Supreme Court Judgments

R. v. Cinous, [2002] 2 S.C.R. 3, 2002 SCC 29

Her Majesty The Queen 

                                                                                                                           Appellant

v.

Jacques Cinous                                                                                              Respondent

and

The Attorney General of Canada and

the Attorney General for Ontario                                                                 Interveners

Indexed as:  R. v. Cinous

Neutral citation:  2002 SCC 29.

File No.:  27788. 

2001:  April 18; 2002:  March 21.

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

on appeal from the court of appeal for quebec

Criminal law – Defences – Self‑defence – Accused found guilty of second degree murder in shooting death of criminal accomplice – Whether defence of self‑defence should have been left to jury – Whether defence of self‑defence possessed an “air of reality” – Evidential standard applicable to air of reality test.

Criminal law – Appeals – Powers of Court of Appeal – Accused found guilty of second degree murder in shooting death of criminal accomplice – Trial judge’s charge to jury on defence of self‑defence containing errors – Whether curative proviso should be applied to uphold conviction – Criminal Code, R.S.C. 1985, c. C‑46, s. 686(1) (b)(iii).

The accused was charged with the first degree murder of a criminal accomplice, M.  He testified that he had been involved in the theft and resale of computers along with M and another accomplice, Y.  The accused said that about a month before the killing, convinced that M had stolen his gun, he decided that he would have no more contact with either Y or M and told them that there would be no more thefts.  The accused also testified that he began to hear rumours that Y and M wanted to kill him, and that he was warned by a friend to watch out for them.  One morning Y and M called the accused and asked him to participate in a computer theft and the accused agreed to meet with them that evening at his apartment.  The accused testified that when Y and M arrived, they kept their jackets on and whispered to one another as they sat in the living room and he saw Y constantly placing his hand inside his coat, which made the accused suspicious that the two were armed.  The accused said he decided to participate in the theft to see if they really intended to kill him.  They left the apartment and got into the accused’s van.  The accused said that he knew M and Y wanted to kill him when he saw the gloves that they were wearing.  Y had on different gloves than the ones he had arrived with at the apartment and M got into the van wearing surgical latex gloves.  The accused said he associated this type of gloves with situations where bloodshed was expected.  The accused testified that in the van Y avoided making eye contact with him and kept touching his jacket as if he had a gun.  He said he interpreted Y’s hand inside his jacket as a threat.  The accused said he was sure that he was going to be killed and that the shot would more than likely come from behind – from M.  However, since he was driving, he could not get to his own gun quickly enough, were anything to happen.  He pulled into a populated and well‑lit gas station, where he bought a bottle of windshield washer fluid.  After returning to the van, he opened the back door, “saw the opportunity”, pulled out his gun and shot M in the back of the head.  The accused testified that this was an instinctive reaction to a situation of danger.  He said that it did not occur to him to run away or to call the police.  At trial, the judge allowed the defence of self‑defence to be put to the jury.  The accused was nonetheless found guilty of second degree murder.  The Court of Appeal held that the defence was not properly explained to the jury.  It overturned the conviction and ordered a new trial.

Held (Iacobucci, Major and Arbour JJ. dissenting):  The appeal should be allowed and the accused’s conviction restored. 

Per McLachlin C.J. and L’Heureux‑Dubé, Bastarache, and LeBel JJ.:  A defence should be put to a jury if, and only if, there is an evidential foundation for it.  A trial judge must thus put to the jury all defences that arise on the facts, whether or not they have been specifically raised by an accused, but he has a positive duty to keep from the jury defences lacking an evidential foundation — or air of reality.  This is so even if the defence is the only defence open to the accused.  The air of reality test imposes a burden on the accused that is merely evidential, rather than persuasive.  In applying the air of reality test, a trial judge considers the totality of the evidence, and assumes the evidence relied upon by the accused to be true.  The threshold determination by the trial judge is not aimed at deciding the substantive merits of the defence.  That question is reserved for the jury.  The trial judge does not make determinations about the credibility of witnesses, weigh the evidence, make findings of fact, or draw determinate factual inferences.  Nor is the air of reality test intended to assess whether the defence is likely to succeed at the end of the day.  The question for the trial judge is whether the evidence discloses a real issue to be decided by the jury, and not how the jury should ultimately decide the issue.

The pre‑ and post‑Pappajohn authorities support a two‑pronged question for determining whether there is an evidential foundation warranting that a defence be put to a jury.  The question is whether there is (1) evidence (2) upon which a properly instructed jury acting reasonably could acquit if it believed the evidence to be true.  The terms “no evidence”, “some evidence” or “any evidence” can be used to describe the applicable evidential standard, provided these terms are understood as elliptical references to the full question.  The second part of this question can be rendered by asking whether the evidence put forth is reasonably capable of supporting the inferences required to acquit the accused.  This is the current state of the law, uniformly applicable to all defences.

The air of reality test must be applied to each of the three elements of self‑defence under s. 34(2)  of the Criminal Code , which have both a subjective and an objective component.  With regard to the first element it would be possible for the jury reasonably to conclude that the accused believed that he was going to be attacked, and that this belief was reasonable in the circumstances.  There is an air of reality to the subjective component of the defence as there is direct evidence on the accused’s beliefs, in the form of the accused’s testimony.  A jury acting reasonably could draw an inference from the circumstances described by the accused, including particularly the many threatening indicators to which he testified, to the reasonableness of his perception that he was going to be attacked.  With respect to the second element of self‑defence, reasonable apprehension of death or grievous bodily harm,  for the same reason there is also an air of reality to the accused’s perception that the attack would be deadly.  The accused’s testimony is unambiguously to the effect that he feared a deadly attack.  A jury acting reasonably could draw an inference from the circumstances described by the accused, including particularly the indications that Y and M were armed, the rumours of a plan to assassinate him, the suspicious behaviour, and the wearing of the gloves, to the reasonableness of his perception that he was in mortal danger.  With respect to the third element of self‑defence, namely a reasonable belief in the absence of alternatives to killing, it must be established both that the accused believed that he could not preserve himself except by shooting the victim, and that he held this belief on reasonable grounds.  There is an air of reality to the accused’s claim that, at the time he shot the victim, he actually believed that he had no alternative.  The accused’s extensive direct testimony regarding his subjective perceptions at the relevant time amounts to more than a “mere assertion” of the element of the defence.  However, the belief that the accused had no other option but to kill must have been objectively reasonable.  Section 34(2) requires that the accused have believed on reasonable grounds that there was no alternative course of action open to him at that time, so that he reasonably thought he was obliged to kill in order to preserve himself from death or grievous bodily harm.  In this case, there is absolutely no evidence from which a jury could reasonably infer the reasonableness of a belief in the absence of alternatives.

As the three conditions of self‑defence were not all met on the facts of this case, the defence lacked the “air of reality” required and should never have been put to the jury.  Any errors in the charge to the jury relating to it are therefore irrelevant.  The curative proviso of s. 686(1)(b)(iii) should be applied and the conviction upheld.

Per Gonthier and Binnie JJ.: The reasons of the majority were agreed with.  In this case, the only way the defence could succeed is if the jury climbed into the skin of the accused and accepted as reasonable a sociopathic view of appropriate dispute resolution.  There is otherwise no air of reality, however broadly or narrowly defined, to the assertion that the accused believed on reasonable grounds that he could not otherwise preserve himself from death or grievous bodily harm, as required by s. 34(2) (b) of the Criminal Code .  The objective reality of his situation would necessarily be altogether ignored, contrary to the intention of Parliament as interpreted in our jurisprudence.  A criminal code that permitted preemptive killings within a criminal organization on the bare assertion by the killer that no course of action was reasonably available to him while standing outside a motor vehicle other than to put a shot in the back of the head of another member sitting inside the parked vehicle at a well‑lit and populated gas station is a criminal code that would fail in its most basic purpose of promoting public order. 

Per Iacobucci, Major and Arbour JJ. (dissenting):  The test upon which a trial judge must decide whether a defence has an “air of reality” so as to be left to the jury only involves a determination of whether there is “some evidence” in support of the defence, and should not otherwise involve a measure of the sufficiency of that evidence.  In other words, when examining whether  there is evidence upon which a reasonable jury, properly instructed and acting judicially, could convict, it is the “no evidence” test, rather than the “sufficient evidence” test, which must be applied.  The test should be substantially the same as the one applied to cases of directed verdicts of acquittal or motions for non‑suit, as well as committal for trial under s. 548(1)  of the Criminal Code .  The test for committal for trial, directed verdicts of acquittal and the withdrawing of a defence from the jury strives to respect the long‑standing divisions of tasks between judges and jury, and favours great deference to the wisdom of the jury.  Discharges at the preliminary inquiry and directed verdicts of acquittal also promote judicial economy and may serve as an early barrier to the danger of a wrongful conviction.  Not so where a defence is withdrawn from the jury.

The “air of reality” test was never meant to lead to directed verdicts of conviction, but was mostly designed to avoid confusing the jury, particularly in cases of inconsistent alternative defences.  Where only one defence is raised and guilt is otherwise admitted, if any real meaning is to be given to the right to a trial by jury, the application of the test should be strictly limited to situations where a technical evidentiary requirement necessary to fulfill the accused’s evidential burden for a specific defence is lacking, or when there is a complete absence of evidence on one or more of the essential ingredients of the defence.  Only in those cases can it be said that the defence is not available in law, and only in this manner can the proper role of the jury be respected.  It is important to distinguish cases where there is a complete lack of evidence from those cases where there is some evidence, but the court does not consider it strong enough to raise a reasonable doubt.  Especially where appellate review is involved, the “air of reality” analysis, when applied to the sole defence raised and available to the accused, should focus only on the presence or absence of evidence, as opposed to its quality, sufficiency, or weight. 

The only principled and practical justification for withdrawing a defence is to avoid confusing the jury.  When the requisite legal elements of a single defence are properly explained to the jury, there is little risk of confusion on the part of the jury solely because the evidence in support of the defence is weak and unpersuasive.  The cost of risking a wrongful conviction and possibly violating the accused’s constitutionally protected rights by inadvisably withdrawing a defence from the jury is a high one.  Since this Court has consistently been of the view that the possible advantages that would be gained by adopting a higher threshold for the test respecting directed verdicts of acquittal are not sufficient to justify a change in the test, then there is no possible justification for the adoption of a higher threshold in the “air of reality” context, where such an adoption would involve fewer advantages and more risks.

While reasonableness constitutes a legal requirement of self‑defence under s. 34(2)  of the Criminal Code , the law is clear that it is for the jury to decide whether an accused’s perceptions were reasonable.  The fact that reasonableness involves an objective, rather than a purely subjective, assessment does not authorize judges to substitute their own appreciation of that critical factual element for that of the jury.  For there to be some evidence of reasonableness, there must merely be some evidence of the circumstances surrounding the conduct at issue, so that its reasonableness can be assessed in context.  Since reasonableness is a question for the jury, an accused is entitled to have a properly instructed jury assess his reasonableness when the defence of self‑defence is put forward.  In this case the factual record is complete and the accused explained, and was cross‑examined about, the events, his thoughts, his feelings, assumptions and rationale for acting as he did.  Given the evidence, whether or not he acted reasonably, subjectively and objectively, is a matter of judgment, and that judgment is the raison d’être of the jury system.  Where a reasonableness requirement is involved in a defence, the “air of reality” test must focus on assessing whether there is any evidence of an explanation for the accused’s perceptions and conduct.  The court should not embark upon an assessment of the reasonableness, or potential reasonableness, of this explanation for that is precisely the task that the law reserves for the jury.

For the purpose of assessing whether the accused’s defence had “an air of reality” in this case, it must be determined whether there is any evidence relevant to the three elements of self‑defence.  The accused introduced evidence that he believed  he was being assaulted and he also provided an explanation for this belief.  His testimony detailed the circumstances which gave rise to his state of mind on the night of the offence, as well as the broader context which he said led him to believe that M and Y were assaulting him, including the rumours that he had previously heard and his belief that M had stolen his gun.  The accused’s belief that the behaviour of Y and of M was increasingly unusual as the evening progressed was also evidence of an explanation for the accused’s ultimate perception that he was in fact being threatened or that an attempt to kill him was in progress.  Whether on these facts, the accused’s perception was reasonable is for the jury.  Secondly, the factors which led the accused to believe that he was being assaulted also caused him, according to his evidence, to apprehend death.  There were numerous factors and their relevance to the accused was clearly explained in his testimony.  They amount to some evidence upon which the jury may make its own assessment of reasonableness, since the jurors were provided with the full background and explanation for the accused’s perceptions.  Lastly, the accused testified that he believed that shooting M was the only way to preserve his life on the night in question.  His testimony was clearly to the effect that he believed himself to be in a situation of “kill or be killed”, so that shooting M was, in his mind, the only way to get out of the situation.  The accused explained why he believed that he could not simply walk away from the situation.  Whether or not that explanation is compelling, and whether or not the accused’s belief was reasonable, will once again be a matter for the jury.

Cases Cited

By McLachlin C.J. and Bastarache J.

Referred to:  R. v. Pétel, [1994] 1 S.C.R. 3; R. v. Lavallee, [1990] 1 S.C.R. 852;  R. v. McConnell, [1996] 1 S.C.R. 1075; R. v. Vaillancourt (1999), 136 C.C.C. (3d) 530; R. v. Lifchus, [1997] 3 S.C.R. 320; Reilly v. The Queen, [1984] 2 S.C.R. 396; R. v. Charlebois, [2000] 2 S.C.R. 674, 2000 SCC 53; R. v. Starr, [2000] 2 S.C.R. 144, 2000 SCC 40; R. v. Beauchamp, [2000] 2 S.C.R. 720, 2000 SCC 54; R. v. Russell, [2000] 2 S.C.R. 731, 2000 SCC 55; R. v. Avetysan, [2000] 2 S.C.R. 745, 2000 SCC 56; R. v. Osolin, [1993] 4 S.C.R. 595; R. v. Park, [1995] 2 S.C.R. 836; R. v. Davis, [1999] 3 S.C.R. 759; Wu v. The King, [1934] S.C.R. 609; R. v. Squire, [1977] 2 S.C.R. 13; Pappajohn v. The Queen, [1980] 2 S.C.R. 120; R. v. Finta, [1994] 1 S.C.R. 701; R. v. Latimer, [2001] 1 S.C.R. 3, 2001 SCC 1; R. v. Schwartz, [1988] 2 S.C.R. 443; R. v. Ewanchuk, [1999] 1 S.C.R. 330; R. v. Bulmer, [1987] 1 S.C.R. 782; R. v. Esau, [1997] 2 S.C.R. 777; R. v. Robinson, [1996] 1 S.C.R. 683; R. v. Lemky, [1996] 1 S.C.R. 757; R. v. Ruzic, [2001] 1 S.C.R. 687, 2001 SCC 24; R. v. Thibert, [1996] 1 S.C.R. 37; Brisson v. The Queen, [1982] 2 S.C.R. 227; R. v. Hebert, [1996] 2 S.C.R. 272; Parnerkar v. The Queen, [1974] S.C.R. 449; Perka v. The Queen, [1984] 2 S.C.R. 232; Kelsey v. The Queen, [1953] 1 S.C.R. 220; Workman v. The Queen, [1963] S.C.R. 266; R. v. Robertson, [1987] 1 S.C.R. 918; R. v. Arcuri, [2001] 2 S.C.R. 828, 2001 SCC 54; Commonwealth v. Webster, 59 Mass. (5 Cush.) 295 (1850); Metropolitan Railway Co. v. Jackson (1877), 3 App. Cas. 193.

By Arbour J. (dissenting)

R. v. Lavallee, [1990] 1 S.C.R. 852; R. v. Biniaris, [2000] 1 S.C.R. 381, 2000 SCC 15; R. v. Robertson, [1987] 1 S.C.R. 918; Woolmington v. Director of Public Prosecutions, [1935] A.C. 462; Mancini v. Director of Public Prosecutions, [1942] A.C. 1; Holmes v. Director of Public Prosecutions, [1946] A.C. 588; R. v. Lobell, [1957] 1 Q.B. 547; Wu v. The King, [1934] S.C.R. 609; Latour v. The King, [1951] S.C.R. 19; R. v. Proudlock, [1979] 1 S.C.R. 525; R. v. Tripodi, [1955] S.C.R. 438; R. v. Côté, [1964] S.C.R. 358; R. v. Nelson, [1968] 2 C.C.C. 179; Parnerkar v. The Queen, [1974] S.C.R. 449; Morgentaler v. The Queen, [1976] 1 S.C.R. 616; R. v. Squire, [1977] 2 S.C.R. 13; Alward v. The Queen, [1978] 1 S.C.R. 559; Linney v. The Queen, [1978] 1 S.C.R. 646; Mazza v. The Queen, [1978] 2 S.C.R. 907; Landry v. The Queen, [1979] 1 S.C.R. 552; Olbey v. The Queen, [1980] 1 S.C.R. 1008; R. v. Louison, [1975] 6 W.W.R. 289, aff’d [1979] 1 S.C.R. 100; Kwaku Mensah v. The King, [1946] A.C. 83; R. v. Porritt, [1961] 1 W.L.R. 1372; Leary v. The Queen, [1978] 1 S.C.R. 29; Pappajohn v. The Queen, [1980] 2 S.C.R. 120; Kelsey v. The Queen, [1953] 1 S.C.R. 220; Workman v. The Queen, [1963] S.C.R. 266; R. v. Bulmer, [1987] 1 S.C.R. 782; R. v. Reddick, [1991] 1 S.C.R. 1086; R. v. Trottier (1981), 58 C.C.C. (2d) 289; R. v. Cook (1985), 46 C.R. (3d) 129; R. v. White (1986), 24 C.C.C. (3d) 1; R. v. Osolin, [1993] 4 S.C.R. 595; R. v. Park, [1995] 2 S.C.R. 836; Sansregret v. The Queen, [1985] 1 S.C.R. 570; R. v. Dickson, [1994] 1 S.C.R. 153, aff’g (1993), 81 C.C.C. (3d) 224; R. v. Livermore, [1995] 4 S.C.R. 123; R. v. Esau, [1997] 2 S.C.R. 777; R. v. Ewanchuk, [1999] 1 S.C.R. 330; R. v. Davis, [1999] 3 S.C.R. 759; R. v. M.O., [2000] 2 S.C.R. 594, 2000 SCC 49, rev’g (1999), 138 C.C.C. (3d) 476; R. v. Silva (1994), 31 C.R. (4th) 361; R. v. Rarru (No. 2) (1995), 60 B.C.A.C. 90; R. v. Stolz (1996), 71 B.C.A.C. 127; R. v. Seaboyer, [1991] 2 S.C.R. 577; R. v. Malott (1996), 110 C.C.C. (3d) 499, aff’d [1998] 1 S.C.R. 123; R. v. Stewart (1995), 41 C.R. (4th) 102; R. v. Hebert, [1996] 2 S.C.R. 272; R. v. McConnell, [1996] 1 S.C.R. 1075, rev’g (1995), 32 Alta. L.R. (3d) 1; Brisson v. The Queen, [1982] 2 S.C.R. 227; R. v. Lemky, [1996] 1 S.C.R. 757; R. v. Robinson, [1996] 1 S.C.R. 683; R. v. Caron (1998), 126 C.C.C. (3d) 84, 16 C.R. (5th) 276; R. v. Fisk (1996), 108 C.C.C. (3d) 63; R. v. Taillefer (1995), 100 C.C.C. (3d) 1, 40 C.R. (4th) 287, leave to appeal denied, [1996] 1 S.C.R. x; R. v. McKinnon (1989), 70 C.R. (3d) 10; R. v. Martin (1980), 53 C.C.C. (2d) 425; R. v. Latimer, [2001] 1 S.C.R. 3, 2001 SCC 1; R. v. McKay (1992), 13 C.R. (4th) 315; R. v. Ruzic, [2001] 1 S.C.R. 687, 2001 SCC 24; R. v. Finta, [1994] 1 S.C.R. 701; R. v. Aalders, [1993] 2 S.C.R. 482; Young v. The Queen, [1981] 2 S.C.R. 39; R. v. Duclos, [1995] Q.J. No. 678 (QL); R. v. Stone, [1999] 2 S.C.R. 290; R. v. Thibert, [1996] 1 S.C.R. 37; R. v. Daviault, [1994] 3 S.C.R. 63; R. v. Parks, [1992] 2 S.C.R. 871; R. v. Sheridan, [1991] 2 S.C.R. 205, rev’g (1990), 55 C.C.C. (3d) 313; Perka v. The Queen, [1984] 2 S.C.R. 232; Bergstrom v. The Queen, [1981] 1 S.C.R. 539; R. v. Murray (1994), 93 C.C.C. (3d) 70; R. v. Bazinet (1986), 25 C.C.C. (3d) 273; R. v. Faid, [1983] 1 S.C.R. 265; R. v. Gee, [1982] 2 S.C.R. 286; R. v. Holmes, [1988] 1 S.C.R. 914; R. v. Bernard, [1988] 2 S.C.R. 833; R. v. Charemski, [1998] 1 S.C.R. 679; United States of America v. Shephard, [1977] 2 S.C.R. 1067; Mezzo v. The Queen, [1986] 1 S.C.R. 802; R. v. Monteleone, [1987] 2 S.C.R. 154; R. v. Litchfield, [1993] 4 S.C.R. 333; R. v. Arcuri, [2001] 2 S.C.R. 828, 2001 SCC 54; Skogman v. The Queen, [1984] 2 S.C.R. 93; R. v. Collins (1993), 79 C.C.C. (3d) 204; R. v. Find, [2001] 1 S.C.R. 863, 2001 SCC 32; R. v. Pan, [2001] 2 S.C.R. 344, 2001 SCC 42; R. v. G. (R.M.), [1996] 3 S.C.R. 362; CAIMAW v. Paccar of Canada Ltd., [1989] 2 S.C.R. 983; Lester (W.W.) (1978) Ltd. v. United Association of Journeymen and Apprentices of the Plumbing and Pipefitting Industry, Local 740, [1990] 3 S.C.R. 644; Pointe‑Claire (City) v. Quebec (Labour Court), [1997] 1 S.C.R. 1015; Canada Safeway Ltd. v. RWDSU, Local 454, [1998] 1 S.C.R. 1079; Ajax (Town) v. CAW, Local 222, [2000] 1 S.C.R. 538, 2000 SCC 23; Ivanhoe Inc. v. UFCW, Local 500, [2001] 2 S.C.R. 565, 2001 SCC 47; Sept‑Îles (City) v. Quebec (Labour Court), [2001] 2 S.C.R. 670, 2001 SCC 48; Centre communautaire juridique de l’Estrie v. Sherbrooke (City), [1996] 3 S.C.R. 84; Toronto (City) Board of Education v. O.S.S.T.F., District 15, [1997] 1 S.C.R. 487; Battlefords and District Co‑operatives Ltd. v. RWDSU, Local 544, [1998] 1 S.C.R. 1118; Nanaimo (City) v. Rascal Trucking Ltd., [2000] 1 S.C.R. 342, 2000 SCC 13; R. v. Pétel, [1994] 1 S.C.R. 3; R. v. Charlebois, [2000] 2 S.C.R. 674, 2000 SCC 53; R. v. Hibbert, [1995] 2 S.C.R. 973.

Statutes and Regulations Cited

Canadian Charter of Rights and Freedoms , s. 11 ( d ) , (f).

Criminal Code , R.S.C. 1985, c. C‑46 , ss. 17 , 34(2) , 232 , 265(4) , 548(1) , 686(1) (a)(ii) [am. 1991, c. 43, s. 9 (Sch., item 8)], (b)(iii) [idem].

Authors Cited

Braithwaite, W. J.  “Developments in Criminal Law and Procedure: The 1979‑80 Term” (1981), 2 Supreme Court L.R. 177.

Keane, Adrian N.  The Modern Law of Evidence, 5th ed.  London: Butterworths, 2000.

McCormick on Evidence, 5th ed. by John W. Strong, General Editor. St. Paul, Minn.: West Group, 1999.

Merriam‑Webster’s Dictionary of Law.  Springfield, Mass.:  Merriam‑Webster, Inc., 1996.

Roach, Kent. Criminal Law, 2nd ed. Toronto:  Irwin Law, 2000.

Sopinka, John, Sidney N. Lederman and Alan W. Bryant.  The Law of Evidence in Canada, 2nd ed.  Toronto:  Butterworths, 1999.

Stuart, Don.  Canadian Criminal Law: A Treatise, 3rd ed.  Scarborough, Ont.: Carswell, 1995. 

Watt, David.  Watt’s Manual of Criminal Evidence.  Scarborough, Ont.: Carswell, 2001.

Weiser, Irit.  “The Presumption of Innocence in Section 11(d)  of the Charter  and Persuasive and Evidential Burdens” (1988‑89), 31 Crim. L.Q. 318.

Williams, Glanville. Textbook of Criminal Law, 2nd ed.  London: Stevens & Sons, 1983.

Williams, Glanville.  The Proof of Guilt: A Study of the English Criminal Trial, 3rd ed. London: Stevens & Sons, 1963.

Williams, John M.  “Mistake of Fact: The Legacy of Pappajohn v. The Queen” (1985), 63  Can. Bar Rev. 597.

APPEAL from a judgment of the Quebec Court of Appeal (2000), 143 C.C.C. (3d) 397, [2000] Q.J. No. 6 (QL), allowing the accused’s appeal from his conviction for second degree murder and ordering a new trial.  Appeal allowed, Iacobucci, Major and Arbour JJ. dissenting.

Lori Renée Weitzman and Manon Ouimet, for the appellant.

Christian Gauthier and Louis Gélinas, for the respondent.

Bernard Laprade and François Lacasse, for the intervener the Attorney General of Canada.

Michael Bernstein, for the intervener the Attorney General for Ontario.

