him. He mentioned concerns about involvement of the Children’s Aid Society with his family and complaints about the witness protection programme. He said that if a new trial were ordered he would not testify.

[36] This court made an order to have Emmerson cross-examined before a special examiner. Emmerson, who was now out of custody, attended but claimed complete amnesia about the events testified to at trial or his conversations with Mr. Barnes or Superintendent Davidson. He blamed his amnesia on the drug overdose.

[37] Something must be said at this point about the roles played by Detective Dougherty and John Last in this case. Detective Dougherty was not involved in the case when the police first contacted Emmerson in September 1993 but was involved by the time Emmerson decided to become a witness in November 1993. After the appellants’ trial, in late 1997, Gaudreault and Jack Trudel were living together. On September 11, 1997, Gaudreault telephoned Detective Dougherty to say that Jack Trudel was in possession of a handgun. Dougherty consulted with Detective Inspector MacCharles who told him to visit Gaudreault with Detective George Snider and seize the gun. Dougherty took the gun and gave it to Snider who threw it into a lake. The officers made no record of the transaction and did not lay charges against Jack Trudel. They attempted to cover up the event even though Trudel was a witness at the pending trial of Stewart and Mallory. The cover-up began to unravel when Gaudreault told an officer with the witness protection programme about the events. From late 1997 to July 1998 the three officers lied about the events and attempted to suppress Gaudreault’s story. Dougherty eventually confessed his misconduct and was disciplined. The gun incident is obviously also important in respect of the Jack Trudel fresh evidence and we will make further reference to it when we discuss that evidence.

[38] John Last was a member of an outlaw motorcycle gang in Ottawa. As we noted earlier, some point before trial, a videotape made by the police at the crime scene began to circulate in the Ottawa underground. John Last surrendered the tape to police in February 1993. It was the theory of the defence at trial that Emmerson had a number of sources, other than his alleged conversations with Sauvé, for the information he testified to at the trial in 1996. For example, Emmerson’s father had attended portions of the preliminary inquiry. The defence also argued that it was possible he had seen the videotape. The videotape was important because it briefly (about five seconds) shows the pillow on Ms. Bourdeau’s head, a fact not released to the media, but referred to by Emmerson in his testimony. Emmerson denied seeing the videotape.

[39] At trial, Emmerson was asked if he knew John Last. He said, “Heard of him, don’t know him.” Last died of a drug overdose before trial. It will be recalled that Emmerson told Mr. Barnes that Last was one of the people connected to the case who was dying. When interviewed by the Toronto officers, Emmerson described Last as the partner of “Gramps” and he was aware of how Last had died. When he was cross-examined, Emmerson identified “Gramps” as his best friend. He said that his father had told him that Last used to hang around his father’s bar with Gramps. He claimed not to know Last personally.

[40] Thus, the Emmerson fresh evidence may be summarized as follows. First, it consists of a brief hearsay recantation in 1998 to Mr. Barnes. Emmerson has now recanted the recantation and claims no memory of the events. Second, there is a slightly stronger link between Emmerson and the videotape, which could be a source for some of the information Emmerson testified to at trial.

[41] In our view, the proposed Emmerson fresh evidence does not meet the test for the admission of fresh evidence. We accept Mr. Barnes’ version of the telephone calls from Emmerson. We are satisfied that Emmerson is not telling the truth when he denies having made the calls and in his claim of amnesia. Mr. Barnes’ evidence, however, would have only limited probative value. It is hearsay and is not admissible for its truth. The requirements laid down in R. v. B. (K.G.), [1993] 1 S.C.R. 740, even modified to take into account that this is evidence sought to be adduced by the defence rather than the prosecution, have not been met. Emmerson was not under oath nor warned about the consequences of lying and there is no complete accurate audio recording of the telephone calls. Thus, at best, Mr. Barnes’ evidence might have value to impeach Emmerson should he testify at a new trial and maintain his first trial evidence.

[42] The recantation meets the due diligence and relevancy parts of the test in R. v. Palmer, [1980] 1 S.C.R. 759. The real issues are whether the evidence is credible in the sense that it is reasonably capable of belief and is such that, if believed, could when taken with the other evidence adduced at trial, reasonably be expected to have affected the result. As this court noted in R. v. Babinski (1999), 135 C.C.C. (3d) 1, the question of credibility is problematic in a case such as this where there is little dispute that the person to whom the trial witness recanted is credible but the recantation, now retracted, may not be credible. We have considerable doubt that the recantation itself is credible. The defence and Crown offer competing theories for why the recantation might be true or false respectively. In light of Emmerson’s subsequent feigned amnesia, it is simply not possible to test either theory. There is, however, little on this record that provides a solid foundation for preferring the defence explanation. The idea that someone like Emmerson suddenly wanted to clear his conscience or that he momentarily underwent some kind of religious or moral conversion seems highly unlikely. The theory that the discredited police officer, Detective Dougherty, was instrumental in assisting Emmerson to concoct his evidence fails to take into account that Emmerson first told his story, admittedly not the whole story, to Superintendent Davidson who was not implicated in the Jack Trudel gun cover-up.

[43] In any event, as in Babinski, we prefer to deal with the admissibility question by considering the fourth Palmer factor: whether the proposed evidence could have affected the result. The appellants must demonstrate that the proposed evidence, when taken with the other evidence, including the retraction of the recantation, could be expected to have affected the result. We accept for the purpose of considering this argument that while Gaudreault was central to the Crown’s case, the trial testimony of Emmerson, himself a very suspect witness, was nevertheless capable of confirming the credibility of a highly suspect witness.

[44] The reason that we are not satisfied the recantation could have affected the result is that it was already before the jury that Emmerson had attempted to recant his evidence. About one year after Emmerson testified, but still in the course of the trial, Superintendent Davidson testified that Emmerson had been removed from the witness protection programme and was serving a penitentiary sentence for robbery and break and enter. Davidson had heard that Emmerson wanted to recant. Davidson spoke to Emmerson, who denied intending to recant but admitted that he may have told his lawyer otherwise out of fear. The recantation to Mr. Barnes is strikingly similar to the events testified to by Davidson at the trial. The fact that Emmerson had once again briefly recanted while in custody but retracted his recantation after speaking to Davidson would have added nothing to the jury’s assessment of his credibility. Given everything else the jury knew about Emmerson—his lies to police, his perjury, his attempts to obstruct justice, his criminal record and history of criminal activity—it is not reasonable to think this additional piece of evidence, admissible only for the limited purpose of further impeaching Emmerson’s credibility, if believed, could reasonably, when taken with the other evidence adduced at trial, be expected to have affected the result.

[45] In making this finding, we also think it is proper to take into account that defence counsel did not seek to call Emmerson to testify about the alleged recantation. They were content to rely upon Davidson’s hearsay version. This tactical decision suggests that defence counsel did not think confronting Emmerson with these types of recantations was of any significant forensic value.

[46] We are also not satisfied that the evidence concerning John Last should be admitted as fresh evidence. While the evidence is relevant, it is fair to take into account the lack of due diligence in pursuing the matter. The record shows that in the lengthy cross-examination of Emmerson he was asked the single question about John Last. With reasonable diligence the defence could have uncovered Last’s relationship with Mallory’s father’s bar and thus the alleged link to “Gramps”. At the time Emmerson was cross-examined, Mallory was still part of the trial.

[47] In any event, the proposed evidence concerning Emmerson’s relationship with Last is not such as to reasonably have affected the verdict. The fact that Emmerson had sources, other than his alleged conversation with Sauvé, was already before the jury. The proposed fresh evidence, while modestly strengthening the inference that a source for the information could have been the videotape that ended up in Last’s hands, still does not show that Emmerson actually saw the videotape. In all of his various statements since the trial, Emmerson has never contradicted his trial evidence that he did not see the videotape.

[48] Accordingly, for these reasons we would not admit the Emmerson fresh evidence.

Jack Trudel

[49] It will be recalled that in September 1997, officers Dougherty and Snider seized a handgun from Trudel and then tried to cover up the event. From January to July 1998, other police officers were investigating the incident and in May 1998 one of them contacted a friend of Jack Trudel’s. Jack Trudel became upset about the inquiries and, while drinking, telephoned Susan Mulligan, counsel for Robert Stewart, whose trial was pending in Ottawa. She taped the telephone call. Trudel began to make threats to kill Gaudreault, who he blamed for the investigation. He also said that he had lied at the appellants’ trial. Ms. Mulligan contacted the Crown’s office about the threats. In earlier conversations between Ms. Mulligan and Jack Trudel, Trudel had admitted that he had lied at the trial but never indicated a willingness to formally recant because he feared prosecution for perjury.

[50] In February 1999, Susan Mulligan contacted appellate counsel for Trudel and advised him that Jack Trudel wished to recant his trial testimony. Jack Trudel then met with Andras Schreck and Catherine Glaister at Ms. Mulligan’s office. Trudel refused to be videotaped but did sign a statement. He would not sign the statement under oath until he spoke to a lawyer. In the statement, he says that he had no knowledge of who killed Bourdeau and Giroux, that his brother and Sauvé never said anything about the murders and his testimony at trial about their involvement was fabricated. He said that he had falsely implicated the appellants to the police because he was angry with his brother over the drug business and because his brother had put a contract out on his life. He also wanted to ensure that the police would keep their bargain. As indicated, Trudel had made an agreement with the Crown that his ten-year sentence for certain drug charges would be reduced to seven years. He was also upset with Sauvé because he had slapped his younger brother who committed suicide later the same day.

[51] Trudel said that his motive for coming forward after the trial was because he had not been treated fairly by the police and the witness protection programme. After his release from prison, Trudel applied to be reinstated in the witness protection programme. In June 1997, he was told that he would be supported for a short time by the O.P.P. but that he would not be given witness protection financing. In December 1997, Trudel told Snider that he had been manoeuvred into testifying and was going to court to do the “right thing” and that previously he had done the “wrong thing”. These statements to officer Snider were disclosed to counsel for Stewart and Mallory and as indicated, Stewart’s counsel then contacted the appellants’ appellate counsel. It is unclear whether Trudel thought he would be a witness at the Stewart/Mallory trial or was referring to this appeal.

[52] In January 1999, R.C.M.P. officers attempted to interview Trudel about the gun incident. He refused to talk to them about the incident and to implicate Dougherty, Snider and MacCharles and said that he would be a defence witness at the Stewart and Mallory trial.

[53] On May 12, 1999, Jack Trudel telephoned both Mr. Schreck and Ms. Glaister and told them that he had lied to them and that his trial testimony was true. The appellants suggest that this call was probably a result of influence exerted by Snider or Dougherty that day. This suggestion is based on the following chain of events. On May 6 and 11 respectively, Dougherty and Snider gave statements to the R.C.M.P. admitting to their role in the gun incident and cover-up. On May 11, the R.C.M.P. asked Snider to arrange for Trudel to give a statement and take a polygraph. Snider said he would do so. Although they had been ordered not to, Snider and Dougherty had remained in contact with Trudel during the R.C.M.P. investigation. On May 13, Dougherty contacted the R.C.M.P. and told them that Trudel would cooperate.

[54] On November 16, 2001, the Toronto police officers investigating the proposed fresh evidence met with Trudel. Trudel confirmed that he had made the April 25, 1999 statement and that it was true. He would not say why he had lied at the trial and refused to discuss the statement further. The officers asked Trudel about his relationship with the witness protection programme and his grievance with the programme. Trudel said it was all in the April 25, 1999 statement. The officers asked Trudel about the May 12, 1999 calls when he retracted the recantation. He said that he had no recollection of those calls. After the interview was over and the tape recorder was turned off, one of the officers suggested that he thought Trudel had received a “raw deal” from the programme. At this point, Trudel pointed at the officer and told him to tell the head of the programme that “what goes around comes around and I’ll go to jail if I have to”.

[55] In January 2002, Trudel was brought from jail to the Ottawa courthouse to be examined by counsel about his recantation. Over two days, Trudel refused to be sworn. The evening of the second day, Trudel asked to see the Toronto officers. They visited him about a week later. There was discussion about Trudel’s refusal to be sworn. He then told the police that he would only talk if he were given immunity from perjury. He also wanted his pending charges “cleared up”.

[56] As with the Emmerson recantation, we accept that the proposed Trudel fresh evidence meets the tests for relevancy and due diligence. We therefore turn to the credibility and effect on the verdict parts of the test.

[57] On the credibility aspect of the case, the Crown submits that Jack Trudel’s recantation bears a marked similarity to the fresh evidence sought to be introduced in the Palmer decision. In Palmer, the Supreme Court of Canada held that the British Columbia Court of Appeal properly refused to admit as fresh evidence statements by one Ford, the principal witness against the accused, that his trial testimony was a fabrication. Ford made his recantations after his request from the authorities after the trial for a payment of $60,000 was refused. Instead, the authorities agreed to give him $25,000. The Supreme Court considered Ford’s trial testimony to be true and the recantation to be incredible.

[58] The chronology of events in this case is similar. Trudel recanted his testimony to Ms. Mulligan and then more formally in the April 1999 statement after his attempts at reinstatement in the witness protection programme were rebuffed. It is apparent that before, during, and after the trial, Trudel has been dissatisfied with the programme and the authorities’ refusal to pay him a large sum of money ($300,000). While he has recanted his testimony, he has also, albeit briefly, retracted that recantation and then reaffirmed it. He has, however, refused to be examined under oath about the recantation.

[59] While there is much force to the Crown’s argument, we are of the view that the Trudel fresh evidence cannot be rejected as incredible. In Palmer, the Supreme Court of Canada was satisfied that Ford’s trial evidence was credible based on an examination of the entire record. The fresh evidence, particularly given Ford’s evident motive for recanting, was incredible. This case is quite different. There are many frailties with Jack Trudel’s trial evidence that make it highly suspect. That testimony was given in response to express promises from the authorities, such as the reduction in his prison sentence. Trudel himself was an unstable and violent individual. He falsely implicated the appellants in another murder. There are highly credible explanations for why Trudel may have lied at the trial.

