Mezzo v. The Queen, [1986] 1 S.C.R. 802

Michael Douglas Joseph Mezzo Appellant;

and

Her Majesty The Queen Respondent.

File No.: 17958.

1985: February 20; 1986: June 26.

Present: Dickson C.J. and Beetz, Estey, McIntyre, Chouinard, Lamer, Wilson, Le Dain and La Forest JJ.

on appeal from the court of appeal for manitoba

Criminal law ‑‑ Directed verdict ‑‑ Charge of rape ‑‑ Sufficiency of identification evidence ‑‑ Complain­ant’s observations of assailant made in difficult conditions ‑‑ Improper police identification procedures ‑‑ Trial judge finding identification evidence unreliable to convict ‑‑ Whether trial judge erred in directing a verdict of acquittal ‑‑ Test for a directed verdict.

The only issue raised at the accused's trial for rape was identification. Complainant, the only witness with respect to this issue, was attacked on a street shortly after midnight and dragged behind a shed where she was sexually assaulted. The incident took approximately twenty minutes. She testified that although it was dark behind the shed she was still able to see her assailant's face. She described him to the police on three different occasions with some minor variations. Two weeks after the attack, the complainant took part in a courtroom identification. She reacted visibly when she saw the accused and told the police that he looked like her assailant, but she was not sure because her view was obstructed by a sheriff. On a second courtroom identification two days later, she positively identified the accused as her assailant. She did so again at the preliminary hearing and at the trial. At the close of the Crown's case, the trial judge weighed and considered the quality of the evidence and granted a motion for a directed verdict. The majority of the Court of Appeal allowed the Crown's appeal and ordered a new trial. This appeal is to determine whether the trial judge erred in directing a verdict of acquittal despite the fact that there was some evidence as to the identification of the accused.

Held (Lamer and La Forest JJ. dissenting): The appeal should be dismissed.

Per Beetz, Estey, McIntyre, Chouinard and Le Dain JJ.: In deciding whether to grant a motion for a directed verdict, the trial judge must determine "whether or not there is any evidence upon which a reasonable jury properly instructed could return a verdict of guilty". A trial judge is not entitled to withdraw a case from a jury where there is admissible evidence which could, if believed, result in a conviction, notwithstanding the frailties of the evidence. It is not the trial judge's function to weigh or consider the quality of the evidence. The weight that should be given to evidence, like questions of credibility, are matters for the jury only. In the case at bar, there was direct evidence of identification and the case should have been left to the jury with a proper caution.

Per Dickson C.J. and Wilson J.: The governing test in Canada for a directed verdict is "whether or not there is any evidence upon which a reasonable jury properly instructed could return a verdict of guilty". This test may be susceptible of more than one interpretation, but an examination of its antecedents support the view that at this stage of the trial, the basis on which a trial judge can withdraw a case from the jury is extremely narrow. Although some cases suggest that the trial judge can direct a verdict only where there is no evidence at all, the words "no evidence", in this context, are not to be taken literally as meaning a total absence of evidence but rather as meaning no evidence capable of supporting a conviction.

In determining whether or not to direct a verdict in cases which turn on eyewitness testimony, the trial judge should address his mind to the factors going to the quality of the identification evidence and, where the frailties in the evidence can be remedied by a caution, he should leave the matter to the jury.

In the present case, the trial judge erred in law in directing a verdict of acquittal on the basis solely of the quality of the initial identification evidence. The substantial consistency in the complainant's description of the accused given on three separate occasions prior to any improper police procedures required him to put the evidence to the jury with a caution as to the inherent frailty of the identification evidence, coupled with an instruction to consider carefully the conditions under which the identification is made. A jury is in just as good a position as the trial judge to assess the witness' opportunity for observation and the strength of his evidence based on that opportunity.

The courtroom identification neither destroyed the complainant's fairly reliable initial evidence nor weakened it beyond the point where it could not be taken care of by a properly worded caution. The conduct of the police line‑up, while irregular, was not so irregular by Canadian standards as to be beyond the capabilities of a jury, properly directed, to balance the overall strengths and weaknesses of her evidence against the prejudicial impact of the improper identification process and the difficulties inherent in the complainant's initial observation of her assailant.

Per Lamer and La Forest JJ., dissenting: The test to be applied by a judge when deciding a motion for a directed verdict is "whether or not there is any evidence upon which a reasonable jury properly instructed could return a verdict of guilty". When assessing the sufficiency of the evidence, the trial judge cannot assess the witness' honesty which is a matter for the jury but may assess the witness' capacity to be accurate and his resulting reliability. Thus, when applied to testimonial evidence, the test involves an assessment by the judge of the circumstances under which a witness makes his observations. These circumstances may undermine the reliability of the witness' accuracy.

