Le simple passage du temps ne rend pas automatiquement une information périmée, car sa pertinence doit être évaluée de manière contextuelle en tenant compte notamment de la nature continue des activités criminelles visées

R. v. James, 2019 ONCA 288

Lien vers la décision

[55] There is no rule as to how recent information has to be in order to be relevant: R. v. Dionisi, [2012] A.J. No. 65, 2012 ABCA 20, 285 C.C.C. (3d) 502, at para. 22. In my view, the trial judge fundamentally misunderstood the nature of the offences being investigated and the respondent's possible involvement in them. The information from the investigation showed that Primo was not some low level drug dealer operating in small quantities that might be sold on a street corner or in a back alley. Rather, the information showed that Primo was dealing in much larger quantities of cocaine, up to the kilogram level. The activities of large-scale drug dealers are not transitory.

[57] Not only do these events support this conclusion, so too does the criminal record of the respondent. Prior criminal activity -- including prior involvement in the drug trade -- is a relevant factor in determining whether there are reasonable and probable grounds for a search: R. v. Pasian, [2017] O.J. No. 2858, 2017 ONCA 451, 349 C.C.C. (3d) 144, at para. 6.

[60] Against that background, and the contemporaneous evidence regarding the drug activity involving the respondent and MD, the affiant's failure to mention that the April 2012 charges were stayed is an entirely minor error that does not, in any way, detract from the overall picture presented to the authorizing justice. It certainly would not have unduly influenced the justice of the peace in terms of whether or not to issue the search warrant. And, on review, whether you excise the April 2012 charges, or amplify the record by including the stay of those charges, neither would result in a conclusion that the justice of the peace could not have issued the search warrant.

[61] Returning to the issue of the currency of the information, the only authority to which the trial judge made reference on the issue of stale-dated information was R. v. Silveira (1995), 1995 CanLII 89 (SCC), 23 O.R. (3d) 256, [1995] 2 S.C.R. 297, [1995] S.C.J. No. 38. It is not clear how that authority would assist on the issue here since that case [page338] dealt with a search in exigent circumstances. I do note, however, that Cory J., speaking for the majority, discussed the problem that the police had in terms of whether they could obtain a search warrant in the exigent circumstances that they were confronted with. He suggested that one solution to the problem would have been for the police to have obtained the search warrant based on information that they had from a week or so earlier. In making that suggestion, Cory J. made the following observation, at para. 154, which is particularly apt to the situation here:

In this case, evidence existed upon which a search warrant might have been obtained before the arrests were made. It may be that it would have been preferable for the police to have obtained a search warrant based on the earlier transactions prior to that made on the day of the arrests . . . At the time of trial, if the search warrant was attacked on the grounds that it was outdated, evidence could be adduced of the difficulty of providing up-to-date material in circumstances like these and that, in light of the necessity of protecting the evidence, the police found it necessary to obtain a warrant based on the earlier transactions and observations. In the absence of an unreasonable delay between the observations and the application, it would be difficult to imagine that the warrant could be successfully attacked on the grounds that it was stale-dated. That is the way the police should have proceeded.

[62] A more pertinent authority on this subject is this court's decision in R. v. Breton, 1994 CanLII 939 (ON CA), [1994] O.J. No. 2097, 74 O.A.C. 99, 93 C.C.C. (3d) 171 (C.A.). In that case, there had been drug transactions in an apartment on September 16 and October 8. A search warrant was executed on October 15. One of the challenges to the warrant was that there was no information to believe that drugs would be found in the apartment on October 15. While that argument found favour with the trial judge, it did not find favour with this court. This court concluded that it was open to the issuing justice to infer from the earlier transactions that there was a probability that drugs would be found in the apartment on the later date. In so concluding, this court made reference to an earlier decision where a search warrant was issued on July 12 based on a drug transaction from May 3.¹ This court noted that, in that earlier case, "[n]otwithstanding the time lapse this court held that the information was sufficient": Breton, at para. 48.