[60] There is actually little to choose between Trudel’s trial version of events and the version as described in the April 25th statement. Unlike the recanting witness in Palmer, Trudel’s testimony incriminating the appellants consists of nothing more than brief statements by the appellants. There are almost no details, and what details are provided, are largely inconsistent with Gaudreault’s evidence. In Palmer and in a similar case in this court, Reference re: Kelly (1999), 135 C.C.C. (3d) 449, the recanting witness was the principal Crown witness who gave a detailed account of the events rich with detail.

[61] It seems to us that the statement would, at least, have substantial value for impeachment purposes. The evidence should not be rejected at the credibility stage: Babinski at paragraphs 55 to 59.

[62] Thus, as in Babinski, the admissibility of this evidence turns on the fourth Palmer criterion. The appellants argue that the evidence of the recantation could have affected the verdict because it undermines the Crown position at trial that Trudel had never changed his story about what his brother told him, and was a man of principle who was influenced to come forward by the enormity of the crime committed by his brother and his co-accused. See Babinski at para. 72. They rely upon various portions of Crown counsel’s jury address such as the following:

Jack Trudel threatened to refuse to testify many times, and he in fact refused to testify many times. He used whatever means he had at his disposal to enforce the promises which had been made to him, not to make a new deal, just to enforce the old one. It just never worked out. He never lied, though. He just refused to testify.

There is nothing being dangled in front of Jack Trudel. There is not carrot out there. The Prosecution had nothing to offer him when he decided he would testify before you at this Trial. Absolutely no pressure was brought to bear on him, no promises were made. Come if you want to, we would like to have you, but if you are not coming, that’s okay too. How many times did he say to you, “I still got no deal. I got absolutely nothing”?

With Jack Trudel what you see is what you get and what you got was a man with a long criminal past, yet in a strange way he was a principled man. His principles could not accept that his own brother and the others would kill a pregnant woman, that to him is not the way to do business, and if anyone was going to fold in this massive conspiracy surely it was Jack. He hates the way the OPP Witness Protection Program handled him, so much so that he may sue them, but he knows what he was told and now through his courage and his determination so do you. You can believe him without corroboration. In my submission it’s very easy to understand Jack if you step back and realize that not everyone lives by the same moral standards as you or I do, but Jack could not be moved on what he was told by these two men. You can depend completely on that [emphasis added].

[63] The jury address is also replete with references to Jack Trudel having decided to do “the right thing” for once. We also take into account that while the trial judge made a fulsome review of Trudel’s evidence he concluded that review with the following, summarizing Trudel’s testimony:

He did not testify out of revenge against anyone, including Sauvé, Mallory or Stewart, notwithstanding his run-in with Stewart. As for his brother, he said he knows Rick broke his agreement but that’s as far as it goes. He is not testifying out of revenge or spite.

He admitted in chief that if he had not been charged with trafficking he probably would not have come forward to testify. His decision to testify itself required him “to weigh all the consequences”. As he said: “You just don’t jump into a decision like that. It’s not an easy decision to make. There’s a lot of things you have to take into consideration here, you know.” He would have preferred to continue his criminal life and the lifestyle it afforded him. In a sense his “involvement” in this case has not “benefited him.” He said: “If anything I lost, I ruined my life in a sense.” As far as getting a good deal on his conspiracy charges, he said even after the appeal reduction of three years he still got the most of anyone sentenced under that operation. As far as he is concerned he has no understandings or commitments to help him after he gets out of jail; he is on his own. He does not think he owes anyone [emphasis added].

[64] The appellants point out that in his recantation Jack Trudel reveals his real reasons for coming forward; his grudge against his brother over the drug business, his desire to get a lesser sentence on the drug charges he was facing and payments from the police. On the other hand, his recantation must be read as a whole. It also shows that Trudel was unhappy because he was not sent to medium security as promised, had not received early parole, and had not been treated fairly by the police and by the witness protection programme. It should also be noted that in his review of the defence positions, the trial judge did point out the many reasons why the jury should not accept Trudel’s evidence, including his grievance against his brother and the reduction in his sentence because of his cooperation.

[65] In our view, the appellants have met the final part of the Palmer test. The fresh evidence of Trudel’s recantation, if believed, could reasonably, when taken with the other evidence adduced at trial, be expected to have affected the result. We recognize that the Crown’s case principally depended on Gaudreault’s testimony. But, the Crown relied on Trudel to support that evidence, notwithstanding the discrepancies between his evidence and the evidence of Gaudreault. It is probable that the jury would disbelieve Trudel once they learned of his recantation. The balance of the Crown’s case is not so strong that the verdict would inevitably have been the same. There was some evidence to support Gaudreault, but much of it was suspect and subject to its own substantial frailties, some of which we have set out earlier. In fact, in her jury address, Crown counsel suggested that Gaudreault’s evidence in the final analysis may not have been needed:

You can arrive at the guilt of Richard Trudel and James Sauvé for First Degree Murder without Denis Gaudreault. That is not a concession that I think you should disregard. I am just suggesting another approach because the approach that the Defence have taken is to concentrate on Denis Gaudreault.

When I review the evidence with you, I am going to leave him out altogether at first. And when you see that you can be convinced without him, then we will overlay his evidence on top of everything else we already know. You will see how well it all fits together. It is like a marking template on a perfect exam; every box is going to be checked off just as it should be.

[66] Crown counsel then embarked on a review of the evidence including a lengthy review of the evidence of Emmerson and Jack Trudel.

[67] Since we have decided to admit the fresh evidence, on that ground alone, the appeal must be allowed. We will, however, consider the other grounds of appeal.

Grounds of Appeal Relating to the Conduct of the Trial

Introduction

[68] Before dealing with the various errors alleged by the appellant we wish to say something about the conduct of this trial. The investigation of the offences was lengthy and difficult and this produced a long and difficult trial. Many of the witnesses were deeply involved in the Ottawa criminal underground and the fair presentation of their evidence posed serious problems. The trial judge’s approach to the case was thoughtful and even-handed. As just one example, his charge to the jury was in many respects masterful. He did not simply review the evidence of the witnesses seriatim or leave the evidence to the jury en masse but integrated their stories to present a coherent explanation of the case. We have attempted to approach this case bearing in mind the many difficulties faced by the trial judge and counsel at the trial.

[69] This court does, however, have an obligation to ensure that the law is properly applied so that the appellants obtained a trial that did not produce a substantial wrong or miscarriage of justice. That obligation does not disappear because a trial, like this one, was unusually long and complex, or because a retrial may be taxing to the administration of justice.

The Vetrovec warning

[70] The appellants’ principal ground of appeal relating to the conduct of the trial concerns the alleged inadequacy of the trial judge’s warning to the jury about the frailties in the evidence of the three Crown witnesses Gaudreault, Jack Trudel, and Emmerson. The appellants submit that having regard to the position occupied by these three witnesses in the Crown’s case, their unsavoury backgrounds and other factors, a Vetrovec warning (R. v. Vetrovec (1982), 67 C.C.C. (2d) 1 (S.C.C.)) was mandatory. We agree with that submission. To the extent that some of the trial judge’s comments suggest that this was a matter of discretion, he was in error. The evidence of these three witnesses, especially Gaudreault, was central to the Crown’s proof of guilt. All three witnesses had such significant frailties that a warning was mandatory. All three lied to the police, selectively disclosed their stories, sought benefits of various kinds for their testimony, were part of the Ottawa criminal underground, had lengthy criminal records, and were heavy users of illicit drugs. It was open to the jury to find that they had lied under oath either at the preliminary inquiry or the trial. The evidence of Emmerson had the added defect that he was in jail with Sauvé when he claimed to have had the incriminating conversations with him. These factors and their central role in the case mandated the warning. See R. v. Bevan (1993), 82 C.C.C. (3d) 310 (S.C.C.) and R. v. Brooks (2000), 2000 SCC 11, 141 C.C.C. (3d) 321 (S.C.C.).

[71] The appellants’ submission focuses on the adequacy of the Vetrovec warning that the trial judge did give. The trial judge gave the following warning near the beginning of his lengthy jury charge before he began his review of the evidence:

Credibility is important in this case. There are witnesses whose criminal records are quite extensive, and whose records do not include all their criminal activity. Some witnesses have admitted to lying under oath whether or not a charge of perjury can be successfully laid. Obviously at least three of those witnesses are important to the Crown’s case. They are Denis Gaudreault, Scott Emmerson and Jacques Trudel. Where witnesses have such unsavoury backgrounds, it is prudent to examine their evidence carefully, and to look at the evidence of other witnesses or exhibits to see if they support the evidence of witnesses like Gaudreault, Trudel and Emmerson. It will be for you to determine how much supporting evidence you would require to make any particular part of their evidence acceptable. There may well be parts of their evidence which you find acceptable as given because of the way it was given or the common sense supporting such an inference.

Of course, you must remember that simply because a person is unsavoury does not mean that his or her evidence must be rejected without it being adequately considered. It is not unusual for witnesses of unsavoury character to be called as witnesses when the facts of the case require an examination of an illegal business. Who are the employees of an illegal business likely to be? They are often likely to have criminal records and have had other brushes with law enforcement. As with any witness, a trier of fact must dig into his or her evidence and consider it adequately. As part of that consideration it may be prudent for you, where you consider it advisable, to look for evidence which tends to support the evidence of those three particular witnesses. The Crown and the defence have referred you to the particular parts of the evidence that they think are supportive and non-supportive respectively. The ultimate decision is for you to make. Remember that as the triers of the facts you may choose to accept all, part or none of what any witness says, whether that witness be unsavoury or not. The criteria on which you base these decisions lie entirely in your hands. You are the judges of the facts [emphasis added].

[72] Some of the other directions the trial judge gave are also important, including the following concerning the use of criminal records.

The mere fact that any witness in this case has criminal convictions does not, of itself, destroy or impair his or her credibility, but it may indicate a lack of moral responsibility to tell the truth. The fact that a witness has a criminal record is something that you weigh in the balance in assessing his or her trustworthiness as a witness. They may be used to assess the credibility of such witnesses on the basis that having one does not destroy credibility automatically, but it may indicate more irresponsibility towards the truth. Within those criteria it can be taken into account in weighing the credibility of any witness who has a criminal record.

[73] Counsel for the appellants objected at trial to the adequacy of the directions. They argued that the trial judge was required to give a much stronger warning and that much of the force of the warning was diluted by the trial judge’s observation that it was not unusual for the prosecution to depend on these kind of witnesses. The trial judge refused to recharge the jury and gave this explanation:

I don’t intend to change the Vetrovec warning. The Court spent a lot of time on that. I considered it in relation to the background of the total case, the amount of time spent on credibility issues and lying. It’s front and center with the jury. I did it the way I did it, to be blunt, because I didn’t want the jurors to have an emotional reaction to the witnesses and walk out and take a day and come back in and throw up their hands. What I wanted them to do was to get into the case, to get their hands dirty, to use their power to believe all, part or none of what a witness said and to do an analysis of what the people are saying, and that’s why it’s worded exactly as it is and that’s why I wouldn’t change one word of it.

The warning is there, they know the problem, but they have to be intellectual about it and they have to be rational about it and that’s why I did it that way [emphasis added].

[74] The appellants submit that this warning had the following defects:

(1) Failed to focus the jury on the witnesses’ characteristics that made their evidence suspect.

(2) Failed to tell the jury that it would be “dangerous” to act on the unconfirmed evidence of these witnesses.

(3) Instructed the jury that it was open to them to act on the unconfirmed evidence of the witnesses “because of the way it was given”.

(4) Stated that it is “usual” for such witnesses to be called when illegal activities are the subject matter of a trial.

[75] We will discuss each of these submissions in turn. However, we first wish to set out some general comments about the context of this trial and the purpose of the Vetrovec warning. This was a very unusual trial. As we have said, it lasted for fifteen months and a good part of that time was taken up with the testimony from these three witnesses. Specifically, Gaudreault testified for 26 days; 23 of those days were taken up with cross-examination. Jack Trudel testified for 16 days; 12 days in cross-examination. Emmerson testified for 7 days; 5 days in cross-examination. A relatively small portion of the time spent in cross-examination concerned the actual incriminating evidence given by these witnesses. Most of the cross-examination was taken up with questions concerning the witnesses’ characters, their various illegal activities, their dealings with the police and their motives to lie. By the end of this almost 50 days of cross-examination the jury had as complete a picture as is possible in a courtroom setting of the character of these witnesses and what brought them to court to tell their stories.

[76] The purpose of the Vetrovec warning is to alert the jury that there is a special need for caution in approaching the evidence of certain witnesses whose evidence plays an important role in the proof of guilt. The caution is of particular importance where there are defects in the evidence of a witness that may not be apparent to a lay trier of fact. Perhaps the most important of these is the jailhouse informer. Recent experience has shown that jailhouse informers are a particularly dangerous type of witness. The Report of the Commission on Proceedings Involving Guy Paul Morin (Toronto: Ontario Ministry of the Attorney General, 1998) and The Report of the Inquiry Regarding Thomas Sophonow (Winnipeg, Man.: Manitoba Justice, 2001) have shown that these witnesses can be very convincing liars and are capable of fabricating evidence. The Morin Inquiry Report was released in 1998 and the Sophonow Inquiry Report was released in 2001. The trial judge therefore did not have the benefit of these reports. This recent experience also shows that the motives of these witnesses may not always be apparent and that their expressed purposes for testifying, such as a distaste for the accused’s particular crime, or to tell the truth and make a clean break from their criminal past are simply untrue. Their claims that they neither sought an advantage nor received one have been shown to be patently false. The Honourable Peter deC. Cory summed up the problem presented by these witnesses in his report on the Sophonow prosecution in these words:

It is true that Justice Dickson, in Vetrovec v. The Queen, 1982 20 (SCC), [1982] 1 S.C.R. 811, cautioned against placing witnesses in pigeon holes so that only some classes of witnesses would require warnings regarding their testimony. Nonetheless, jailhouse informants are in a special class with the demonstrated ability to mislead and deceive the most discerning and experienced observers. They have, as a class, established a unique record of consistently giving false testimony. They must be given special attention and their evidence should generally be excluded and only be admitted in very rare cases. On those rare occasions that it is admitted, it must be approached with the greatest caution.