In the case at bar, the trial judge did not err in directing a verdict of acquittal. It is clear from his reasons that he applied the proper test and that he addressed his mind to an assessment of the quality of the identification evidence. He found that the difficult conditions under which the observations were made and the improper identification procedures of the police‑‑procedures that could tend to induce the witness into mistakenly identifying someone‑‑made it too dangerous to rest a conviction on her testimony, honest as it might well have been.

The majority of the Court of Appeal found that it would not have been unsafe to convict, while the dissenting judge shared the trial judge's view. This did not raise a question of law alone in the Court of Appeal, and the Crown's appeal should have been dismissed.

Cases Cited

By McIntyre J.

United States of America v. Shephard, 1976 CanLII 8 (SCC), [1977] 2 S.C.R. 1067, applied; R. v. Comba, 1938 CanLII 7 (SCC), [1938] S.C.R. 396, aff'g 1938 CanLII 14 (ON CA), [1938] O.R. 200; R. v. Turnbull, [1976] 3 All E.R. 549, considered; R. v. Duhamel, 1980 ABCA 247, [1981] 1 W.W.R. 22; R. v. Atwood and Robbins (1788), 1 Leach 464, 168 E.R. 334; Metropolitan Railway Co. v. Jackson (1877), 3 App. Cas. 193; R. v. Morabito, 1949 CanLII 1 (SCC), [1949] S.C.R. 172; R. v. Dowsey (1866), 6 N.S.R. 93; R. v. Lloyd (1890), 19 O.R. 352; Girvin v. The King (1911), 1911 CanLII 42 (SCC), 45 S.C.R. 167; Fraser v. The King, 1936 CanLII 42 (SCC), [1936] S.C.R. 1; Fraser v. The King, 1936 CanLII 25 (SCC), [1936] S.C.R. 296; R. v. Knox, 1967 CanLII 692 (BC CA), [1968] 2 C.C.C. 348; R. v. Sawrenko (1971), 1971 CanLII 1218 (YK CA), 4 C.C.C. (2d) 338; R. v. Paul, 1975 CanLII 185 (SCC), [1977] 1 S.C.R. 181; Lavoie v. The Queen, 1977 CanLII 212 (SCC), [1977] 1 S.C.R. 193; R. v. Cooper, 1977 CanLII 11 (SCC), [1978] 1 S.C.R. 860, referred to.

By Wilson J.

United States of America v. Shephard, 1976 CanLII 8 (SCC), [1977] 2 S.C.R. 1067, applied; R. v. Turnbull; R. v. Roberts; R. v. Whitby, [1976] 3 All E.R. 549; R. v. Comba, 1938 CanLII 7 (SCC), [1938] S.C.R. 396; R. v. Morabito, 1949 CanLII 1 (SCC), [1949] S.C.R. 172, aff'g (1948), 1948 CanLII 11 (ON CA), 91 C.C.C. 210; R. v. Faryna (1982), 1982 CanLII 3796 (MB CA), 18 Man. R. (2d) 185; R. v. Sutton, 1969 CanLII 497 (ON CA), [1970] 2 O.R. 358; R. v. Smith (1975), 1975 CanLII 2371 (NS CA), 12 N.S.R. (2d) 289; R. v. Engel (1981), 1981 CanLII 3519 (MB CA), 9 Man. R. (2d) 279; R. v. Simpson (No. 2) (1981), 1981 CanLII 3284 (ON CA), 58 C.C.C. (2d) 122; Marcoux v. The Queen, 1975 CanLII 10 (SCC), [1976] 1 S.C.R. 763, considered; R. v. Duhamel, 1980 ABCA 247, [1981] 1 W.W.R. 22; Hodge’s Case (1838), 1838 CanLII 1 (FOREP), 2 Lewin 227, 168 E.R. 1136; John v. The Queen, 1970 CanLII 199 (SCC), [1971] S.C.R. 781; R. v. Paul, 1975 CanLII 185 (SCC), [1977] 1 S.C.R. 181; Lavoie v. The Queen, 1977 CanLII 212 (SCC), [1977] 1 S.C.R. 193; R. v. Kyling, 1970 CanLII 159 (SCC), [1970] S.C.R. 953; Feeley v. The Queen, 1952 CanLII 11 (SCC), [1953] 1 S.C.R. 59; Commonwealth of Puerto Rico v. Hernandez, [1973] F.C. 1206; R. v. Smierciak (1946), 1946 CanLII 331 (ON CA), 87 C.C.C. 175; R. v. Wray, 1970 CanLII 2 (SCC), [1971] S.C.R. 272, referred to.