[63] Another example is found in R. v. Lucas (2014), 121 O.R. (3d) 303, [2014] O.J. No. 3471, 2014 ONCA 561, leave to appeal to S.C.C. dismissed [2014] S.C.C.A. No. 460, 2015 CanLII 1812. In that case, there was a challenge to the ITO relied upon to obtain an authorization to intercept private communications. The target, Lucas, was believed to be trafficking in firearms. Lucas [page339] challenged the information in the ITO on the basis that it was stale because it referred to activities that had taken place some three years earlier. This court rejected that challenge. This court said, at para. 140:

[W]e agree with the trial judge that in this case, the fact that guns had allegedly been supplied to Lucas some years before the investigation into the offences referred to in the ITO did not undermine the grounds for issuing the authorization to intercept his communications. The ITO provided reasonable grounds to believe that Lucas had not disposed of all the guns delivered to him by Cooke, that he had supplied guns to the Doomstown Crips in the past, and the surveillance evidence from January 2006 indicated that Lucas may have been transporting firearms. The ITO thus provided reasonable grounds to believe that Lucas had been and continued to be involved in the organized trafficking of a substantial number of weapons.

(Emphasis added)

[64] The approach to the question of whether information is so dated such that it cannot be relied upon for the purpose of seeking a judicial authorization must be undertaken on a common sense and practical basis, taking all of the prevailing circumstances into account. As the British Columbia Court of Appeal observed in R. v. Ballendine, [2011] B.C.J. No. 838, 2011 BCCA 221, 271 C.C.C. (3d) 418, at para. 54:

Merely because information is "dated" does not mean it is "stale". While the length of time that has passed is to be taken into account in a reasonable-grounds determination, it is but one factor.

[65] As the court in Ballendine also observed, on these issues, a court is entitled to draw common sense inferences regarding the activities of persons: at paras. 53, 55, 57. That common sense point was echoed by this court in R. v. Beauchamp, [2015] O.J. No. 1939, 2015 ONCA 260, 326 C.C.C. (3d) 280, at para. 113:

[I]t would have been open to the authorizing judge to infer current criminality from past criminality in the circumstances disclosed by the evidence in the affidavit. In considering the capacity of the evidentiary predicate to ground a finding of probable cause, it is important to keep in mind the investigative objective and the nature of the alleged criminality involved. Ongoing criminal enterprises do not come into existence, nor do they disappear overnight. Information that seems dated at first blush can retain its relevance.

(Emphasis added)

[66] The suggestion, implicit in the respondent's position, that three weeks after the respondent was observed driving from London to Windsor in order to deliver a significant quantity of cocaine to MD, he became disassociated from, and unconnected with, drug trafficking finds no common sense foundation.

[67] In support of his position, the respondent relies on the decision in R. v. Morelli, 2010 SCC 8 (CanLII), [2010] 1 S.C.R. 253, [2010] S.C.J. No. 8, [page340] 2010 SCC 8, where a warrant to search for evidence of possession of child pornography based on four-month-old information was held to have been improperly issued. In my view, the decision in Morelli does not assist the respondent. The decision in that case is very much fact driven. A particularly salient fact is that the links on the accused's computer to two possible child-pornography websites, that a technician had noticed, were gone when the technician returned to view the computer the next day. As Fish J. said, at para. 95:

At best, this may be a ground for suspicion, but surely the deleted links afford no reasonable and probable grounds to believe that the appellant was in possession of child pornography, and still less that evidence of that crime would be found upon a search of his computer.

[68] The decision in Morelli does not preclude reliance on a pattern of behaviour as part of the evidentiary foundation for the issuance of a search warrant. It merely requires that there be sufficient facts to provide that foundation. In Morelli, the court found that there were not sufficient facts. Here, as I have already explained, there is a pattern of conduct by the respondent based on the information that the police had gained from their investigation and the criminal record of the respondent. Put simply, there was offence specific information establishing a pattern of drug dealing involving the respondent.