It is not unduly difficult for jailhouse informants to obtain information, particularly in high profile cases, which would appear to come only from the perpetrator of the crime. As a result, they appear to be reliable and credible witnesses. This case demonstrates that experienced police officers considered very unreliable informants to be credible and trustworthy. Crown Counsel obviously thought that they were credible witnesses who should be put forward. If experienced police officers and Crown Counsel can be so easily taken in by jailhouse informants, how much more difficult it must be for jurors to resist their blandishments. How difficult, if not impossible, it is for jurors to appreciate the polished and practiced facility with which they deliver false testimony. Jailhouse informants are, indeed, a dangerous group. Their testimony can all too easily destroy any hope of holding a fair trial and severely tarnish the reputation of Canadian justice.

The findings [concerning jailhouse informants] can be summarized in the following manner:

1. Jailhouse informants are polished and convincing liars.

2. All confessions of an accused will be given great weight by jurors.

3. Jurors will give the same weight to “confessions” made to jailhouse informants as they will to a confession made to a police officer.

4. “Confessions” made to jailhouse informants have a cumulative effect and, thus, the evidence of three jailhouse informants will have a greater impact on a jury than the evidence of one.

5. Jailhouse informants rush to testify particularly in high profile cases.

6. They always appear to have evidence that could only come from one who committed the offence.

7. Their mendacity and ability to convince those who hear them of their veracity make them a threat to the principle of a fair trial and, thus, to the administration of justice.

[77] Of the three witnesses with whom we are concerned in this case technically only Emmerson falls within the jailhouse informer category since only he claims to have received a confession from one of the appellants while in jail with him. Even at that, Emmerson was slightly different from the usual jailhouse informer since when he provided the information to the police he was no longer in jail and the police approached him. He did not approach them seeking favours, although he certainly sought them afterwards. That said, all three witnesses to one degree or another exhibited characteristics that required special care. All three witnesses had some clearly expressed motives for testifying but there may have been other motives that would have been difficult to detect. For example, Emmerson and Jack Trudel claimed that they came forward, in part, because of their horror over the killing of the pregnant Ms. Bourdeau. In one way or another, Emmerson and Trudel also claimed some kind of epiphany that led them to testify. For Emmerson, it was his meeting with Reverend Main and his near death experience from the drug overdose. For Trudel, it was his realization following his 1990 arrest for conspiracy that crime does not pay. Judicial experience instructs that triers of fact should be cautioned to treat these claims with scepticism.

[78] Informers, especially jailhouse informers, have means to obtain information other than from the accused. The prosecution will often take the position that this evidence must have come from the accused since the informer had no other means of obtaining it. Informers are resourceful and well capable of obtaining information that they later plant into the mouth of the accused. This was a particular problem with Emmerson who seemed to have details of the crimes and of Sauvé’s background that could only have come from Sauvé or one of the other accused. Yet, he had deep roots in the Ottawa criminal underground and therefore access to the information attributed to his conversations with Sauvé.

[79] There were other problems with these witnesses that required special care. For example, it was shown that Gaudreault had lied under oath and had actually manufactured evidence in order to bolster his credibility. He purchased a blank disk and presented it to the police claiming it contained Stewart’s drug information. He maintained this lie under oath at the preliminary inquiry. Yet, when he testified at the appellants’ trial, he claimed to be telling the truth. The jury needed to understand that he and the other two principal Crown witnesses were quite capable of lying and manipulating the truth to an astonishing degree, but also present as confident and honest before the jury.

[80] Gaudreault presented a special problem because his close association with the murders would enable him to provide details of the crime and falsely implicate the accused in order to cover up his own involvement. On his story, he was merely the driver, an unwitting dupe who fully realized what happened only after the fact, perhaps when he described the events of the evening to Declare who told him that he had just participated in a “hit”. Again, it was important that the jury understand that such witnesses can have particular motives to lie.

[81] In sum, the need for this special care and for informing the jury of the reasons for the special care rests on the concern that the lay members of the jury simply do not have the necessary experience to adequately assess the credibility of these types of witnesses. This is not to say that such witnesses are incapable of telling the truth or that their evidence can never safely constitute an acceptable basis for a conviction. Rather, we say only that this kind of evidence must be approached with caution.

[82] The cases establish four characteristics of a proper Vetrovec warning:

(1) the evidence of certain witnesses is identified as requiring special scrutiny;

(2) the characteristics of the witness that bring his or her evidence into serious question are identified;

(3) the jury is cautioned that although it is entitled to act on the unconfirmed evidence of such a witness, it is dangerous to do so; and

(4) the jury is cautioned to look for other independent evidence which tends to confirm material parts of the evidence of the witness with respect to whom the warning has been given.

[83] See Vetrovec at p. 17; Brooks at para. 93 to 97 (per Major J. dissenting on another ground); and R. v. Suzack (2001), 141 C.C.C. (3d) 449 (Ont. C.A.) at para. 187. In our view, the warning given by the trial judge did not meet these minimum requirements. We will discuss our concerns by reference to these minimum requirements. We will also deal with the other issues raised by the appellants; the use of demeanour and instructing the jury that these types of witnesses are not unusual in a prosecution of this type.

(1) Special scrutiny

[84] The direction given by the trial judge did fulfill this part of the Vetrovec warning. The three witnesses Gaudreault, Emmerson and Trudel were singled out for special attention. See R. v. Bevan (1991), 63 C.C.C. (3d) 333 (Ont. C.A.) at 362 per Osborne J.A. dissenting (the decision of the majority was overturned by the Supreme Court of Canada (1993), 82 C.C.C. (3d) 310). In any event, it would have been obvious to the jury that the evidence of these witnesses required special attention given the focus of the trial.

(2) Reasons for special scrutiny

[85] In our view, the directions given by the jury did not meet this requirement of the Vetrovec warning. The trial judge referred to the fact that the three witnesses had extensive criminal records that did not include all of their criminal activity and that they had lied under oath. But this was only part of the problem. What was required was, in the words of Dickson J. in Vetrovec at p. 17, “a clear and sharp warning to attract the attention of the juror to the risks of adopting, without more, the evidence of the witness”. Those risks have been catalogued above. Without understanding the reason for the need for special scrutiny, the jury would not be able to accurately assess the risk of acting on the evidence of these three witnesses.

(3) “Dangerous” to act on unconfirmed evidence

[86] The usual language employed in a Vetrovec warning refers to the danger of acting on the unconfirmed evidence of the suspect witnesses. There is, however, no particular magic in the use of the words “danger” or “dangerous”. Thus, depending on the circumstances, the nature and the content of the Vetrovec warning are a matter of discretion. See R. v. Harriott (2002), 161 C.C.C. (3d) 481 (Ont. C.A.) at para. 32, affirmed (2003), 2003 SCC 5, 171 C.C.C. (3d) 351 (S.C.C.). This court explained the rule in R. v. Jones (2001), 146 O.A.C. 118 at para. 3:

While it certainly would have been open to the trial judge to have given a stronger warning relating to Graves' evidence, he committed no error in approaching the matter as he did. He was very clear on the points which made Graves' evidence suspect, her prior criminal record, the fact that she was also charged with the murder and stood to gain from inculpating the appellant, and the fact her evidence was inconsistent with prior statements. In our view, this warning was sufficient to bring home to the jury the dangers her evidence posed. Vetrovec itself holds at p. 17 that no particular formula is required and that the matter lies within the discretion of the trial judge. In our view the trial judge did not go outside the acceptable limits of his discretion [emphasis added].

[87] Thus, it is not the particular phrase used by the trial judge, but whether the language he used conveyed the need for the level of caution required by the circumstances of this case. The issue then is whether the trial judge’s decision to structure the Vetrovec warning in the way that he did went “outside the acceptable limits of his discretion”. As we have pointed out, the trial judge’s decision to structure the Vetrovec warning in the way that he did was deliberate. To repeat, he gave this reason for doing so:

I did it the way I did it, to be blunt, because I didn’t want the jurors to have an emotional reaction to the witnesses and walk out and take a day and come back in and throw up their hands. What I wanted them to do was to get into the case, to get their hands dirty, to use their power to believe all, part or none of what a witness said and to do an analysis of what the people are saying, and that’s why it’s worded exactly as it is and that’s why I wouldn’t change one word of it.

The warning is there, they know the problem, but they have to be intellectual about it and they have to be rational about it and that’s why I did it that way [emphasis added].

[88] In our view, this was not a reasonable exercise of discretion. The trial judge’s rationale, that he did not want the jurors to have “an emotional reaction to the witnesses”, was wrong. It was based on the faulty premise that if given a strong direction, the direction required by the circumstances of the case, the jurors would not obey their oath. Such an approach is inconsistent with the law as laid down by the Supreme Court of Canada. For example in R. v. Corbett (1988), 41 C.C.C. (3d) 385 (S.C.C.) at 400-1, Dickson C.J.C. rejected the argument that jurors would be unable to understand the limited use of an accused’s criminal record:

In my view, it would be quite wrong to make too much of the risk that the jury might use the evidence for an improper purpose. This line of thinking could seriously undermine the entire jury system. The very strength of the jury is that the ultimate issue of guilt or innocence is determined by a group of ordinary citizens who are not legal specialists and who bring to the legal process a healthy measure of common sense. The jury is, of course, bound to follow the law as it is explained by the trial judge. Jury directions are often long and difficult, but the experience of trial judges is that juries do perform their duty according to the law. We should regard with grave suspicion arguments which assert that depriving the jury of all relevant information is preferable to giving them everything, with a careful explanation as to any limitations on the use to which they may put that information. So long as the jury is given a clear instruction as to how it may and how it may not use evidence of prior convictions put to an accused on cross-examination, it can be argued that the risk of improper use is outweighed by the much more serious risk of error should the jury be forced to decide the issue in the dark [emphasis added].

[89] Virtually without exception our law permits the parties to call any witness with relevant evidence to give no matter how disreputable, unsavoury, or unreliable the witness may be. But, the law requires that the trial judge give the jury the directions called for by the circumstances. It is wrong to proceed on the basis that the jury cannot be trusted to give proper attention to the evidence if given appropriate instructions. This case demanded the strongest possible warning about acting on the evidence of the three central witnesses. It was not sufficient to simply instruct the jury that it may be prudent, where they considered it advisable, to look for supporting evidence. The warning in this case was mandatory. The jury, of course, had the option to act on the evidence of the three suspect witnesses, but treating their evidence with the utmost care was not optional. This case required a direction that it was “dangerous” to act on the unconfirmed evidence of the three principal witnesses. Despite the broad discretion that a trial judge has in structuring the Vetrovec warning, and the deference that this court must accord that decision, this is one of those few cases where the trial judge’s exercise of discretion was unreasonable.

(4) Other independent evidence

[90] As Major J. explained in Brooks at para. 95, the Vetrovec warning “should also be accompanied by a reference to the evidence capable of providing independent confirmation of the unsavoury witness’s testimony. The independent confirmation relates to other evidence that would support the credibility of the unsavoury witness.” The trial judge reviewed at very considerable length the evidence of all the important witnesses. He did not single out for special attention the testimony that was capable of supporting the testimony of the three unsavoury witnesses. The appellants did not argue that they were prejudiced by this omission.

[91] While there was not a lot of independent evidence supporting the three main Crown witnesses, there was some and it would have been preferable for the judge to have referred to some examples along with an appropriate caution about the frailties attached to that confirmatory evidence. An example is the evidence of Jamie Declare. Declare testified that he saw Gaudreault in the driver’s seat of a white Cadillac and was talking to him when Stewart pulled up. This could be important confirmatory evidence except that he was unable to identify Mallory as being in the car and his identification of the appellants was uncertain. Further, he remembered seeing Gaudreault in the car only several years after he first gave a statement to the police.

(5) Demeanour

[92] The trial judge directed the jury that they might find parts of the suspect witnesses’ evidence “acceptable as given because of the way it was given or the common sense supporting such an inference”. The appellants submit that this direction was erroneous in so far as it suggests that the jury could make a determination of the veracity of the witnesses’ stories by their demeanour. In a normal trial, this would be an error. These kinds of witness are notorious for being convincing liars and demeanour would be a very uncertain guide by which to measure the honesty of their evidence. However, this was a very unusual case. The jury saw these witnesses for extended periods where they were subjected to the most rigorous and exhaustive cross-examination. Moreover, the demeanour direction was accompanied by the much more helpful direction that the jury measure the story against common sense. As well, later in the charge the trial judge returned to the question of demeanour in relation to Gaudreault in these terms:

So it seems, members of the jury, that notwithstanding Mr. Orr’s [counsel for Sauvé] assiduous search for the key to the alchemy of Denis Gaudreault’s lying, you will be left to apply your own experience and your own common sense. There is no simple litmus test for nose rubbing, ear scratching, or eyes clouding to ferret out truth from lies. The decision will be for you, aided by your own appreciation of all the evidence and your own powers of observation and analysis. You saw the major witnesses testify for many days. You had a chance to study their demeanour over all those days and to appreciate, in relation to all of their evidence, whether their evidence has internal consistency and if it helps to support or is supported by the evidence of other witnesses. As I will tell you again, you may believe all, part or none of what any witness has told you. You may pick and choose after analysis as to what evidence fits or does not fit, as you see fit.