By Lamer J. (dissenting)

United States of America v. Shephard, 1976 CanLII 8 (SCC), [1977] 2 S.C.R. 1067; R. v. Comba, 1938 CanLII 7 (SCC), [1938] S.C.R. 396; Commonwealth of Puerto Rico v. Hernandez, [1973] F.C. 1206; R. v. Turnbull, [1976] 3 All E.R. 549; R. v. Sutton, 1969 CanLII 497 (ON CA), [1970] 2 O.R. 358; R. v. Spatola, 1970 CanLII 390 (ON CA), [1970] 4 C.C.C. 241; R. v. Duhamel, 1980 ABCA 247, [1981] 1 W.W.R. 22; R. v. Atfield (1983), 1983 ABCA 44, 42 A.R. 294; R. v. Faryna (1982), 1982 CanLII 3796 (MB CA), 18 Man. R. (2d) 185; R. v. Smierciak (1946), 1946 CanLII 331 (ON CA), 87 C.C.C. 175.

Statutes and Regulations Cited

Criminal Code, R.S.C. 1970, c. C‑34, ss. 475 [am. R.S.C. 1970 (2nd Supp.), c. 2, s. 8; 1985 (Can.), c. 19, s. 101(1)], 618(2)(a) [rep. & subs. 1974‑75‑76 (Can.), c. 105, s. 18(2)].

Authors Cited

Devlin, P. Trial by Jury, London, Stevens & Sons Ltd., 1966.

Morris, J. Fish. "Committal for Trial: ‘Some’ Evidence is Not ‘Sufficient’ " (1979), 39 R. du B. 607.

APPEAL from a judgment of the Manitoba Court of Appeal (1983), 1983 CanLII 3697 (MB CA), 35 C.R. (3d) 272, 22 Man. R. (2d) 223, allowing the Crown's appeal from the accused's acquittal on a directed verdict on a charge of rape and ordering a new trial. Appeal dismissed, Lamer and La Forest JJ. dissenting.

John Scurfield, for the appellant.

Brian Kaplan, for the respondent.

The reasons of Dickson C.J. and Wilson J. were delivered by

1. Wilson J.‑‑I have had the benefit of reading the reasons of my colleague, Lamer J. He would allow the appeal on the basis that the majority of the Court of Appeal had no jurisdiction to interfere with the trial judge's decision because no question of law arose. With respect, I have a different view. I think two questions of law arose.

2. The first question of law concerns the test to be applied in determining when a case should be withdrawn from the jury because of the quality of the identification evidence. Although there is a considerable amount of jurisprudence on the test for a directed verdict, some confusion has been generated by the use of different terms to describe what may or may not be different tests. An examination of the existing authorities is, I believe, helpful in eliciting the proper principles.

3. The second question of law raised by the trial judge's decision is when subsequent improper police procedures so impair the quality of earlier identification evidence that the case should be withdrawn from the jury. It would appear that the trial judge would have directed the verdict on either of these grounds. However, he clearly felt that the second ground reinforced his conclusion on the first and this is not surprising since the two are obviously linked. It would seem that the weaker the initial identification evidence the less impropriety it would take to undermine it totally. The stronger the initial identification evidence the higher the degree of subsequent impropriety it might take to undermine it.

4. The Courts Below

5. There is some disagreement in the courts below as to the basis on which the evidence was withheld from the jury. It is clear that Wright J. at trial felt that the complainant's testimony met the test of extreme insufficiency required for a directed verdict. What is unclear is whether this was because of the potential for inaccuracy in the complainant's initial observation of her assailant or because of the irregularities in her subsequent identification of Mr. Mezzo as the accused. Wright J. analyzed the problem before him in terms of the English Court of Appeal's decision in R. v. Turnbull, [1976] 3 All E.R. 549. The key passage from Wright J.'s reasons reads:

I have concluded that these facts fall squarely within the directions in Turnbull, approved by the Alberta Court of Appeal, and even where there has been a proper or adequate identity parade, require that the case be taken from the jury.

In the present case there is the additional factor that I do not believe the procedure followed by the police in the subsequent identification of the accused by the complainant at the police station was proper and that fact has further damaged the complainant's identification evidence.