[69] My colleague draws support for the trial judge's conclusion from R. v. Rocha (2012), 112 O.R. (3d) 742, [2012] O.J. No. 4991, 2012 ONCA 707. However, that case also does not assist. Rocha involved the sufficiency of an ITO based on an informer's tip, and whether that tip was sufficiently credible and corroborated to tie drugs to a particular residence. The decision does not turn on whether the information was "dated", nor does it turn on any pattern of conduct.

[70] The trial judge also rejected the Crown's explanation for the delay in obtaining the warrant based on the fact that, had the police moved on the respondent's involvement earlier, it would have curtailed the broader ongoing police investigation. In response to this point, the trial judge said:

If they were after the bigger fish with respect to the firearms they should have left [the respondent] alone until they could achieve their purpose and go after him on the evidence that they had at the time.

[71] It is unclear to me what the trial judge meant by "on the evidence that they had at the time". The evidence that the police had was the evidence they put in the ITO. If the trial judge meant that, if the police had nothing further regarding the respondent when they moved on the broader investigation, they ought to have [page341] left the respondent alone, then that view reflects a fundamental misunderstanding of large-scale police investigations, and an extremely problematic one at that.

[72] As numerous large-scale police investigations have demonstrated, such investigations take time. During the course of those investigations, different individuals committing different offences may be identified at different times. It is common practice to address all such individuals, and all such offences, at the time that the broader investigation is completed -- the so-called "take-down day". This is done as a practical matter to avoid undermining the broader investigation. This necessary approach does not, of course, remove the need for the police to have the requisite grounds for the authorization at the time that they obtain it.

[73] If the thrust of the trial judge's decision here is, as I fear, that the police must either move immediately, when each of those offenders and offences is identified, or otherwise walk away from the prosecution of such persons, the result will seriously undermine the effectiveness, indeed the viability, of these types of investigations. The police will be left with the insolvable dilemma of either moving on the more peripheral but still serious offences immediately, and thus risk disrupting the larger investigation that often reflects more widespread and ongoing offences, or complete their investigation and walk away from the earlier offences revealed. Neither of those results is palatable in terms of public safety or proper law enforcement. With respect, no evidence is required to take into account such common sense real-world considerations.

[74] Further, any suggestion that the police should act on such offences and just hope that doing so will not jeopardize the larger investigation, imposes on the police a requirement to assume an unacceptable risk. It also potentially involves placing any police agents and undercover officers who might be involved in the investigation, as was the case here, at risk for their safety. No proper interpretation of s. 8 Charter rights compels such intolerable choices.

*** Note de l'auteur de ce blog: cette dissidence de la Cour d'appel de l'Ontario est avalisée par la Cour suprême - voir R. c. James, 2019 CSC 52 ***

La pertinence d'une information à la base de motifs raisonnables d'un policier n'est soumise à aucune règle stricte de récence

R. v. Dionisi, 2012 ABCA 20

Lien vers la décision

[20]           We conclude that the trial judge erred in her approach as to whether there was a section 8 Charter breach in the first place. In a voir dire as to a section 8 Charter breach, the onus is not upon the claimant to merely show that there were defects in the ITO, the claimant must establish the breach of section 8 on a balance of probabilities. This can be done only by showing that it is likely that the search warrant would not have issued based upon the content of the ITO. It is not enough to persuade the trial judge that she would not have been inclined to issue the warrant on the grounds thus provided. The trial judge must be satisfied that the warrant is invalid, not that the trial judge would not have authorized it. The trial judge does not re-try the warrant: R v Schindler, 2001 ABCA 22, 277 AR 98 at para 18, citing R v Araujo, 2000 SCC 65 (CanLII), [2000] 2 SCR 992.