[93] While, in general, the demeanour direction should be avoided, it was not an error to give that direction in the unusual circumstances in this case.

(6) “Usual” type of witnesses

[94] The trial judge told the jury that it is “not unusual for witnesses of unsavoury character to be called as witnesses when the facts of the case require an examination of an illegal business”. The appellants submit that this direction undermined the caution since it would have led the jury to believe that there was nothing special about these witnesses and they are the types of witnesses one could expect in this kind of case. We do not consider this to be an error. The trial judge was entitled to provide the jury with the benefit of his experience that the prosecution is often required to call witnesses who are involved in illegal activities. That said, a more balanced instruction could also have provided the jury with the information that these witnesses were unusual in the sense that there were many frailties attached to their evidence.

(7) Conclusion on the Vetrovec warning

[95] We have found two fundamental errors with the Vetrovec warning in this case. The trial judge did not explain to the jury the reasons for the need for special scrutiny and did not give a sufficiently strong warning about the danger of acting on the unconfirmed evidence of the suspect witnesses. This was an error in law. We will consider whether the proviso in s. 686(1)(b)(iii) of the Criminal Code can be applied at the end of our discussion of the other alleged errors. At this stage, we merely wish to place the Vetrovec error in some context.

[96] We have already noted that these witnesses were subjected to lengthy and exhaustive cross-examination that covered almost every aspect of their criminal life and their motives for falsely implicating the appellants. Thus, the cross-examination of Gaudreault covered the following topics:

• His use of drugs, especially cocaine, and especially around the time of the murders;

• His difficulties with his memories;

• His flashbacks, paranoia, blackouts and hallucinations;

• His lying to the police;

• His lying at the preliminary inquiry;

• His lies to other agencies such as welfare authorities;

• His deals with the police and the suggestion that he was testifying for money;

• His many criminal convictions, criminal behaviour, drug dealing, and stealing;

• His fabricating evidence such as the note book and the Stewart electronic diary;

• His dealings with the witness protection programme;

• His threats to kill others over drug debts;

• That he owed Stewart a large amount of money and had stolen hashish from him before going out to British Columbia;

• His many inconsistent statements;

• His difficulties in remembering details such as the vehicle he was driving;

[97] The charge to the jury also contains a complete review of the evidence. That review included reference to many of the frailties with the evidence of Gaudreault, Emmerson and Jack Trudel. For example, the trial judge referred to the following with respect to Gaudreault:

• He thought he may have been hallucinating and asked for assurance from Rhonda Nelson that it really happened;

• His use of drugs and its effect on his memory;

• His continued criminal activity even after he began cooperating with the police;

• His holding back on information from the police to use as bargaining ploys;

• His obvious interest in obtaining money in exchange for his cooperation with the authorities;

• His concern that he might be charged with the murder depending on what he told the police;

• His lies under oath about Stewart’s electronic organizer;

• His many “outright lies” to the police; and

• His many early attempts to minimize his involvement in the crimes.

[98] As well, the trial judge’s review of the defence positions focused on the unsavoury nature of the witnesses. For example, the trial judge put the defence position respecting Gaudreault, in part, as follows:

The defence’s position is that Gaudreault is not credible. He had reasons to fabricate given his own actions as driver and custodian of the weapons, in order to get Stewart off the streets, or for money, or he may well have been hallucinating, and that he is a congenital liar. All of these reasons come under the and/or rubric for the defence.

[99] And he put the appellant Trudel’s position in part as follows:

It is the position of the defence that Richard Trudel had absolutely no involvement in the deaths of Manon Bourdeau and Michel Giroux. The only evidence that directly implicates Richard Trudel comes from two extremely unsavoury individuals, Denis Gaudreault and Jack Trudel. Denis Gaudreault’s evidence is totally unworthy of belief because he fails every credibility test: lengthy criminal record, serious long-term addiction to crack, admitted perjurer, admitted numerous lies to the police, the convoluted and manipulative method which he used to develop his story over a period of months has resulted in his story being sprinkled with inconsistencies and untruths.

Gaudreault maintained the police trust by manufacturing two pieces of physical evidence – the black book and the disk – both of which were pieces of physical evidence that turned out to be utter frauds. Gaudreault’s third ace in the hole, the utterances attributed to Richard Trudel prior to entering Stewart’s residence, are only disclosed to the police on the eve of Gaudreault’s testimony after he was aware that Stewart had been severed. Other transparent manipulative devices that a child could see through: Gaudreault owed a large amount of money to the accused Stewart. The only way he could extinguish his debt was by having Stewart locked up for good.

In addition to inconsistencies vis-à-vis other witnesses, there are numerous inconsistencies in Gaudreault’s own statements and testimony. It refers here to the two hundred and fourteen thousand and still counting with respect to witness protection. There is also reference to the numerous fraudulent acts even while in the Witness Protection Program, - the welfare scam, the bag of marijuana on the plane – shows his complete sense of immunity. He knows that the police have invested so much in his story that as long as he continues to maintain his value by not recanting, the Crown will continue to rain benefits on him.

The argument goes on. Jack Trudel is equally untrustworthy and equally as sophisticated at being fraudulent, only not as successful. He has a lengthy criminal record showing a total disregard for honesty. He admits to shooting 11 persons and yet claims that it is the moral repugnance of the shooting of a pregnant woman that forced him to do the honourable thing by turning in his own brother. He discloses his story when he’s in a jam with the law. His statement to the jury “I love my brother” was a most manipulative ploy that is transparent to all but the most gullible. He insulted all of our intelligence with that comment. He hated his brother for having modernized and expanded his drug business and then squeezed him out while he was in jail. He claims that his brother owes him $500,000. That is the real reason why he incriminated his brother.

[100] The trial judge also gave instructions on assessing credibility generally that they jury would recognize as having special application to the three suspect witnesses. For example, he referred to the particular problem of perjury:

It’s possible that some witnesses may have lied to you. Falsehood can be a very serious matter, and it is a very serious matter in front of the court, that could be very serious and it might affect the witness’ evidence considerably.

Reasonable doubt

[101] The appellants submit that the direction to the jury concerning the definition of reasonable doubt was inadequate. The trial judge gave the following direction:

Now I will turn to some concepts of law – you’ve heard about them before – and they are the presumption of innocence and the onus of proof. You cannot be too emphatic about the presumption of innocence, and that presumption is stated that the accused is presumed innocent until the Crown has proven his guilt beyond a reasonable doubt. Now, that is the central concept in our criminal justice system. It is not for the accused, for example, to prove his innocence. The onus is on the Crown to prove its case and they must prove that case beyond a reasonable doubt.

Now, what does beyond a reasonable doubt mean? Well, in some sense it’s easier to say what it is negatively than positively, but let’s start with some of the negative examples. First of all it’s not proof beyond any possible or any imaginary doubt. It’s not proof to a mathematical certainty. I mean, when you finish this case you’re not going to get the feeling you get when you say two plus two equals four; you know that. It’s not that kind of mathematical certainty because we’re not dealing in mathematics to start with. And it’s not a doubt which might be conjured up by a juror to escape responsibility, an any port in the storm kind of doubt. Beyond a reasonable doubt is when you feel sure or when you are sure of the accused’s guilt.

So to summarize this area, then, there’s no burden on the accused to prove innocence. The burden is on the Crown throughout, the burden begins with the Crown and ends with the Crown, and the burden of proof never shifts and that burden is to prove the guilt of the accused persons beyond a reasonable doubt. If you have a reasonable doubt at the end of the case, you must give the benefit of that doubt to the accused and you must acquit. On the other hand, if you’re satisfied beyond a reasonable doubt, then you must find the accused guilty.

[102] This direction was given at the beginning of the legal instructions after a lengthy review of the evidence and a similar direction was given again on the fifth day of deliberations in response to a question from the jury asking that the reasonable doubt instruction be repeated. The trial judge gave his charge to the jury in this case some 16 months before the decision of the Supreme Court of Canada in R. v. Lifchus (1997), 118 C.C.C. (3d) 1. Not surprisingly, there was no objection by trial counsel. Counsel for the appellants now submit that the instructions suffered from the following defects:

(1) The jury was not told that proof beyond a reasonable doubt is a higher standard than proof on a balance of probabilities.

(2) The jury was not told that a reasonable doubt may be based on the evidence or an absence of evidence.

(3) The jury was instructed that they may convict if they are “sure” before they were provided with a proper definition of the term “reasonable doubt”.

(4) The jury was told not to “conjure up” a doubt to “escape responsibility”.

[103] In considering whether there has been substantial compliance with Lifchus it is helpful to review the summary provided by Cory J. in that case at paras. 36‑37:

Perhaps a brief summary of what the definition should and should not contain may be helpful. It should be explained that:

· the standard of proof beyond a reasonable doubt is inextricably intertwined with that principle fundamental to all criminal trials, the presumption of innocence;

· the burden of proof rests on the prosecution throughout the trial and never shifts to the accused;

· a reasonable doubt is not a doubt based upon sympathy or prejudice;

· rather, it is based upon reason and common sense;

· it is logically connected to the evidence or absence of evidence;

· it does not involve proof to an absolute certainty; it is not proof beyond any doubt nor is it an imaginary or frivolous doubt; and

· more is required than proof that the accused is probably guilty -- a jury which concludes only that the accused is probably guilty must acquit.

On the other hand, certain references to the required standard of proof should be avoided. For example:

· describing the term "reasonable doubt" as an ordinary expression which has no special meaning in the criminal law context;

· inviting jurors to apply to the task before them the same standard of proof that they apply to important, or even the most important, decisions in their own lives;

· equating proof "beyond a reasonable doubt" to proof "to a moral certainty";

· qualifying the word "doubt" with adjectives other than "reasonable", such as "serious""substantial" or "haunting", which may mislead the jury; and

· instructing jurors that they may convict if they are "sure" that the accused is guilty, before providing them with a proper definition as to the meaning of the words "beyond a reasonable doubt".

[104] We note the following about the charge in this case. The charge makes clear that proof beyond a reasonable doubt is intertwined with the “central concept in our criminal justice system” of the presumption of innocence. The direction also states that the burden of proof rests on the prosecution throughout. The charge omits any direction that a reasonable doubt is not a doubt based upon sympathy or prejudice, but this would inure to the benefit of the appellants. The charge does not instruct the jury that a reasonable doubt is not an imaginary or frivolous doubt but it uses language that would convey a similar message by directing the jury that it is not “any port in the storm kind of doubt”.

[105] With respect to the failure to inform the jury that reasonable doubt is not made out by proof on a balance of probabilities, it was significant in this case that the trial judge contrasted the two standards when dealing with the co-conspirator’s exception to the hearsay rule. Thus, the trial judge directed the jury as follows:

The main thing you must remember in this three-stage approach is that if you find a probability that the accused participated in the common enterprise, that finding does not make a conviction automatic. You must go on to consider whether on all of the evidence the guilt of the accused is established beyond a reasonable doubt. In the end you must be satisfied beyond a reasonable doubt on both issues: (1) that a common enterprise existed and (2) that the accused, or either of them, participated in it.

[106] There could be no doubt that as a result the jury understood that proof beyond a reasonable doubt requires more than proof on a balance of probabilities.

[107] The failure to tell the jury that a reasonable doubt can also be based on an absence of evidence was not fatal in this case. In the end, this case turned primarily on the credibility of the three principal Crown witnesses. Moreover, the trial judge gave the jury a correct instruction in accordance with R. v. W. (D.) (1991), 63 C.C.C. (3d) 397 (S.C.C.) in relation to the appellants’ alibis.

[108] The charge also largely omitted language that was disapproved of in Lifchus. The trial judge did not describe reasonable doubt as an ordinary expression and did not invite the jurors to apply the same standard of proof that they would apply to other important decisions in their own lives. The trial judge did not use the phrase “moral certainty”, nor use adjectives such as “serious” or “substantial” to qualify the term “doubt”. The charge did refer to a reasonable doubt not being one conjured up to “escape responsibility”. That instruction resembles the “timid juror” instruction disapproved of by this court in R. v. Karthiresu (2000), 129 O.A.C. 291. However, it does not carry the negative connotation that makes the timid juror instruction so troubling by implying that jurors who acquit are timid and may be avoiding their responsibilities, while courageous jurors convict.

[109] The only significant defects in the reasonable doubt charge were in failing to direct the jury that a reasonable doubt is a doubt based upon reason and common sense and is logically connected to the evidence or absence of evidence. As a result, the jury did not have a complete definition of reasonable doubt before they were told that proof beyond a reasonable doubt is when they “feel sure or when you are sure of the accused’s guilt”. In summary, while reasonable doubt was never defined, this was not a case where the instructions given served to actually weaken the content of the reasonable doubt standard. The jury would also have understood that proof beyond a reasonable doubt was situated between mathematical certainty and a balance of probabilities, although not necessarily that it was closer to the former than the latter. Another important aspect of the charge was the instruction on circumstantial evidence that served to emphasize the high standard of proof. Thus, the trial judge told the jury the following:

Now where the Crown’s case relies considerably on circumstantial evidence it’s for you to determine whether you find the circumstantial evidence convincing, and let me say something about that. We have an expression in our language which says that “Oh, it’s only circumstantial evidence.” Well, that’s really a conclusion, it’s not really what circumstantial evidence is. You’ve already decided the central issue when you say it’s only circumstantial evidence. What you have decided is that the circumstances are weak. It doesn’t mean that this methodology is wrong because it may be quite possible to prove something circumstantially which is so compelling that there’s no way out for anybody as a result of all the circumstances which are proven. So there’s no inherent weakness in circumstantial evidence as a methodology per se; the weakness only exists when you decide at the end of your reflections that the circumstances here were not very compelling so therefore I wouldn’t believe or I wouldn’t do, or whatever.