The judgment of McLachlin C.J. and L’Heureux-Dubé, Bastarache and LeBel JJ. was delivered by

1                                   The Chief Justice and Bastarache J. — The narrow issue on this case is whether the defence of self-defence should have been left to the jury.  On the law, this depends on whether the defence possessed an “air of reality”.

2                                   We conclude that on the authorities, a defence possesses an air of reality if a properly instructed jury acting reasonably could acquit the accused on the basis of the defence.  Applying this test, we find that the defence of self-defence did not possess an air of reality on the evidence presented at trial.  It follows that the trial judge’s error in the charge to the jury on self-defence was immaterial and that the accused’s conviction must be affirmed.

I.        Facts

3                                   The accused, Jacques Cinous, was charged with the first degree murder of a criminal accomplice, Michaelson Vancol (“Mike”), at a gas station in Montréal on the evening of February 3, 1994.  Despite his claim to having acted in self-defence, the accused was found guilty of second degree murder on October 21, 1995.

4                                   The accused testified that he and a friend, Josue Laforest (“Kent”), had been involved in the theft and resale of computers since August 1993, an activity from which he earned approximately $5,000 per week.  Another accomplice, Yves Louis (“Ice”), and the victim were also involved in some of these thefts.

5                                   Approximately one month before the killing, at the beginning of January 1994, the accused described an incident during which his gun disappeared.  According to the accused, one night that January, he, Ice and Mike went out to steal some computers.  The presence of a security guard and other employees on the targeted premises forced them to abandon the theft.  The accused testified that when he returned home that evening, he found that his gun was missing.  This was a 9-millimetre gun the accused said he had acquired some weeks after his release from jail, in April 1993, and had carried with him every day since.  Convinced that Mike had stolen the weapon, the accused said that he decided that from then on he would have no more contact with either Ice or Mike.  This gun has never resurfaced.  There is no evidence that it was stolen or that Mike in fact stole it.  It was not on Mike the night of the killing and is not alleged to have been involved in the events of that night.

6                                   The accused testified that he spoke with Ice and Mike on the phone the day after the gun disappeared.  Since he wanted no more to do with them, he said he would leave the money he owed them with his cousin ($800 for each of them).  He also told them that there would be no more thefts because the police had him under surveillance.  The accused testified that Ice and Mike nevertheless called him every day, several times a day, and would show up at the door at unexpected times wanting to do thefts with him.  He also testified that four or five days after the gun went missing, he began to hear rumours that Mike and Ice wanted to kill him, and that he was warned by a friend to watch out for them.  He had his cousin, whom he lived with, block their phone calls.  He testified that he managed to avoid them for three weeks to a month, but, on cross-examination, admitted to some contact with them during that period.

7                                   On the morning of February 3, 1994, at 7:00 a.m., Mike and Ice called the accused and asked him to participate in a computer theft.  The accused agreed to meet with them that evening at his apartment, at 7:00 p.m.; nevertheless, he claimed that he had no intention of going out to steal the computers and thought they might not even come.  That evening, Kent arrived at the accused’s apartment by 6:00 p.m., and Mike and Ice arrived at approximately 7:00 or 7:15 p.m.  The accused testified that Mike and Ice kept their jackets on and whispered to one another as they sat in his living room watching television.  The accused said he saw Ice constantly placing his hand inside his coat and that made him suspect the two were armed.  He asked, and they told him they were not armed.  He said that he decided to participate in the theft with them to see if they really intended to kill him, to find out whether or not the rumours were true.

8                                   Although the accused described what had gone on as suspicious, he did not argue that coercion was used to obtain his consent to participate in this theft or any of the previous thefts.  There was no evidence led to indicate that Mike had previously assaulted the accused or threatened him in order to obtain his consent to participate in thefts, nor was there a suggestion that Mike had assaulted or threatened to assault the accused on any other occasion, for any reason.

9                                   The four left the apartment and got into the accused’s van in order to drive to the location of the computer theft.  The accused said that he knew Mike and Ice wanted to kill him when he saw the gloves that they were wearing.  Ice was wearing different gloves than the ones he had arrived with at the apartment; he changed from leather to grey cloth gloves, and the accused thought it significant that he put these gloves on before entering the van.  The accused testified that Mike tarried in a nearby alley before coming to the van and got in wearing surgical latex gloves.  Neither of these gloves were the black woolen gloves that the accused said were kept in the glove compartment of the van for thefts.  Moreover, the surgical latex gloves Mike wore were the type of gloves that the accused said he associated with situations where bloodshed was expected and that he had twice before seen used on “burns” — attacks on criminals by other criminals.

10                               There was no evidence led to suggest that surgical latex gloves carried any particular significance as a gang practice in their case or in general, e.g. that their use would indicate (i) bloodshed (as opposed to preventing fingerprints from being left during a theft), and (ii) bloodshed that would be specifically directed against a co-criminal (a “burn”).  There was evidence that the surgical latex gloves had been stolen by Mike at an opportune moment from a hospital.  The accused said that he had seen them being worn by hit men in movies.

11                               In the van, Kent was sitting directly behind the accused, who was driving.  Ice was next to the accused in the front passenger seat, and Mike was behind Ice.  The accused testified that Ice avoided making eye contact with him and kept touching his jacket as if he had a gun.  The accused also said that no one spoke of the theft and in fact no one was speaking at all, things that were unusual on the way to a theft.  The accused also said that Mike snapped the surgical latex gloves he was wearing, once. Kent testified that there was normal talking and laughing on the way to the van and that the atmosphere in the van was relaxed.

12                               The accused admitted that no threats in the form of overt words or actions were made to him in the apartment, in the van or later at the gas station.  However, he said that he interpreted Ice’s hand inside his jacket as a threat.

13                               In his testimony, the accused said he was sure that he was going to be killed and that the shot would more than likely come from behind — from Mike.  However, since he was driving, he could not get to his own gun quickly enough, were anything to happen.  He said that he tried to drive along well-lit streets and avoid red lights so as to minimize the chance that Mike or Ice would kill him then and there.  He also said that he felt trapped.  He pulled into a populated and well-lit gas station “to release the pressure” and get himself out of this bad situation.  He exited the vehicle and headed towards the cashier cabin to get a bottle of windshield washer fluid. Unable to locate his own money, he returned to the van in order to ask Ice for $5 to pay for the fluid.  He went back to pay for the bottle, returned to the van, popped the hood of the van from the inside, poured the fluid in under the hood and brought the bottle around to the back of the van.  He opened the back door of the van, “saw [the] opportunity”, as he put it, pulled out his gun and shot Mike in the back of the head.

14                               The accused testified that this was an instinctive reaction to a situation of danger and, more specifically, that he felt like a bell was ringing in his head.  He said that it did not occur to him to run away or to call the police.  He mentioned in addition to this that the police would have required him to work as an informant in exchange for their help and that he should not have to run away and leave his van.

15                               When asked whether he felt scared at the gas station, the accused stated “scared in a sense, yes”.  He said that he did not feel safe there or that he was out of danger.  He said he felt like Ice and Mike were going to kill him, that they were just waiting for the right moment.

16                               The accused testified to chasing Ice after he shot Mike.  Ice fell to the ground, but the accused said he did not shoot him because the danger was gone and he was in control of the situation.  Ice fled.  The accused returned to the van which he and Kent tried to push.  These two ultimately also fled the scene, abandoning the van with the body inside.  No gun was found on Mike’s body, although he was carrying a knife.  The accused was arrested some time later.

17                               At trial, the judge allowed the defence of self-defence to be put to the jury.  The Court of Appeal held that the defence was not properly explained to the jury, both in terms of the explanation of the burden of proof in relation to the defence and the explanation of the defence itself.

18                               The appellant Crown concedes, as it did before the Court of Appeal, that errors in the charge were made, but asks this Court to apply the curative proviso in s. 686(1)(b)(iii) of the Criminal Code , R.S.C. 1985, c. C-46 , to uphold the conviction on the basis that any errors in the directions on self-defence are immaterial since the accused’s claim to self-defence lacked any “air of reality” and should never have been put to the jury.  According to the Crown, no retrial is required since, without the defence, the verdict would necessarily have been the same:  the accused would have been convicted of murder.

19                               The accused responds by saying that there was an air of reality to the defence since this threshold test requires only “some” evidence in order for the defence to be put to the jury and the accused’s own testimony satisfies this requirement.  A retrial is necessary, according to the respondent, in order for the defence to be properly explained to the jurors who must assess whether or not they believe the accused’s claim to self-defence.

II.    Legislation

20                               Criminal Code , R.S.C. 1985, c. C-46 

34.  . . .

(2)  Every one who is unlawfully assaulted and who causes death or grievous bodily harm in repelling the assault is justified if

(a)  he causes it under reasonable apprehension of death or grievous bodily harm from the violence with which the assault was originally made or with which the assailant pursues his purposes; and

(b)  he believes, on reasonable grounds, that he cannot otherwise preserve himself from death or grievous bodily harm.

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

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

                                                                   . . .

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

                                                                   . . .

(b)  may dismiss the appeal where

                                                                   . . .

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

III.   Judicial History

A.        Quebec Superior Court

21                               In his charge to the jury, Viau J. began his explanation of the specific law to be applied in the case with the five conditions for first degree murder:  (i) that the victim died a violent death; (ii) that the death was the result of an unlawful act, not an accident; (iii) that the person who caused him his injuries was the accused; (iv) that the accused wanted to cause the victim’s death or to cause him bodily injury that he knew was likely to cause death or that he was reckless whether death ensued or not; and (v) that the killing was planned and deliberate.

22                               Viau J. then turned to the law of self-defence in s. 34(2)  of the Criminal Code  and gave the jury the definition of assault in s. 265.  Dealing with the three conditions of s. 34(2), Viau J. began with the existence of an assault.  Here, he said, “[t]he whole milieu, facts, circumstances like hand in the jacket, surgical gloves, flick of this glove, flick of [that] glove, head turns when he was looking at them, were gestures that were assault and provocation towards him, according to Mr. Cinous, and convincing him he was about to be killed by Ice and Mike”.  Here he emphasized that the relevant question was not “was Mr. Cinous unlawfully assaulted, but did Mr. Cinous reasonably believe in the circumstances that he was being unlawfully assaulted”.  He identified this as a subjective test.

23                               As for the second condition of s. 34(2), whether the accused was acting under the reasonable apprehension of death or grievous bodily harm, the trial judge emphasized that this was to be assessed from an objective point of view:  “Would a reasonable person — not Mr. Cinous — but a reasonable person, taking into account that Mr. Cinous is include[d] in all [the] population, would a reasonable person in the same circumstances have reasonably apprehended fear, death or serious harm?”

24                               The jury was then told, with regard to the third and last condition, whether the accused believed he could not otherwise preserve himself from death or grievous bodily injury, that this test is subjective.  Here, Viau J. said that the question is:  “Was the accused really believ[ing] his life or his physical security were in danger and that the only thing he could do would be to shoot Michaelson, Mike Vancol?”  He noted that the model is not the average citizen but the accused “as he is”.

25                               Counsel for the defence asked the judge to clarify two points to the jury that were contained in the charge.  First, the jury members must be told to determine how the accused perceived the relevant facts for all three conditions in s. 34(2).  Counsel cited this Court’s decision in R. v. Pétel, [1994] 1 S.C.R. 3, summarizing the law by saying that all three conditions “are objective determinations based on the subjective frame of mind”.  The trial judge returned to the explanation of the three conditions to the jurors and told them the following:  “In all three (3) cases, you must seek to determine how Mr. Cinous perceived the relevant facts and whether his perception was reasonable”.

26                               The second problem brought to the judge’s attention by defence counsel related to the explanation of the term “unlawful assault”.  Counsel pointed out that both R. v. Lavallee, [1990] 1 S.C.R. 852, and Pétel, supra, establish that “a person need not be in the process of being attacked.  All you need is an attack that is more or less imminent.  You don’t have to wait to be attacked before you hit back, and that is even before Lavallee and Pétel”.  This is not an issue that Viau J. decided to address again and clarify for the jury.

27                               After the jurors had begun their deliberations, they sent a note asking the trial judge to explain s. 34(2) and s. 265.  Subsequently, they sent a note asking about the second condition for first degree murder, namely, whether an act committed in self-defence would be considered an unlawful act.  They also asked for an explanation of the notion of “reasonable belief” in relation to s. 34(2).  The trial judge clarified the first issue by telling the jury that there would be no question of an unlawful act if they decided to accept the defence of self-defence, since the accused would then be acquitted.  On the second issue, Viau J. repeated that the question to be asked is whether Mr. Cinous believed that his perception was a reasonable one, not what would be the belief or perception of the average citizen.

28                               Finally, the jury asked for clarification of the term “planned and deliberate” in the first degree murder charge.  The trial judge gave this explanation without incident.

29                               Defence counsel asked the judge to tie the defence of self-defence directly to his direction on the burden of proof, namely, to tell the jury that there is no burden on the accused to prove that he acted in self-defence and that the burden is on the Crown to prove beyond a reasonable doubt that the accused did not act in self-defence.  Viau J. did not do this.

B.        Quebec Court of Appeal (2000), 143 C.C.C. (3d) 397

30                               Writing for the three members of the Quebec Court of Appeal, Biron J. (ad hoc) was of the view that the errors in the charge to the jury made by the trial judge necessitated overturning the conviction and ordering a new trial.

31                               The first error identified by Biron J. that he thought would have created confusion in the minds of the jurors was the trial judge’s failure to make reference to self-defence when explaining the five conditions of first degree murder, particularly with respect to the requirement of an unlawful act.  On this point, Viau J. said to the jury:  “as a matter of law, to discharge a firearm at any person with intent to wound or to endanger the life of a person is an unlawful act”.  However, he did not tell the jury that, if the defence was accepted by them, Cinous’ shooting of Mike would not be an unlawful act.  Biron J. noted that this error was however corrected when the jury asked for clarification of this issue and the trial judge said that, if the jurors reached the conclusion that the defence applied, the verdict would be not guilty; there would be no question of an unlawful act.

32                               Biron J. found that there were two other errors in the trial judge’s charge to the jury that were not corrected.

33                               The first of these uncorrected errors was that the jurors were never told that the burden was on the Crown to prove beyond a reasonable doubt that the elements of the defence were not established.  Biron J. said that this, in conjunction with the fact that the accused testified and the trial judge’s use of phrases that give the impression that the accused has something positive to prove (e.g., “the accused wishes to repel an assault” (para. 47 (emphasis added))), would have left the jury with the impression that the burden was on the accused to prove he acted in self-defence rather than on the Crown to prove that he did not.

34                               The second uncorrected error identified by Biron J. was Viau J.’s failure to clearly tell the jury that under the law of self-defence there is no formal requirement that the danger that the accused is claiming to protect him or herself against be imminent.  Biron J. relied here on this Court’s acceptance of the dissenting opinion of Conrad J.A. in the decision of the Alberta Court of Appeal in R. v. McConnell, [1996] 1 S.C.R. 1075, the Quebec Court of Appeal’s decision in R. v. Vaillancourt (1999),136 C.C.C. (3d) 530, and the decisions of Pétel, supra, and Lavallee, supra.  While imminency is one of the factors that the jury should use to determine whether the accused’s apprehension of danger was reasonable, and whether he or she reasonably believed there were no other alternatives available than to kill the aggressor, the second and third conditions of the defence, it is not a formal requirement.

35                               On this point, Biron J. noted that Viau J. said things that would lead a jury to believe that imminency was a requirement.  For example, in relation to the second condition of the defence, Viau J. said that “the accused [must be] acting under the reasonable apprehension of death or grievous bodily harm from the violence with which Michael Vancol, in these circumstances, was assaulting him” (Court of Appeal decision, para. 47 (emphasis added)).  Biron J. said of this, at para. 72:  [translation] “[i]n my view, the condition formulated in that way could have induced the jurors into error.  How could one talk about the violence with which an attack was carried out if a victim is seated, his hands in his pockets and his back to the accused?”  In his view, the jury should not have been left with the impression that imminency was required in this way.

36                               Biron J. rejected the Crown’s claim that the errors in the charge were irrelevant and could be cured by s. 686(1)(b)(iii) because, in any event, the defence lacked an “air of reality”.  Stating that the trial judge must have believed that the defence of self-defence had an “air of reality” because he submitted it to the jury, Biron J. also noted that the length of the jury deliberations and the questions asked by the jury suggest that the defence passed this threshold test (para. 77).  Conceding that the verdict was not unreasonable, Biron J. nonetheless stated, at para. 79:  [translation] “I am incapable of finding that if the law had been correctly explained  to the jury and if the jurors believed the accused, a verdict of acquittal was not legally possible”.

37                               Biron J. added that the trial judge made reference to “moral certainty” five times in his directions to the jury on reasonable doubt, contrary to this Court’s pronouncement in R. v. Lifchus, [1997] 3 S.C.R. 320.

IV.   Analysis

38                               While we agree with the Court of Appeal that the trial judge made several errors in the charge to the jury, we do not agree that any of these, viewed individually or cumulatively, warrant overturning the conviction and ordering a new trial.  It is our view that since the three conditions of self-defence were not all met on the facts of this case, the defence lacked the “air of reality” required in order to warrant leaving it with the jury.  Since the defence should never have been put to the jury, any errors made in the charge to the jury relating to it are irrelevant.  These errors of law can be safely set to one side, and s. 686(1)(b)(iii) should be applied in order to uphold the conviction.

A.        The Jury Charge

39                               The charge to the jury contained several errors.  First, the trial judge failed to specifically tie the question of the burden of proof to self-defence, despite counsel for the defence bringing this problem expressly to his attention.  General warnings about the presumption of innocence and the fact that the burden of proof lies with the Crown are insufficient in these circumstances, i.e. in the course of a long and complicated jury charge involving self-defence.  In a situation where the accused relies on this defence, and perhaps especially when the accused testifies, the jury must be told that the burden of proof in relation to this defence is on the Crown, who must prove beyond a reasonable doubt that the defence does not apply.  Otherwise, there is a risk that the jury will be left with the impression that the burden is on the accused to prove that the defence applies.  An express direction on this is required in order to eliminate this risk.

40                               Second, the charge failed to clearly state that imminence of attack is not a formal requirement but merely a factor to be taken into consideration in self-defence in determining the reasonableness of the accused’s response: see Lavallee and Pétel.  Biron J. was correct when he said that imminency is a factor to be taken into consideration in an assessment of the reasonableness of the accused’s apprehension of danger and belief in the absence of other alternatives to killing the aggressor.  In Pétel, supra, Lamer C.J. stated, at pp. 13-14:

. . . Lavallee, supra, rejected the rule requiring that the apprehended danger be imminent.  This alleged rule, which does not appear anywhere in the text of the Criminal Code , is in fact only a mere assumption based on common sense.  As Wilson J. noted in Lavallee, this assumption undoubtedly derives from the paradigmatic case of self-defence, which is an altercation between two persons of equal strength.  However, evidence may be presented (in particular expert evidence) to rebut this presumption of fact.  There is thus no formal requirement that the danger be imminent.  Imminence is only one of the factors which the jury should weigh in determining whether the accused had a reasonable apprehension of danger and a reasonable belief that she could not extricate herself otherwise than by killing the attacker.  [Emphasis added.]

The failure to identify imminency as a factor rather than a formal requirement was not a reversible error.

41                               Third, the charge failed to properly set out the three conditions of self-defence under s. 34(2)  of the Criminal Code .  Pétel expressly stated that the same test applied to all three conditions:  “the jury must seek to determine how the accused perceived the relevant facts and whether that perception was reasonable” (p. 12).  Relying on Reilly v. The Queen, [1984] 2 S.C.R. 396, at p. 404, Lamer C.J. called this “an objective determination” because there must be an objectively verifiable basis for the accused’s perception.  Recently, in R. v. Charlebois, [2000] 2 S.C.R. 674, 2000 SCC 53, at para. 13, Bastarache J. for the majority characterized this test as “simultaneously subjective and objective”.  Hence, the trial judge was wrong to distinguish initially between the three parts of the defence, calling the first and third conditions subjective and the second objective.  However, he ultimately corrected this statement of the law in his clarification of s. 34(2) and in his clarification of the term “reasonable belief”.  Hence, this error, like the one concerning the explanation of unlawful act in the five conditions of first degree murder, was corrected.

42                               Fourth, the charge is faulted for use of the term “moral certainty” in connection with reasonable doubt.  Biron J. was correct to point out that this Court looked with disapprobation upon the use of “moral certainty” in the explanation of reasonable doubt in Lifchus, supra.  At para. 25, writing for the majority, Cory J. said that the problem with the use of this phrase was that “different jurors may have different ideas about the level of proof required before they are ‘morally certain’ of the accused’s guilt”.  However, as Biron J. acknowledged, at para. 82, when he described the use of this phrase as one [translation] “to be avoided, if not [ruled out entirely]”, Cory J. went on immediately to say, at para. 25:  “this expression, although not necessarily fatal to a charge on reasonable doubt, should be avoided”.  In situations in which the jury charge predates this Court’s comments in Lifchus, as is the case with this charge, other recent cases of this Court have held that the standard to be applied is that of “substantial compliance” with the principles of Lifchus; see R. v. Starr, [2000] 2 S.C.R. 144, 2000 SCC 40; R. v. Beauchamp, [2000] 2 S.C.R. 720, 2000 SCC 54; R. v. Russell, [2000] 2 S.C.R. 731, 2000 SCC 55; R. v. Avetysan, [2000] 2 S.C.R. 745, 2000 SCC 56.

43                               In Starr, Iacobucci J., writing for the majority, pointed out, at para. 237, that the principles in Lifchus (e.g. what the jury charge should include and should avoid) were “not intended to suggest that a new trial is warranted for all previous convictions obtained following jury charges that were not in strict compliance”.  For example, in Beauchamp, a pre-Lifchus charge using the “moral certainty” formulation was found by this Court to be nonetheless in substantial compliance with the Lifchus principles.

44                               As Iacobucci J. stated in Starr, supra, at para. 233:

A charge must be examined in its entirety to determine whether the essential elements of a fair and accurate instruction on reasonable doubt are present and have been properly explained.  The question in every case in which a trial judge’s instructions on reasonable doubt are impugned is whether there is a reasonable likelihood that the jury was under a misapprehension as to the correct standard of proof to apply.  If the charge, when read as a whole, could not have placed the jury under a misapprehension as to the correct standard of proof, then the jury verdict should not be disturbed . . . .

Whatever the problems in this charge as they relate to self-defence and the burden of proof with regard to that defence, the explanation given by Viau J. concerning the standard of proof in the explanation of “reasonable doubt” is a different thing.  While the references to “moral certainty” are numerous (like Biron J., we count five), and the concept is central to the explanation provided there, we are unconvinced, reading this portion of the charge, that these references are sufficient to establish that it was reasonably likely that the jury misunderstood what standard of proof they were to apply.

45                               Recent case law of this Court supports the view that mere use of the discontinued language of “moral certainty” in the jury charge on reasonable doubt will not necessarily lead to the conclusion that the trial was unfair.  This is, in our view, what Major J. described in Avetysan, at para. 12, as a charge “where the language used, although no longer preferred, meets the substantially correct test”.  As in Beauchamp, where, as we have said, references to “moral certainty” were also relied on, this is a situation where the charge on reasonable doubt is nonetheless in substantial compliance with the principles articulated in Lifchus and would not have misled the jury on the standard of proof to be applied.

46                               In summary, most of the alleged errors in the charge were corrected or insufficient to support overturning the conviction.  Only the error with respect to the charge on imminence of attack remains problematic.  The Crown argues that this error was immaterial since on the evidence, there was no air of reality to the defence of self-defence in any event.  This brings us to the critical issue on this appeal — whether the defence possessed an air of reality.

B.         Air of Reality

47                               The key issue is whether there was an air of reality to the defence of self-defence in this case.  It is our view that there is no air of reality to the defence: a properly instructed jury acting reasonably could not acquit the accused on the ground of self-defence, even if it accepted his testimonial evidence as true. Since the defence should never have been put to the jury, any errors made in the charge to the jury relating to that defence are irrelevant.  The curative proviso of  s. 686(1)(b)(iii) should be applied, and the conviction upheld.

48                               This Court has considered the air of reality test on numerous occasions. The core elements of the test, as well as its nature and purpose, have by now been clearly and authoritatively set out. See R. v. Osolin, [1993] 4 S.C.R. 595; R. v. Park, [1995] 2 S.C.R. 836; R. v. Davis, [1999] 3 S.C.R. 759.  Nevertheless, a controversy has arisen in this case concerning the extent of a trial judge’s discretion to keep from a jury defences that are fanciful or far-fetched. More narrowly, the contentious issue is the correct evidential standard to be applied in determining whether there is an air of reality to the defence of self-defence on the facts of this case.

49                               In our view, the controversy can be resolved on the basis of existing authority, which we consider to be decisive. The correct approach to the air of reality test is well established. The test is whether there is evidence on the record upon which a properly instructed jury acting reasonably could acquit. See Wu v. The King, [1934] S.C.R. 609; R. v. Squire, [1977] 2 S.C.R. 13; Pappajohn v. The Queen, [1980] 2 S.C.R. 120; Osolin, supra; Park, supra; R. v. Finta, [1994] 1 S.C.R. 701. This long-standing formulation of the threshold question for putting defences to the jury accords with the nature and purpose of the air of reality test.  We consider that there is nothing to be gained by altering the current state of the law, in which a single clearly-stated test applies to all defences. See Osolin, supra;  Park, supra; Finta, supra.  There is no need to invent a new test, to modify the current test, or to apply different tests to different classes of cases.

(1)  The Basic Features of the Air of Reality Test

50                               The principle that a defence should be put to a jury if and only if there is an evidential foundation for it has long been recognized by the common law. This venerable rule reflects the practical concern that allowing a defence to go to the jury in the absence of an evidential foundation would invite verdicts not supported by the evidence, serving only to confuse the jury and get in the way of a fair trial and true verdict. Following Pappajohn, supra, the inquiry into whether there is an evidential foundation for a defence is referred to as the air of reality test. See Park, supra, at para. 11.