 

[21]           Even if it is established that there are flaws in the ITO, such as that information is inaccurate, or a material fact is not disclosed, this will not necessarily detract from the existence of the statutory pre-conditions: see R v Pires, 2005 SCC 66 at para 30, [2005] 3 SCR 343. As the trial judge noted, the fact that the confirmatory information in the ITO related to details other than the actual observations of drugs in the house or the fact of appellant’s trafficking, that confirmation still has value on the totality of the circumstances: R v Caissey, 2007 ABCA 380 at para 23, 227 CCC (3d) 322, aff’d 2008 SCC 65, [2008] 3 SCR 451; R v Hillgardener, 2010 ABCA 80, 252 CCC (3d) 486. Here the ITO said that the informant had provided reliable information in the past and that he had been compensated for it, and also that he had a relationship with the handler. The fact that the informant had given specific details about non-offence particulars was still consistent with the informant providing information on which reliance could be placed. There would in such a context be support for the idea that the informant had no grudge against the appellant but a motive to work with the police.

 

[22]           The trial judge appears to have felt that the information from the informant might have been dated. There is no rule as to how recent information has to be in order to be relevant. In any event, the ITO alleged the information in the present tense. These averments provided no room for the conjecture that the information provided was not recent enough to be relevant. The trial judge also speculated as to whether the information provided by the ITO was information that the informant learned rather than seeing personally. An ITO is to be read in a practical, non-technical, common-sense basis: see R v Whitaker, 2008 BCCA 174 at paras 41-42, 254 BCAC 234, leave refused [2008] 3 SCR x. There was no foothold for the assumption that the information was second-hand. Even if it were second-hand, the police experience with the informant was such that his information had been found to be reliable in the past.

[23]           More generally, it would not be correct for a trial judge to assume that any observations made by that trial judge as to possible questions about the content of the ITO from looking at the text would be such as would have escaped the notice of the authorizing judge. For example, at one point she questioned whether the grammar in the ITO was identifying more than one informant. This court does not share her concern in that respect, but in any event the law does not assume that the trial judge is smarter than the authorizing judge in that sense. Rather, the search warrant is presumed to be valid: R v Pitre, 2011 NBCA 106 at para 34 citing Quebec (Attorney General) v Laroche, 2002 SCC 72 at para 68, [2002] 3 SCR 708.

 

[24]           The case law has said repeatedly that the function of the trial judge is not to substitute her opinion for that of the authorizing judge: R v Campbell, 2011 SCC 32 at para 14, [2011] 2 SCR 549 (“The question for a reviewing court is ‘not whether the reviewing court would itself have issued the warrant, but whether there was sufficient credible and reliable evidence’ to permit an issuing justice to authorize the warrant (Morelli, at para. 40).”). Also see R v Lee, 2011 ABCA 310 at para 14 (“ The function of the reviewing judge is not to conduct a rehearing; rather, that judge should not interfere if the authorizing judge could have granted the authorization: R v Garofoli, 1990 CanLII 52 (SCC), [1990] 2 SCR 1421 at 1452, 60 CCC (3d) 161; R v Araujo, 2000 SCC 65, [2000] 2 SCR 992 at 1016‑1017.”).

 

[25]           A search warrant should issue only on a credibly-based probability, having regard to the whole of the circumstances, but as pointed out in Lee:

 

15        The proof sufficient to meet the standard of credibly based probability is whether the totality of the relevant circumstances support the inference. Here the appellant invites us to examine individual circumstances in isolation, and suggests other inferences are possible. We are satisfied that a collective assessment of the circumstances supports the finding of the reviewing judge that there was sufficient evidence that the authorizing judge could properly have granted the search warrant in this case.