So I want you to remember that circumstantial evidence can prove something just as effectively and just as completely as direct evidence of someone who actually saw something being done.

So direct evidence is evidence which, if accepted as true, proves a fact in issue without the necessity of drawing an inference. In other words, someone sees it and says this is what happened and you accept that witness’ evidence, therefore that becomes proof. Circumstantial evidence, on the other hand, is evidence which does not directly prove a fact in issue but which may give rise to an inference of the existence of the fact in issue. Any inference from circumstantial evidence must be based on a fact or facts proved by the evidence and not on a mere suspicion or conjecture.

So in this kind of a case where the Crown relies to some degree on circumstantial evidence, I will tell you that before you can base a verdict on circumstantial evidence you must be satisfied beyond a reasonable doubt that the guilt of the accused is the only reasonable inference to be drawn from the proven facts [emphasis added].

[110] The Supreme Court relied upon a direction about circumstantial evidence in R. v. Rhee (2001), 2001 SCC 71, 158 C.C.C. (3d) 129 (S.C.C.) at paras. 35 and 36 to find that there was substantial compliance with the Lifchus requirements. In our view, when the various instructions are considered including the circumstantial evidence instruction, the W. (D.) instruction, and the conspiracy instruction, there was substantial compliance with Lifchus in this case and we would not give effect to this ground of appeal.

Co-conspirators exception to the hearsay rule

[111] The appellants submit that the trial judge erred in instructing the jury that the co-conspirators exception to the hearsay rule applied in this case and, in any event, erred in instructing the jury as to its application. Since the accused were not charged with conspiracy, the trial judge used the term “common enterprise” in explaining this rule of evidence to the jury. We will use that term or the phrase “common design” in our discussion of this ground of appeal.

[112] The appellants’ principal submission on this aspect of the case concerns the trial judge’s definition of the common enterprise. The trial judge defined the common enterprise in these terms:

The Crown says that the accused together with Mallory and Stewart attended at the residence of Bourdeau and Giroux, or to use them as guinea pigs of what happens to drug customers who do not pay their debts.

Here the common enterprise alleged is that Stewart, Mallory, Trudel and Sauvé set out to make Giroux and Bourdeau examples of what happens to people who do not pay their debts or who fool around with suppliers of drugs.

[113] The appellants submit that this definition of the common enterprise was vague and an artifice resorted to by the Crown to extend the scope of what was, at worst, a conspiracy to commit murder, so that certain crucial statements by Stewart would be admissible against the appellants and that a statement by the appellant Trudel would be admissible against Sauvé.

[114] In our view, there was some evidence from which the jury could find that the killings of the deceased were in pursuit of a broader common design than a simple conspiracy to murder. That common design was to intimidate Stewart’s other dealers and ensure that they paid their debts. Admittedly, the evidence of that common design was not compelling and rested primarily on Stewart’s acts and declarations before the killing. There was, however, a body of evidence showing that Stewart was concerned about unpaid debts and that he intended to use the killings as an example to his other dealers. It was open to the jury to find that this was not a simple revenge killing in which Stewart was unhappy with one of his dealers and wanted to eliminate him. In other words, the fact that Giroux owed Stewart money was not simply the motive for the killing; it was the excuse but also provided an opportunity to teach others. We can see no reason why this could not serve as a common design to invoke the evidentiary rule.

[115] The appellants submit, as a matter of law, that for the purpose of the hearsay exception the common design must either be the offence charged or a conspiracy to commit it. Thus, they submit that even if there was a common design it was completed with the murder and any of the declarations after the killings were not capable of meeting the “in furtherance” requirement for the application of the exception. The appellants submit that as a matter of policy the hearsay exception must be so limited to prevent a dangerous and unwarranted expansion of the exception.

[116] We do not accept the appellants’ submission on this issue. In our view, the matter is settled by this court’s decision in R. v. Baron and Wertman (1976), 31 C.C.C. (2d) 525. In that case, the court had to consider whether the trial judge erred in instructing the jury that the conspiracy to murder included a conspiracy to “cover up” the murder. Martin J.A. dealt with the issue as follows at p. 550:

To engraft a conspiracy to avoid detection and prosecution, as a matter of law, on every conspiracy to commit a crime would have far reaching implications. The effect of such a doctrine would be to extend the duration of the conspiracy and to make the attempted bribery of a police officer or the subornation of a witness by one conspirator admissible against the other as an act done in furtherance of the presumed subsidiary conspiracy to escape detection and punishment. Such a conspiracy could, of course, like any other conspiracy, be established by evidence [emphasis added].

[117] To a similar effect is the decision of this court in R. v. Vrany (1979), 46 C.C.C. (2d) 14 at 26:

However, those who conspire to commit crime do not do so with any nice or careful delineation of the offences intended and a single conspiracy may encompass a number of closely linked crimes which may be essential to the success of the conspiracy and form part of the common design. This fact may then lead to the admission into evidence of acts and declarations of one conspirator against another which occurred after the principal object of the conspiracy had taken place if those acts and declarations were in furtherance of the common design…. [emphasis added].

[118] In other words, the scope of the conspiracy, or in this case the common enterprise, does not depend on the definition of the particular crime charged but the nature of the agreement. A court must be careful not to artificially graft on to a conspiracy to murder a conspiracy to avoid detection, or as in this case, a common design to use the murder as a means of intimidating others. However, if the evidence establishes such a conspiracy or common design, it can serve as a foundation for the hearsay exception.

[119] Accordingly, the real issue was whether the statements after the killings were in furtherance of the conspiracy. The two important declarations after the killings were Trudel’s report to Stewart within minutes of the killing and Stewart’s statements and conduct a few days later with the newspaper clipping. The trial judge gave the jury express instructions about those two statements and emphasized that they were only admissible against the appellants if in furtherance of the common design. The trial judge also provided the jury with explanations that could lead them to find that they were not. For example, he told the jury that they might find that Stewart’s statements in connection with the newspaper article were nothing more than a “swaggering ego trip … or for some other purpose not in furtherance of the common enterprise”.

[120] Finally, the appellants submit that the jury could have used the threats made by Stewart and Mallory towards Gaudreault’s family as evidence against the appellants by application of the co-conspirators exception. We do not think this was a realistic possibility. The trial judge never suggested that these statements could be in furtherance of the common enterprise and the jury would have realized that the common enterprise had achieved its objectives by the time Stewart and Mallory were making these threats. The examples referred to by the trial judge, Stewart’s use of the newspaper clipping and Trudel’s report to Stewart, were closely tied to the defined common enterprise.

[121] We would not give effect to this ground of appeal.

Frailties of identification evidence

[122] In their factum, the appellants raised as a ground of appeal the failure of the trial judge to charge the jury on the frailties of identification evidence. In oral argument, the appellants abandoned this ground of appeal, and it was not argued before us.

The “double hearsay” evidence of Dan Charbonneau

[123] The time of death was a contentious issue at trial and, as we have pointed out, was important for the appellants’ alibis, especially for Sauvé. The Crown contended that Giroux and Bourdeau were killed shortly after 10:00 p.m. on January 16th. The defence argued that the death occurred some time after 9:30 p.m. on January 17th.

[124] In support of its position on the time of death issue, the Crown called Michael McFadden, who gave evidence that he arrived at the Giroux home at approximately 10:20 p.m. on January 16th, saw the injured Giroux lying on the floor, did nothing, and simply left the premises.

[125] Two days later, on January 18th, McFadden visited the Carlsbad Springs Hotel Bar. The Crown led evidence that, on that occasion, McFadden told a man named Dan Charron that, on January 16th, he had seen Giroux lying injured on the floor of his home. Charron allegedly related his conversation with McFadden to a third man, Dan Charbonneau, who was called as a witness at trial to testify to what Charron had told him. McFadden and Charbonneau were both cross-examined on McFadden’s alleged conversation with Charron. Charron was not called as a witness.

[126] Defence counsel objected to Charbonneau’s evidence on the ground that it was inadmissible hearsay. The Crown responded that the evidence was being led for the fact that the declarant, McFadden, made the statement on January 18th, before news of the murders became public. The trial judge rejected the defence objection on the basis that Charbonneau’s evidence was not being tendered for the truth of its content. He intervened during Charbonneau’s testimony and instructed the jury:

Members of the jury, I might indicate that you cannot accept this statement for proof of the truth of what was being said but merely for the fact that it was said that night, in that place. You’ve already had the other evidence from McFadden.

[127] In R. v. Smith (1992), 75 C.C.C. (3d) 257 (S.C.C.) at 264, Lamer C.J.C. said, after referring to the “helpful” formulation of the hearsay rule in Subramanian v. Public Prosecutor, [1956] 1W.L.R. 965 (P.C.) at 970:

This statement of the “hearsay rule” is a useful illustration of the circumstances in which statements made by persons who are not called as witnesses have traditionally been considered inadmissible. When such statements are introduced to prove the truth of their contents, they have generally been considered to be inadmissible. However, when introduced to prove that they were made, they have traditionally been regarded as admissible, either under an “exception” to the hearsay rule, or more correctly from an analytical point of view because they fall outside the definition of hearsay. What is important is that the evidentiary dangers traditionally associated with statements by persons not called as witnesses – principally, the unavailability of the declarant for cross-examination – are not present, or are present to a far less significant degree, when the relevance of such statements lies simply in the fact that they were made.

[128] The “double-hearsay” statement from Dan Charbonneau comprises three propositions: (a) that Charbonneau was told by Charron; (b) that McFadden told Charron; (c) that McFadden had seen Giroux’s body on the Tuesday night. The Crown argued that the crucial proposition was McFadden’s statement that “I saw Giroux’s body on Tuesday night” because, even if false, it was relevant in fixing McFadden with knowledge that the murders had been committed before news about them became public.

[129] With respect, the focus of the inquiry is not on whether what McFadden said to Charron was relevant, even if not true, but rather on whether what Charron said was relevant, even if not true. In other words, the subject of the analysis should be Charron’s statement to Charbonneau: is it true that Charron had that conversation with McFadden on the Thursday, not is it true that McFadden had seen the body of Giroux on the previous Tuesday.

[130] There was no suggestion before us that Charbonneau’s testimony about his alleged conversation with Charron was anything other than hearsay, and the Crown did not argue before us that his testimony fell under any hearsay exception or that it was admissible under the principled exception. It ought not to have been put in evidence at the trial.

[131] However, no substantial wrong or miscarriage of justice was occasioned by the admission of Charbonneau’s evidence. Although Charron was not called to testify as to his conversation with Charbonneau, McFadden was called to testify as to his conversation with Charron. The evidence was tendered primarily to confirm McFadden’s testimony that he had seen the body of Giroux on the Tuesday night and the principal hearsay danger, that the declarant was unavailable to be cross-examined, was thus greatly attenuated. Quite apart from the hearsay evidence of Charbonneau, the Crown adduced a substantial body of circumstantial evidence that confirmed McFadden’s evidence regarding the date on which Giroux and Bourdeau were murdered.

[132] We would not give effect to this ground of appeal.

Liability for lesser included offences

[133] The appellants raised two issues with respect to liability for lesser offences. Both arise from the charge to the jury. They submitted, first, that the trial judge erred in his instruction regarding the availability of alternative verdicts with respect to the killing of Michel Giroux; and second, that the trial judge failed to instruct the jury regarding both the potentially exculpatory and the potentially inculpatory impact of alleged statements made by Richard Trudel to Jack Trudel regarding the murders.

(1) Availability of alternate verdicts

[134] In his charge to the jury on the subject of alternative verdicts, which outlined the legal requirements for a verdict of first degree murder, of second degree murder and of manslaughter, the trial judge referred numerous times to “Richard Trudel and/or James Sauvé”. The relevant portion of the charge in this context includes the following passages:

Members of the jury, in the case of the counts affecting Michel Giroux we have the evidence of Jacques Trudel in which he relates what Richard Trudel told him about Rob Stewart having shot Michel Giroux after a fit of pique and embarrassment caused by his being an object of ridicule in his attempts to intimidate Michel Giroux. If you accept that version in the face of all of the other evidence in the case, then for the Giroux count the analysis will begin with Rob Stewart being the principal offender and the roles of Richard Trudel and James Sauvé would have to be looked at as to whether either or both of them aided or abetted Rob Stewart in committing the offence of murder. The law would be the same as I have outlined where James Sauvé was treated as the principal offender.

The three-step process would be, in that situation: You would have to be satisfied beyond a reasonable doubt that Rob Stewart murdered Michel Giroux. You would have to be satisfied beyond a reasonable doubt that Richard Trudel and/or James Sauvé did something and that act actually aided or abetted Rob Stewart in committing the act of murder. And, finally, you would have to be satisfied beyond a reasonable doubt that Richard Trudel and/or James Sauvé knew or intended that his or their acts would aid and abet Rob Stewart to commit the offence of murder. You must be satisfied beyond a reasonable doubt that Richard Trudel and/or James Sauvé intended to kill, or meant to cause bodily harm of a kind likely to cause death, and were reckless whether death ensued or not to Michel Giroux.

[135] The appellants submit that this instruction was misleading as it suggested, contrary to the evidence, that there existed a basis upon which Richard Trudel could be convicted as a principal in the murder of Michel Giroux. They argue that the trial judge was obliged to give clear direction regarding the basis on which each of the appellants could be found guilty and, further, that Richard Trudel should not have been considered at all with respect to the question of actually causing death because the evidence disclosed no basis on which the jury could so find.