51                               The basic requirement of an evidential foundation for  defences gives rise to two well-established principles. First, a trial judge must put to the jury all defences that arise on the facts, whether or not they have been specifically raised by an accused.  Where there is an air of reality to a defence, it should go to the jury.  Second, a trial judge has a positive duty to keep from the jury defences lacking an evidential foundation. A defence that lacks an air of reality should be kept from the jury.  Wu, supra; Squire, supra; Pappajohn, supra; Osolin, supra; Davis, supra. This is so even when the defence lacking an air of reality represents the accused’s only chance for an acquittal, as illustrated by R. v. Latimer, [2001] 1 S.C.R. 3, 2001 SCC 1.

52                               It is trite law that the air of reality test imposes a burden on the accused that is merely evidential, rather than persuasive. Dickson C.J. drew attention to the distinction between these two types of burden in R. v. Schwartz, [1988] 2 S.C.R. 443, at p. 466:

Judges and academics have used a variety of terms to try to capture the distinction between the two types of burdens. The burden of establishing a case has been referred to as the “ major burden,” the “primary burden,” the “legal burden” and the “persuasive burden.” The burden of putting an issue in play has been called the “minor burden,” the “secondary burden,” the “evidential burden,” the “burden of going forward,” and the “burden of adducing evidence.” [Emphasis added.]

The air of reality test is concerned only with whether or not a putative defence should be “put in play”, that is, submitted to the jury for consideration. This idea was crucial to the finding in Osolin that the air of reality test is consistent with the presumption of innocence guaranteed by s. 11( d )  of the Canadian Charter of Rights and Freedoms .

53                               In applying the air of reality test, a trial judge considers the totality of the evidence, and assumes the evidence relied upon by the accused to be true. See Osolin, supra; Park, supra.  The evidential foundation can be indicated by evidence emanating from the examination in chief or cross-examination of the accused, of defence witnesses, or of Crown witnesses. It can also rest upon the factual circumstances of the case or from any other evidential source on the record. There is no requirement that the evidence be adduced by the accused. See Osolin, supra;  Park, supra; Davis, supra.

54                               The threshold determination by the trial judge is not aimed at deciding the substantive merits of the defence. That question is reserved for the jury. See Finta, supra; R. v. Ewanchuk, [1999] 1 S.C.R. 330. The trial judge does not make determinations about the credibility of witnesses, weigh the evidence, make findings of fact, or draw determinate factual inferences. See R. v. Bulmer, [1987] 1 S.C.R. 782;  Park, supra.  Nor is the air of reality test intended to assess whether the defence is likely, unlikely, somewhat likely, or very likely to succeed at the end of the day.  The question for the trial judge is whether the evidence discloses a real issue to be decided by the jury, and not how the jury should ultimately decide the issue. 

55                               Whether or not there is an air of reality to a defence is a question of law, subject to appellate review. It is an error of law to put to the jury a defence lacking an air of reality, just as it is an error of law to keep from the jury a defence that has an air of reality. See Osolin, supra; Park, supra; Davis, supra. The statements that “there is an air of reality” to a defence and that a defence “lacks an air of reality” express a legal conclusion about the presence or absence of an evidential foundation for a defence. 

56                               The considerations discussed above have led this Court to reject unequivocally the argument that the air of reality test licenses an encroachment by trial judges on the jury’s traditional function as arbiter of fact.  As Cory J. stated in Osolin, supra, at p. 682:

This is no more than an example of the basic division of tasks between judge and jury. It is the judge who must determine if evidence sought to be adduced is relevant and admissible. In the same way, it is the judge who determines if there is sufficient evidence adduced to give rise to the defence. If there has been sufficient evidence put forward, then the jury must be given the opportunity to consider that defence along with all the other evidence and other defences left with them in coming to their verdict.

Indeed, the air of reality inquiry has been found not only to be consistent with the traditional division of labour as between judge and jury, but actually to enhance the jury’s ability to carry out its task. Again, Cory J.’s statement in Osolin, supra, at p. 683 is apposite:

The jury system has in general functioned exceptionally well. Its importance has been recognized in s. 11( f )  of the Charter . One of the reasons it has functioned so very well is that trial judges have been able to direct the minds of jurors to the essential elements of the offence and to those defences which are applicable. That process should be maintained. The charge to the jury must be directed to the essential elements of the crime with which the accused is charged and defences to it. Speculative defences that are unfounded should not be presented to the jury. To do so would be wrong, confusing, and unnecessarily lengthen jury trials. [Emphasis added.]

57                               This Court has held on many occasions that a single air of reality test applies to all defences. Osolin, supra; Park, supra, at para. 12.  The test has been applied uniformly to a wide range of defences over the years. These include the defence of honest but mistaken belief in consent in sexual assault cases (Pappajohn, supra; Bulmer, supra; Osolin, supra; Park, supra; R. v. Esau, [1997] 2 S.C.R. 777; Ewanchuk, supra; Davis, supra), and other defences such as intoxication (R. v. Robinson, [1996] 1 S.C.R. 683; R. v. Lemky, [1996] 1 S.C.R. 757), necessity (Latimer, supra), duress (R. v. Ruzic, [2001] 1 S.C.R. 687, 2001 SCC 24), provocation (R. v. Thibert, [1996] 1 S.C.R. 37), and self-defence (Brisson v. The Queen, [1982] 2 S.C.R. 227; R. v. Hebert, [1996] 2 S.C.R. 272).  Adopting different evidential standards for different classes of cases would constitute a sharp break with the authorities.

(2)  The Evidential Standard Applicable to the Air of Reality Test

58                               It remains to consider the test itself. The difficulty lies in defining the evidential standard a judge must apply in determining whether a defence possesses an air of reality.  

59                               It is argued by Arbour J. that McIntyre J.’s formulation of the air of reality test in Pappajohn, supra, did not just introduce new terminology (“air of reality”), but also substantively altered the law.  On this view, the applicable standard before Pappajohn was the “no evidence” threshold alleged to have been set out in Parnerkar v. The Queen, [1974] S.C.R. 449, and followed thereafter. Provided there was “some evidence” or “any evidence” going to the elements of a defence — or, stated negatively, unless there was “no evidence” — a defence had to be put to the jury. It is argued that the Court’s holding in Pappajohn, supra, changed all that by introducing a new and more onerous “sufficiency of the evidence” requirement. This requirement is alleged to have licensed a judicial incursion into the traditional province of the jury. Finally, it is argued that the “sufficient evidence” standard from Pappajohn ought to be understood as a special rule applicable to the defence of mistaken belief in consent in sexual assault cases, and that the purportedly traditional “no evidence” rule ought to apply to all other defences, notably self-defence under s. 34(2). For the reasons that follow, we respectfully conclude this view of the law cannot be reconciled with the authorities.

60                               The air of reality test as treated in Pappajohn, supra,  and subsequent cases is entirely consistent with the cases that preceded it. The issue has remained the same, namely: whether a defence rests upon an evidential foundation warranting that it be put to a jury. Most significantly, the specific question to be asked by a trial judge in determining whether the threshold evidential burden for putting a defence to a jury is met, has remained constant. Cory J. stated the question in Osolin, supra, at p. 682:

The term “air of reality” simply means that the trial judge must determine if the evidence put forward is such that, if believed, a reasonable jury properly charged could have acquitted.

The question as formulated by Cory J. rested squarely on the previous authorities, going back to Wu, supra.

(3)  The Evidential Standard Before Pappajohn

61                               In Wu, supra, as in the present case, the issue was whether a defence of self-defence merited being put to the jury. Lamont J. held that the trial judge in that case was under no obligation to put the defences of self-defence or provocation to the jury. He set out the reasoning of the Court, at p. 617:

There is no evidence in the record from which a jury could reasonably infer that the accused when he shot the complainant did so under a reasonable apprehension of death or bodily harm to himself, or that he reasonably believed that he could not otherwise save himself from bodily injury. The rule, therefore, that an accused person at trial is entitled to have the jury pass upon all his alternative defences is limited to the defences of which a foundation of fact appears in the record. [Emphasis added.]

It is clear from the above-quoted paragraph that the requirement has two components: (1) there must be evidence  (2) upon which a jury could reasonably draw the inferences necessary to acquit the accused. The question is not just whether there is evidence in some general sense, but whether there is evidence capable of forming the basis for an acquittal. The two-pronged statement of the test for determining whether a defence rests upon an evidential foundation — in contemporary terms, whether it has an air of reality — applies to this day.

62                               In Parnerkar, supra, Fauteux C.J. stated the question to be asked, in the context of the defence of provocation, at p. 454:

If, then, the record is denuded of any evidence potentially enabling a reasonable jury acting judicially to find a wrongful act or insult of the nature and effect set forth in s. 203(3)(a) and (b), it is then, as a matter of law, within the area exclusively reserved to the trial judge to so decide and his duty to refrain from putting the defence of provocation to the jury. [Emphasis added.]

Fauteux C.J. did not limit the question to whether the record was “denuded of any evidence”.  Rather, the question was whether the record was denuded of any evidence having the potential to enable a reasonable jury acting judicially to acquit. Whether or not the evidence has this potential is a question for the trial judge. While courts have often spoken of an inquiry into whether there is “some evidence” or “any evidence” without qualification, this must be understood as a short-form reference to the full question. The full question is whether there is evidence (some evidence, any evidence) upon which a properly instructed jury acting judicially could acquit. If there is any or some such evidence, then the air of reality hurdle is cleared. If there is no such evidence, then the air of reality hurdle is not cleared.  In view of the frequent resort by courts to short-form expressions, it may be tempting to focus on the first part of the question (“there must be some evidence”) to the exclusion of the crucial second part (“upon which a properly instructed jury acting reasonably could acquit”).  That this temptation must be resisted is made clear by the authorities.

63                               In Squire, supra, the issue was whether the defence of provocation should have been put to the jury. The accused had been convicted at trial.  The Court of Appeal ordered a new trial on the ground that the defence of provocation should have been put to the jury. The Court allowed the appeal and restored the jury’s conviction. Spence J., for the Court, stated the familiar principle that it is the duty of the trial judge to put to the jury all defences arising from the evidence, whether or not they were argued by counsel to the jury. He then stated, at p. 19:

It is, however, equally plain that a trial judge is under no duty to invite the jury to consider defences of which there is no evidence or which cannot reasonably be inferred from the evidence: . . . [Emphasis added.]

This statement makes it clear that the question of whether there is “any evidence” is incomplete. As Spence J. explicitly states, it is necessary to consider whether the evidence is reasonably capable of supporting the inferences necessary for the defence to succeed.  To the extent that the inferences required for the defence to succeed cannot reasonably be supported by the evidence, the defence lacks an evidential foundation — it has no air of reality — and should therefore not be put to the jury.

64                               It cannot plausibly be claimed that Squire, supra, presents a case where there was unambiguously “no evidence”.  On the contrary, there was some evidence which the Court of Appeal thought sufficient to warrant leaving the defence of provocation with the jury. There was evidence, in other words, capable of bearing an interpretation different from that suggested by Spence J.  The contentious issue was not the bare existence or absence of “any evidence” in some general sense, but rather whether that evidence was reasonably capable of supporting the inferences necessary to acquit the accused.  Indeed, Spence J.’s explanation of his reasoning, at pp. 21-22, demonstrates the way courts have analysed evidence in determining whether the threshold evidential burden is met:

As I have said, it is the duty of this Court to come to a decision whether those circumstances amount to any evidence that a reasonable jury acting judicially could find a wrongful act or insult sufficient to deprive an ordinary person of the power of self‑control (s. 215(2)), and I am strongly of the view, with great respect to the opinion expressed by Martin J.A., that no jury acting judicially could come to such a conclusion. It must be remembered that on the evidence the respondent was, throughout the disgraceful incidents of the evening, a bad‑tempered aggressor and that he seized on the slightest confrontation by Tremblay to again become the aggressor. If, during the fight which followed, and which I am of the opinion was a fight between two men fully consenting to the battle, he suffered a couple of kicks of indefinite violence, the provocation resulting therefrom could not possibly have caused a police officer to draw his gun and fire five shots at his assailant. To put it very shortly, I agree with the view of the learned counsel for the respondent at trial who, upon the basis of this evidence, decided that there was simply nothing on provocation which he could present to a jury and who, therefore, depended on the one available defence of drunkenness. There was much to be said in favour of that defence but very evidently the jury rejected it and that, of course, is not the subject of any review in this Court. [Emphasis added.]

It is impossible to read the above-quoted paragraph as limiting a judge to the examination of whether or not there is “any evidence” in a general sense. Clearly there was some evidence for the trial judge to consider. The question was whether that evidence, taken in its context, reasonably permitted the inference necessary to support the case.

65                               We conclude that the pre-Pappajohn authorities support a two-pronged question for determining whether there is an evidential foundation warranting that a defence be put to a jury. The question is whether there is (1) evidence (2) upon which a properly instructed jury acting reasonably could acquit if it believed the evidence to be true. The second part of this question requires the trial judge to determine whether the evidence relied upon is reasonably capable of supporting the inferences required to acquit the accused.

(4)  The Evidential Standard After Pappajohn

66                               It is argued by Arbour J. that Pappajohn did not just introduce new terminology, but substantively altered the law by introducing a “sufficiency of the evidence” requirement, which is alleged to have been different from, and more onerous than the traditional “some evidence” requirement. The following passage from Pappajohn, supra, at p. 127,  per McIntyre J., is cited as evidence of the break with the authorities:

What is the standard which the judge must apply in considering this question? Ordinarily, when there is any evidence of a matter of fact, the proof of which may be relevant to the guilt or innocence of an accused, the trial judge must leave that evidence to the jury so that they may reach their own conclusion upon it. Where, however, the trial judge is asked to put a specific defence to the jury, he is not concerned only with the existence or non‑existence of evidence of fact. He must consider, assuming that the evidence relied upon by the accused to support a defence is true, whether that evidence is sufficient to justify the putting of the defence. [Emphasis added.]

It is argued that whereas the traditional rule required only that a judge ascertain whether there was some evidence (or, put negatively, no evidence), the new rule imposed the additional requirement that the trial judge consider whether this evidence, if true, would be sufficient to warrant an acquittal.

67                               We must respectfully disagree with this analysis. The air of reality test as formulated in Pappajohn, supra, rests squarely on the previous authorities. First, as discussed above, the evidential threshold requirement before Pappajohn, was that there be some evidence on the basis of which a properly instructed jury acting reasonably could acquit.  The proposition that the test before Pappajohn was whether there is “some evidence” is correct only insofar as the expression “some evidence” is an ellipsis referring to the complete expression “some evidence on the basis of which a properly instructed jury acting reasonably could acquit”.  There is no support in the pre-Pappajohn authorities for the proposition that the evidential threshold could be satisfied by pointing to some evidence on the basis of which a properly instructed jury acting reasonably would be incapable of acquitting. That has never been the law.

68                               We are equally unconvinced by the second component of the argument. McIntyre J. in Pappajohn plainly did not suggest that the air of reality test asks whether the evidence is sufficient to justify an acquittal. Rather, the inquiry was stated to be into whether there was evidence sufficient to warrant putting the issue of acquittal to the jury.  The inquiry has always been about the sufficiency of the evidence in this narrow sense.

69                               It is argued that asking whether “the evidence is sufficient” is precisely what Parnerkar had ruled out for the trial judge as being within the exclusive province of the jury.  The argument relies on the following passage from Parnerkar, supra, at p. 454:

There is nothing, either expressed or necessarily implied, in the language of s. 203(3) to indicate an intention of Parliament to modify the principle according to which the sufficiency of evidence, which is an issue only where there is some evidence, is a question of fact for the jury and the absence of evidence is a question of law for the trial judge.  [Emphasis added.]

We respectfully conclude that this argument ascribes too much significance to the use of the terms “sufficient” and “sufficiency”, and in so doing misreads both Parnerkar, supra, and Pappajohn, supra.

70                               The key point about “sufficiency” is rooted not in a legal text, but in ordinary language. The statement that “the evidence is sufficient” on its own is meaningless, in the same way that the statement “two dollars is sufficient” is meaningless. The question arises: sufficient for what purpose? Evidence, or anything else for that matter, is sufficient or insufficient in relation to a specific standard or criterion. In the criminal law, evidence is measured against two distinct standards, corresponding to two distinct issues.  First, evidence can be sufficient or insufficient to meet the evidential burden imposed by the air of reality test. Second, evidence can be sufficient or insufficient to establish guilt beyond a reasonable doubt or to warrant an acquittal. The first issue, which is the subject matter of the evidential burden (or “minor burden”) relating to defences, is reserved for the judge. The second issue, which is the subject matter of the persuasive burden (or “major burden” or “primary burden” or “legal burden”)  is reserved for the jury.

71                               Fauteux C.J.’s statement in Parnerkar, supra, that the sufficiency of the evidence is for the jury to decide clearly referred to the second issue, namely the question of whether the evidence was sufficient to warrant an acquittal.  It seems to us that he could not have meant that the word “sufficient” should not be used to describe the evidential threshold applicable to whether a defence should be put to a jury. Indeed, the words “sufficient” and “sufficiency” have often been used, appropriately, to describe the evidential threshold. See for example Perka v. The Queen, [1984] 2 S.C.R. 232, at p. 259. The following basic explanation of the concept of an evidential burden illustrates the use of the term "sufficient" in this context:

The obligation on a party to adduce sufficient evidence on a fact in issue to justify a finding on that fact in his favour, is referred to as ‘the evidential burden’. A party discharges an evidential burden borne by him by adducing sufficient evidence for the issue in question to be submitted to the jury (tribunal of fact). Whether there is sufficient evidence is a question of law for the judge.

(A. Keane, The Modern Law of Evidence (5th ed. 2000), at p. 37)

72                            A further example is to be found in Finta, supra, at pp. 846-47, in which Cory J. gives a clear explanation of the idea of “sufficient evidence” in the context of the threshold test:

It is for the trial judge to decide whether the evidence is sufficient to give rise to the defence as this is a question of law (Parnerkar v. The Queen, [1974] S.C.R. 449; Dunlop v. The Queen, [1979] 2 S.C.R. 881). There is thus a two‑stage process to be followed. The trial judge must look at all the evidence to consider its sufficiency. Then, if the evidence meets the threshold, it should be put before the jury which will weigh it and decide whether it raises a reasonable doubt. See: Wigmore on Evidence (1983), vol. IA, at pp. 968‑69; R. v. Faid, [1983] 1 S.C.R. 265, at p. 276. This is all that is meant by the requirement of sufficient evidence. [Emphasis added.]

73                            In writing that the judge must ascertain whether the evidence is sufficient to warrant putting it to a jury, McIntyre J. in Pappajohn, supra, did not introduce a new standard for the threshold evidential test. Indeed, he did not thereby introduce any standard.  Rather, he indicated the standard  in the next sentence, at p. 127, in which he cited Wu, supra, and then quoted from Kelsey v. The Queen,  [1953] 1 S.C.R. 220, and  Workman v. The Queen, [1963] S.C.R. 266:

The test to be applied has, in my opinion, been set down by Fauteux J., as he then was, in Kelsey v. The Queen.

The allotment of any substance to an argument or of any value to a grievance resting on the omission of the trial judge from mentioning such argument must be conditioned on the existence in the record of some evidence or matter apt to convey a sense of reality in the argument and in the grievance.

In addition, I would refer to the words of Judson J., speaking for the majority, in R. v. Workman and Huculak where he said:

. . . I can see no possible ground for any instruction that, on any view of the evidence Huculak could be an accessory after the fact and not a principal. Before this could be done, there must be found in the record some evidence which would convey a sense of reality in the submission (Kelsey v. The Queen, 105 C.C.C. 97 at p. 102, [1953] 1 S.C.R. 220 at p. 226, 16 C.R. 119 at p. 125). Failure of counsel to raise the matter does not relieve the trial judge of his duty to place a possible defence before the jury but there must be something beyond fantasy to suggest the existence of the duty. [Emphasis added.]

74                               We have already considered Wu, supra, and the test set out in that case. The remaining question is whether by specifying that there must be some evidence apt to convey a sense of reality to the defence, and that there must be evidence which would convey a sense of reality to the defence, McIntyre J. was imposing an additional requirement, in excess of the traditional requirement set out in Wu, supra.  In Pappajohn, supra, at p. 128, McIntyre J. set out just what was required to sustain the “sense of reality”, to move “beyond fantasy”:

In this case, to convey such a sense of reality, there must be some evidence which, if believed, would support the existence of a mistaken but honest belief that the complainant was in fact consenting to the acts of intercourse which admittedly occurred. [Emphasis added.]

This is precisely the test stated in Wu, supra, and the other authorities considered above. The statement of the test in Pappajohn, supra, can hardly be regarded as heralding a break from the authorities. Rather, the adoption of the phrase “air of reality” served to emphasize what had already long been established:  the question is not whether there is some evidence, but whether there is some evidence reasonably capable of supporting an acquittal.

75                               In Pappajohn, supra, McIntyre J. explained his conclusion that the defence of mistaken but honest belief in consent lacked an air of reality, at p. 132:

In reaching this conclusion, I am not unmindful of the evidence of surrounding circumstances which were said to support the appellant's contention. I refer to the absence of serious injury suffered by the complainant and the absence of damage to clothing, as well as to the long period of time during which the parties remained in the bedroom. These matters may indeed be cogent on the issue of actual consent but, in my view, they cannot by themselves advance a suggestion of a mistaken belief. [Emphasis added.]

McIntyre J.’s reasoning was that the evidence put forward could not serve as the basis for the inferences required by the defence. The evidence was incapable of  “advancing” the conclusions that would have been necessary for the defence to succeed. Just as in Squire, supra, the issue was not whether there was “some evidence” in a general sense, but whether there was some evidence upon which a jury could reasonably have based an acquittal. This is precisely what a trial judge is called upon to decide, following the authorities.

76                               McIntyre J. wrote for the majority in Pappajohn. Martland J. wrote concurring reasons, in which he did not disagree with McIntyre J.’s description of the test.  But the reasons of Dickson J., as he then was, dissenting, are especially helpful in understanding the view of the Court concerning the approach to determining whether a defence merits being put to the jury. It is apparent that Dickson J., for himself and Estey J., affirmed precisely the same approach to air of reality as McIntyre J. had.  Dickson J. wrote, at p. 158:

If there was “some” evidence to “convey a sense of reality” to a defence of mistake as to consent, then the jury ought to have been instructed to consider that plea. Kelsey v. The Queen, [1953] 1 S.C.R. 220, at p. 226.

Evidently, the above-quoted sentence cannot be understood as saying that the presence of “some evidence” will in itself give an air of reality to a defence. Rather, the plain meaning is that some evidence that is capable of giving an air of reality to the defence is required. Dickson J. then went on to state how that determination ought to be made:

As stated in Dunlop and Sylvester v. The Queen, [1979] 2 S.C.R. 881, at p. 890:

The question of whether there is sufficient evidence to go to the jury, i.e. any evidence upon which a jury, properly instructed, could find the appellants guilty . . . is a question of law, which can found an appeal to this Court under s. 618(1)(a).  [Emphasis added.]

Just as McIntyre J. had done, Dickson J. first stated the question to be whether there was some evidence that could give the defence a “sense of reality”, and then specified that the answer to this question would depend on whether a properly instructed jury could base the required inferences on the evidence. The standard applied by Dickson J. was the same as that put forth by McIntyre J., and the same as that in Wu: is there evidence upon which a properly instructed jury could acquit?

77                               While Dickson J. agreed with McIntyre J. about the applicable test, he came to a different conclusion on its application to the facts of that case. Once again, it is instructive to consider Dickson J.’s reasoning. He wrote in Pappajohn, supra, at p. 163:

There is circumstantial evidence supportive of a plea of belief in consent: (1) Her necklace and car keys were found in the living room. (2) She confirmed his testimony that her blouse was neatly hung in the clothes closet. (3) Other items of folded clothing were found at the foot of the bed. (4) None of her clothes were damaged in the slightest way. (5) She was in the house for a number of hours. (6) By her version, when she entered the house the appellant said he was going to break her. She made no attempt to leave. (7) She did not leave while he undressed. (8) There was no evidence of struggle, and (9) She suffered no physical injuries, aside from three scratches. [Emphasis added.]

Dickson J. did not just point to “some evidence” in some abstract sense, but to  evidence that he considered to be supportive of the inferences required for the defence to succeed. His disagreement with the majority turned upon his view that the evidence on the record was evidence upon which a reasonable jury could acquit.  The crucial point for present purposes is that, while Dickson J. dissented on the application of the test to the facts, the whole Court in Pappajohn was  agreed as to the correct approach for determining whether a defence has an air of reality.

78                               It is true that Pappajohn attracted criticism. But for the most part, that criticism was directed at the corroboration requirement that had been introduced in that case. In our view, the corroboration requirement was never thought to be a core principle of the air of reality test itself. Rather, the corroboration requirement was initially seen as the practical consequence of applying the traditional air of reality principles to the defence of mistaken but honest belief in consent to sex.  Subsequent cases of this Court on air of reality have made it clear that there is no corroboration requirement. See Osolin, supra; Park, supra; Davis, supra. Therefore, criticism of Pappajohn that objected to the corroboration requirement, or that objected to the air of reality test on the basis of the corroboration requirement, must be regarded as having lost its relevance to the assessment of the air of reality test.

79                               In Brisson, supra, decided shortly after Pappajohn, the key issue was whether there was sufficient evidence to leave the defence of self-defence with the jury. Though Pappajohn, supra, had been about sexual assault, there was no suggestion that the air of reality test was for that reason inapplicable, or was required to be applied differently, to the defence of self-defence. In Brisson, supra, McIntytre J. stated for the majority, at p. 235:

A trial judge must put to the jury all defences which arise from the evidence. There must, however, be some evidence sufficient to give an air of reality to a defence before the obligation to put a defence can arise. [Emphasis added.]