 

[26]           The Court in Pitre explained that it is not for the trial judge to decide if the warrant would have necessarily issued on the ITO, but whether it could have been authorized by a judge acting judicially: see likewise R v Ebanks, 2009 ONCA 851, 249 CCC (3d) 29 at paras 21, 33. (Ebanks was applied by ONCA majority in its decision later affirmed in Campbell.) It is apparent from the reasons of the trial judge, notably at paras 35 to 37, that she adopted a totality test, contended for by counsel for the appellant, that the ITO had to provide “compelling” information. Her concerns about the conclusory nature of what the ITO contained overlooked the specific detail of the manner and price at which the appellant allegedly was selling cocaine. The suggestion that this may be “mere rumour or gossip” is not grounded in anything but the argument of counsel. She dismissed the averments about the 2004 incident as unhelpful. She concluded her analysis with the observation that “I find that the edited ITO contains insufficient grounds ...” by reference to the “compelling” test. That was not the question. The question was whether no judge, acting judicially, could have chosen to grant the authorization on the information remaining after redaction.

La règle des confessions dérivées vue par la Cour d'appel de l'Ontario

R. v. M.D., 2012 ONCA 841

Lien vers la décision

[53]      The derived confessions rule is a common law rule that governs the admissibility of a confession that has been preceded by an involuntary, thus inadmissible confession. The derived confessions rule is not a per se or bright line rule that excludes all subsequent confessions on the ground that they are tainted, irrespective of the degree of their connection to the prior inadmissible statement: R. v. I (L.R.) and T. (E.), 1993 CanLII 51 (SCC), [1993] 4 S.C.R. 504, at p. 526; and R. v. Hobbins, 1982 CanLII 46 (SCC), [1982] 1 S.C.R. 553, at p. 558.

[54]      To determine whether a subsequent statement will be excluded under the derived confessions rule because of the taint left by its involuntary and thus inadmissible predecessor, a trial judge must examine all the relevant circumstances to determine the degree of the connection between the two statements: T. (E.), at p. 526.  The Supreme Court of Canada has set out some of the relevant circumstances or factors to consider in determining the degree of connection between the two statements, and thus the influence of the antecedent taint: see T. (E.), at p. 526; Hobbins, at p. 558; and R. v. G. (B.), 1999 CanLII 690 (SCC), [1999] 2 S.C.R. 475, at para. 21. These include but are not limited to:

•                    the time span between the statements;

•                    advertence to the earlier statement during questioning in the subsequent interview;

•                    discovery of additional information after completion of the first statement;

•                    the presence of the same police officers during both interviews; and

•                    other similarities between the two sets of circumstances.

 

[55]      The application of these factors will render a subsequent statement involuntary if either the tainting features that disqualified the first continue to be present, or if the fact that the first statement was made was a substantial factor that contributed to the making of the second statement: T. (E.), at p. 526; G. (B.), at paras. 21 and 23. It will generally be easier to establish that tainting affected the first when both these conditions are present. In the end, however, what matters most and mandates exclusion is that the connection is sufficient for the second to have been contaminated by the first: G. (B.), at para. 23.

[56]      The inquiry required when the derived confessions rule is invoked to exclude a subsequent statement is essentially a causation inquiry that involves a consideration of the temporal, contextual, and causal connections between the proffered and earlier statements: R. v. Plaha (2004), 2004 CanLII 21043 (ON CA), 188 C.C.C. (3d) 289 (Ont. C.A.), at para. 46. The inquiry is a case-specific factual inquiry: R. v. Simon, 2008 ONCA 578, 269 O.A.C. 578, at para. 69.

[57]      Despite its origins as a common law rule where lack of voluntariness is the contaminating factor, the derived confessions rule is of more general application.  The contaminating factor may be constitutional infringement, say a breach of s. 10(b) of the Charter. There, the subsequent statement is tainted if the breach and impugned statement can be said to be part of the same transaction or course of conduct. The admissibility analysis in these cases is performed under s. 24(2) of the Charter: R. v. Wittwer, 2008 SCC 33, [2008] 2 S.C.R. 235, at para. 21; Simon, at para. 69; and Plaha, at paras. 42-45.

[58]      The derived confessions rule may also be engaged where the contaminant in the prior statement is a failure to comply with s. 146(2) of the YCJA, or its predecessor, s. 56 of the Young Offenders Act, R.S.C. 1985, c. Y-1, as repealed by Youth Criminal Justice Act, S.C. 2002, c. 1: T. (E.), at p. 527; R. v. S.G.T., 2010 SCC 20, [2010] 1 S.C.R. 688, at para. 28.