[136] While it would have been preferable for the trial judge to draw the jury’s attention to the lack of evidence upon which Richard Trudel could be found guilty as a principal, the issue is, in our view, of little moment. Read in its entirety, the charge provided clear instructions to the jury regarding the possible bases on which each of the appellants could be found guilty of each of the alternative verdicts. The trial judge specifically disabused the jury of any miscommunication in the expression “Richard Trudel and/or James Sauvé”: see para. 161, below. In any event, having regard to the law on party liability and the evidence before the jury, the appellants’ submission has no practical consequence. Where, as here, there is evidence of concerted action in the commission of the offence, it is open to a jury to convict all of the accused either as principals or as aiders and abettors: Thatcher v. The Queen (1987), 32 C.C.C. (3d) 481 at 510-511 (S.C.C.); R. v. Wood (1989), 51 C.C.C. (3d) 201 at 220 (Ont. C.A.). Any lack of clarity in the trial judge’s charge in this respect would thus have no significant impact.

(2) Richard Trudel’s alleged statement to Jack Trudel

[137] Jack Trudel testified that, following the murders, Richard Trudel told him that he had entered the Giroux/Bourdeau home for the purpose of intimidating Giroux, that James Stewart had shot Giroux in a fit of pique, and that Sauvé had then been obliged to kill Bourdeau in order to eliminate her as a witness.

[138] The appellants submit that this evidence had both potentially exculpatory and inculpatory impact, and that the trial judge erred in failing to instruct the jury accordingly. In counsel’s submission, the statement, if believed, constituted a denial of planning and deliberation and of the mens rea required for murder. However the trial judge focused only on the potentially inculpatory effect of the statement, without alerting the jury to the possibility that the statement constituted a denial of knowing participation in murder. The absence of such an instruction, in their submission, was unfair to both Sauvé and Trudel.

[139] Neither the Crown nor the defence accepted that the appellant Richard Trudel gave a true recounting of his role in the killing when talking to his brother. At trial, the Crown took the position that Richard Trudel was concerned to minimize his involvement in the killing of a pregnant woman. The defence position was that Richard Trudel never made the statement and that Jack Trudel was lying. Although (as appears in para. 134, above) the trial judge made reference to this statement when dealing with the subject of alternative verdicts, he made no reference to it in his review of the Crown and defence theories of the case and of the evidence elsewhere in his charge.

[140] While it might have been appropriate for the trial judge to direct the jury’s attention to the potentially exculpatory impact of this statement, his failure to do so caused no prejudice to the defence, given the positions taken at trial.

[141] We would not give effect to this ground of appeal.

Evidence of Detective-Constable Lamarche

[142] In their factum, the appellants objected to the introduction, by way of narrative, of the evidence of Detective-Constable Lamarche, the officer in charge of the investigation. In particular, they objected to her evidence detailing how the investigation proceeded, what various Crown witnesses had told the police and why the appellants were ultimately arrested. In oral argument, the appellants abandoned this ground of appeal, and it was not argued before us.

“No matter what, we stick together”

[143] Denis Roy committed suicide on November 17th, 1989. At trial, the Crown sought to adduce evidence that, some time after November 17th, Linda Beland (who was Rob Stewart’s wife at the time of the Giroux/Bourdeau murders) overheard Richard Trudel tell Stewart something to the effect of, “No matter what, we stick together”. Beland was uncertain whether the statement was made before or after Christmas. She testified that she heard no other part of the conversation and that she was not aware of what was referred to in the fragmented statement she overheard.

[144] Defence counsel at trial objected to the admission of this statement. The trial judge ruled the evidence admissible on the basis that it was relevant to demonstrate a relationship between Stewart and Richard Trudel. He said:

No, I think I’ll allow this in given the proximity of time, it’s less than two months before the events that concern us on January the 16, 1990. I think it shows the nature of the relationship and I agree that the defence can argue that it could be that they were going to not tell their wives they were cheating or their girlfriends, but nevertheless even that is some kind of a relationship of closeness I suppose.

[145] Counsel for appellants contended that the words, a mere fragmented utterance, were incapable of any definitive meaning and could therefore not be relevant evidence at the trial. The Crown argued that, having regard to the fact that the nature of the relationship among Stewart, Mallory, Trudel and Sauvé was in issue at trial, the “utterance” was relevant in supporting the existence of a relationship between Trudel and Stewart. The Crown further noted that, in any event, the jury was given a limiting instruction respecting the use of this evidence and that, even if the utterance was improperly admitted, there was considerable other evidence attesting to the existence of a relationship between Stewart and Richard Trudel.

[146] In our view, the trial judge did not err in law in admitting the utterance into evidence. Unlike R. v. Ferris (1994), 1994 ABCA 20, 27 C.R. (4th) 141 at 152-154 (Alta. C.A.), aff’d, 1994 31 (SCC), [1994] 3 S.C.R. 756, where the utterance in question was held to be inadmissible because no meaning could be extracted from the fragment overheard, it was possible to ascertain some meaning from the utterance in the present case. Given that the relationship among the accused was in issue, evidence tending to show that a bond existed before the murders took place was relevant and probative, although the evidentiary force of the utterance in question was admittedly not strong.

[147] We would not give effect to this ground of appeal.

“Arrangement killings” and individual intents

[148] The appellants raised two issues with respect to the charge to the jury on the issue of the classification of the two killings as first degree murder. Neither complaint was made at trial, either during an extensive hearing specifically convened to discuss the trial judge’s draft charge, or at the conclusion of the delivery of his charge to the jury. The absence of an objection by defence counsel at trial, while not determinative, is an indication that the charge was proper in the context of the trial: R. v. Jacquard (1997), 113 C.C.C. (3d) 1 at 19 (S.C.C.).

[149] The two issues raised for the first time on appeal were that the trial judge’s charge was fatally defective because:

(a) the instruction on section 231(3), the so-called “contract killing” provision of the Criminal Code, invited convictions for first degree murder in the absence of any prior arrangement; and

(b) it failed to differentiate between the positions of the two appellants and the need for the jury to find planning and deliberation on the part of each of them before they could be convicted of first degree murder.

[150] We deal first with the charge relating to s. 231(3) of the Criminal Code.

[151] The appellants submitted that that the charge relating to this section effectively directed that the appellants should be found guilty of first degree murder if they received money from Rob Stewart after, and because of, the murder. In making this submission, they emphasized the following passage in the charge:

In this case there was evidence that the accused were paid money by Rob Stewart after the killings were completed. There is a special rule covering this situation.

Section 231(3) of the Criminal Code says:

Without limiting the generality of subsection (2), murder is planned and deliberate when it is committed pursuant to an arrangement under which money or anything of value passes or is intended to pass from one person to another, or is promised by one person to another, as consideration for that other’s causing or assisting in causing the death of anyone or counselling another person to do any act causing or assisting in causing that death.

[152] Standing alone, the two sentences that preceded the trial judge’s reference to s. 231(3) support the suggestion that first degree murder could be made out if the appellants received money from Rob Stewart after and because of the murders. But the full extract from the trial judge’s charge that follows those introductory sentences make clear that the section applies only where there is a pre-existing arrangement for the payment of money to compensate someone for a killing to take place in the future. After quoting s. 231(3), the trial judge’s charge continued:

This means that a person commits first degree murder when he or she pays to have someone killed or when he or she kills someone for money.

Therefore, if you are satisfied beyond a reasonable doubt that Richard Trudel and/or James Sauvé are guilty of murder as I have just outlined it to you, and that they, or either of them, was paid money to kill Manon Bourdeau and Michel Giroux, or either of them, you must return a verdict of guilty on the charge of first degree murder. If, however, you are not satisfied beyond a reasonable doubt that the accused, or either of them, were paid money to kill Manon Bourdeau and Michel Giroux you should go on to consider whether or not the alleged murder was planned and deliberate in the sense that I defined those words to you a few moments ago.

[153] In our view, the full charge relating to s. 231(3) made clear the requirement that the money must pass “pursuant to an arrangement” for a contemplated murder and that the gratuitous payment of money after the fact does not suffice to bring an accused person within the subsection.

[154] We would not give effect to this ground of appeal.

[155] We turn to the ground relating to individual intent.

[156] The appellants submitted that the trial judge failed to instruct the jury that, in order to be convicted, each of the two accused must have intended a planned and deliberate homicide. In their submission, the trial judge effectively suggested to the jury that each accused could be found guilty of first degree murder through participation in a murder in which the other accused had the necessary mental element of planning and deliberation.

[157] In support of their submission, the appellants relied in particular upon the following extracts from the charge to the jury.

You should consider the following evidence of whether or not the alleged murder was planned and deliberate. You must consider the evidence of Denis Gaudreault of how the five men came together seemingly under Rob Stewart’s direction on January 16th, 1990 after a prior meeting at Romeo and Juliet’s; how he ordered the guns to be handed out by Gaudreault. You should also consider Stewart’s telephone conversation with the “bitch from Cumberland” which preceded January 16th. …You should also consider Gaudreault’s evidence of Stewart’s role. … You should also consider the evidence of Jamie Declare and Linda Beland and Denis Gaudreault as to Stewart’s complaints to them of people not paying their bills. You should consider the evidence of Linda Beland as to how hard up Rob Stewart was.

After consideration of all that evidence and the rest of the evidence in the case, including the alibi evidence, you can then decide if the Crown has proved beyond a reasonable doubt that the alleged murder was planned and deliberate as I have defined those terms to you. Remember, members of the jury, that the alleged murder must be both planned and deliberate. The existence of only one of them is not sufficient to prove it was planned and deliberate.

[158] These passages, in the appellants’ submission, directed the jury to consider not whether the appellants planned and deliberated on the murder but rather whether the murder was planned and deliberate. They failed to bring home to the jury the need to determine specifically whether the two appellants individually planned and deliberated on the murder, or instead simply adhered to a plan conceived by Stewart but not shared with them.

[159] While mere participation in a planned and deliberate murder will not, standing alone, render an accused guilty of first degree murder, it is not correct to suggest that each of the accused must have planned and deliberated. The relevant test is not whether the accused was the person who originally planned and deliberated but whether he knew about, adopted and executed a plan to commit such a murder: R. v. Brown (1995), 1995 NSCA 186, 102 C.C.C. (3d) 422 at 437-38 (N.S.C.A.).

[160] In any event, a reading of the charge to the jury as a whole conveys a different impression than the extracts on which the appellants relied. In speaking of the mental element required to convict of first degree murder, the trial judge told the jury:

Before you can find that the accused had the necessary criminal intent, you must be satisfied beyond a reasonable doubt that this intent is the only reasonable inference to be drawn from the proven facts. Please remember that the question for you to decide is what did Richard Trudel and/or James Sauvé in fact individually intend.

[161] In response to an observation by Crown counsel during a break in the charge to the jury, the trial judge specifically directed the jury, when it returned to the courtroom, about his use of the phrase “either/or”. He said:

Members of the jury, one of the counsel, as a matter of fact it was the Crown counsel, indicated to me that the use of the phrase “either/or” continually throughout the piece may leave you with the impression that the intention of one person, either of them, is enough to convict both of them of having that intention. That’s not what is intended. The intention must be proven for each person individually. I mean, no one is responsible for somebody else’s intention, if I put it that way.

[162] Having regard to this specific clarification, we do not agree that there was a danger that the jury would find both Sauvé and Trudel guilty of a planned and deliberate first degree murder without finding that each of them had the requisite intent for first degree murder.

[163] We would not give effect to this ground of appeal.

The Oath-Helping evidence

(1) The Evidence of Reverend Bill Main

[164] On September 30, 1993, Scott Emmerson called the police after going on a cocaine binge and becoming paranoid. The police put him in the cells until the drugs wore off. Officers Davidson and Turner then came to see Emmerson. As noted earlier, Davidson had left instructions that he be notified if any of the Emmerson family was in custody because he knew that Mallory was a lifelong friend of Scott Emmerson’s father. The officers asked Emmerson if he knew anything about the Cumberland murders. He said yes, but lied and said that Mallory was not involved and was water-skiing at Emmerson’s parents’ cottage (the murders, of course, had occurred in the middle of winter). Emmerson claimed that he gave information to the police because he wanted to turn his back on the way he used to live. Davidson raised the possibility of witness protection and asked Emmerson to think about it. At the time, Emmerson was facing a fraud charge. He did not ask for Davidson’s help with that charge. About one month later Emmerson was charged with more property offences and he asked to see Davidson and asked for his help in getting bail. Davidson took a statement from him. Emmerson said he wanted to make an official statement about what he knew; that he was at the point of either going straight or dying. Emmerson also called Harvest House to get admission to their drug programme. Emmerson was released into the custody of Harvest House.

[165] Reverend Bill Main was a minister involved with Harvest House. Emmerson testified that his motivation for testifying was that Reverend Main told him he had to deal with certain areas of his life. One of Emmerson’s problems was knowing what he knew about the Cumberland killings and so he decided to come forward as a witness. Emmerson was extensively cross-examined about his motivation for becoming a witness and that his real motive was to get a reduced sentence on outstanding charges.

[166] Over objection by the defence, the Crown was permitted to call Reverend Main. Reverend Main testified that in the first week of his stay at Harvest House, Emmerson came to see him because he had a problem to solve and did not know what to do. Reverend Main told him he should do as he was being asked, and tell the truth and cooperate with the police. He gave him this advice because he felt Emmerson had to start with a clean slate and be honest with himself and others. This is part of the Alcoholics Anonymous philosophy.