He then quoted from Kelsey, supra, and cited Wu, supra, and the other authorities cited in Pappajohn, supra, in order to flesh out the substantive criterion for determining whether a defence has an air of reality.  In other words, McIntyre J. once again applied the test as traditionally formulated, but couched the analysis in the new air of reality language. After considering the evidence, McIntyre J. concluded that there was no air of reality to the defences claimed by the accused.

80                               The concurring reasons of Laskin C.J., writing for himself and Ritchie J., help to shed further light on the approach of the Court to determining whether a defence rests upon an evidential foundation sufficient to leave it with a jury.  Laskin C.J. stated, at pp. 234-35, that while he had initially thought there was sufficient evidence to warrant putting the defences of self-defence and provocation, he had subsequently changed his mind, having been persuaded by McIntyre J.’s analysis of the evidence.  The approach set out by McIntyre J. in Pappajohn, supra, and affirmed in Brisson, supra, was subsequently followed in other cases of this Court. See for example Bulmer, supra; R. v. Robertson, [1987] 1 S.C.R. 918.

81                               Any doubt as to the correct evidential standard must be taken to have been dispelled by the clear statement of the air of reality test by Cory J. in Osolin, supra. That case made it clear that the question to be asked by the trial judge in applying the air of reality test is whether there is evidence upon which a properly instructed jury acting reasonably could acquit if it accepted the evidence as true; see Osolin, supra, per Cory J., at p. 682. This statement of the test was subsequently affirmed by this Court in numerous cases, covering a wide range of defences. See Finta, supra, per Cory J., at pp. 846-47, in the context of a defence of mistake of fact and obedience to superior orders to charges of war crimes and crimes against humanity; Park, supra, at paras. 30-31, in the context of sexual assault; Thibert, supra, per Cory J., at para. 7, in the context of provocation; Robinson, supra, at para. 71, per Lamer C.J., in the context of an intoxication defence; Latimer, supra, at para. 35, in the context of a defence of necessity. In Lemky, supra, which concerned the defence of intoxication,  McLachlin J. (as she then was) stated the test in terms of whether the evidence was reasonably capable of supporting the inferences required for the defence to succeed, at para. 15. This formulation is entirely consistent with Wu, supra, Parnerkar, supra, Squire, supra, and Osolin, supra.

82                               We conclude that the authorities after Pappajohn continue to support a two-pronged question for determining whether there is an evidential foundation warranting that a defence be put to a jury. The question remains whether there is (1) evidence (2) upon which a properly instructed jury acting reasonably could acquit if it believed the evidence to be true. The second part of this question can be rendered by asking whether the evidence put forth is reasonably capable of supporting the inferences required to acquit the accused. This is the current state of the law, uniformly applicable to all defences.

(5)  “Some Evidence”, “No Evidence”, “Sufficiency of the Evidence”: A Restatement

83                               There is nothing inherently objectionable about using the “no evidence”, “some evidence” or “any evidence” terminology when describing the evidential standard applicable to the air of reality test, provided that these terms are understood as elliptical references to the full question, which is whether there is evidence (some evidence, any evidence) on the basis of which a properly instructed jury acting reasonably could base an acquittal if it believed the evidence to be true.  The corollary of this point is that it is an error of law to use the terms “some evidence”, “no evidence” and “any evidence” to refer to a question other than the full question stated above.  See Osolin, supra, at pp. 678-79; Park, supra, at para. 37.  This approach was affirmed for directed verdicts of acquittal and committal for trial following a preliminary inquiry in R. v. Arcuri, [2001] 2 S.C.R. 828, 2001 SCC 54, at para. 25, per McLachlin C.J. for the Court.  There is no authority for a threshold that could be satisfied by pointing to evidence that is incapable of reasonably supporting the inferences necessary to acquit an accused.  Before putting a defence to a jury, it is the trial judge’s duty to ask not just whether there is evidence in some general sense, but whether there is evidence that is reasonably capable of supporting an acquittal. This requires an assessment of whether the evidence relied upon is reasonably capable of supporting the inferences required for the defence to succeed.

(6)  Determining the Range of Reasonable Inferences

84                               The whole edifice of air of reality analysis is built upon the premise  that putting outlandish defences to the jury would be confusing and would invite unreasonable verdicts.  This is the practical basis for the trial judge’s duty to keep from the jury defences that are fanciful or far-fetched in relation to the evidence adduced at trial. As L’Heureux-Dubé J. stated for the Court in Park, supra, at para. 11:

The common law has long recognized that a trial judge need not put to the jury defences for which there is no real factual basis or evidentiary foundation. Courts must filter out irrelevant or specious defences, since their primary effect would not be to advance the quest for truth in the trial, but rather to confuse finders of fact and divert their attention from factual determinations that are pertinent to the issue of innocence or guilt.

The role of the jury is of paramount importance in the common law tradition. It is in keeping with the solemnity of the jury’s task that frivolous questions are not put to it. Inviting the jury to solemnly consider that which is plainly unworthy of solemn consideration would in effect be to mislead the jury. The jury must not be diverted from the real issues in a case by defences that the evidence cannot reasonably support.

85                               We do not agree that the more outlandish a defence, the less likely a jury is to be deceived by it. At bottom, this is not really an argument for a particular evidential standard, but an argument against having an air of reality test in the first place. In our view, the long-standing concern with ensuring that specious defences are kept from the jury is well founded, and reflects the practical wisdom of the common law.

86                               The approach to be followed by a trial judge in ensuring that only defences arising on the facts are put to the jury is well established. The question is whether there is evidence upon which a properly instructed jury acting reasonably could acquit if it accepted it as true. We have seen that this question requires the trial judge to consider whether the inferences required to be established for the defence to succeed can reasonably be supported by the evidence. Where evidence does not permit a reasonable inference raising a reasonable doubt on the basis of the defence, the defence must be kept from the jury. This duty of the trial judge is at the very heart of the air of reality analysis.

87                               The trial judge must review the evidence and determine whether, if  believed, it could permit a properly instructed jury acting reasonably to acquit. It follows that the trial judge cannot consider issues of credibility. Further, the trial judge must not weigh evidence, make findings of fact, or draw determinate factual inferences.

88                               If there is direct evidence as to every element of the defence, whether or not it is adduced by the accused, the trial judge must put the defence to the jury. By definition, the only conclusion that needs to be reached in such a case is whether the evidence is true: see Watt's Manual of Criminal Evidence (2001), at § 8.0: “[d]irect evidence is evidence which, if believed, resolves a matter in issue”; McCormick on Evidence (5th ed. 1999), at p. 641; J. Sopinka, S. N. Lederman and A. W. Bryant, The Law of Evidence in Canada (2nd ed. 1999), at p. 39, citing Commonwealth v. Webster, 59 Mass. (5 Cush.) 295 (1850), at p. 310 (direct evidence is witness testimony as to “the precise fact which is the subject of the issue on trial”).  It is for the jury to say whether and how far the evidence is to be believed. However, the mere assertion by the accused of the elements of a defence does not constitute direct evidence, and will not be sufficient to put the defence before a jury. See Park, supra; Davis, supra.

89                               The judge’s task is somewhat more complicated where the record does not disclose direct evidence as to every element of the defence, or where the defence includes an element that cannot be established by direct evidence, as for example where a defence has an objective reasonableness component.  In each of these cases, the question becomes whether the remaining elements of the defence — that is, those elements of the defence that cannot be established by direct evidence — may reasonably be inferred from the circumstantial evidence.  Circumstantial evidence is “evidence that tends to prove a factual matter by proving other events or circumstances from which the occurrence of the matter at issue can be reasonably inferred” (Merriam‑Webster’s Dictionary of Law (1996), at p. 172).

90                               The approach to be followed by the judge in determining whether the evidence put forth in a case is reasonably capable of supporting the requisite inferences  was discussed in Arcuri, supra.  In that case, the Court clarified the principles governing a preliminary inquiry judge’s assessment of the evidence in determining whether “there is any evidence upon which a reasonable jury properly instructed could return a verdict of guilty” (para. 21).  That question is essentially the same as the question applicable to air of reality analysis. Moreover, the same prohibition against assessing credibility, weighing evidence, making findings of fact and drawing factual inferences applies to both air of reality and preliminary inquiries. McLachlin C.J., for the Court, coined the term “limited weighing” to characterize the nature of the inquiry traditionally undertaken by the trial judge, at para. 23:

Answering this question inevitably requires the judge to engage in a limited weighing of the evidence because, with circumstantial evidence, there is, by definition, an inferential gap between the evidence and the matter to be established — that is, an inferential gap beyond the question of whether the evidence should be believed: . . . (circumstantial evidence is "any item of evidence, testimonial or real, other than the testimony of an eyewitness to a material fact. It is any fact from the existence of which the trier of fact may infer the existence of a fact in issue”); . . . (“[c]ircumstantial evidence . . . may be testimonial, but even if the circumstances depicted are accepted as true, additional reasoning is required to reach the desired conclusion”).  The judge must therefore weigh the evidence, in the sense of assessing whether it is reasonably capable of supporting the inferences that the [accused would ask] the jury to draw.  This weighing, however, is limited.  The judge does not ask whether she herself would conclude that the accused is guilty. Nor does the judge draw factual inferences or assess credibility. The judge asks only whether the evidence, if believed, could reasonably support an inference of guilt. [Emphasis added; emphasis in original deleted.]

91                               The expression “limited weighing”, adopted by the full Court in Arcuri, was intended as a more accurate description of the judge’s traditional function.  The “limited weighing” description accords just as well with the approach to air of reality as originally set out in Wu, supra, and followed to this day. In Arcuri, McLachlin C.J. quoted at para. 24 from Metropolitan Railway Co. v. Jackson (1877), 3 App. Cas. 193 (H.L.),  to emphasize the consonance of the limited weighing approach with the traditional distinction between the respective tasks of the judge and jury:

The Judge has a certain duty to discharge, and the jurors have another and a different duty. The Judge has to say whether any facts have been established by evidence from which [the matter in issue] may be reasonably inferred; the jurors have to say whether, from those facts . . . [the matter in issue] ought to be inferred. [Emphasis added.]

This distinction is crucial. The judge does not draw determinate factual inferences, but rather comes to a conclusion about the field of factual inferences that could reasonably be drawn from the evidence.

C.        Application to Self-Defence Under Section 34(2)

92                                  This brings us to the application of the air of reality test to the facts of this case. The question to be asked is whether there is evidence on the record upon which a properly instructed jury acting reasonably could acquit if it believed the evidence to be true.

93                                  In Pétel, supra, at p. 12, Lamer C.J. stated the three constitutive elements of self-defence under s. 34(2): “(1) the existence of an unlawful assault; (2) a reasonable apprehension of a risk of death or grievous bodily harm; and (3) a reasonable belief that it is not possible to preserve oneself from harm except by killing the adversary”.  All three of these elements must be established in order for the defence to succeed. The air of reality test must therefore be applied to each of the three elements. If any of these elements lacks an air of reality, the defence should not be put to the jury. See Hebert, supra; Latimer, supra.

94                                  Each of the three elements under s. 34(2) has both a subjective and an objective component. The accused’s perception of the situation is the “subjective” part of the test.  However, the accused’s belief must also be reasonable on the basis of the situation he perceives.  This is the objective part of the test.  Section 34(2) makes the reasonableness requirement explicit in relation to the second and third conditions.  Pétel held that the same standard applies to the first component of the defence, namely, the existence of an assault.  With respect to each of the three elements, the approach is first to inquire about the subjective perceptions of the accused, and then to ask whether those perceptions were objectively reasonable in the circumstances.

95                                  The air of reality analysis must be applied to each component of the defence, both subjective and objective. Evidence capable of supporting a particular finding of fact with respect to one component of the defence will not necessarily be capable of supporting other components of the defence. In the case of a  defence of self-defence under s. 34(2), the testimony of the accused as to his perceptions does not necessarily constitute evidence reasonably capable of supporting the conclusion that the perception was reasonable.

96                                  The difficult issue in this case is whether there is some evidence upon which a properly instructed jury acting reasonably could have concluded that the accused’s purported perceptions were reasonable under the circumstances. Since reasonableness is inherently incapable of being established by direct evidence, the key question is whether there is evidence on the basis of which reasonableness could reasonably be inferred by a jury. If a jury could not have reasonably come to the conclusion that the accused’s perceptions were reasonable, even accepting that his testimonial evidence was true, then the defence should not have been put to the jury.

97                                  There is no authority for the proposition that reasonableness is exempt from the air of reality test, or that evidence satisfying the air of reality test as to the subjective component of defence will automatically confer an air of reality upon the whole defence. Moreover, we consider that the introduction of such a requirement would constitute an unwarranted and illogical break with the rationale underlying air of reality analysis. The long-standing requirement is that the whole defence must have an air of reality, not just bits and pieces of the defence.  See Hebert, supra, at para.16, per Cory J.,  holding that a defence of self-defence lacked an air of reality precisely in that the reasonableness of an accused’s purported perception could not be supported by the evidence. See also Thibert, supra, per Cory J., at paras. 6-7.

(1)   The Putative Evidential Basis for the Defence

98                                  The evidence relied upon in this case emanates from the accused’s own testimony. While this Court has made it clear that a mere assertion by the accused of the elements of a defence will not be sufficient to clear the air of reality hurdle, that principle does not have any application to the present case. The accused’s testimony goes beyond merely asserting the elements of the defence, and provides a comprehensive account of his perceptions and his explanation for them.  As was stated above, credibility is not an issue in air of reality analysis. The issue is not whether the accused (or any other witness) should be believed. Rather, the question is whether, if the jury were to accept the construction of the evidence most favourable to the accused’s position, the requisite inferences could reasonably be drawn.

99                                  With these considerations in view, we now turn to a review of the relevant evidence.

100                              In his testimony, the accused pointed to many things that he perceived as indications that he was about to be attacked. The accused testified that he had heard rumours that the victim, Mike, and his companion, Ice,  planned to kill him. He claimed to have received a specific warning to that effect by a friend of his.

101                              He testified that on the night in question, Mike and Ice did not take off their jackets when they came to his home. The accused testified that Ice put his hand under his coat, in what the accused took to be a suspicious gesture.  The accused claimed that these facts led him to believe that Mike and Ice were armed, though they denied it when he asked them about it.  They whispered to each other throughout the evening, which the accused also found suspicious.

102                              The accused testified that his suspicions were further aroused when he entered the van. He ascribed significance to the fact that Ice removed the gloves that he had initially been wearing and replaced them with a different pair, putting them on  before getting in the van. He testified that this was unusual. He also testified that Mike sat behind the accused, on the passenger side, and was wearing latex surgical gloves. The accused testified that he associated wearing  gloves of this type with a “burn” (i.e. an attack on a criminal by another criminal). He testified that this was based on limited personal experience of his own, and on movies that he had seen, in which hit men wore such gloves. The accused testified that Mike loudly snapped the gloves at some point during the ride, which he interpreted as an overt threat.  The accused testified that the gloves indicated that a plan to kill him existed and would be executed that very night. 

103                              The accused further testified that the sudden change in the routine of the criminal group had meaning.  He testified that Ice avoided eye contact with him, and that everyone in the van was uncharacteristically silent. He testified that Ice repeated the suspicious gesture that he had initially made in the apartment, placing his hand under his coat.  He said that Ice did this both in the van and in the gas station. The accused testified that in the gang culture to which he belonged, this gesture is meant to communicate to rival gang members that one is armed.  When pressed to say whether the purpose of Ice’s gesture was to show he had a gun or to signal to the accused that he was going to kill him, the accused was reluctant to say it was a message that he would kill him. He nevertheless indicated that this gesture had added significance to him because there was a rumour that he was about to be killed.

104                              The accused testified that he felt trapped, and that he was convinced that Ice and the victim, Mike, were just waiting for the right moment to kill him. He testified that he felt that Ice and Mike had set a trap for him. He testified that he thought that Mike, who was sitting behind him, would be the one to kill him. He testified that he knew that Mike had used firearms before.

105                              When asked why he did not run away or call the police, the accused claimed not to have thought of these options.  When pressed on cross-examination about not running away, he indicated that he felt he should not have to leave the van.  With respect to calling the police, he also said:  “I’m not used to calling the police, you know.  I’m just not used to that.  I never called the police in my life.  People have been calling the police [on] me all my life.  You know, I’ve been running away from the police all my life”.  He also stated that the police would have arrived too late to save him.  He added that asking the police for help would have meant having to work for them as an informant.  He said of being an informant:  “that’s the only way  you can get protection, you know?  And since I wasn’t gonna do that, and I never will – you know what I’m saying? – so there was no way the police was going to protect me, no way!”

106                              It must now be determined whether a properly instructed jury acting reasonably could base an acquittal on the evidence reviewed above, assuming that evidence to be true. In order to make this determination, the evidence must be considered in relation to each of the three elements of self-defence under s. 34(2).

(2)   The Existence of an Assault

107                              Lamer C.J. stated in Pétel, supra, that the existence of an actual assault is not a  prerequisite for a defence under s. 34(2).  Rather, the starting point is the perspective of the accused. Lamer C.J. stated at p. 13:

The question that the jury must ask itself is therefore not “was the accused unlawfully assaulted?” but rather “did the accused reasonably believe, in the circumstances, that she was being unlawfully assaulted?”

Of course, in applying the air of reality test, the judge should not try to answer the question stated by Lamer C.J.  The focus of the trial judge in air of reality analysis is narrower. The question is whether there is evidence upon which a jury acting reasonably could conclude that the accused reasonably believed he was about to be attacked,  not whether the jury should so conclude.  Assuming there is an air of reality to the whole defence, it will be up to the jury to decide whether or not the accused actually believed that he was about to be attacked, and whether or not that perception was reasonable.

108                              There is an air of reality to the subjective component of the defence.  There is direct evidence on the accused’s beliefs, in the form of the accused’s testimony. It is open to the jury to believe this testimony. It is open to the jury to believe that the accused interpreted the various items pointed out in the evidence reviewed above as indicating that the victim and Ice were going to attack him.

109                              Whether a jury could reasonably infer on the basis of the evidence that the accused’s perception of an attack was reasonable in the circumstances presents a more difficult issue. Nevertheless, here again it seems to us that the threshold test is met. A jury acting reasonably could draw an inference from the circumstances described by the accused, including particularly the many threatening indicators to which he testified, to the reasonableness of his perception that he was going to be attacked.

110                              We conclude that it would  be possible for the jury reasonably to conclude that the accused believed that he was going to be attacked, and that this belief was reasonable in the circumstances. In coming to this conclusion, we do not express any opinion as to the substantive merits of the defence with respect to the first element of self-defence under s. 34(2).  That question is reserved for the jury.

111                              The inquiry does not end here. In order for the defence to be put to the jury, there must also be an “air of reality” to the remaining two elements of self-defence under s. 34(2), namely the accused’s reasonable perception of the risk of death or grievous bodily harm and his belief on reasonable grounds that there was no  alternative to killing the victim.

(3)   Reasonable Apprehension of Death or Grievous Bodily Harm

112                              The analysis as it relates to this second prong of self-defence under s. 34(2) follows substantially the same path as for the first prong. In order for this element of self-defence to clear the air of reality hurdle, it must be  possible for the jury reasonably to infer from the evidence not only that the accused reasonably believed that he was facing an attack, but that he faced death or grievous bodily harm from that attack.

113                              The accused’s testimony is unambiguously to the effect that he feared a deadly attack. It is open to the jury to accept this testimony, that is, to accept that he did in fact have this perception. There is therefore an air of reality to the subjective component of the defence.

114                              There is also an air of reality to the objective component of this element of the defence. On the particular facts of this case, this conclusion goes hand in hand with the determination that there is an air of reality to the first element of self-defence. That is, for the same reason that there is an air of reality to the reasonableness of the accused’s perception he was going to be attacked, so too is there an air of reality to the accused’s perception that the attack would be deadly.  The accused’s whole story is that he thought Ice and Mike were carrying out a plan to kill him, and that at least one of them was armed. The jury could not reasonably accept the accused’s testimony that he believed that he was going to be attacked, but simultaneously disbelieve his claim that he thought the attack would be deadly. Similarly, the evidential basis for inferring the reasonableness of the accused’s perception that he was going to be attacked is also the evidential basis for inferring the reasonableness of his perception that the attack would be deadly.

115                              A jury acting reasonably could draw an inference from the circumstances described by the accused, including particularly the indications that Mike and Ice were armed, the rumours of a plan to assassinate him, the suspicious behaviour, and the wearing of the gloves, to the reasonableness of his perception that he was in mortal danger.

116                              Once again, we wish to stress that the conclusion that there is an air of reality to the second prong of this defence does not involve an appraisal of the substantive merits of the defence. This conclusion rests upon our assessment that a properly instructed jury acting reasonably could infer the reasonableness of the accused’s perception that he faced a deadly attack. Whether a jury should come to such a conclusion is an entirely different question, which is entirely irrelevant to the air of reality analysis.

(4)   Reasonable Belief in the Absence of Alternatives to Killing

117                              We now come to the third and final element of self-defence under s. 34(2).  This requirement too has both a subjective and an objective component. The inquiry starts with the subjective perceptions of the accused at the relevant time, and then asks whether those perceptions were reasonable. It must be established both that the accused believed that he could not preserve himself except by shooting the victim, and that he held this belief on reasonable grounds.

118                              The inquiry into the inferences reasonably capable of being drawn by a jury must focus on the following sequence of events leading up to the accused’s killing of  Mike, the victim. The accused testified that he feared a deadly attack from Ice and, more particularly, from Mike, who was sitting behind him. The accused testified to the following sequence of events. Reasoning that Mike and Ice would be less likely to carry out their murderous plan in public view, the accused  pulled into the well-lit parking lot of a service station. He then set about replenishing the van’s supply of windshield washer fluid. He got out of the van, in which Ice and Mike continued to sit. He entered the service station. There, he had an exchange with the cashier, and realized that he did not have enough money to pay for the fluid.  He exited  the service station, and returned to the van, in which Mike and Ice still waited. He borrowed money from Ice. He then re-entered the service station, and bought the washer fluid. He exited the service station again. He regained the van, popped open the hood, and replenished the supply of fluid. When he was done, he put down the container, walked to the back door of the van, opened the door, and shot Mike in the head.

119                              The first question is whether there is an air of reality to the accused’s claim that, at the time he shot the victim, he actually believed that he had no alternative. We believe that there is. The starting point in air of reality analysis is that the accused’s evidence is assumed to be true.  The accused’s extensive direct testimony regarding his subjective perceptions at the relevant time amounts to more than a “mere assertion” of the element of the defence. Provided there is an air of reality to the whole defence, a jury is entitled to make a determination as to credibility, and to decide whether the accused really did believe that he could not preserve himself from death or grievous bodily harm except by killing the victim.

120                              A final issue remains. The question is whether there is anything in the testimony of the accused on the basis of which a properly instructed jury acting reasonably could infer the reasonableness of the accused’s belief that he had no alternative but to kill the victim, at the end of the sequence of events described above. We conclude that there is no such evidence to be found in the accused’s testimony, or in any other source.

121                              By specifying that an accused must believe on reasonable grounds that he had no alternative, Parliament injected an element of objectivity into the defence of self-defence. It is not enough for an accused to establish a subjective conviction that he had no choice but to shoot his way out of a dangerous situation. Nor is it enough for an accused to provide an explanation setting out just why he believed what he did was necessary. The accused must be able to point to a reasonable ground for that belief. The requirement is not just that the accused be able to articulate a reason for holding the belief, or point to some considerations that tended, in his mind, to support that belief. Rather, the requirement is that the belief that he had no other option but to kill must have been objectively reasonable.

122                              The accused testified that calling the police from within the service station would have been ineffective, as he believed that they would not have arrived in time to save him.  This part of the accused’s testimony may provide an evidential basis from which a jury acting reasonably could infer the reasonableness of the accused’s belief that he could not have preserved himself from death by calling the police.  There may even be an evidential basis from which a jury acting reasonably could infer a reasonable belief by the accused that it was unsafe to return to his apartment. But this is not an evidential foundation capable of supporting the defence of self-defence under s. 34(2). 

123                              Section 34(2) does not require that an accused rule out a few courses of action other than killing. The requirement is that the accused have believed on reasonable grounds that there was no alternative course of action open to him at that time, so that he reasonably thought he was obliged to kill in order to preserve himself from death or grievous bodily harm. In this case, there is absolutely no evidence from which a jury could reasonably infer the reasonableness of a belief in the absence of alternatives. There is nothing in the evidence to explain why the accused did not wait in the service station rather than go back to the van. There is absolutely nothing to explain why he did not flee once he had left the van. Indeed, there is nothing to suggest the reasonableness of his conclusion that he needed to walk back to the van and shoot the victim.

124                              Self-defence under s. 34(2) provides a justification for killing. A person who intentionally takes another human life is entitled to an acquittal if he can make out the elements of the defence. This defence is intended to cover situations of last resort.  In order for the defence of self-defence under s. 34(2) to succeed at the end of the day, a jury would have to accept that the accused believed on reasonable grounds that his own safety and survival depended on killing the victim at that moment. There is no evidence in the case on the basis of which a properly instructed jury acting reasonably could come to that conclusion. The inferences required for the defence to succeed are simply not capable of being supported by the accused’s testimony.

125                              Since there is no evidential foundation for the third element of  self-defence under s. 34(2), the defence as a whole lacks an air of reality.

126                              We conclude that the defence of self-defence under s. 34(2) should never have been put to the jury.  The appeal should be allowed.

The reasons of Gonthier and Binnie JJ. were delivered by

127                              Binnie J. — I concur with the Chief Justice and Bastarache J., and with the reasons they have given, that the appeal should be allowed.  I add these paragraphs on what I think is the decisive point.

128                              My colleagues have mobilized considerable scholarship for and against all aspects of the issues.  When the smoke clears, this appeal comes down to a simple proposition.  A criminal code that permitted preemptive killings within a criminal organization on the bare assertion by the killer that no course of action was reasonably available to him while standing outside a motor vehicle other than to put a shot in the back of the head of another member sitting inside the parked vehicle at a well-lit and populated gas station is a criminal code that would fail in its most basic purpose of promoting public order. 