[59]      To determine whether the derived confessions rule will warrant exclusion of a subsequent statement, a trial judge must follow a contextual and fact-based approach: S.G.T, at para. 29. The nature of the inquiry and the findings required in the derived confessions analysis have implications for the scope of appellate review. The admissibility of a confession that has been preceded by an involuntary (or otherwise) inadmissible confession, in other words, the application of the derived confessions rule, involves a factual determination based on factors designed to ascertain the degree of connection between the two statements: T. (E.), at p. 526. This determination, like a determination of whether a statement is voluntary or compliant with s. 146(2) YCJA, is largely a question of fact. Appellate review of the judge’s decision is limited to deciding whether the judge erred in her assessment of the evidence, failed to consider relevant circumstances, or failed to apply the correct principles: T. (E.), at p. 526; R. v. McIntosh (1999), 1999 CanLII 1403 (ON CA), 141 C.C.C. (3d) 97 (Ont. C.A.), at paras. 21-22.

Les limites de l'interventionnisme judiciaire : comment l'usurpation du rôle d'avocat par le juge détruit l'apparence d'équité d'un procès selon le critère de l'observateur raisonnable

R. v. Stucky, 2009 ONCA 151

Lien vers la décision

1)        The role of a trial judge

[61]         The role of a trial judge is often very demanding, owing not only to the inherent nature of the case, but also to the particular conduct of the litigants: R. v. Brouillard, 1985 CanLII 56 (SCC), [1985] 1 S.C.R. 39, at p. 42.  In the case at bar, the trial lasted 69 days and involved the examination of multiple witnesses, including experts and the accused himself.  Notwithstanding the length and complexity of a particular trial, a trial judge must exercise restraint and maintain impartiality so as to act within the scope of his or her neutral role.  As cautioned by Lamer J. in Brouillard, at pp. 42-43:

Like anyone, a judge may occasionally lose patience.  He may then step down from his judge’s bench and assume the role of counsel.  When this happens, and, a fortiori, when this happens to the detriment of an accused, it is important that a new trial be ordered, even when the verdict of guilty is not unreasonable having regard to the evidence, and the judge has not erred with respect to the law applicable to the case and has not incorrectly assessed the facts.

The reason for this is well-known.  It is one of the most fundamental principles of our case law [citation omitted]

… that justice should not only be done, but should manifestly and undoubtedly be seen to be done.

[62]         The principles that limit the permitted interventions by trial judges during the course of a trial and, specifically, during the examination-in-chief and cross-examination of witnesses, are well established.  We review them below.

2)        Permitted interventions by a trial judge

 

[63]         In Brouillard, at p. 44, Lamer J. acknowledged that a trial judge may intervene to ask questions, and, where necessary, he or she has a duty to ask questions where justice requires it.  However, at the same time, he expressly warned that there are definite limits on this right: Brouillard at p. 46.  A trial judge “should confine himself as much as possible to his own responsibilities and leave to counsel…[his or her] function”: R. v. Torbiak and Campbell (1974), 1974 CanLII 1623 (ON CA), 18 C.C.C. (2d) 229 (Ont. C.A.), at pp. 230-231.

[64]         In R. v. Valley (1986), 1986 CanLII 4609 (ON CA), 26 C.C.C. (3d) 207 (Ont. C.A.), at p. 230, leave to appeal refused, [1986] 1 S.C.R. xiii, Martin J.A. set out three situations in which questions put by a trial judge to a witness may be justified, namely: to clear up ambiguities and call a witness to order; to explore some matter which the witnesses’ answers have left vague; or, to put questions which should have been asked by counsel in order to bring out some relevant matter, but which were nonetheless omitted.  He noted, however, that questions put by a trial judge to a witness should generally be put after counsel has completed his or her examination of the witness and, further, that the witness should not be cross-examined by the trial judge during examination-in-chief: Valley at p. 230.  These comments provide guidance as to the timing and nature of interventions that a trial judge may make.