[167] In our view, this evidence was admissible. We agree with the appellants that this evidence was not admissible to rebut an allegation of recent fabrication. Where a witness’s evidence is attacked as a recent fabrication, it is open to the party calling the witness to lead evidence that on a prior occasion the witness had made a prior consistent statement, before the witness had a motive to falsely testify. That rule had no application here. The evidence does not show that Emmerson made his statements to Reverend Main before he had the motive to implicate the appellants. However, as the Crown points out, Reverend Main was not permitted to testify as to what Emmerson told him and therefore the recent fabrication rule was not engaged. We do not accept the appellants’ submission that the jury would assume that what Emmerson told Reverend Main was consistent with his trial testimony. There is simply nothing in Reverend Main’s brief testimony to support that conclusion. We also do not accept the submission that Reverend Main was vouching for the accuracy of the testimony. Again, there is nothing in Reverend Main’s testimony to suggest that. He told Emmerson to be honest with himself and others but there is nothing to indicate that Reverend Main believed that what Emmerson was saying or would say was true.

[168] In the end, the only purpose of this evidence was to confirm an aspect of Emmerson’s testimony that he did have a discussion with Reverend Main and this was part of his motivation for coming forward. Since Emmerson’s true motives were very much in dispute, it was open to the Crown to lead evidence to confirm that there was such a conversation. That is as far as Reverend Main’s evidence went. This evidence did not offend the rule against oath-helping. That rule prohibits a party from leading evidence to vouch for the sincerity of one of their witnesses, before the witness’s credibility has been attacked. Reverend Main did not purport to vouch for the truth of Emmerson’s testimony or even to vouch for Emmerson’s credit generally. His evidence was led in response to a particular allegation by the defence about Emmerson’s motives for testifying.

(2) Confessions of other inmates to Emmerson

[169] Emmerson was cross-examined on the improbability that Sauvé would confess to him, a stranger, in jail. In response, Emmerson testified that two other witnesses had confessed to him. In re-examination, over objection by the defence, Crown counsel elicited the details of those confessions and the fact that those two persons had pleaded guilty.

[170] In our view, this evidence should not have been admitted. Whatever minimal probative value this evidence might have had was outweighed by its prejudicial effect. Whether or not two other persons confessed to Emmerson in no way established whether Sauvé would have confessed to him. The fact that these other inmates pleaded guilty, according to Emmerson, also was not any indication of whether Emmerson was telling the truth with respect to Sauvé. See R. v. McNeill (2000), 144 C.C.C. (3d) 551 (Ont. C.A.) at paragraphs 54 to 60.

[171] Most importantly, the effect of this evidence was to improperly minimize the danger of evidence from jailhouse informers. The reports from the Morin and Sophonow Inquiries demonstrate the unreliability of evidence of alleged confessions to jailhouse informers, and point out that it is not unusual for informers to claim that they have received many confessions from other inmates. The fact that they make these kinds of claims does not mean that their evidence is any more reliable. To the contrary, these reports would suggest that these informers who claim to have received multiple confessions are probably highly unreliable.

[172] We will consider the effect of the error in admitting this evidence below when dealing with the cumulative effect of the errors.

(3) The videotape of Gaudreault’s visit to the scene

[173] Over objection from the defence, the Crown was permitted to play a videotape made by the police the first time that Gaudreault went with the police to try and find the scene of the killings. The trial judge held that the jury should see the videotape as a “scientific experiment”. We are satisfied that the evidence was properly admitted. The videotape shows the police driving Gaudreault around the Cumberland area as he tries to jog his memory as to where he let off Sauvé, Trudel and Mallory. A substantial attack was made upon Gaudreault’s reliability based in part on his admissions that he had been using large quantities of cocaine in the days up to and on the day of the killings and had asked his spouse whether he had hallucinated about the events. The fact that, as demonstrated by the videotape, Gaudreault was able to find the place of the killings and identify where he had let off the appellants and Mallory was admissible to bolster this part of his story. It showed that he had an independent memory of the events. The contention that Gaudreault could have learned this information from the press reports, in these circumstances, went only to the weight of the evidence. The evidence had some probative value and little prejudicial effect.

(4) Evidence from Gaudreault’s relatives

[174] In cross-examination, Richard Gravelle, Gaudreault’s brother-in-law, was asked to give his opinion about Gaudreault’s honesty. Gravelle agreed that Gaudreault was “a big time liar” but added that he would only lie if he would profit by it. Over objection, Crown counsel was permitted to asked Mr. Gravelle to explain what he meant by that, and in particular whether being in the witness protection programme would be the kind of thing that would make Gaudreault lie. Mr. Gravelle testified that he did not think so.

[175] The defence then pursued a similar line of questioning with Gaudreault’s sister, Sylvie Gravelle. She agreed that her brother would lie for profit. However, in re-examination, she testified that she did not think that he would lie just to be in the witness protection programme because he could make much more money in less time if he was not “on the law’s side”.

[176] We are satisfied that the impugned re-examination was proper. The defence opened up this area by asking for the witness’s opinion about Gaudreault’s credibility. It was open to the Crown to explore the area and to complete the picture.

(5) Gaudreault’s British Columbia handler

[177] Staff Sergeant Richardson of the Saanich Police Department was asked by the O.P.P. to act as a “handler” for Gaudreault while Gaudreault was in Victoria. He was not an investigator on the case. He was with Gaudreault for several months in 1990 and had several conversations with him about the Cumberland killings. With the permission of the O.P.P., Richardson also used Gaudreault as an informer for criminal activities on Vancouver Island.

[178] Some of Sergeant Richardson’s evidence was admissible to rebut an allegation that the O.P.P. investigators had supplied Gaudreault with details about the killings. However, Richardson also testified that in his opinion, Gaudreault was the best informant he had worked with in his 31 years of handling informants; that Gaudreault was “uncanny” in his knowledge and powers of recollection and many times he thought Gaudreault was lying but it turned out he was telling the truth. Crown counsel led this evidence in the course of a lengthy examination of Sergeant Richardson and it was a very minor part of his evidence.

[179] Gaudreault’s credibility was attacked by the defence during a prolonged (almost 30 days) cross-examination. It was open to the Crown to attempt to rehabilitate his evidence by, for example, resorting to the recent fabrication exception to the rule against prior consistent statements. As we have said, some of Richardson’s evidence fell into this category. Further, once Gaudreault’s credibility was attacked, it would have been open to the Crown to attempt to rehabilitate him by calling evidence of his reputation for veracity, assuming such evidence was available. See R. v. Clarke (1998), 129 C.C.C. (3d) 1 (Ont. C.A.). We were not referred to any authority, however, that would permit the kind of oath-helping evidence or personal opinion given by Richardson in this case. It was not suggested that Richardson’s evidence was admissible as expert opinion evidence. In any event, expert evidence concerning credibility of a witness is rarely admissible. See R. v. Béland (1987), 36 C.C.C. (3d) 481 (S.C.C.) at 493-96.

[180] However, while this evidence should not have been led by the Crown, its admission in the context of this trial was harmless. The evidence came out almost in passing without objection and played such a minimal role in this case that no prejudice was occasioned to the appellants.

Sauvé’s conviction for manslaughter

[181] It will be recalled that Scott Emmerson testified that in July 1991 while he was in jail with Sauvé and Mallory, Sauvé told him that he had killed the two victims. Emmerson testified that Sauvé told him that after he killed the man, he chased the woman. Mallory put a hand on his shoulder but Sauvé told Mallory, “I’ve learned my lesson. I’m not gonna make any more mistakes”. Sauvé then put the pillow over her head and shot her. He also said that he was not leaving anyone or any witnesses behind. Emmerson testified that the next day he overheard Sauvé tell another inmate that he had learned his lesson by leaving a witness, that he had done four years out of seven for a taxi driver killing in Montreal. Emmerson claimed that he only learned about Sauvé’s conviction for manslaughter a couple of years after this statement, from his father.

[182] At trial the defence objected to the Crown being allowed to lead that part of Emmerson’s testimony referring to Sauvé’s conviction for manslaughter. The trial judge held that the evidence was admissible. He also allowed the Crown to lead extrinsic evidence about the manslaughter conviction. That evidence was placed before the jury in the form of an agreed statement of facts and evidence from Sauvé’s parole officer and may be summarized as follows. On October 14, 1987, Sauvé was found guilty of manslaughter for the shooting death of a taxi driver in Hull. The taxi driver died from a gunshot wound while he and Sauvé were in the front seat of a moving taxi. The date of the offence was July 25, 1987 and Sauvé was in jail shortly after the offence. As a result of the shooting, Sauvé’s clothing was blood stained. Sauvé received a seven-year sentence and was granted day parole in April 1989 and full parole in August 1990. His parole was revoked once he was arrested for the Cumberland murders.

[183] The trial judge held that the evidence concerning the manslaughter conviction went to motive, identity and “probably to narrative”. The evidence showed more than Sauvé’s bad character and “it was the accused himself who raised his prior bad character in relation to his experience and to the lessons he has learned”. He concluded that the evidence tends to corroborate Emmerson’s testimony and its probative value outweighs its prejudicial effect. The trial judge could not see any way to edit Emmerson’s testimony to make it less prejudicial without significantly undermining its probative value.

[184] The trial judge warned the jury at the time Emmerson testified and when the extrinsic evidence was led that the evidence of the manslaughter conviction could not be used to show that because Sauvé committed that crime he was more likely to have committed the murders for which he was on trial. He also warned the jury in his charge to the jury about misusing this evidence. The appellants submit that the trial judge erred in permitting the Crown to lead evidence of Sauvé’s prior manslaughter conviction and, in any event, that the instructions given to the jury about the use of that evidence were inadequate. A complicating factor in this case is that the Crown theory concerning the motive for killing Ms. Bourdeau changed during the trial. We begin with the question of the admissibility of this evidence.

[185] Evidence that the appellant Sauvé had a prior conviction for manslaughter and the details surrounding that conviction was extremely prejudicial. It was a conviction for a similar offence for which he was on trial and was close in time to the trial. The fact that he received what the jury might consider was a lenient sentence and was on parole when he was alleged to have committed these murders exacerbated the prejudicial effect. The evidence carried most of the dangers associated with the introduction of bad character evidence. As Sopinka J. explained in R. v. D. (L.E.) (1989), 50 C.C.C. (3d) 142 (S.C.C.) at 161-62, in the context of admission of similar fact evidence, the prejudicial effect of this kind of evidence is felt by the jury in a number of ways including the following:

The first is that the jury, if it accepts that the accused committed the prior "bad acts", may therefore assume that the accused is a "bad person" who is likely to be guilty of the offence charged. … [T]his assumption might raise “... in the minds of the jury sentiments of revulsion and condemnation which might well deflect them from the rational, dispassionate analysis upon which the criminal process should rest”. The second effect on the jury might be a tendency for the jury to punish the accused for past misconduct by finding that accused guilty of the offence charged.

[186] This evidence was not tendered by the Crown for classic similar fact reasons. The Crown did not suggest that there was such a striking similarity between the killing of the taxi driver in Hull and the killings of the deceased in this case to identify Sauvé as the perpetrator of the Cumberland murders. Rather, the evidence was potentially probative on the following theories. First, the conversation suggested a motive for killing Ms. Bourdeau, that is, to eliminate her as a witness. Second, the conversation suggested that Sauvé had used the pillow as a shield. Third, the details of the manslaughter conviction were capable of confirming Emmerson’s testimony because it showed that Emmerson had knowledge that he must have obtained as a result of a conversation with Sauvé, a matter hotly disputed at trial.

[187] We will deal with each of these theories. On its own, the motive theory was not a sufficient basis to allow for admission of this evidence. The central feature of this case was identity of the killer or killers. A motive to eliminate a witness was not sufficiently unique to accurately identify Sauvé or anyone else as the perpetrator. More troubling, in this case, as we mentioned, the Crown’s theory about the motive for killing Ms. Bourdeau evolved during the trial. When the Crown sought to lead evidence of the manslaughter conviction, the theory was that Ms. Bourdeau happened to be at the house when the accused came to execute Mr. Giroux and was killed by Sauvé to eliminate a witness. By the end of the trial, Crown counsel went to the jury and suggested that this was a fabrication and that in fact the accused were planning on killing both of the deceased and in particular on killing Ms. Bourdeau because she had threatened to go to the police. This was consistent with Gaudreault’s testimony of the planning of the killing, including the killing of the “bitch on Cumberland”. In any event, as we have said, it is our view that the motive theory was not sufficiently probative or cogent to warrant admission of this highly prejudicial evidence.

[188] The reference to the pillow was also not a basis for admitting the manslaughter evidence. The probative value of the reference to the pillow lay in the fact that potentially it was a detail that only the killer or killers would know, because it had not been reported in the press. Sauvé’s reason for using the pillow, that he had learned a lesson in the past, did not appreciably add to the probative value of this evidence. Thus, the Crown could have obtained the necessary probative value from this aspect of the evidence without any reference to the manslaughter conviction. The reference to the pillow part of the Sauvé conversation, without reference to the manslaughter conviction, carried no prejudicial effect.

[189] This leaves the evidence of the manslaughter conviction as confirming Emmerson’s testimony that he had a conversation with Sauvé and later overheard a conversation between Sauvé and another inmate. The probative value of this evidence depended on the theory that Emmerson could only have obtained the details about the conviction from Sauvé. Those details were that he killed a taxi driver in Montreal, that he learned his lesson and was not leaving any witnesses behind, that he had done four years out of seven for the killing. There were substantial difficulties with this theory. First, there was no evidence that Sauvé had left a witness behind in the taxi driver killing. Second, the killing took place in Hull, not Montreal. Third, in the statement to Emmerson, Sauvé did not claim that he used the pillow as a shield to prevent his being covered with blood as happened in the Hull shooting. Fourth, there was actually no evidence that the pillow had been used as a shield during the shooting. While a pillow was found on Ms. Bourdeau’s head, there was no forensic evidence such as blood or gunshot residue to connect it to the shooting. Finally, the evidence was fairly convincing that Emmerson could have learned about Sauvé’s manslaughter conviction from his father before he spoke to the police. In summary, in our view, the evidence of the manslaughter conviction itself had very limited probative value. In any event, whatever probative value it did have was far outweighed by the extreme prejudice from introducing that evidence.