129                              The respondent says he did not consider going to the police, although he was outside the car and in a position to flee the scene.  He said “I never called the police in my life”.  Even if the police unexpectedly got there before a shoot-out, they would ask for some information in return for protection.  “That’s how it works”, he said.  Accordingly, there was evidence that subjectively, as a self-styled criminal, he felt his only options were to kill or be killed.  He wishes the jury to judge the reasonableness of his conduct by the rules of his criminal subculture, which is the antithesis of public order.

130                              A trial judge should be very slow to take a defence away from a jury.  We all agree on that.  Here, however, the only way the defence could succeed is if the jury climbed into the skin of the respondent and accepted as reasonable a sociopathic view of appropriate dispute resolution.  There is otherwise no air of reality, however broadly or narrowly defined, to the assertion that on February 3, 1994, in Montréal, the respondent believed on reasonable grounds that he could not otherwise preserve himself from death or grievous bodily harm, as required by s. 34(2) (b) of the Criminal Code , R.S.C. 1985, c. C-46 .  The objective reality of his situation would necessarily be altogether ignored, contrary to the intention of Parliament as interpreted in our jurisprudence.

131                              If, in these circumstances, jurors gave effect to the plea of self-defence, the Crown could be expected to successfully attack the judge’s erroneous instruction that left self-defence to their consideration.  Even the most patient jurors are entitled to expect that if they are asked to consider a defence, and accept it, the verdict will not be reversed on appeal on the ground that as a matter of law there was no objective basis in the evidence for the judge to have put self-defence to them in the first place.

The reasons of Iacobucci, Major and Arbour JJ. were delivered by

Arbour J. (dissenting) —

I.        Introduction

132                                   The respondent was convicted of second degree murder by a jury.  His sole defence was self-defence.  The question before us, in substance, is whether that defence was available to him.  More specifically, the issue is whether it would have been unreasonable for a properly instructed jury to have a reasonable doubt as to whether the respondent had an honest and reasonable belief that the only way to protect his own life was to act as he did.  We are not concerned with the respondent’s guilt or innocence, nor are we concerned with the reasonableness of his perceptions and actions. The question is not whether a “pre-emptive strike” can be justified as self-defence.  Rather, what this Court has to decide is whether, on the evidence introduced at trial, the task of assessing the reasonableness of the respondent’s perceptions and actions belonged to the jury, the trial judge, or the members of this Court.  For the reasons that follow, I conclude that this is a case where the respondent was entitled to have his reasonableness assessed by a properly instructed jury. 

133                              Because the appellant concedes that errors which could have prevented the jury from properly understanding the law of self-defence were made by the trial judge, this case turns on the application of s. 686(1)(b)(iii) of the Criminal Code , R.S.C. 1985, c. C-46 , commonly referred to as “the curative proviso” or “the proviso”, which reads:

686. (1) On the hearing of an appeal against a conviction or against a verdict that the appellant is unfit to stand trial or not criminally responsible on account of mental disorder, the court of appeal

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

                                                                   . . .

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

                                                                   . . .

(b) may dismiss the appeal where

                                                                   . . .

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

The Crown argues that the curative proviso should be applied here because the trial judge should not even have put self-defence to the jury in the first place.  An assessment of this argument will require careful consideration of the concept of the “air of reality” that a defence must have before it can be left for consideration by the jury.

134                              As indicated by my colleagues the Chief Justice and  Bastarache J., the trial judge in this case made at least two errors of law in his instructions on self-defence.  First, he did not make it clear to the jury that the accused had no burden to prove the defence and that the Crown had to disprove its existence beyond a reasonable doubt.  Second, he did not properly explain the objective and subjective nature of the assessment required for the three essential elements of self-defence.  My colleagues found no reversible error in the trial judge’s failure to specify that imminency does not constitute a formal requirement of self-defence.  Since the error with respect to the burden of proof must, in my view, necessarily require that a new trial be ordered if it is found that the defence of self-defence possessed the requisite “air of reality”, I find it unnecessary to decide that last issue.  In any event, it is clear that this case turns on the assessment of whether the accused’s defence has an “air of reality”.  Clearly, if there was no basis upon which self-defence could be left with the jury,  then any errors made with respect to it become irrelevant and the proviso properly applies.  Conversely, if the defence had an evidentiary foundation, it had to be adequately explained to the jury, which it was not, and a new trial must be ordered.  I do not propose to review the facts in detail.  They are set out in my colleagues’ reasons.

135                              In short, the respondent Cinous killed Michaelson Vancol (“Mike”) by shooting him in the back of the head while Mike was sitting in the back seat of a van that the respondent had been driving.  Convinced that Mike and his friend Yves Louis (“Ice”) were armed and set to kill him, the respondent drove to a service station, got out of the van, moved behind the van, opened the back door and killed Mike.

136                              The respondent testified in his defence and claimed that he feared for his life and saw no alternative but to kill his perceived assailant.  He explained the circumstances that brought him to that conclusion.

137                              The Crown did not argue at trial that self-defence had no air of reality and should not be left to the jury.  The trial judge instructed the jury on self-defence and the jury convicted the respondent.  The Court of Appeal also concluded that there was some evidence upon which the jury could entertain a reasonable doubt as to whether the respondent was acting in self-defence as that defence is defined in the Criminal Code .  I share that view.

138                              As I will set out to explain below, I conclude that the test upon which a trial judge must decide whether a defence has an “air of reality”, so as to be left to the jury only involves a determination of whether there is “some evidence” in support of the defence, and should not otherwise involve a measure of the sufficiency of that evidence.  Since in a case like the present one, the withdrawal of the defence is akin to a directed verdict of conviction — the accused having admitted the elements of the offence — the test should be substantially the same as the one applicable to a directed verdict of acquittal.

139                              Reasonableness is a question for the jury (R. v. Lavallee, [1990] 1 S.C.R. 852, at p. 891).  The fact that reasonableness involves an objective, rather than a purely subjective, assessment does not authorize judges to substitute their own appreciation of that critical factual element for that of the jury.  For there to be some evidence of reasonableness, there must merely be some evidence of the circumstances surrounding the conduct at issue, so that its reasonableness can be assessed in context.  I do not believe that one can usefully speak of a threshold of reasonableness, as in whether it would be unreasonable to say that something was reasonable.

140                              My colleagues have concluded that in this case the accused, who may have reasonably believed that he was about to be killed, and who may have honestly thought that he had no other option but to strike first, cannot ask a jury to decide whether it was reasonable for him to do so even after having explained, under oath, all the circumstances surrounding his actions.  They come to that conclusion because they think that it would be unreasonable for the jury to be left in doubt on that point, and on that basis they hold that the law requires that the question not be put to them.

141                              On that reasoning, had the absence of any “air of reality” to the defence of self-defence been raised at trial, as it should have been, I suppose that the trial judge would have had to charge the jury to disregard the evidence.  He could have done so, presumably, in one of two ways.  First, without any explanation to the jury, the trial judge could have stated as a legal directive that it was to ignore the evidence.  Alternatively, he could have explained to the jury the analysis contained in paras. 107 to 125 of my colleagues’ reasons.  Now, hypothetically, if the jury had nonetheless proceeded to acquit the accused after its deliberations, the Crown would have had no right of appeal from that verdict as there would have been no error of law in the instructions, and there is no such thing as an “unreasonable acquittal” as per the unanimous judgment of this Court in R. v. Biniaris, [2000] 1 S.C.R. 381, 2000 SCC 15, at para. 33.

142   This would be, at best, a strange result.  In another sense, this result serves as a reminder, in my view, of the importance of yielding to the constitutional imperative of giving the accused the benefit of a trial by jury by leaving the critical issue of reasonableness to the jury.  The requirement for a defence to have an “air of reality” should be understood in that context.

143   I now turn to the “air of reality” test, and to its application to the reasonableness requirement in self-defence.  

II.    Air of Reality

B.        History of the Concept

1.  The Classic Formulation

144                              The idea that a defence can only be left to the jury if it has an “air of reality” finds its source, as was explained by Wilson J., for the Court, in R. v. Robertson, [1987] 1 S.C.R. 918, at pp. 936-38, in the very well-established principle of criminal law, derived from English law, that an accused bears no burden to prove his innocence, but bears an evidential burden to raise a defence, where the Crown’s case might otherwise satisfy a jury beyond a reasonable doubt of the accused’s guilt.

145                              In Woolmington v. Director of Public Prosecutions, [1935] A.C. 462, the House of Lords reiterated the centrality of the presumption of innocence in English criminal law, and ruled that an accused never has any burden to prove his innocence.  Therefore, any defence raised by the accused does not have to be demonstrated by him, but only has to raise a reasonable doubt.  It is for the prosecution to prove beyond a reasonable doubt that the defence is not made out.  To hold otherwise, according to Viscount Sankey L.C., at p. 480, would amount to allowing judges to decide alone  the case, if they concluded, at the end of the Crown’s case, that the onus had shifted to the accused and that the accused had not presented any evidence.  In 1941, the House of Lords clarified, in Mancini v. Director of Public Prosecutions, [1942] A.C. 1, that this principle was not intended to mean that every possible defence must always be put to the jury in every case.  Viscount Simon L.C. stated, at p. 8:

To avoid all possible misunderstanding, I would add that this is far from saying that in every trial for murder, where the accused pleads not guilty, the judge must include in his summing-up to the jury observations on the subject of manslaughter.  The possibility of a verdict of manslaughter instead of murder only arises when the evidence given before the jury is such as might satisfy them as the judges of fact that the elements were present which would reduce the crime to manslaughter, or, at any rate, might induce a reasonable doubt whether this was, or was not, the case.

The House of Lords held in Mancini that it was not necessary to deal with the defence of provocation, in addition to the self-defence argued by the accused, since there was no evidence “which might, if believed, be regarded as amounting to sufficient provocation” (p. 9).  Similarly, in Holmes v. Director of Public Prosecutions, [1946] A.C. 588, the House of Lords held that even where provocation was argued, the trial judge had to instruct the jury that a verdict of manslaughter was not possible on the evidence, because there was no evidence which could support the view that the provocation was available in law.

146                              This was described by G. Williams, Textbook of Criminal Law (2nd ed. 1983), at pp. 52-59, as the accused’s evidential burden.  At p. 53, Professor Williams indicated that imposing an evidential burden on the defendant who wishes to raise a specific defence was designed to simplify the task of the jury and most importantly, to “prevent the defendant submitting no case if the prosecution have led no evidence on that issue”.  Indeed, without an evidential burden on the accused, the Crown would have to “disprove, in the initial presentation of its case, the existence of every conceivable defence or extraneous factor that might affect a finding of guilt” (I. Weiser, “The Presumption of Innocence in Section 11(d)  of the Charter  and Persuasive and Evidential Burdens” (1988-89), 31 Crim. L.Q. 318, at p. 344), or the accused would be entitled to a directed verdict of acquittal.  In the case of R. v. Lobell, [1957] 1 Q.B. 547 (C.C.A.), Lord Goddard C.J. explained, at p. 551:

It must, however, be understood that maintaining the rule that the onus always remains on the prosecution does not mean that the Crown must give evidence-in-chief to rebut a suggestion of self-defence before that issue is raised, or indeed need give any evidence on the subject at all.  If an issue relating to self-defence is to be left to the jury there must be some evidence from which a jury would be entitled to find that issue in favour of the accused, and ordinarily no doubt such evidence would be given by the defence.  But there is a difference between leading evidence which would enable a jury to find an issue in favour of a defendant and in putting the onus upon him.

147                              In Canada, the concept of the evidential burden imposed on the accused in order to raise a defence was already given some recognition in 1934, when this Court held  in  Wu v. The King, [1934] S.C.R. 609, at p. 616, that an accused is entitled to rely on any defence “for which a foundation of fact appears in the record”.  In that case, it was held that the alternative defence of self-defence was not available where the defence of alibi negatived the possibility of self-defence.  In  Latour v. The King, [1951] S.C.R. 19, at p. 24, this Court also recognized the “necessity there may be for an accused in the course of the enquête to introduce evidence in order to explain away the case being made by the Crown”.  This was referred to as the “minor burden” and contrasted with the burden of proof, properly speaking, which must always remain with the Crown (p. 25).  The accused’s evidential burden was also explained in the following terms in R. v. Proudlock, [1979] 1 S.C.R. 525, at p. 549:

The accused does not have to “establish” a defence or an excuse, all he has to do is to raise a reasonable doubt.  If there is nothing in the evidence adduced by the Crown from which a reasonable doubt can arise, then the accused will necessarily have the burden of adducing evidence if he is to escape conviction.  However, he will not have the burden of proving his innocence, it will be sufficient if, at the conclusion of the case on both sides, the trier of fact has a reasonable doubt.

148                              R. v. Tripodi, [1955] S.C.R. 438, was one of the first instances in which this Court applied the evidential burden requirement in a way similar to the modern “air of reality” test.  The Court held that the question of whether there was any evidence going to the requirements of the defence of provocation was a question of law.  In that case, even assuming the accused’s testimony to be true, it was held that since the accused had admitted that he knew of his wife’s infidelity, her confession to him, which was alleged to constitute provocation, did not give rise to the defence, since the provocation could not, in law, be said to be sudden.  Similarly, in R. v. Côté, [1964] S.C.R. 358, this Court found that there was no evidence in the record capable of supporting the submission that the accused was sufficiently intoxicated to be able to raise the defence of drunkenness, since there was no evidence respecting the effect on the accused of the alcohol allegedly consumed.  The same principle was applied to self-defence by the British Columbia Court of Appeal, in R. v. Nelson, [1968] 2 C.C.C. 179, where it was held, at p. 182, that while the trial judge must not weigh the evidence, only defences for which a foundation of fact appears in the record have to be put to the jury.

149                              In Parnerkar v. The Queen, [1974] S.C.R. 449, Fauteux C.J., for the majority, restated and clarified the applicable law, in the context of the defence of provocation, and explained it in terms of the respective functions of the trial judge and of the jury.  He emphasized the “principle according to which the sufficiency of  evidence, which is an issue only where there is some evidence, is a question of fact for the jury and the absence of evidence is a question of law for the trial judge” (p. 454).  Therefore, the question which the Court had to answer was whether there was some evidence of a wrongful act or insult within the meaning of the provocation defence, and it was found that there was no such evidence in that case.  This principle, focussing on the lack of evidence, as opposed to its sufficiency, was then applied consistently by this Court and by courts of appeal in a number of subsequent cases, without reference to the “air of reality” language (see, for example, Morgentaler v. The Queen, [1976] 1 S.C.R. 616; R. v. Squire, [1977] 2 S.C.R. 13; Alward v. The Queen, [1978] 1 S.C.R. 559; Linney v. The Queen, [1978] 1 S.C.R. 646; Mazza v. The Queen, [1978] 2 S.C.R. 907; Landry v. The Queen, [1979] 1 S.C.R. 552; Olbey v. The Queen, [1980] 1 S.C.R. 1008; R. v. Louison, [1975] 6 W.W.R. 289 (Sask. C.A.), aff’d [1979] 1 S.C.R. 100).     

150                              Another consequence of the principle according to which an accused only has an evidential burden to raise a defence is that when the evidence puts a defence in issue, the trial judge has a duty to instruct the jury on that defence whether or not it was argued by the accused (Mancini, supra, at p. 7; Kwaku Mensah v. The King, [1946] A.C. 83 (P.C.); R. v. Porritt, [1961] 1 W.L.R. 1372 (C.C.A.); G. Williams, The Proof of Guilt (3rd ed. 1963), at p. 36; Wu, supra; Squire, supra, at p. 19; Mazza, supra, at p. 911).  In fact, cases concerning the accused’s evidential burden, and later cases respecting the “air of reality” test, often arose in a context where, while it was argued on appeal that an alternative defence was available on the evidence, this defence was not argued by the accused at trial (see Landry, supra; Mazza, supra; Leary v. The Queen, [1978] 1 S.C.R. 29; Squire, supra).  In other cases, the trial judge had recognized his duty to put to the jury a defence not advanced by the accused; it was  then argued on appeal that this defence should not have been put to the jury (see Olbey, supra). 

151                              The availability of a defence not relied upon at trial raises different concerns.  The decision not to raise an available defence may be essentially strategic, designed to avoid prejudicing the main defence (see Mancini, supra, at p. 7).  W. J. Braithwaite, “Developments in Criminal Law and Procedure: The 1979-80 Term” (1981), 2 Supreme Court L.R. 177, at p. 230, explained:

It is trite law that a trial judge is under an obligation to charge the jury in relation to every defence which arises on the evidence, even if the accused has not raised the defence.  So, for example, in Olbey, where the accused argued the defence of self-defence, the trial judge on his own initiative raised the defence of provocation.  Such developments should not be regarded as unusual, for some defences are more appropriately suggested by the court; in a case like Olbey the accused’s raising the defence of provocation rather than relying upon the trial judge, as Lord Salmon explains in D.P.P. v. Walker, [[1974] 1 W.L.R. 1090 (P.C.),] “might prejudice the accused by offering the jury an option of returning a verdict of manslaughter, which could result in the accused being sentenced to a long term of imprisonment instead of being acquitted as he would be, should self-defence be accepted by the jury”.  [Emphasis added.]

The possibility of a defence leading to a less favourable result for the accused should not prevent a judge from instructing the jury on that defence if it clearly arose on the evidence.  However, in borderline cases, the fact that experienced counsel for the accused made a tactical decision at trial not to raise a defence may be taken into account by a court of appeal that is subsequently asked to rule that the trial judge was in error in not putting the alternative defence to the jury.  For example, in Leary, supra, Pigeon J., for the majority, said, at p. 60, that even if there had been some slight evidence to support the defence of drunkenness, which in any case was not available for the crime of rape as it was a crime of general intent, the fact that counsel chose not to raise a weak alternative defence that contradicted the main defence, so as to not prejudice the accused’s position, was a relevant factor.  Where, as in the present case, the inquiry concerns the only defence raised and relied upon by the accused at trial, those concerns are not present, and the analysis must be envisaged from a different perspective.

2.  The “Air of Reality” Terminology

152                              The expression “air of reality” was used for the first time  in Pappajohn v. The Queen, [1980] 2 S.C.R. 120, in the context of an assessment of the availability of the defence of honest but mistaken belief in consent to the crime of rape, which has since become sexual assault.  In Pappajohn, McIntyre J., for the majority, used as the starting point of his analysis the well-established principle discussed above that the accused bears an evidential burden in order to raise a defence, and that the trial judge “must put before the jury any defences which may be open to the accused upon the evidence whether raised by the accused’s counsel or not” (p. 126).  He reiterated that this obligation arises only where there is an evidentiary basis upon which the defence can rest, and that “[o]rdinarily, when there is any evidence of a matter of fact, the proof of which may be relevant to the guilt or innocence of an accused, the trial judge must leave that evidence to the jury so that they may reach their own conclusion upon it” (p. 127).  However, McIntyre J. also stated that the standard was different where an accused sought to have a specific defence put to the jury.  At p. 127, he said:

Where, however, the trial judge is asked to put a specific defence to the jury, he is not concerned only with the existence or non-existence of evidence of fact.  He must consider, assuming that the evidence relied upon by the accused to support a defence is true, whether that evidence is sufficient to justify the putting of the defence.  [Emphasis added.]

This seemed to constitute a departure from the previous authorities, especially from the statement in Parnerkar, supra, quoted above, according to which the sufficiency of the evidence was a matter for the jury while only the absence of evidence was a matter for the judge.  It is in that context that the phrase “air of reality”, which has now become a term of art, was first used by this Court.  McIntyre J., at p. 127, was quoting from the previous case of Kelsey v. The Queen, [1953] 1 S.C.R. 220, where the term “sense of reality” had been used.  In Kelsey, the term was used not to assess the availability of a defence, but rather the validity of a defence argument to the effect that the trial judge should have mentioned in his charge some of the evidence which was said to contradict the veracity of the accused’s confession.  It was in that context that Fauteux J., as he then was, for the majority, had said, at p. 226 (also quoted in Pappajohn, supra, at p. 127):

The allotment of any substance to an argument or of any value to a grievance resting on the omission of the trial Judge from mentioning such argument must be conditioned on the existence in the record of some evidence or matter apt to convey a sense of reality in the argument and in the grievance.

The term “sense of reality” had also been used, as pointed out by McIntyre J., at p. 127, in Pappajohn, in Workman v. The Queen, [1963] S.C.R. 266, where it was argued, on appeal, that the judge should have instructed the jury on the “defence” that one of the co-accused was only involved in the murder as an accessory after the fact.  This possibility was found not to arise on the evidence. 

153                              Whatever its origins, it seems that the use of the phrase “air of reality” in Pappajohn was used to convey a meaning different from the well-established evidential burden principle, albeit derived from it.  In order that a defence of honest but mistaken belief in consent  have the requisite “air of reality”, the majority of the Court imposed the sufficiency requirement discussed above, and held that the defence was not available where the accused’s and the complainant’s versions of the facts were diametrically opposed and where nothing in the rest of the evidence could suggest a mistaken belief in consent (p. 132).  Thus, when the complainant spoke of lack of consent, and of terror and resistance, and the accused spoke of actual consent and cooperation, there was no room for the suggestion that the complainant might not have been consenting, but that the accused might have nevertheless honestly believed that she did.  The only defence available to the accused, which was in fact raised as the main defence in Pappajohn, was that the complainant in fact consented.  Also, McIntyre J. indicated, at p. 133, that  the defence of honest but mistaken belief in consent did not have to be put to the jury every time a “complainant denies consent and an accused asserts it”.  There had to be some evidence beyond the “mere assertion of belief in consent by counsel for the appellant”, and such evidence had to “appear from or be supported by sources other than the appellant”.

154                              As will be seen below, the context of sexual assault had significant influence on the adoption of this new “air of reality” test and the test has been, to this day, mostly often used in that context.  Following Pappajohn, the modified “air of reality” test for the evidential burden of the accused was applied consistently, in sexual assault cases, by this Court as well as by the courts of appeal (see R. v. Bulmer, [1987] 1 S.C.R. 782; Robertson, supra; R. v. Reddick, [1991] 1 S.C.R. 1086; R. v. Trottier (1981), 58 C.C.C. (2d) 289 (B.C.C.A.); R. v. Cook (1985), 46 C.R. (3d) 129 (B.C.C.A.); R. v. White (1986), 24 C.C.C. (3d) 1 (B.C.C.A.)).  Shortly after the Pappajohn decision, the Criminal Code  was also amended to include s. 244(4) (now s. 265(4)), which applies to all forms of assault, including sexual assault, and which reads:

(4) Where an accused alleges that he believed that the complainant consented to the conduct that is the subject-matter of the charge, a judge, if satisfied that there is sufficient evidence and that, if believed by the jury, the evidence would constitute a defence, shall instruct the jury, when reviewing all the evidence relating to the determination of the honesty of the accused’s belief, to consider the presence or absence of reasonable grounds for that belief.  [Emphasis added.]

This Court found that the section amounted to a codification of the principles elaborated in Pappajohn (see Robertson, supra, at p. 938; Bulmer, supra, at p. 792).  

155                              However, the test laid down in Pappajohn, and most importantly its departure from the principles previously applied with respect to the accused’s evidential burden, was also subject to criticism.  It was said that imposing a corroboration requirement on an accused, in order to raise a defence, was not consistent with generally accepted principles of criminal law (see D. Stuart, Canadian Criminal Law: A Treatise (3rd ed. 1995), at p. 260; J. M. Williams, “Mistake of Fact: The Legacy of Pappajohn v. The Queen” (1985), 63 Can. Bar Rev. 597, at p. 617).  On this subject, Lamer J. (as he then was) indicated, in concurring reasons in  Bulmer, supra, at p. 798:

The old common law rule in sexual assault cases that the trial judge must instruct the jury that it is unsafe to convict in the absence of corroboration of the complainant’s testimony (which was abolished by s. 246.4 of the Criminal Code ) would in effect be replaced by a rule requiring corroboration of the accused’s testimony.  Such a requirement will often work an injustice to the accused.  Clearly the best, and quite often the only, evidence of the accused’s subjective belief will be his or her testimony, and there is no basis in law or in principle for requiring corroboration.  [Emphasis added.]

Also, the principle that the defence of honest but mistaken belief in consent should not be available, in the absence of other evidence, where the accused’s and the complainant’s versions are diametrically opposed, was criticized as ignoring the well-established principle according to which the trier of fact is not bound to deal with the evidence of any witness as a whole, but can accept it or reject it in part only (see J. M. Williams, supra, at pp. 611-12).

156                              The “air of reality” doctrine was revisited in R. v. Osolin, [1993] 4 S.C.R. 595.  The concerns discussed above were addressed and Cory J., writing for the majority on this point, specified, at p. 686, that the “air of reality” test does not impose a requirement of corroboration on the accused.  While the totality of the evidence must be considered in order to determine whether the defence of honest but mistaken belief in consent is available, there is no need for evidence independent of the accused confirming his belief in order to put the defence to the jury.  Also, McLachlin J. (as she then was), dissenting, but writing for a differently constituted majority on this issue, held at p. 652 that the presence of two diametrically opposed versions of the facts did not make it logically impossible for the defence of mistake to arise, because it was possible to accept a witness’ evidence in part only.  Overall the Court in Osolin established that “[t]he term ‘air of reality’ simply means that the trial judge must determine if the evidence put forward is such that, if believed, a reasonable jury properly charged could have acquitted” (per Cory J., at p. 682).  The Court was then unanimous in finding that the codification of the “air of reality” principles found in s. 265(4)  of the Criminal Code  was constitutional and did not violate the right to be presumed innocent recognized by s. 11( d )  of the Canadian Charter of Rights and Freedoms  or the right to a trial by jury recognized by s. 11(f).