[65]         The first two situations of permitted interventions by the trial judge set out in Valley are self-explanatory.  The third situation in which a trial judge is permitted to intervene, namely, to ask questions that should have been asked by counsel, is not an open-ended invitation to the trial judge to usurp the role of Crown counsel.  The judge cannot leave his or her position of neutrality as a fact-finder and become the cross-examiner: R. v. W.(A.) (1994), 1994 CanLII 218 (ON CA), 94 C.C.C. (3d) 441 (Ont. C.A.) Brooke J.A. in dissent, reversed for the reasons given by Brooke J.A., 1995 CanLII 83 (SCC), [1995] 4 S.C.R. 51.

[66]         Where the appearance of fairness is not maintained at trial, the verdict reached cannot stand and a new trial must be ordered.  In deciding whether or not the appearance of fairness has been compromised, one factor that warrants consideration is whether the trial judge gave counsel an opportunity to ask questions that arise out of the trial judge’s questioning of a witness, in particular, the accused.  An additional factor is whether counsel objected to the trial judge’s questioning of a witness.  The absence of an objection, however, is not in itself determinative.

[67]         We turn now to the test for determining when this unfairness threshold is met.

3)        The test for determining whether the trial judge’s interventions have compromised the appearance of trial fairness

[68]         The test is an objective one.  As stated by Martin J.A. in Valley, at p. 232:

The ultimate question to be answered is not whether the accused was in fact prejudiced by the interventions but whether he might reasonably consider that he had not had a fair trial or whether a reasonably minded person who had been present throughout the trial would consider that the accused had not had a fair trial. [Emphasis added.]

[69]         The appearance of fairness and the trial judge’s corresponding duty to exercise restraint and remain neutral is especially critical in the criminal context where the accused takes the stand: Brouillard at p. 48.  Since a criminal trial is an adversarial process between the prosecution and defence, and not an investigation by the trial judge, the examination and cross-examination of witnesses is, for the most part, the responsibility of counsel: Valley at p. 231.  Although the trial judge is justified in occasionally intervening for one of the legitimate purposes indicated above, the trial judge must be careful not to usurp the role of counsel because otherwise the overall impression created may be fatal to the appearance of trial fairness.

[70]         The effect of interventions by the trial judge on the appearance of trial fairness in a given case must be assessed in relation to the unique facts and circumstances of the particular trial: Valley at p. 231, citing Torbiak at p. 231. 

[71]         In Valley at pp. 231-32, Martin J.A. listed types of interventions by trial judges which have resulted in the quashing of criminal convictions:

1.         Questioning an accused or a defence witness to such an extent or in a manner which conveys the impression that the trial judge has placed the authority of his or her office on the side of the prosecution and conveys the impression that the trial judge disbelieves the accused or the witness;

2.                  Interventions which have effectively made it impossible for defence counsel to perform his or her duty in advancing the defence; and

3.         Interventions which effectively preclude the accused from telling his or her story in his or her own way.

[72]         Interventions by a trial judge which can reasonably be said to create the appearance of an unfair trial may be of more than one type, and trial fairness may be undermined by one or more types of interventions: Valley at p. 232.  However, it is important to emphasize that no trial is perfect.  Accordingly, the record must be assessed in its totality and the interventions complained of in a given case must be evaluated cumulatively, not as isolated occurrences, from the perspective of a reasonable observer present throughout the trial.  As stated by Doherty J.A. in R. v. Stewart (1991), 1991 CanLII 11753 (ON CA), 62 C.C.C. (3d) 289 (Ont. C.A.), at p. 320:

It is a question of degree.  At some point, incidents which, considered in isolation, may be excused as regrettable but of no consequence, combine to create an overall appearance which is incompatible with our standards of fairness.