[190] While a trial judge’s decision in balancing the probative value against the prejudicial effect of bad character evidence is entitled to deference, we are of the view that the trial judge’s decision in this case was unreasonable. We also think, with respect, that some of the reasoning is circular. The trial judge referred in his ruling to the fact that Sauvé himself had introduced the manslaughter conviction into his conversations with Emmerson and the other inmate. In this sense, he was, it would seem, the author of his own misfortune. However, this reasoning fails to take into account that the purpose of leading this evidence was to confirm whether in fact Emmerson was telling the truth that he even had these conversations. In our view, this evidence should not have been admitted.

[191] The appellants also submit that the trial judge did not adequately direct the jury as to the use to be made of this evidence. The jury received instructions from the trial judge about the use of the evidence about the manslaughter conviction on three occasions. Immediately after Emmerson gave the evidence about the conversation with Sauvé the trial judge directed the jury as follows:

Members of the jury, would you come back please? You’ve heard some evidence today about a killing of a taxi driver and so on. I wanted to make it very clear to you from the outset that that evidence can never be used by you as a jury at any point in your deliberations in this case, or thinking about it, or anything else. It can never be used to reason like this: because the accused may have done that, therefore he’s likely the kind of person who would commit this offence – I’m talking about Mr. Sauvé, all right? It can never be used with that sort of reasoning. If it has a use, and you’ll see more clearly later on, I will explain what use it will have, but I want you to know from the outset you can never use it to reason because he did it once, he did it twice, okay? That’s poisonous reasoning and not permitted in the criminal law. Thank you very much.

[192] Then, prior to the trial judge reading the agreed statement of fact concerning the manslaughter conviction, he cautioned the jury as follows:

Now, before I do that, I want to tell you that the statement I’m going to read to you can only be used for certain purposes and there are certain warnings which go with it. The first thing is that, when you hear the statement, I want you to realize that the facts that are proven in it cannot be used to conclude that Mr. Sauvé is a bad person, or a bad man, and therefore is likely to have committed this offence with which he was charged. You cannot use it for that purpose. No one under our system can be convicted on the basis that the jury concludes that they are bad persons, and therefore, for that reason alone, are likely to have committed the offence [emphasis added].

[193] Finally, in the charge to the jury, the trial judge instructed the jury as follows:

There is evidence before you, members of the jury, that the accused may be persons of bad character in that they may have been in prison, and that both of them may have been involved in the drug business. In reference to the evidence of Scott Emmerson and Nathalie Mayer, you have heard evidence that James Sauvé was convicted of manslaughter in Hull in 1987. You can only use that evidence of the manslaughter conviction and the circumstances relating to it, if you so choose, as some confirmation of the evidence of Scott Emmerson and of the conversation in the yard of the Regional Detention Centre which he said he had with Sauvé in July of 1991. What you cannot do is to reason in this fashion: Both Richard Trudel and James Sauvé are persons of bad character, therefore it is likely that either or both of them committed the offences for which they are charged. Any evidence which you find may link the accused to a drug business can only be used to establish that the state of the business then may have rendered it more or less likely in those particular circumstances that they may have committed the offences with which they are charged. In other words, that you are convinced that their drug business was worth saving, and that they or either of them were prepared to commit these offences to save that business. … Always remember that our criminal law punishes persons who commit particular offences; it does not punish types or kinds of people [emphasis added].

[194] The appellants make two submissions concerning this instruction. First, they argue that in the second instruction set out above, the trial judge erred in telling the jury that no one can be convicted on the basis that they are bad persons “and therefore, for that reason alone, are likely to have committed the offence”. The appellants suggest that this may have left the jury with the impression that in some circumstances they could reason to guilt from the fact that the appellants were persons of bad character. Second, they submit that the trial judge did not adequately explain to the jury the permissible use of the evidence.

[195] We would not give effect to the first submission. In our view, the instructions sufficiently conveyed the message that the jury could not use the evidence of the manslaughter conviction to reason that since Sauvé was a bad man, he was more likely to have committed the offence. We think the jury would have understood that there was a permissible and impermissible use of the evidence and that they were not to reason from bad character to guilt.

[196] We are more concerned about the second submission. The evidence of the manslaughter conviction was so prejudicial that the trial judge was required to give explicit instructions to the jury about the permissible use of that evidence. Explicit instructions about the proper use of the evidence would have diminished the danger of the jury using it for an improper purpose. The trial judge ought to have instructed the jury how the evidence could confirm the testimony of Emmerson and just as importantly the frailties in that evidence. Those frailties included the fact that the offence was not committed in Montreal, that there was little evidence to support the pillow as a shield theory, that there was no evidence that any witnesses were involved in the Hull incident, and that the evidence had no confirmatory value if Emmerson obtained the information from sources other than Sauvé. Finally, the jury ought to have been expressly instructed that if they concluded that the evidence did not confirm Emmerson’s testimony, they should ignore it all together.

[197] The trial judge did review Emmerson’s evidence at considerable length and that review included many of the matters we have referred to as well as the related problem that the conversation Emmerson initially attributed to Trudel could not have happened because Trudel was not on the range at the relevant time. Standing on its own, the inadequacy of the direction would not constitute reversible error if the manslaughter conviction evidence were otherwise properly admissible. However, the adequacy of the direction must be taken into account in considering the impact on the trial of the error in admitting the manslaughter evidence.

Trudel’s severance application

[198] The appellant Trudel applied for severance on four occasions. While Trudel’s application was made on several bases, the primary basis was that many witnesses, such as Emmerson, would give evidence only admissible against one or more of the other accused. The question whether or not to grant separate trials is a matter for the discretion of the trial judge and this court will interfere only where the trial judge has erred in principle or the joinder operates a manifest injustice towards one of the accused. See R. v. Sternig (1975), 31 C.R.N.S. 272 (Ont. C.A.) at 284. We have not been persuaded that the trial judge committed any reversible error in failing to sever Trudel. Most of the evidence led by the Crown, especially the crucial evidence of Gaudreault and the evidence of Jack Trudel, was admissible against both Trudel and Sauvé. We would not give effect to this ground of appeal.

Conclusion on the errors

[199] We have found the following errors by the trial judge:

(1) The trial judge did not give the jury an adequate Vetrovec warning.

(2) The trial judge erred in admitting evidence of other “true” confessions by inmates to Emmerson.

(3) The trial judge erred in admitting evidence of Sauvé’s manslaughter conviction.

[200] It remains to consider the impact of those errors. These are all legal errors and the proviso in s. 686(1)(b)(iii) of the Criminal Code has potential application. Courts have been reticent to apply the proviso in cases where the trial judge has erred in failing to give a Vetrovec warning. The application of the proviso in such cases has been discussed by the Supreme Court of Canada in Bevan and Brooks and by this court in R. v. Baltrusaitis (2002), 162 C.C.C. (3d) 539 and R. v. Armstrong (2003), 176 O.A.C. 319.

[201] In Bevan, the trial judge’s error in failing to give the Vetrovec warning was not the only error. Thus, Major J. speaking for the majority of the Supreme Court at p. 329 concluded that the court had to consider the cumulative effect of the errors:

While each of the errors made by the trial judge in this case was serious in nature, it is not necessary to reflect upon whether any one of those errors, if it were the sole error by the trial judge, would have been a sufficient basis for directing a new trial. When the cumulative effect of the errors in question is considered, in my view, this is clearly not a case in which it would be appropriate to apply the curative provision. In all the circumstances of this case there is a reasonable possibility that, but for the trial judge's errors, the verdict would have been different.

[202] Then, speaking specifically of the error with respect to the two suspect witnesses, Major J. found that the jury may have treated their evidence less cautiously during their deliberations than they would have had the trial judge given a proper Vetrovec warning.

[203] In Brooks, only Binnie J. applied s. 686(1)(b)(iii). Bastarache J. (Gonthier and McLachlin JJ. concurring) found that the trial judge did not err in failing to give a Vetrovec warning. Major J. (Iacobucci and Arbour JJ. concurring) dissenting held that it was not a proper case to apply the proviso. At para. 137, Binnie J. held that

In most cases, a witness whose testimony is sufficiently important to require the Vetrovec warning in the first place will likely be sufficiently central to preclude application of s. 686(1)(i)(iii), as in R. v. Sanderson (1999), 134 Man. R. (2d) 191 (C.A.) at p. 193 (where the unsavoury witness gave "the only evidence presented by the Crown which implicated" the accused), and R. v. Siu (1998), 124 C.C.C. (3d) 301 (B.C.C.A.) (“the Crown would have had virtually no case without” the unsavoury witness’s evidence). In such cases, application of the curative proviso would clearly be wrong.

[204] In words that apply directly to this case, Binnie J. summarized the problem at para. 138:

I part company from Major J. on whether in this case there is any reasonable possibility that the verdict would have been different had a Vetrovec warning been given. It must be kept in mind that the respondent was not entitled to a trial that excluded altogether the evidence of Balogh and King. The issue under s. 686(1)(b)(iii) is whether there is any reasonable possibility that it would have made any difference to the ultimate verdict if their evidence had been accompanied by a warning instead of merely being heaped with ridicule by defence counsel, as was the case here [emphasis added].

[205] That is the issue in this case. Is there any reasonable possibility that it would have made any difference to the verdict if the evidence of Gaudreault, Emmerson and Jack Trudel had been accompanied by the proper warning instead of merely being heaped with ridicule by defence counsel? To that equation must be added the fact that the jury were exposed to these witnesses, especially Gaudreault, for weeks and weeks of cross-examination. In Brooks, Binnie J. was able to conclude that the verdict would have been the same even if the appropriate warning had been given about the jailhouse informants in that case. He did that by a careful analysis of the evidence and especially the other inculpatory evidence that did not depend upon the suspect witnesses. He also took into account that there was no objection by defence counsel, that a proper warning would have been accompanied by a review of the corroborative evidence and that the charge to the jury dealt with the other issues very fairly, many of them in a manner favourable to the accused.

[206] In Baltrusaitis, this court adopted a similar approach, where the trial judge had failed to give a Vetrovec warning in relation to a jailhouse informant. Moldaver J.A. held, at para. 57, that the Crown will be able to establish that the error was harmless either if the independent evidence capable of confirming the suspect witness was so compelling that the jury would inevitably have accepted his evidence, or the circumstantial case against the accused was so overwhelming that even without the suspect witness’s evidence, the verdict would inevitably have been the same. Moldaver J.A. found that the test had not been met.

[207] In Armstrong, at para 24, this court referring to R. v. Arradi (2003), 2003 SCC 23, 173 C.C.C. (3d) 1 (S.C.C.) at para. 42, held that there are two classes of errors of law that lead to the application of s. 686(1)(b)(iii). The first consists of "harmless or minor errors having no impact on the verdict". The second class encompasses serious errors that would require a new trial except that the evidence is so overwhelming that no substantial wrong or miscarriage of justice occurred. Brooks would be an example of the latter. Cases falling within the former class would be those where the error itself was trivial or there was no prejudice caused by a more serious error of law. In Armstrong, the court found that while the error in failing to give the Vetrovec warning was not trivial it was possible to trace its effect on the verdict and conclude that there was no prejudice. Having regard to what the trial judge did say about the suspect witness and given the confirmatory evidence, it was possible to conclude that the verdict would inevitably have been the same.

[208] Bearing in mind the principles from these cases, the application of the proviso must take into account the following factors. First, the Vetrovec warning was not the only error by the trial judge. The trial judge also erred in admitting evidence of the manslaughter conviction and the other “confessions” to Emmerson. Both of these errors bore on the credibility of Emmerson one of the suspect witnesses for whom a proper Vetrovec warning should have been given. The error with respect to the other confessions could have had the effect of diminishing the impact of what warning the trial judge did give. The error respecting the conviction could have improperly bolstered Emmerson’s credibility, as well as being extremely prejudicial to a fair trial to Sauvé.

[209] Second, in our view, this is not a case where the evidence other than that from the suspect witnesses was so compelling that the verdict would inevitably have been the same. Nor is this a case where there was other compelling confirmatory evidence. There was other confirmatory evidence but much of it, like the evidence of Jamie Declare, was subject to its own particular frailties.

[210] The most substantial submission in favour of applying the proviso rests on the unusual nature of the trial coupled with what the trial judge did say about these witnesses. We have already discussed at some length the context of this trial. To summarize, the three witnesses were cross-examined over many days about many issues that could have impacted on their credibility. As well, in addition to the caution that the trial judge did give, he reviewed at length the evidence of these witnesses as well as the defence submissions about why they should not be believed. The Crown rightly says that by the end of this trial, the jury had a complete picture of these witnesses with all of their frailties exposed. What was missing, however, was a clear and explicit direction from the trial judge that as a matter of law it was dangerous to act on their evidence and an expression from the trial judge, not defence counsel, as to why it was dangerous to do so in this particular case.

[211] Accordingly, while we consider this a close case, having regard to the central role these three witnesses played in this case and the seriousness of the other errors, we cannot say that the verdict would necessarily have been the same. We therefore would not apply the provisions of s. 686(1)(b)(iii). While two of the errors principally apply to Sauvé, given Trudel’s close association with him, we think that those errors would also have impacted on Trudel to the point that it would not be safe to apply the proviso in his case.

DISPOSITION

[212] Accordingly, we would admit the fresh evidence concerning Jack Trudel. On that basis alone the appeal must be allowed and a new trial ordered. In any event, because of the cumulative impact of the errors in the conduct of the trial we would allow the appeals, quash the convictions and order a new trial for both appellants.

Signed: “M.A. Catzman J.A.”

“M. Rosenberg J.A.”

“S. Borins J.A.”

RELEASED: “MAC” JANUARY 30, 2004