157                              The Osolin approach was confirmed by this Court in R. v. Park, [1995] 2 S.C.R. 836 (hereinafter Park (1995)), where L’Heureux-Dubé J., for the majority, reiterated that no corroboration requirement was imposed on the accused, but that the requirement that the accused present more than a bare assertion of belief in consent meant that the accused had to point to circumstances, put into evidence either by his own testimony or through other sources, which explained why he believed the complainant was consenting (paras. 19-21).  Where the versions provided by the accused and by the complainant are diametrically opposed, L’Heureux-Dubé J. indicated that the defence could still be available where it was possible to “cobble together some of the complainant’s evidence and some of the accused’s evidence to produce a sufficient basis for such a defence” (para. 25).

158                              In other words, if the accused’s version asserting consent and cooperation was entirely believed, then obviously an acquittal would be based on consent, rather than mistaken belief in consent, and the latter defence would become irrelevant.  On the other hand, if the complainant’s evidence of lack of consent, clearly manifested to the accused, was entirely believed, there would be no basis for a suggestion that the accused could have been mistaken as to the complainant’s consent.  However, it would also be possible for a jury, where some circumstances in the evidence so indicated, to believe the complainant’s assertion that she did not consent, but to also have a reasonable doubt as to whether the accused could nevertheless have been mistaken in believing that she did. 

159                              Park (1995) also recognized that where the evidence could only point to either consent, or a lack of consent known to the accused or as to which the accused was reckless or wilfully blind, the defence of mistaken belief would not be available (para. 26).  This was consistent with the Court’s previous decision in Sansregret v. The Queen, [1985] 1 S.C.R. 570, where it was held that the accused’s wilful blindness as to the absence of consent could not form the basis to a defence of honest but mistaken belief in consent.

160                              Those principles were then consistently applied by this Court and the courts of appeal to determine whether the defence of honest but mistaken belief in consent was available in various circumstances (see R. v. Dickson, [1994] 1 S.C.R. 153, aff’g (1993), 81 C.C.C. (3d) 224 (Y.T.C.A.); R. v. Livermore, [1995] 4 S.C.R. 123; R. v. Esau, [1997] 2 S.C.R. 777; R. v. Ewanchuk, [1999] 1 S.C.R. 330; R. v. Davis, [1999] 3 S.C.R. 759; R. v. M.O., [2000] 2 S.C.R. 594, 2000 SCC 49, aff’g the dissenting reasons of Rosenberg J.A. (1999), 138 C.C.C. (3d) 476 (Ont. C.A.); R. v. Silva (1994), 31 C.R. (4th) 361 (Sask. C.A.); R. v. Rarru (No. 2) (1995), 60 B.C.A.C. 90; R. v. Stolz (1996), 71 B.C.A.C. 127).  In  Ewanchuk, supra, the legal requirements of the defence of honest but mistaken belief in consent were further defined, and it was specified, at para. 49, that in order for the defence to be available in law, a belief that the complainant “affirmatively communicated by words or conduct her agreement to engage in sexual activity” was necessary.  Therefore, a belief that the complainant had not communicated a lack of consent, without more, could not form a sufficient basis for the defence, as it would be a mistake of law for an accused to assume that a complainant consented only because no lack of consent was expressed.

161                              Thus while the “air of reality” language is a relatively recent development, the concept itself has its origin in very well-established criminal law principles.  However, the novel “air of reality” language used to describe the evidential burden on an accused to raise a defence created difficulties, as was pointed out by Sopinka J., concurring in Osolin, supra, at pp. 653-54, where he said:

With respect to the defence of mistaken belief, I agree with Cory J. that s. 265(4) “simply sets out the basic requirements which are applicable to all defences” (p. 676) and that it requires no more of the accused than the discharge of an evidentiary burden to adduce or point to some evidence on the basis of which a reasonable jury properly instructed could acquit.  I believe we are all in agreement in this respect.  Indeed, this is the basis for our determination that it is constitutional.  The term “evidentiary burden” and the definition I have set out are well known to trial judges and well accepted.  I cannot understand how the addition of the term “air of reality” helps in understanding the duty of a trial judge with respect to this defence.  I am concerned that when an attempt is made to add to the definition of a basic concept in the criminal law, it only creates confusion.  Just as attempts to refine the meaning of “reasonable doubt” have frequently resulted in reversible error, so too the use of the “air of reality” test encourages trial judges to weigh the evidence rather than apply the legal definition to which I have referred.  [Emphasis added.]

162                              Sopinka J.’s concern, which I largely share, is further discussed in J. Sopinka, S. N. Lederman and A. W. Bryant, The Law of Evidence in Canada (2nd ed. 1999), at pp. 151-53.  The authors indicate that “[t]he courts appear to apply different evidentiary thresholds when applying the air of reality test and it is sometimes difficult to detect which threshold the court has applied in a particular case” (p. 152).  The authors assert that the “no evidence” threshold discussed above is the same as the standard applied for a directed verdict of acquittal, and represents the classical statement of the test used to determine whether the accused’s evidential burden has been fulfilled.  However, they add, at pp. 152-53:

The “air of reality” term is sometimes applied to determine whether there is “sufficient evidence” to raise a defence.  The courts have articulated criteria or standards to determine whether the evidence satisfies this threshold for the air of reality test.  The application of these standards requires the trial judge to weigh the probative value of the evidence to determine if there is a sufficient factual foundation for the proffered defence, that is, whether a reasonable jury properly charged could acquit.

The authors warn that “the higher threshold for the air of reality test risks a conviction where a jury may have acquitted had the jurors been charged on the possible defence” (p. 153).  Searching for a sufficient factual foundation for a defence also seems to contradict the explicit statements of this Court to the effect that an application of the “air of reality” test must not involve the weighing of the evidence (see Davis, supra, at para. 82; Ewanchuk, supra, at para. 57; Park (1995), supra, at para. 13; Osolin, supra, at p. 691; Bulmer, supra, at p. 790; Pappajohn, supra, at p. 132).

163                              While I emphatically agree that no attempt at weighing the evidence should be made by trial judges when determining whether a defence possesses the requisite “air of reality”, I reiterate the warning that if the test adopted is one of “sufficient evidence”, as opposed to “some evidence”, it will become difficult to understand sufficiency as exclusive of weight. In Park (1995), supra, L’Heureux-Dubé J. addressed Sopinka J.’s concern, at para. 19:

Sopinka J. preferred, at pp. 654-55 [of the Osolin case], to characterize the “air of reality” test more simply, as an ordinary evidentiary burden on an accused to adduce sufficient evidence to justify a defence being put to a jury.  In many ways, he is correct.  I think, however, that the unique social and legal context of sexual assault, including the uncommon frequency with which the mistake of fact defence is raised, make it necessary for this Court to elaborate somewhat on that standard.  [Emphasis added.]

3.  The Specificity of Defences in Sexual Assault Cases

164                              I already pointed out that the “air of reality” test was not only elaborated in the sexual assault context, but that it has also mostly been applied in cases respecting the honest but mistaken belief in consent defence.  This was also noted by McLachlin J., dissenting, in Osolin, supra, at p. 648:

While the rule applies generally, it has attracted special attention in the context of trials for sexual assault.  This is because one of the most common defences on such trials, the defence of honest but mistaken belief, is frequently asserted in cases where there is no evidentiary foundation for it, requiring the court to advise the jury that there is no air of reality to the defence.  [Emphasis added.]

In my view, in order to properly understand this development in the law and its applicability to other defences, two distinctive features of the honest but mistaken belief in consent defence, as applied to sexual assault, must be examined. 

165                              First, I observe that in all of the “air of reality” cases examined so far, the honest but mistaken belief in consent defence was not the only defence raised by the accused, but rather constituted an alternative defence which was fundamentally inconsistent with the main defence raised, that of actual consent of the complainant.  This feature of the case law is highly significant.  Indeed, if the accused’s main defence is that the complainant consented, and if he testifies to the fact that she actively participated and expressed her consent, he is negating the suggestion that there is any possibility that his belief in consent could be honestly mistaken, except in a highly hypothetical case where the complainant would entirely admit to the participatory actions and expressions of consent, but would testify that she nevertheless did not consent, either because she was incapable of consenting or because her consent was induced by fear or threats unknown to the accused.  In more common place sexual assault cases, therefore, the assessment of the availability of the honest but mistaken belief in consent defence cannot proceed on the same basis as is required under the classical evidential burden test.  It cannot be assumed that the whole of the accused’s evidence is true, for in that case, the only available defence is consent and there is no room for a mistaken belief.  Therefore, when the two defences of consent and mistake are advanced, the assessment of the “air of reality” of the honest mistake defence would have to proceed, in most cases, on the basis that the accused’s evidence is not entirely true. 

166                              An examination of the facts of the present case makes it immediately apparent that such a concern does not arise here, where the accused raises only one defence and clearly asks this Court to assume that the whole of his evidence is true in order to assess the “air of reality” of his defence, which is indeed the proper approach to deciding, without weighing, whether a defence should go to the jury.  Thus, the principles enunciated in Osolin, supra, and Park (1995), supra, according to which one must find support in the totality of the evidence for the accused’s honest but mistaken belief in consent must be viewed in the context in which they were formulated, that is, in situations where it was impossible to find support for the defence exclusively by assuming the whole of the accused’s evidence to be true, for such an assumption would deny a foundation to the defence, since the accused’s evidence was contrary to it.

167                              Also, the inappropriate use of myths and stereotypes prejudicial to complainants has been a significant concern in the law of sexual assault.  It was held, for example, in R. v. Seaboyer, [1991] 2 S.C.R. 577, that evidence respecting the complainant’s prior sexual behaviour should not be admissible for the purpose of either attacking her credibility as a witness, or inviting the jury to draw the inference that the complainant was more likely to have consented in the circumstances at issue, on the assumption that, as a  sexually active woman, she would be prone to consent to sexual relations on any occasion.  The defence of honest but mistaken belief in consent has often attempted to root itself in similar misconceptions, for example where an accused bases his alleged belief in consent only on his prior belief that the complainant certainly would consent to sexual relations with him (see Park (1995), supra, at paras. 24 and 52-53), or on the complainant’s passivity or lack of resistance (see Ewanchuk, supra) or, worse, on his conception that “no means yes”, or that the complainant is in fact consenting in her own mind even though she is expressing lack of consent.  Neither prosecutions nor defences can properly rest on facts founded exclusively on such unacceptable myths and stereotypes. 

4.  The “Air of Reality” Test for Other Defences              

168                              While the “air of reality” test was developed in the specific context of the crime of sexual assault and of the defence of honest but mistaken belief in consent, and its formulation was therefore influenced by the specific issues that arise in that context, the test was also said to be applicable to all defences (see Osolin, supra, at p. 676).  This is not surprising if one sees the “air of reality” test as an application of the principle according to which the accused bears an evidential burden in order to raise a defence, which was already well established and was always meant to apply to all defences.  The principle is well explained by K. Roach, Criminal Law (2nd ed. 2000), at p. 247:

The judge does not, however, have to instruct the jury about every defence in every case.  The accused has to overcome a threshold evidential burden by pointing to some evidence that justifies consideration of the particular defence.  In other words, the accused has to establish that there is an air of reality that justifies instructing the jury about the defence.  This requires some evidence as to the existence of all the elements of the defence including both their objective and subjective components.  The sufficiency of the evidence is then left to the jury.  [Emphasis added.]

Thus, the “air of reality” language is now being generally used in describing the evidential burden on the accused to raise any defence.  Since the elaboration of the modern test, the principle has in fact been applied to determine the availability of the defences of provocation (see R. v. Malott (1996), 110 C.C.C. (3d) 499 (Ont. C.A.), aff’d on different grounds, [1998] 1 S.C.R. 123; R. v. Stewart (1995), 41 C.R. (4th) 102 (B.C.C.A.)), self-defence (see R. v. Hebert, [1996] 2 S.C.R. 272; R. v. McConnell, [1996] 1 S.C.R. 1075, aff’g the dissenting reasons of Conrad J.A. (1995), 32 Alta. L.R. (3d) 1 (C.A.); Brisson v. The Queen, [1982] 2 S.C.R. 227), intoxication (see R. v. Lemky, [1996] 1 S.C.R. 757; R. v. Robinson, [1996] 1 S.C.R. 683; R. v. Caron (1998), 126 C.C.C. (3d) 84, 16 C.R. (5th) 276 (Que. C.A.); R. v. Fisk (1996), 108 C.C.C. (3d) 63 (B.C.C.A.); R. v. Taillefer (1995), 100 C.C.C. (3d) 1, 40 C.R. (4th) 287 (Que. C.A.), leave to appeal denied, [1996] 1 S.C.R. x; R. v. McKinnon (1989), 70 C.R. (3d) 10 (Ont. C.A.); R. v. Martin (1980), 53 C.C.C. (2d) 425 (Ont. C.A.); Stewart, supra), necessity (see R. v. Latimer, [2001] 1 S.C.R. 3, 2001 SCC 1; R. v. McKay (1992), 13 C.R. (4th) 315 (B.C.C.A.)), duress and compulsion (see R. v. Ruzic, [2001] 1 S.C.R. 687, 2001 SCC 24) and obedience to superior orders and mistake of fact in R. v. Finta, [1994] 1 S.C.R. 701.  The test has also been used to determine whether the trial judge had a duty to instruct the jury on the alternative verdict of manslaughter because of lack of intent to murder (see R. v. Aalders, [1993] 2 S.C.R. 482; Young v. The Queen, [1981] 2 S.C.R. 39).

169                              However, it should be noted that in many of these cases, decided after  Pappajohn, the term “air of reality” is not found, and the test is rather formulated as the classic principle of the accused’s evidential burden (see R. v. Duclos, [1995] Q.J. No. 678 (QL) (C.A.); R. v. Stone, [1999] 2 S.C.R. 290; R. v. Thibert, [1996] 1 S.C.R. 37; R. v. Daviault, [1994] 3 S.C.R. 63; R. v. Parks, [1992] 2 S.C.R. 871 (hereinafter Parks (1992)); R. v. Sheridan, [1991] 2 S.C.R. 205, aff’g the dissenting reasons of Foisy J.A. (1990), 55 C.C.C. (3d) 313 (Alta. C.A.); Perka v. The Queen, [1984] 2 S.C.R. 232; Bergstrom v. The Queen, [1981] 1 S.C.R. 539; Young, supra; R. v. Murray (1994), 93 C.C.C. (3d) 70 (Ont. C.A.); R. v. Bazinet (1986), 25 C.C.C. (3d) 273 (Ont. C.A.); see also R. v. Faid, [1983] 1 S.C.R. 265, where the expression “sense of reality”, derived from the cases of Kelsey and Workman, supra, as explained above, was used, instead of the new “air of reality” test; and also R. v. Gee, [1982] 2 S.C.R. 286, where it was said, at p. 303, that a foundation of fact was necessary to “give a reality” to the defence).  Thus, while this Court has said that the new “air of reality” test is applicable to all defences, it seems that in many instances courts have preferred to continue to use the language that was formerly used to describe the accused’s evidential burden, without reference to the “air of reality” formulation. 

170                              More importantly, it also seems that, regardless of the language used to identify the test, the added requirement focussing on the sufficiency of the evidence, as opposed to its mere existence, has not been applied in all “air of reality” cases concerning defences other than the defence of honest but mistaken belief in consent.  On the contrary, courts often continued to apply the “no evidence” test, and to refer to the need for “some evidence” or “any evidence” relevant to the defence, as opposed to “sufficient evidence” (see Ruzic, supra, at para. 100; Robinson, supra, at para. 71; Lemky, supra, at para. 22;  Parks (1992), supra, at p. 897; Faid, supra, at pp. 278-80; Brisson, supra, at pp. 235-36; Gee, supra, at p. 303; McConnell (C.A. (dissent)), supra, at. para. 84; Malott (C.A.), supra, at p. 521; Sheridan (C.A. (dissent)), supra, at p. 320; Murray, supra, at p. 84).  In Perka, supra, the term used, at p. 257, was “evidence sufficient to raise an issue” (emphasis added), which can be contrasted with evidence sufficient to raise a reasonable doubt, as discussed in Sopinka, Lederman and Bryant, supra, at pp. 152-53, as quoted in para. 162.  In Thibert, supra, at para. 7, Cory J. described the test as follows:

[T]he trial judge must still determine if there is any evidence upon which a reasonable jury properly instructed and acting judicially could find that there had been provocation.  If the trial judge is satisfied that there is such evidence, then the defence must be put to the jury to determine what weight, if any, should be attached to that evidence.  Obviously the trial judge should not weigh the sufficiency of the evidence.  This is the function reserved for the jury.  [Emphasis added.]

In Faid, supra, also in the context of an assessment of the availability of the defence of provocation, Dickson J. (as he then was), for the Court, stated, at pp. 276-77:

Whether the accused was provoked to lose his self-control is a question of fact for the jury.  Where an accused testifies that he killed impulsively in hot blood it must be left to the jury to decide whether he is to be believed or not.  There is, however, the preliminary question to be decided by the judge as a question of law, namely, whether there was any evidence produced on which a jury could decide that the accused acted in the heat of passion.  The question as to whether or not there is any evidence is for the court, but subject to that the following matters are both questions of fact for the jury, namely, (i) the sufficiency of the particular wrongful act or insult to cause an ordinary person to be deprived of self-control, and (ii) whether the accused was actually deprived of his self-control by such act or insult . . . . [Emphasis added.]

171                              A review of the application of the “air of reality” doctrine to defences other than the defence of honest but mistaken belief in consent in the context of sexual assault offers support for the statement quoted above to the effect that “[t]he courts appear to apply different evidentiary thresholds when applying the air of reality test” (Sopinka, Lederman and Bryant, supra, at p. 152).  While courts have sometimes referred to the modified threshold measuring the sufficiency of the evidence which was associated with the introduction of the “air of reality” language (see Latimer, supra, at para. 35; Finta, supra;  Bergstrom, supra, at p. 552; Fisk, supra, at paras. 80-81; Caron, supra, at p. 91 C.C.C.),  it appears that in an important number of cases, the classical evidential burden principles and the “no evidence” test associated with them were preferred, whether or not the new terminology was used. 

172                              Thus, in general, the well-established principle that an accused bears an evidential burden to introduce “some evidence”, as opposed to “sufficient evidence”, relevant to a defence, in order for the jury to be instructed on the defence, remains very much a part of our criminal law, in relation to defences other than the defence of  mistaken belief in consent.  This Court has also recognized, in R. v. Holmes, [1988] 1 S.C.R. 914, that the statutory imposition of an evidential burden on an accused is not inconsistent with the presumption of innocence guaranteed by s. 11( d )  of the Charter , and that such an evidential burden exists with respect to all defences.

173                              The vast majority of cases where an “air of reality” was denied to a defence were cases either involving multiple defences, often inconsistent, or involving defences subject to specific technical or policy requirements, or the legal availability of which was clearly limited to certain types of crimes or specific circumstances.

5.  The Case of Multiple, Alternative Defences

174                              As discussed above, the defence of honest but mistaken belief in consent arises almost always when alternative and inconsistent defences are raised.  However, this Court has also encountered this difficult issue in a variety of other circumstances, well before the “air of reality” test was introduced.  As early as 1934, in Wu, supra, the Court had to determine the availability of the defences of self-defence and provocation, in a case where the main defence was one of alibi, and the alternative defences rested solely on the account of Crown witnesses according to which the accused was being followed by the victim prior to the assault.  Lamont J., for the Court, stated, at p. 617:

The rule, therefore, that an accused person at trial is entitled to have the jury pass upon all his alternative defences is limited to the defences of which a foundation of fact appears in the record.  Even then the rule, in my opinion, is not without exception, and one exception is, that it has no application where the accused, by the defence which he sets up at the trial, has negatived the alternative defence for which he afterwards seeks a new trial.

                                                                   . . .

The defence that the accused was in Victoria at the time of the shooting was not only inconsistent with, but it negatived the defence now sought to be set up.

In Leary, supra, Pigeon J., for the majority, stated, at p. 60:

An accused cannot very well at the same time ask the jury to believe his statement that the complainant did consent if, at the same time, he says he was so drunk as not to know what he was doing.

                                                                   . . .

I am not unmindful of the rule that it is the duty of a trial judge to put to the jury every defence available to the accused.  However, when the latter, with the advice of competent counsel, chooses not to raise a weak alternative defence so as not to prejudice his main defence, it would not in my view be consistent with the proper administration of justice to allow a new trial for the purpose of making it possible to raise the alternative ground of defence.

175                              Nevertheless, this Court has also recognized the possibility of raising inconsistent alternative defences.  For example, in Squire, supra, Spence J., for the Court, indicated, at p. 19:

And, of course, in many cases, there are alternative defences and counsel for the accused feels that his presentation to the jury would only be weakened if he presented alternatives requiring the jury to make two different findings of fact.

Thus it is now well established that this remains a possibility for the accused (see Finta, supra; Sheridan, supra; Taillefer, supra; Martin, supra), which often gives rise to disputes concerning the duty of the trial judge to instruct the jury on defences that were not argued by the accused’s counsel, precisely because he or she might feel that raising a weaker and inconsistent alternative might prejudice the accused’s position.  While this will not mean that the trial judge will be dispensed from his or her duty to put to the jury all defences available to the accused on the evidence, a practical perspective indicates that an inconsistency between the accused’s evidence and the alternative defence argued, often for the first time on appeal, can influence the assessment respecting the “air of reality” of that defence.

176                              The “air of reality” analysis is therefore necessarily influenced by the presence of inconsistent defences, especially where an accused gives evidence in support of only one of the alternative defences put forward.  In such cases, the analysis cannot, as would normally be the case, especially when seeking to apply the classical “no evidence” test, proceed on the basis that the whole of the evidence for the accused must be taken as true, for that would always make the alternative inconsistent defence unavailable.  In contrast, in the present case we can and we must assume that the whole of the respondent’s evidence is true in order to decide whether his defence of self-defence possesses the requisite “air of reality”.

6.    Defences Resting on Technical or Policy Requirements 

177                              Another important factor to be considered in an analysis of the “air of reality” jurisprudence relates to technical or policy requirements applicable to specific defences, which also influence the determination of their availability.  For example, in Stone, supra, at para. 179, the majority of this Court decided that an accused wishing to raise the defence of automatism now bears the legal burden of proof to demonstrate involuntariness on a balance of probabilities.  This was found, in turn, to modify also the accused’s evidential burden, which is now described by Bastarache J., for the majority, at para. 182, as follows: “the defence must satisfy the trial judge that there is evidence upon which a properly instructed jury could find that the accused acted involuntarily on a balance of probabilities”.  In addition, there are specific requirements respecting the type of evidence required for an accused to fulfill the evidential burden related to the defence of automatism, the most important of which is the obligation to present expert psychiatric evidence confirming the accused’s claim (Stone, at paras. 183-84).  Similarly, this Court has held in Daviault, supra, at p. 101, that in the case of the defence of extreme intoxication “akin to automatism or insanity”, the accused bears the legal burden of establishing his or her defence on the balance of probabilities, in addition to the evidential burden of raising the defence (Daviault, at p. 102).  This naturally involves a stricter evidential burden for the accused, and expert evidence is also required (Daviault, at p. 101).  It is clear that no such requirements, or burden of proof, apply with respect to self-defence, and one must therefore be alive to the fact that the “air of reality” test, while in principle the same, cannot be applied in exactly the same manner to all defences. 

178                              There also exists statutory and common law bars to the availability of certain defences.  For example, the defence of intoxication, as opposed to extreme intoxication, is only available for crimes of general intent (see Leary, supra, where the defence was consequently found not to apply to the crime of rape; see also, generally, R. v. Bernard, [1988] 2 S.C.R. 833, affirming and applying the Leary rule; Lemky, supra, at paras. 13-15; Robinson, supra, at para. 75, per L’Heureux-Dubé J., dissenting, but not on this point).  Similarly, the statutory defence of compulsion by threats (provided for in s. 17  of the Criminal Code ), is unavailable for a series of listed offences.  In Ruzic, supra,  LeBel J., for the Court, indicated, at para. 23: “Subject to constitutional review, Parliament retains the power to restrict access to a criminal defence or to remove it altogether”.  In Bergstrom, supra, reference was made to the issue of whether there was sufficient evidence for the defence of compulsion to be raised, but the Court’s main finding was that the defence was not available for the crime of rape, as it was found to be excluded by statute.  Another example of a statutory restriction to the availability of a defence is the fact that the defence of provocation, provided for in s. 232  of the Criminal Code , is only available for the crime of murder, and then only permits, if it raises a reasonable doubt, to return a conviction for manslaughter instead of murder.  So, in Wu, supra, the Court noted, at pp. 616-17, that provocation was not an available defence to the charge of wounding with intent to murder.

179                              Policy considerations peculiar to a specific common law defence may also influence the “air of reality” analysis.  For example, with respect to the defence of necessity, Dickson J., for the majority, indicated, in Perka, supra, at p. 250:

If the defence of necessity is to form a valid and consistent part of our criminal law it must, as has been universally recognized, be strictly controlled and scrupulously limited to situations that correspond to its underlying rationale.  That rationale, as I have indicated, is the recognition that it is inappropriate to punish actions which are normatively “involuntary”. 

Accordingly, Dickson J. stated, at p. 251, that the defence of necessity “was restricted to instances of non-compliance ‘in urgent situations of clear and imminent peril when compliance with the law is demonstrably impossible’”.  The boundaries of the defence of necessity were further complemented by a policy analysis of proportionality, described as follows, at p. 252:

Even if the requirements for urgency and “no legal way out” are met, there is clearly a further consideration.  There must be some way of assuring proportionality.  No rational criminal justice system, no matter how humane or liberal, could excuse the infliction of a greater harm to allow the actor to avert a lesser evil.  In such circumstances we expect the individual to bear the harm and refrain from acting illegally.  [Emphasis added.]

180                              As a result of those policy considerations, Dickson J., indicated in Perka, supra, at p. 259, that the criterion for necessity “is the moral involuntariness of the wrongful action” and that “this involuntariness is measured on the basis of society’s expectation of appropriate and normal resistance to pressure” (emphasis added).  Indeed, in Latimer, supra, at para. 40, this Court indicated that the proportionality requirement of the necessity defence “requires the trial judge to consider, as a question of law rather than fact, whether the harm avoided was proportionate to the harm inflicted” (emphasis added).  It therefore becomes apparent that when applied to the defence of necessity, the “air of reality” analysis becomes more than just the basic evidentiary assessment aimed at determining whether there exists, on the record, any evidence in support of the defence.  Rather, it involves a careful policy exercise of balancing of values, so that the “air of reality” doctrine, in that case, amounts to a definition of the availability, in law, of this particular defence, not just with respect to the fact situation at issue in a given case, but more generally with respect to the societal values underlying the concept of necessity as a defence to a criminal charge.  While questions of policy may be involved in a variety of defences, in most cases general principles respecting the legal requirements of the defence have already been clearly established, and the “air of reality” analysis can be restricted to its purely evidentiary function of determining whether there is any evidence going to those legal requirements.

181                              From the above analysis, it becomes clear that while the “air of reality” test elaborated in the sexual assault context has now gained general acceptance as the means to determine the availability of defences, the contents of the test, and the strictness of its application, might still vary depending on a number of factors.  Thus, the fact that an accused has not raised a defence on the advice of competent counsel, that the defence has sought, or is seeking on appeal, to raise inconsistent alternative defences, as well as specific technical or policy requirements applicable to particular defences, can all influence the application of the test aimed at assessing whether the accused’s evidential burden has been fulfilled with respect to a defence.  In my opinion, the analysis in the present case should focus on the reality of the situation involved, that is, an accused who has raised, throughout the proceedings, only one defence, essentially through his own testimony, while otherwise admitting guilt.  Furthermore, the requirements of the defence of self-defence raised in this case are well defined in law, and the defence involves no specific technical requirements in terms of the nature of the evidence that must be found in the record for the defence to arise.  Nor does the defence involve a legal burden of proof for the accused, who only has to raise a reasonable doubt. 

B.        Application of the “Air of Reality” Test Where the Accused Only Presents One Defence and Otherwise Admits Guilt

182                              Counsel for the Crown conceded in oral argument of this appeal that an application of the “air of reality” test that denies an accused the only defence put forward, where the elements of the offence are otherwise admitted, is akin to a directed verdict of conviction, or a guilty verdict essentially rendered by the trial judge, and not by the jury.  A directed verdict of conviction as such is, of course, a concept unknown to our criminal law, technically impossible at trial, and apparently contrary to all of the fundamental principles of criminal law, in particular the presumption of innocence and the right to a trial by jury, both constitutionally protected rights.  In particular, I would be concerned that a severe curtailment of the opportunity to present a defence to the jury could be challenged as an infringement of the right “to the benefit of  trial by jury” (s. 11( f )  of the Charter ).

183                              It seems to me from the jurisprudence studied above, that the “air of reality” test was never meant to lead to directed verdicts of conviction, but was mostly designed to avoid confusing the jury, particularly in cases of inconsistent alternative defences.  Where only one defence is raised and guilt is otherwise admitted, therefore, if we are to give any real meaning to the right to a trial by jury, the application of the test should be strictly limited to situations where a technical evidentiary requirement necessary to fulfill the accused’s evidential burden for a specific defence is lacking, or when there is a complete absence of evidence on one or more of the essential ingredients of the defence.  Only in those cases can it be said that the defence is not available in law, and only in this manner, in my view, can the proper role of the jury be respected.

184                              It is important to distinguish cases where there is a complete lack of evidence from those cases where there is some evidence, but the court does not consider it strong enough to raise a reasonable doubt.  An example of a true absence of evidence can be found in Brisson, supra, where the accused did not testify at trial, but relied only on a brief statement he made to the authorities, which indicated that the murder was committed in an attempt to [translation] “calm down” the victim, but which did not contain any evidence with respect to the accused’s apprehension of death or grievous bodily harm, nor with respect to the accused’s perception of the availability of alternatives to the course of action taken.  In that case, it could truly be said that there was absolutely no evidence going to at least two of the essential requirements of self-defence.  This was so aside from any assessment of the reasonableness of the accused’s possible perceptions, since the accused did not even provide any evidence of what his perceptions were at the relevant time.  Another example can be found in Duclos, supra, with respect to the defence of intoxication, where while there was evidence of alcohol and drug consumption, there was absolutely no evidence provided with respect to the effect on the accused of such consumption, either through his testimony or through any other source.  Once again, in that case, there was a total absence of evidence concerning an essential element of the defence, that is, the fact that the accused not only consumed intoxicating substances, but was in fact intoxicated at the relevant time. Similarly, in Faid, supra, a case which involved alternative defences and therefore also raised different concerns, it was found, at p. 278, that while there was evidence of an act that the jury could possibly find amounted to provocation, there was absolutely no evidence, in the accused’s testimony or from any other source, that the accused in fact lost control, had his “passions . . . inflamed” or “killed in [the] heat of blood”.

185                              In Ewanchuk, supra, at para. 57, Major J. spoke of the limits of the “air of reality” test.  He indicated:

At this point we are concerned only with the facial plausibility of the defence of honest but mistaken belief and should avoid the risk of turning the air of reality test into a substantive evaluation of the merits of the defence.  [Emphasis added.]

Where the “air of reality” test is applied to a single defence situation, it seems clear that the only possible way to avoid a substantive assessment of the merits of the defence is to limit the application of the test to cases like the ones referred to above, where there truly is a complete lack of evidence going to an essential requirement of the defence.  The application of the test where there is evidence which, in the opinion of the court, is not strong enough to raise a reasonable doubt, should be avoided.  This is especially true where, as in this case, it is the opinion of the Court of Appeal or of this Court which is sought, as opposed to that of the trial judge.  In Thibert, supra, at para. 33, Cory J. indicated: 

It is trite but important to recall that the trial judge had the tremendous advantage of seeing and hearing the testimony of all who testified.  The trial judge was in a very advantageous position to determine if there was such evidence of provocation adduced that the defence should be left with the jury.  He considered the objections of the Crown to leaving the defence with the jury but decided, rightly in my view, that he should instruct the jury on the defence.  Unless there is an absence of any evidence as to the objective and subjective elements of the defence such a decision of a trial judge should not be lightly interfered with by an appellate court. [Emphasis added.]

186                              Therefore, and especially where appellate review is involved, the “air of reality” analysis, when applied to the sole defence raised and available to the accused, should focus only on the presence or absence of evidence, as opposed to its quality, sufficiency, or weight.  An example of the attitude that must, in my opinion, prevail in the “air of reality” analysis in cases like the present is found in the decision of Proulx J.A., for the Quebec Court of Appeal, in Taillefer, supra, with respect to the defence of intoxication, at p. 37 C.C.C.:

[translation] If the appellants, in their written statements, had only mentioned their consumption of alcohol or drugs without any indication of their effects on the acts alleged, then no instruction was required.  However, as the appellants also alluded to their condition caused by the alcohol or the drug, it must be admitted that there is evidence, though not very convincing, since it was only an assertion on which neither elaborated and which also was not tested by cross-examination, nor supported by any other evidence.  Despite the little weight which this evidence has, one cannot say that this ground of defence has no air of reality. [Italics in original; underlining added.]

187                              In other words, in cases like the present one, when examining whether  there is evidence upon which a reasonable jury, properly instructed and acting judicially, could convict,  it is the “no evidence” test, rather than the “sufficient evidence” test, which must be applied.  In fact, in my view, the test should be substantially the same as the one applied to cases of directed verdicts of acquittal or motions for non-suit, as well as committal for trial under s. 548(1)  of the Criminal Code .  In R. v. Charemski, [1998] 1 S.C.R. 679, at paras. 2-3, Bastarache J., for the majority, reiterated that in order to determine whether a directed verdict of acquittal should be granted, the “no evidence” test, derived from the case of United States of America v. Shephard, [1977] 2 S.C.R. 1067, is the applicable one.  This test was formulated as follows by Ritchie J., for the majority, at p. 1080 of the Shephard case: “whether or not there is any evidence upon which a reasonable jury properly instructed could return a verdict of guilty” (emphasis added).  The test was also found to apply to circumstantial evidence (see Mezzo v. The Queen, [1986] 1 S.C.R. 802, at p. 843; R. v. Monteleone, [1987] 2 S.C.R. 154, at p. 161).  In Charemski, supra, Bastarache J. indicated, at para. 4, that the application of the “no evidence” test to circumstantial evidence means that the question of “whether or not there is a rational explanation for that evidence other than the guilt of the accused, is a question for the jury”.

188                              In Monteleone, supra, McIntyre J., for the Court, clearly stated, at p. 160, that when assessing whether a directed verdict of acquittal is appropriate, “it is not for this Court, nor was it for the Court of Appeal, to reach a conclusion or even to consider the issue of guilt or innocence”.  Instead, the Court’s “sole concern is to inquire whether there was a case made by the Crown at trial which ought to have been left to the jury for that determination”.  At p. 161, it was also made clear that the analysis is only concerned with the existence of any evidence.  McIntyre J. indicated:

It is not the function of the trial judge to weigh the evidence, to test its quality or reliability once a determination of its admissibility has been made.  It is not for the trial judge to draw inferences of fact from the evidence before him.  These functions are for the trier of fact, the jury.

The Court concluded that there was, in Monteleone, some evidence which met the Shephard test.  McIntyre J. was careful, however, not to express any opinion as to the strength of this evidence, and rather stated, at p. 167:

In conclusion, I do not suggest that the inculpatory evidence is conclusive or even persuasive.  That is not the function of an appellate court.  The resolution of that question is for the jury upon proper instructions on the law after having heard the evidence.

189                              The principle according to which the application of the “no evidence” test constitutes a question of law which must involve no weighing of the evidence is a well-established principle that has been consistently applied by this Court (see R. v. Litchfield, [1993] 4 S.C.R. 333, at p. 361; Mezzo, supra, at p. 844; Shephard, supra, at p. 1088).  The “limited weighing” permitted in R. v. Arcuri, [2001] 2 S.C.R. 828, 2001 SCC 54, is not a departure from this line of authorities and is specific to cases where the evidence is entirely circumstantial, which is not the case here.  Naturally, it is also clear that no assessment of credibility must be involved in the determination of whether a directed verdict of acquittal should be granted, as was repeatedly emphasized in Shephard, supra, at pp. 1083-84 and 1087.  In addition, it was specified in Litchfield, supra, at p. 362, that “proof beyond a reasonable doubt is not an element of the test for a directed verdict; the proper test is some evidence” (emphasis added).    

190                              As indicated in Shephard, supra, at p. 1080, the test used to determine whether a directed verdict of acquittal is appropriate is the same as the one used to determine whether an accused should be committed for trial under s. 548(1)  of the Criminal Code  (then s. 475(1)).  The “no evidence” test was applied to committal for trial in Skogman v. The Queen, [1984] 2 S.C.R. 93, where Estey J., for the majority, at pp. 107-8, indicated that the evidence in that case “approaches the traditional expression ‘a scintilla of evidence’ but falls short of what may be classified as fanciful” (emphasis added).  It was therefore found, at p. 108, that “[c]onsequently, there can be gleaned from the record ‘some evidence’ to support the action of committal” (emphasis added).

191                              Clearly, the Shephard test to determine whether a verdict of acquittal should be directed and whether the preliminary inquiry judge should refrain from committing the accused for trial is an extremely strict test.  As I indicated in R. v. Collins (1993), 79 C.C.C. (3d) 204 (Ont. C.A.), at p. 213:

The Supreme Court in that case rejected the suggestion that a directed verdict could be granted because the evidence was “manifestly unreliable”, “dubious”, “tainted beyond redemption”, or because it had been eroded by cross-examination.

The rationale behind such a strict test is readily understood when one considers the fundamental importance of the jury system in our law.  In Charemski, supra, Bastarache J., elaborated on this rationale, in the context of the application of the “no evidence” test to cases involving circumstantial evidence only.  He stated, at para. 13:

In my view, the trial judge should have directed the jury according to the requirement that a finding of guilt could only be made where there was no other rational explanation for the circumstantial evidence but that the defendant committed the crime (John v. The Queen, [1971] S.C.R. 781, at pp. 791-92; R. v. Cooper, [1978] 1 S.C.R. 860, at p. 881; Mezzo v. The Queen, [1986] 1 S.C.R. 802, at p. 843).  Making that finding is essentially a factual matter arising from an evaluation of the evidence.  That assessment is properly left to the jury.  Judges should not be hasty to encroach on that time-honoured function, . . . [Emphasis added; emphasis in original deleted.]

192                              In Mezzo, supra, McIntyre J., for the majority, emphasized the importance of the jury’s role as the rationale for the restrictive “no evidence” test.  At p. 836, he stated:

The problem which arises here has its roots in the tendency to overlook the division of duties inherent in a trial by judge and jury.  No authority need be cited for the proposition that in a jury trial all questions of law are for the judge alone and, of equal importance, all questions of fact are for the jury alone.  The distinction is of fundamental importance.  It should be preserved so long as it is considered right to continue the use of the jury in the criminal law.  Much of the difficulty that has arisen on this subject has been caused by a failure to recognize and preserve this distinction.

Thus, McIntyre J. indicated at p. 845, at the conclusion of his reasons:

There may have been a time when a paternalistic approach to unsophisticated jurors was justified.  That time is now past and modern jurors represent a well-educated, well-informed and experienced cross-section of our society.  If it is unsafe to preserve in today’s world the distinction between the functions of a judge and a jury, that fact would count as an argument for the entire abolition of the jury system rather than for a mere change in the law relating to the extent of the jury’s role.  This would be a development that I would much regret.  [Emphasis added.]

193                              Far from being inclined to elaborate rules that would restrict the role of the jury or lead to the abolition of the jury system, this Court has on the contrary recently strongly reiterated its faith in the system and its commitment to maintaining the integrity of the jury’s important role in our criminal law (see R. v. Find, [2001] 1 S.C.R. 863, 2001 SCC 32, at paras. 26-27 and 107; R. v. Pan, [2001] 2 S.C.R. 344, 2001 SCC 42, at paras. 41-43 and 100).  As indicated by Cory J., for the majority, in R. v. G. (R.M.), [1996] 3 S.C.R. 362, at para. 13:

The jury system is clearly a significant factor in many democratic regimes.  This is emphatically true in Canada.  It is extremely important to our democratic society that jurors as representatives of their community may make the decision as to the guilt or innocence of the accused before the court based solely on the evidence presented to them.  There is a centuries-old tradition of juries reaching fair and courageous verdicts.  [Emphasis added.]

At para. 14, Cory J. added that the jury system “has proven itself in the centuries past and continues to do so today”.

194                              Our faith in the jury system explains why we have devised the “no evidence” test in order to assess whether a criminal case should go to trial and if so, whether it should be put to the jury.  We have found ourselves willing and eager to let the jury make the final decision in every case in which there was any evidence at all, no matter how weak or unreliable, going to the essential elements of the offence, even if to do so involved accepting the possibility that innocent persons might be forced to stand trial in the face of very limited evidence against them, just so that the jury, and not the judge, may make the final determination of guilt or innocence.  In my opinion and as has been repeatedly emphasized by this Court, our faith in the jury is not misplaced.  However, I see no reason why the same principles should not govern the application of the “air of reality” test, where the accused risks what is akin to a directed verdict of conviction.  There is no reason, either in law or in policy, to justify a greater level of scrutiny by the court of evidence tendered in support of an acquittal than of evidence introduced to convict.  In both cases, the same level of “sufficiency” is called for.

195                              Moreover, there is no reason to expand to the present context the “limited weighing” of the evidence endorsed in Arcuri, supra.  Arcuri was a committal for trial case, and the evidence was entirely circumstantial.  Inevitably, the court had to determine whether the evidence was capable of supporting all the inferences required so that there would be some evidence on all the essential ingredients of the offence.  Here, the full case was already before the jury, and it was based not only on circumstantial evidence but mostly on direct evidence tendered by the accused himself.

196                              In my opinion, the rationale, based on the division of tasks between judge and jury, behind the “no evidence” test for directed verdicts of acquittal and committals for trial applies with at least equal force to the assessment of the “air of reality” of the  defence put forward by an accused.  If anything, when it is the accused who is asking for his “case” to go to the jury, there are additional justifications for applying strictly a “no evidence” test.  The right to a trial by jury is constitutionally guaranteed to the accused under s. 11( f )  of the Charter .  That right is expressed as the right to the benefit of a jury trial, and it is a self-standing constitutional right, separate and distinct from the right to a fair trial (s. 11(d)).  It is therefore essential to afford the jury the opportunity to pronounce on every factual issue, in all cases where there is “some evidence” supporting the defence.

197                              Furthermore, different judicial policy concerns arise when a court considers, on the one hand, withdrawing a case from the jury, either by discharging the accused at the preliminary inquiry or directing an acquittal at trial and, on the other hand, refusing to put a defence to the jury.  A discharge at the preliminary inquiry allows for the saving of time and resources, since a possibly lengthy, expensive and unnecessary trial is avoided.  Similarly, where an acquittal is directed, the time and resources that would have been necessary for the introduction of the defence evidence, if any, the judge’s charge to the jury, and the jury deliberation, are all avoided, while the rights of the accused are protected and he is not placed in unnecessary jeopardy.  However, in spite of the considerable advantages involved in directed acquittals and refusals to commit for trial, this Court has consistently held that where there is any evidence, no matter how tainted, weak or unreliable, the case must nevertheless go to the jury in order to preserve the basic division of tasks between judge and jury and to avoid the risk of unjustified acquittals. 

198                              An application of the “air of reality” test that withdraws from the jury the only defence adduced at trial by an accused does not involve the same benefits and economy of resources as a directed acquittal or a refusal to commit the accused for trial.  The “air of reality” determination can only be made after all the evidence has been heard, so that no time or resources will be saved during the trial per se.  Since no rule of admissibility could possibly prevent an accused from testifying as to his or her version of the events, the accused’s evidence with respect to the defence raised will have to be heard whether or not the judge finds that there is an “air of reality” to the defence.  In addition, since a directed verdict of conviction as such is not an option, the judge will still have to fully charge the jury on the applicable legal principles, such as the essential elements of the offence and the Crown’s burden of proof, even where the only defence adduced is essentially withdrawn from the jury.  The jury will have to deliberate and render a verdict in the usual manner.  In sum, while it might make the trial somewhat shorter, in the end, little time or resources will be saved by an application of the “air of reality” doctrine.

199                              Therefore, the only principled and practical justification for withdrawing a defence is to avoid confusing the jury (see Pappajohn, supra, at pp. 126-27; Osolin, supra, at p. 648; Park (1995), supra, at para. 11; Sopinka, Lederman and Bryant, supra, at p. 150; G. Williams, Textbook of Criminal Law, supra, at p. 119).  This is not to say that there is no justification in principle for imposing an evidential burden on an accused who wishes to raise a defence.  On the contrary, as explained earlier, such a burden serves to ensure that the Crown does not have to adduce evidence with respect to every possible defence in order to complete its case, but must only negate beyond a reasonable doubt the defences which actually arise on the evidence.  This principle is sound.  However, when the requisite legal elements of a single defence are properly explained to the jury, there is little risk of confusion on the part of the jury solely because the evidence in support of the defence is weak and unpersuasive.

200                              Yet the cost of risking a wrongful conviction and possibly violating the accused’s constitutionally protected rights by inadvisably withdrawing a defence from the jury is a high one.  Since this Court has consistently been of the view that the possible advantages that would be gained by adopting a higher threshold for the test respecting directed verdicts of acquittal are not sufficient to justify a change in the test, then I can see no possible justification for the adoption of a higher threshold in the “air of reality” context, where such an adoption would involve fewer advantages and more risks.

201                              Therefore, I conclude that the “air of reality” test, as a judicial evaluation of evidence that is for a jury to accept or reject, is essentially the same test as the one applicable to the Crown’s case in directed verdicts of acquittal or committals for trial.  That test is whether there is any evidence upon which a jury, properly instructed and acting judicially, could acquit.  In applying the test, trial judges and appellate courts must assume that all the evidence supportive of the defence, including the testimony of the accused, is true.  Furthermore, the courts must avoid weighing the evidence and applying their own assessment of questions of fact.

202                              This Court has clearly indicated that the question of whether a defence possesses the requisite “air of reality” is a question of law (see Esau, supra, at para. 21; Osolin, supra, at p. 691).  As indicated by L’Heureux-Dubé J. in Park (1995), supra, at para. 13:

It is a legal threshold, not a factual one.  It is an error of law for a judge not to put a defence to the jury where an air of reality to that defence exists and it is an error of law to put a defence to a jury where no such air of reality exists. [Emphasis added.]

That being so, it is particularly imperative that the prohibition against weighing the evidence or assessing credibility that applies both in the context of directed verdicts of acquittal (see Litchfield, supra, at p. 361; Monteleone, supra, at p. 161; Mezzo, supra, at p. 844; Shephard, supra, at pp. 1083-84 and 1087-88)  and in the context of the “air of reality” determination (see Davis, supra, at para. 82; Ewanchuk, supra, at para. 57; Thibert, supra, at para. 7; Park (1995), supra, at para. 13; Osolin, supra, at p. 691; Bulmer, supra, at p. 790; Pappajohn, supra, at p. 132), be strictly observed when applying the “air of reality” test to the only defence raised by an accused.  Refraining  from weighing the evidence or assessing credibility, through a strict application of the “no evidence” test, will also serve to ensure that the basic division of tasks between judge and jury is observed.

203                              I add that when making the assessment of whether the defence put forward by the accused possesses the requisite “air of reality”, the testimony of the accused must not be viewed as second-class evidence.  The accused is obviously an interested witness, but this consideration can only be relevant to a determination of the accused’s credibility and of the probative value of his evidence.  Those questions are for the jury and should not enter into the “air of reality” analysis.  When an accused testifies as to the essential elements of his defence, allowing the jury to observe his demeanour and assess his credibility, he is entitled to have a properly instructed jury weigh his evidence, assess the plausibility and the reasonableness of his assertions, and decide whether his defence should be accepted or rejected. 

204                              The legal requirement goes further.  In determining whether a defence should go to the jury, the court must take the accused’s evidence as true.  Unlike the situation that prevails in cases where inconsistent alternative defences are involved, there is no need, in a case like the present, to “cobble together” the versions of the Crown and defence witnesses or to search for support for the accused’s version in the totality of the evidence as was necessary in Osolin, supra, and Park (1995), supra.  In a case such as this one there is no reason why the trial judge should not assume the whole of the accused’s evidence to be true in order to decide if the defence has an “air of reality”.  Skepticism about the truth or the plausibility of the accused’s testimony have no place in that assessment.  That, and contradictions with the rest of the evidence, will be matters for the consideration of the jury when assessing whether the accused’s evidence is in fact to be believed and what weight it should be given.

205                              The sole issue for the court is whether the facts asserted in support of the defence are, as a matter of law, capable of amounting to some evidence on all the essential elements of the defence.  It is clearly not sufficient for an accused to assert the existence of a legal defence (e.g., “I was provoked, or I was drunk”) nor is it sufficient for the accused to assert a fact that does not give rise to a defence (e.g., “I believed that the complainant consented because she did not fight back” — Ewanchuck, supra).

206                              Support for a defence can be entirely found in the accused’s testimony.  But a simple statement, “I acted in self-defence” made by an accused, even under oath, means nothing in law.  If nothing else, there might be significant differences between the popular understanding of the term self-defence and its legal meaning.  In the context of self-defence, what is relevant is the accused’s belief that he was being, or about to be, assaulted or threatened, that his life or bodily integrity was in danger and that he had no alternative but to act as he did.  When an accused testifies to that effect and indicates what he perceived to be the threat to his safety, it cannot be said that this only amounts to a mere assertion that the offence was committed in self-defence.  On the contrary, such testimony constitutes evidence concerning each requirement of the defence, and, in order to assess the “air of reality” of the defence, it must be assumed to be true (Osolin, supra; Brisson, supra; Lemky, supra; Louison, supra; Davis, supra).  The fact that the defence of self-defence includes a reasonableness requirement does not significantly modify the analysis.  The defence must go to the jury every time there is evidence relevant to each of its requirements, including the requirement of reasonableness.

C.        Where the Defence Includes a Reasonableness Requirement

207                              While reasonableness constitutes a legal requirement of self-defence under s. 34(2)  of the Criminal Code , the law is clear that it is for the jury to decide whether an accused’s perceptions were reasonable.  In Lavallee, supra, at p. 891, Wilson J., for the majority, held that “[u]ltimately, it is up to the jury to decide whether, in fact, the accused’s perceptions and actions were reasonable” (emphasis added).  It must therefore be kept in mind that the use of an objective standard, based on the legal concept of reasonableness, does not, in law, allow the trial judge or the appellate courts to make their own determination of the issue, since this determination forms an integral part of the role exclusively reserved to the jury.

208                              Since reasonableness is a question for the jury, an accused is entitled to have a properly instructed jury assess his reasonableness when the defence of self-defence is put forward.  The question then inevitably arises: can a trial judge find that the evidence is incapable of raising a reasonable doubt as to the reasonableness of the accused’s perceptions and actions, without at the same time ruling that the accused’s perceptions and actions are, in the trial judge’s opinion, unreasonable?  Trial judges are skilled and experienced at assessing reasonableness, and some may argue that they would be in a better position than the jury to decide whether an accused’s beliefs and actions are reasonable. As the law stands, this task, however,  rests with the jury.  That being the case, trial judges, when making the “air of reality” determination, must refrain from assessing the substantive merits of the defence, both in its subjective and objective components.

209                              This brings us to the crux of this case: how does the trial judge determine whether there is some evidence upon which a jury, properly instructed and acting judicially, could have a reasonable doubt as to whether the actions of the accused were reasonable?  My colleagues conclude that objective reasonableness cannot be established by direct evidence only, and that it calls for the drawing of reasonable inferences.  They turn to the “limited weighing” exercise described in Arcuri, supra, to delineate the role of the trial judge.

210                              With respect, I do not agree with that approach which will inevitably lead to a substitution of the judge’s views of reasonableness for that of the jury.  A finding of reasonableness is not a matter of drawing inferences, it is a matter of judgment.  The process of drawing permissible inferences (under the supervision of the trial judge) is required to complete the factual record.  Here, the factual record in that sense is complete.  The accused explained, and was cross-examined about, the events, his thoughts, his feelings, assumptions and rationales for acting as he did.  Given the evidence, whether or not he acted reasonably, subjectively and objectively, is a matter of judgment, and that judgment, in a case like this one, is the raison d’être of the jury system.

211                              My colleagues assert that “[i]t is in keeping with the solemnity of the jury’s task that frivolous questions are not put to it” (para. 84).  In my view, what is unworthy of the jury is that it be deprived of the opportunity to decide the only question of fact that is truly central to the guilt or innocence of the accused, having of course heard all the evidence on the point.  Again my colleagues express the concern that “[t]he jury must not be diverted from the real issues in a case by defences that the evidence cannot reasonably support” (para. 84).  There is no risk of diverting the jury from anything in this case since the only issue is whether the accused acted in self-defence; everything else he admitted.

212                              Under these circumstances, I respectfully disagree that the approach adopted by my colleagues manages to avoid, completely, consideration of the substantive merits of the accused’s defence.  I find some support in this regard by comparing their conclusions regarding the first two elements of self-defence (paras. 107-16) to their conclusions regarding the third element (paras. 117-24).  Despite the differing approach favoured by my colleagues, they effectively determine under the first two inquiries that the circumstances described by the accused could provide a basis upon which the jury could found its judgment.  Regarding the third component of the defence, however, my colleagues would apply a much more stringent standard.  Practically speaking, they would require some evidence regarding each of the alternative courses of action that were conceivably open to the accused, and why he believed these all to be unavailable to him.

213                              As I have already stated, I do not believe that one can usefully speak of a threshold of reasonableness.  Nor do I believe that reasonableness, as it is considered in applying the “air of reality” test to the defence of self-defence under s. 34(2) (b) of the Criminal Code , is best measured by an ex post facto identification of the full panoply of alternate courses of action, and a concomitant requirement that there be evidence from which the jury could determine whether the accused held a reasonable belief in the non-viability of each.  Again, reasonability is a judgment call, and one properly reserved for the jury.  It would seem to me that, considering the evidentiary threshold demanded by my colleagues in terms of the third element of the defence, they de-emphasize this fact to the point where their analysis treads a fine line between what a jury acting reasonably could conclude, and what it should conclude.

214                              The proper approach, in my opinion, is as follows.  Where a reasonableness requirement is involved in a defence, the “air of reality” test must focus on assessing whether there is any evidence of an explanation for the accused’s perceptions and conduct.  The court should not embark upon an assessment of the reasonableness, or potential reasonableness, of this explanation for that is precisely the task that the law reserves for the jury.  We have seen earlier that a mere assertion by an accused that he acted in self-defence will not suffice to give an “air of reality” to the defence, but that an assertion by the accused under oath as to what he perceived and understood with regard to the circumstances, that the law makes relevant to his defence, will be sufficient.  Similarly, where a reasonableness requirement is involved, a mere assertion of subjective perceptions will not suffice to give the objective component of the defence an “air of reality”, but the assertion under oath of the reasons and circumstances explaining why the accused perceived the situation the way he says he did will constitute some evidence relevant to the reasonableness requirement of self-defence.  As indicated earlier, in the case of a defence involving no specific technical requirement as to the type of evidence needed to fulfill the accused’s evidential burden, some evidence relevant to each requirement of the defence will always be sufficient to give the defence the requisite “air of reality”.

215                              Therefore, where, as here, an accused explains, in his testimony, why he thought he was being, or about to be, assaulted or threatened; why he apprehended death or grievous bodily harm; and why he believed that there was no alternative but to act as he did, the defence of self-defence has to be left to the jury, who will in turn have to decide whether to believe the accused and, if so, whether his perceptions and actions were reasonable in the circumstances that the jury will have found existed at the time of the offence.  In this way, the basic division of tasks between the judge and the jury will be observed, the jury’s function will not be usurped by the judge, and  the issue of reasonableness will properly remain in the province of the jury.

216                              Before essentially taking the case away from the jury, leaving the jury with no alternative but to convict, courts will be prudent to remember that their own assessment of reasonableness, even of patent unreasonableness, is at times elusive and controverted.  For example, in the context of administrative law, experienced jurists often disagree on the application of the reasonableness or patent unreasonableness standard of review.  Members of this Court have disagreed amongst themselves, or with court of appeal judges, on the question of whether a particular decision was patently unreasonable (see, for example, CAIMAW v. Paccar of Canada Ltd., [1989] 2 S.C.R. 983; Lester (W.W.) (1978) Ltd. v. United Association of Journeymen and Apprentices of the Plumbing and Pipefitting Industry, Local 740, [1990] 3 S.C.R. 644; Pointe-Claire (City) v. Quebec (Labour Court), [1997] 1 S.C.R. 1015; Canada Safeway Ltd. v. RWDSU, Local 454, [1998] 1 S.C.R. 1079; Ajax (Town) v. CAW, Local 222, [2000] 1 S.C.R. 538, 2000 SCC 23; Ivanhoe Inc. v. UFCW, Local 500, [2001] 2 S.C.R. 565, 2001 SCC 47; Sept-Îles (City) v. Quebec (Labour Court), [2001] 2 S.C.R. 670, 2001 SCC 48; Centre communautaire juridique de l’Estrie v. Sherbrooke (City), [1996] 3 S.C.R. 84; Toronto (City) Board of Education v. O.S.S.T.F., District 15, [1997] 1 S.C.R. 487; Battlefords and District Co-operatives Ltd. v. RWDSU, Local 544, [1998] 1 S.C.R. 1118; Nanaimo (City) v. Rascal Trucking Ltd., [2000] 1 S.C.R. 342, 2000 SCC 13).  It therefore appears that reasonableness findings are not necessarily as self-evident as the term might suggest, and courts are not always able to make unequivocal and unanimous findings that something was or was not reasonable.  I think that it is important to keep this in mind in a case where the only issue is basically whether it would be unreasonable for a jury to have a reasonable doubt as to whether the accused’s perceptions were reasonable.

217                              In the case at bar, four experienced jurists were of the view that it would have been possible for a reasonable jury to acquit the respondent on the basis of self-defence.  The three Quebec Court of Appeal judges expressly stated their opinion that the defence of self-defence raised by the respondent possessed the requisite “air of reality”.  As for the trial judge, while he did not make an explicit ruling on this issue, probably because he was never asked, the fact that he instructed the jury on the law of self-defence must be taken to mean that he believed that the defence had an “air of reality”.  Contrary to the submissions of the Crown, I am of the view that where a trial judge charges the jury with respect to a defence, appellate courts should not simply assume that he did so “just in case” and that it is possible that he did not really think that the defence had an “air of reality”.  Except where a trial judge expressly states otherwise, we must assume that he conducted himself in accordance with the law and did not deliberately choose to make an error of law by putting a defence to the jury that was not available on the evidence just to insulate himself from appellate review or to ensure that his decision did not prejudice the accused.

218                              Therefore, in the present case, all the judges below believed that there was some evidence upon which a properly instructed jury, acting reasonably, could have found the respondent’s perceptions and actions to be reasonable in the circumstances, had the jurors believed the respondent’s evidence.  The jury in this case rejected the defence.  Another jury may very well reject it again.  As set out earlier, that is beside the point.

III.   “Air of Reality” of the Respondent’s Defence

219                              In order to assess whether the respondent’s defence had an “air of reality” in the present case, we must determine whether there is any evidence relevant to the three elements of self-defence.  Those elements, as indicated by Lamer C.J., in R. v. Pétel, [1994] 1 S.C.R. 3, at p. 12, are the existence of an unlawful assault, a reasonable apprehension of a risk of death or grievous bodily harm, and a reasonable belief that it is not possible to preserve oneself from harm except by killing the adversary.  As Pétel also makes clear, at p. 13, an accused may make an honest but reasonable mistake as to each of those three requirements, including the existence of an assault, and still succeed in his defence.  Evidence relevant to the existence of each of these requirements is necessary for the defence to possess the requisite “air of reality”.  As for the reasonableness requirement, the “air of reality” analysis must not invite an assessment of the reasonableness of the accused’s perceptions, but rather a determination of whether there is any evidence of an explanation as to why the accused perceived things the way he did, so that the jury can decide whether the accused’s perceptions and actions were reasonable in the circumstances.

220                              I emphasize that, at this stage of the analysis, the fact that the accused’s testimony is unsupported by other evidence, or even that it is contradicted by other evidence, is irrelevant.  Consequently, the fact, for example, that the accused’s testimony as to the unusual silence in the van was contradicted by another witness, while it will undoubtedly be relevant when the time comes for the jury to assess the accused’s credibility and to decide whether any weight should be given to his testimony, can have no impact on the “air of reality” analysis.  This assessment, as explained above, is concerned only with the existence of some evidence, as opposed to its weight or credibility. 

221                              The accused’s testimony must be taken as entirely true in order to make the “air of reality” determination and the evidence must be “considered most favourably for the accused” (Lemky, supra, at para. 20).  There is also no need, as recognized by this Court in Osolin, supra, for the accused’s version to be corroborated or confirmed by other evidence, although this may become relevant to the jury’s assessment of weight and credibility.  Similarly, since self-defence is only concerned with the accused’s perceptions and their reasonableness, the fact, for example, that no independent evidence was led as to the significance of latex gloves, apart from the accused’s testimony, that no firearm was in fact found on the victim, and that no evidence was led to demonstrate that the victim had in fact stolen the accused’s gun as the accused believed he had, are equally irrelevant to an assessment of the existence of any evidence respecting the accused’s perceptions.

222                              My colleagues have concluded that there was an “air of reality” to all but one of the ingredients of self-defence in this case. They find that there is an “air of reality” to (1) the existence of an assault, both subjectively and objectively, to (2) the reasonable apprehension of death or grievous bodily harm, again both from a subjective and an objective perspective and to (3) the subjective belief in the absence of alternatives to killing.  In their view, the whole defence however had to be withdrawn from the jury because there was no evidence of the objective component of the accused’s belief.  Even though I agree with them on all but the last point, I will review the evidence on each issue to situate the contentious one in its proper context.

A.        The Existence of an Assault

223                              There was evidence that the respondent believed that he was being assaulted.  An assault can take the form of a physical attack, of threats communicated to the accused or of an attempt to assault him.  In all three cases, however, what is relevant for our purposes is only what the accused perceived as to the existence of any of these possibilities, and why.  Here, the respondent testified that he believed that the victim had stolen his gun; that he heard rumours that the victim and his companion, Ice, wanted to kill him and that he was specifically warned to that effect by a close friend.  He testified that the victim kept trying to make contact with him after he had clearly indicated that he wanted no more contact, for example, by coming to his door unannounced and unexpected.  The respondent also indicated that the victim and Ice, when they arrived at his apartment on the night in question, kept their jackets on and whispered to each other, and that Ice put his hand inside his coat.  The respondent said that he suspected, because of those facts, that Ice and the victim were armed, but that they denied it when he asked them.  He also testified that when he entered the van with Ice, and was joined by the victim shortly afterwards, he became convinced that his suspicions were true and that he was going to be killed that night.

224                              The respondent also explained why his suspicions became for him a certainty when he was in the van.  He indicated that the fact that Ice had put on different gloves than the ones he previously wore, and, most importantly, that he had put them on before entering the van, rather than waiting until they were on their way to commit the intended break and enter, was unusual and constituted an indication that a plan to murder him was about to be carried out.  More significantly, he testified that the fact that the victim was wearing surgical gloves was a clear sign to him that he was about to be killed.  The respondent explained the special significance of those gloves by reference to the fact that only twice before had he seen such gloves being used, both times in situations where bloodshed was expected, particularly in the context of attacks on criminals by other criminals.  In addition to his personal experience, the respondent also testified that he had seen movies in which latex gloves were used when bloodshed was expected.

225                              Throughout his testimony, the respondent asserted that he was convinced that the victim and Ice were just waiting for the right moment to kill him and that they were bound to do it at any time.  More specifically, the respondent testified that he believed that the victim would be the one to execute the plan, since he was sitting in the back.  The accused also testified that he knew that the victim had used firearms before.          

226                              The respondent testified that the unusual silence in the van further confirmed his belief that the victim and Ice were about to kill him, in addition to the fact that Ice was avoiding eye contact.  Also he said that Ice kept touching his jacket as if he had a gun, and remained in that position for most of the time spent in the van.  This was significant for the respondent in two respects.  First, it made him feel trapped, as it was clear to him that if Ice decided to use his gun, he would be able to do so before the respondent could reach for his own weapon, as Ice already had his hand ready while the respondent was driving.  Secondly, the respondent interpreted Ice’s gesture as a threat to him and said that in his understanding, such a gesture is usually meant to indicate to others that one is armed and to threaten them.  The fact that the victim snapped his surgical gloves while in the van was also a confirmation to the respondent that he was going to be killed.  The respondent stated that while they were at the gas station, every time he came back towards the van, Ice would put his hand back inside his jacket, which led him to feel that he was still under threat and that Ice and the victim were still just waiting for the right moment to kill him.

227                              It is not clear whether the respondent thought that the victim and Ice had made a plan to kill him that they tried to keep secret until it was carried out, or whether they made gestures, once in the van, that were meant to communicate to him their intention to kill him, such as Ice placing his hand inside his jacket and the victim snapping his gloves.  In fact, it is not even clear whether the victim ever intended to kill or threaten the respondent at all.  Whether the victim or Ice could have been found guilty of a criminal offence for threatening the respondent, and particularly whether it could have been shown that they intended to communicate to him their intention to assault him, is highly doubtful.  However, for our purposes, the victim’s true intentions are not relevant; we are only concerned with the accused’s perceptions and the explanation he provides for them (see Pétel, supra, at p. 13).  Also, whether or not the respondent reluctantly indicated that he perceived Ice’s gesture as a threat will be a matter for the jury.  In my view, at the stage of the “air of reality” analysis, it is of no import.  

228                              Since the respondent testified that he interpreted Ice’s gestures as a threat to him, and explained that he took this gesture as a signal by Ice that he was armed, it cannot in my view be said that there is no evidence that the respondent believed that he was being threatened.  Nor can it be said, when all the other circumstances testified to by the respondent are also taken into account, that there was no evidence explaining his perception and thereby allowing the jury to assess its reasonableness.  While the rumours and warning that the respondent alleges to have heard about the intentions of the victim were not interpreted by the respondent, according to his testimony, as constituting threats in themselves, they were clearly provided by the respondent as an explanation as to why he would believe that Ice and the victim were threatening him on the night of the offence, and such an explanation allows the jury to assess whether or not the accused’s belief was reasonable in the circumstances.  If the jury were to find that the respondent took these rumours seriously, rightly or wrongly, it could serve to assess the reasonableness of his subsequent actions. Furthermore, the respondent testified that he believed that the victim and Ice were attempting to kill him, in that they had “set a trap” for him in order to be able to kill him as soon as there was an opportunity.  He said that he interpreted the victim’s and Ice’s actions as being part of an attempt to kill him.   

229                              In short, in my view the accused introduced evidence that he believed  he was being assaulted and he also provided an explanation for this belief.  His testimony detailed the circumstances which gave rise to his state of mind on the night of the offence, as well as the broader context which he said led him to believe that the victim and Ice were assaulting him, including the rumours that he had previously heard and his belief that the victim had stolen his gun.  The accused’s belief that the behaviour of Ice and of the victim was increasingly unusual as the evening progressed was also evidence of an explanation for the accused’s ultimate perception that he was in fact being threatened or that an attempt to kill him was in progress.  Whether on these facts, the accused’s perception was reasonable is for the jury to decide.  The sole question before us is whether there was evidence explaining the accused’s belief, enabling the jury to make the determination of its reasonableness.  Again, the evidence in support of reasonableness does not have to be persuasive.

230                              The respondent also testified about the special significance that he attached to the fact that the victim was wearing latex gloves, and why.  Being provided with an explanation of why the respondent attached such significance to what might otherwise seem like an  innocuous indicator, the jury was well positioned to determine whether it was consistent with the respondent’s perception or not, and whether his belief was a reasonable one.  It matters little, it seems to me, where the source of his understanding lies, whether in his own criminal past or in movies.  Had he received a dead fish wrapped in newspaper or found a horse’s head in his bed, he might have also thought that his life was under threat.  I think a jury would be able to determine whether, in such circumstances, the respondent’s understanding was a reasonable one.

231                              The respondent testified that he believed that a physical attack on him was imminent.  He repeated that he believed he was going to be killed at any moment and that the victim and his companion were just waiting for the right moment to shoot him.  In cross-examination, he maintained that, although he went to the gas station to reduce the chances of being shot by staying in a crowded and well-lit area and although he got out of the van where he could not use his weapon as quickly as his adversaries, he was still not certain that he would not be shot at the gas station, and still believed that he could be killed at any time.

232                              The law of self-defence does not require the accused to wait until a firearm is actually pointed at him before defending himself.  That is, in my view, why the so-called imminency requirement was not only relaxed, but unequivocally rejected by this Court as an essential element of the defence.  In Pétel, supra, at p. 13, Lamer C.J., for the majority, clearly indicates that “[m]oreover, Lavallee, supra, rejected the rule requiring that the apprehended danger be imminent” (emphasis added).  Unlike suddenness in the defence of provocation, which is made an integral part of the defence’s requirements by the very text of the section setting out the defence (see s. 232  of the Criminal Code ), imminency of the danger is not, as such, a legal requirement of self-defence.  I do not understand the reference by the majority in R. v. Charlebois, [2000] 2 S.C.R. 674, 2000 SCC 53, at para. 16, to a requirement of imminency to be other than an application of reasonableness to the facts of that case.

233                              When it is said that imminency remains a factor for the jury to assess, and that evidence can be led to rebut this presumption of fact (Pétel, supra, at p. 13), it must be understood that there exists no legal presumption, in the form of a burden that would be imposed on the accused to demonstrate, where the attack was not imminent, that he or she nevertheless reasonably believed to be in immediate danger.  The evidential burden imposed on the accused to raise the defence remains unchanged, whether or not the attack was, or was perceived to be, imminent.  The so-called presumption of imminency simply means that where the accused does not claim to have been in immediate danger, but rather says that he seized an opportunity to avoid a distant but certain danger, the jury will be entitled, though never obligated, to draw an inference of fact that the accused’s actions were not really motivated by self-defence, since the accused did not reasonably apprehend serious harm or believe that he had no alternative.

234                              In this case, the accused repeatedly testified that he felt trapped in the situation and he believed that the only way out was to first get out of the van and then shoot before he got shot.  The accused also gave an explanation as to why he did not walk away or call the police and as to why he believed that the only way out of the situation was to shoot the victim.  It will be for the jury to decide whether this explanation is compelling and whether it can displace the common sense inference resulting from the lack of imminency, and from the fact that the victim had his back turned to the accused at the time of the shooting.  Imminency will be but one factor for the jury to take into account in its assessment of reasonableness, and, since it does not constitute a formal legal requirement of the defence, its absence alone may not be used by trial judges or appellate courts to withdraw self-defence from the jury.

B.        Reasonable Apprehension of Death or Grievous Bodily Harm

235                              The respondent has testified that he apprehended death on the night of the offence.  He said that he was convinced that the victim would be the one to kill him, because he was sitting in the back and he believed that the shot would come from behind.  This was evidence of his apprehension of serious harm.  The respondent also provided, in his testimony, an explanation for his apprehension, which would allow the jury to assess its reasonableness.  The factors referred to above, which led the respondent to believe that he was being assaulted, also caused him, according to his evidence, to apprehend death.  There were numerous factors and their relevance to the respondent was clearly explained in his testimony.  They may not appear very persuasive that the respondent’s apprehension of death was reasonable in the circumstances, and they may not prove sufficient to raise a reasonable doubt in the jury’s mind as to the reasonableness of the respondent’s apprehension.  However, they amount to some evidence upon which the jury may make its own assessment of reasonableness, since the jurors were provided with the full background and explanation for the respondent’s perceptions.

C.        Subjective Belief in the Absence of Alternatives to Killing

236                              Finally, the respondent testified that he believed that shooting the victim was the only way to preserve his life on the night in question.  His testimony was clearly to the effect that he believed himself to be in a situation of “kill or be killed”, so that shooting the victim was, in his mind, the only way to get out of the situation.  The respondent explained first that he did not call the police because, having spent his life running away from them, he did not believe that the police would be willing to help him.  He also believed that the police would force him to work as an informant if he were to seek protection.  He said that he thought that the police would arrive too late to help him if he were to call.  As for running away, the respondent indicated that it did not enter his mind at the time.  He explained that he did not want to leave his van with the three others in it, and that he believed he had nowhere to run, as the victim and Ice knew where he lived.  He stated that he thought that the only way out of the situation was to make Ice and the victim leave the van.  Again, all this constituted evidence in this case which provided an explanation as to why the respondent believed that he could not simply walk away from the situation.  Whether or not that explanation is compelling, and whether or not the respondent’s belief was reasonable, will once again be a matter for the jury. 

237  Reasonableness can hardly be established by a single item of evidence; rather, it calls for an appreciation of all the circumstances, including the credibility of the witnesses who relate the events.  Despite its objective component reasonableness remains a question for the jury.  In the present case, the accused’s testimony detailed the circumstances surrounding the offence, as well as the broader context in which his perceptions arose.  The evidence required to give a defence an “air of reality” may come from the accused’s testimony alone, and need not be confirmed by independent evidence.  Therefore, it will be for the jury to assess whether, in the context described by the accused of having two men in his van that he believed to be armed and intent on killing him that very night, it was reasonable for him not to consider running away or calling the police.

238  I must add that the accused did not forfeit his right to rely on self-defence merely because he entered his van with the victim and Ice without being coerced to do so, while he suspected, but was not yet certain, that they wanted to kill him.  What is relevant for the purposes of the law of self-defence is the reasonableness of the accused’s actions and perceptions at the time of the offence.  The background circumstances and the prior actions of the accused will of course be relevant to this assessment, but the jury is not asked to pass judgment on the reasonableness of the accused’s past actions, or of his whole lifestyle.

239  In Lavallee, supra, referring to the fact that the accused had not left her partner before in spite of the intolerable violence that she was subjected to, Wilson J. indicated, at p. 884, that “[t]his question does not really go to whether she had an alternative to killing the deceased at the critical moment”.  While this evidence was of course relevant to provide the background circumstances necessary to assess the reasonableness of the accused’s belief that she had no alternative at the time of the offence, the jury was not to assess whether it was reasonable for the accused not to leave her partner at a previous time.  Wilson J. further explained, at p. 888:

I emphasize at this juncture that it is not for the jury to pass judgment on the fact that an accused battered woman stayed in the relationship.  Still less is it entitled to conclude that she forfeited her right to self-defence for having done so.  [Emphasis added.]

240  The respondent is not asking this Court to take into account, in deciding whether there is any evidence relevant to the reasonableness of his perceptions and actions, a criminal “world-view”, or a specific set of values that would advocate, for example, choosing to fight instead of running away in order to preserve one’s honour.  Such values would certainly not be controlling.  In R. v. Hibbert, [1995] 2 S.C.R. 973, at paras. 59-60, Lamer C.J., for the Court, indicated that the standard to be used to assess the reasonableness of the accused’s belief in the lack of alternative should be similar for the defences of necessity, duress and self-defence.  In Ruzic, supra, at para. 61, LeBel J. described the standard as follows, in the context of the defence of duress:

The test requires that the situation be examined from the point of view of a reasonable person, but similarly situated.  The courts will take into consideration the particular circumstances where the accused found himself and his ability to perceive a reasonable alternative to committing a crime, with an awareness of his background and essential characteristics.

241  The fact that the respondent’s background in this case, as well as the circumstances surrounding the offence, involve criminal activity, does not mean that it cannot be taken into account by the jury.  On the contrary, the assessment remains similar, regardless of the type of circumstances involved.  In all cases, the jury must take into account the accused’s circumstances and background to the extent that they shed light on and explain his perception of the existence of an assault, his apprehension of death and his belief in the lack of alternative.  Reprobation for the accused’s lifestyle or past actions is properly expressed in a prosecution for any crime that he might have committed  in the past, but it should not be expressed by an effective denial of the defence of self-defence through the refusal to take into account circumstances that are related to a criminal background.

IV.    Conclusion

242  The test for committal for trial, directed verdicts of acquittal and the withdrawing of a defence from the jury strives to respect the long-standing divisions of tasks between judges and jury, and favours great deference to the wisdom of the jury.  Discharges at the preliminary inquiry and directed verdicts of acquittal also promote judicial economy and may serve as an early barrier to the danger of a wrongful conviction.  Not so where a defence is withdrawn from the jury.  The only rational purpose served by withdrawing a defence is to avoid confusing the jury who may then wrongly acquit on the basis of that confusion.  When the critical issue in a criminal case is whether a person who otherwise committed a crime acted reasonably in the circumstances, we ought not to misconstrue the risk of the jury being confused with the possibility that we may disagree with its verdict.

243  Consequently, I would dismiss the appeal.

Appeal allowed, Iacobucci, Major and Arbour JJ. dissenting.

Solicitor for the appellant:  The Attorney General of Quebec, Montréal.

Solicitors for the respondent:  Pasquin, Bibeau, Brouillard, Gariépy & Associés, Montréal.

Solicitor for the intervener the Attorney General of Canada:  The Attorney General of Canada, Ottawa.

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