(Emphasis added.)

5. There is nothing in Wright J.'s reasons to indicate that he considered the possibility of counteracting the frailties in the complainant's initial observation of her assailant with a strongly worded caution to the jury. His failure to do so may have been due to the fact that he collapsed the issue of the quality of the initial identification with the issue of the effect of the improper police procedures. This seems to be the interpretation favoured by Matas J.A., dissenting in the Court of Appeal (1983), 1983 CanLII 3697 (MB CA), 35 C.R. (3d) 272. He expressed the view that Wright J. could not possibly have decided to direct a verdict on the basis of the complainant's initial identification evidence alone as that would be to take the principles in Turnbull too far.

6. With respect, I think Matas J.A. may be wrong in his assessment of what the trial judge did. I think that the trial judge in the passage already quoted from his reasons did decide that the frailties of the initial identification evidence were reason enough for directing a verdict. He specifically states that "even where there has been a proper or adequate identity parade" the case must be taken from the jury. However, he went on to find that the "additional factor" of the improper police procedures "further damaged" the complainant's initial identification evidence. The question whether Wright J. erred in law in his assessment of the complainant's initial identification evidence is therefore validly raised. Indeed, as I hope to show in the ensuing analysis of the authorities, Wright J. erred in taking the case from the jury on this basis. The substantial consistency in the complainant's descriptions of the accused given on three separate occasions prior to any improper police procedures required him, in my view, to put the evidence to the jury with an appropriate caution.

7. Matas J.A. in his dissenting reasons concluded that the trial judge was correct in directing a verdict. He thought that "the difficulty with the several descriptions furnished by the complainant was compounded by the improper show‑up identifications" (p. 284). He concluded that it was impossible to say "if the complainant identified the accused on the second occasion only as the man she saw in court the first time or whether she was making a valid identification of the assailant" (p. 285). Accordingly, for this reason Matas J.A. thought that the only course open to the trial judge was to direct a verdict of acquittal.

8. Hall J.A., speaking for the majority, had a different view. He considered that the identification evidence of the complainant should have gone to the jury for an assessment of its quality in light of the opportunity she had to see him provided proper directions were given by the trial judge. He found that the applicable principle was that enunciated in R. v. Comba, 1938 CanLII 7 (SCC), [1938] S.C.R. 396, and adopted by the Alberta Court of Appeal in R. v. Duhamel, 1980 ABCA 247, [1981] 1 W.W.R. 22, rather than that put forward in Turnbull which the trial judge had applied. In other words, the proper test in Hall J.A.'s view was whether in light of the "dubious nature of the evidence it would be unsafe to find the prisoner guilty".

9. The Test for a Directed Verdict

10. In Canada the governing test for a directed verdict is set out by Ritchie J. in United States of America v. Shephard, 1976 CanLII 8 (SCC), [1977] 2 S.C.R. 1067 at p. 1080, as follows:

... whether or not there is any evidence upon which a reasonable jury properly instructed could return a verdict of guilty. The "justice", in accordance with this principle, is, in my opinion, required to commit an accused person for trial in any case in which there is admissible evidence which could, if it were believed, result in a conviction.

I agree with Lamer J. that the division of views expressed by the majority and the minority in Shephard is of little relevance here. It is clear that in the present case the issue was not the credibility of the complainant but the quality of her evidence based on her opportunities for observation. However, I disagree with my colleague's view that " "quality" of evidence is truly the realm of the trial judge". A jury is in just as good a position as a trial judge to assess the witness' opportunity for observation and the strength of his or her evidence based on that opportunity. Indeed, this is something that lay people are peculiarly well equipped to do.

10. In my view, both the law and the traditional importance of the jury require a large measure of deference to be accorded the jury even where quality as opposed to credibility is at issue. Although by itself Ritchie J.'s test in Shephard may be susceptible of more than one interpretation, an examination of its antecedents supports the view that at this stage of the trial the judge's function is narrowly circumscribed by the central role of the trier of fact.

11. In R. v. Morabito, 1949 CanLII 1 (SCC), [1949] S.C.R. 172, affirming (1948), 1948 CanLII 11 (ON CA), 91 C.C.C. 210 (Ont. C.A.), this Court had occasion to expand on the dual role of a trial judge sitting alone on a motion for a directed verdict. In support of the motion the defence had argued that the evidence adduced by the Crown should leave in the mind of the trial judge a reasonable doubt as to the guilt of the accused. The trial judge proceeded to weigh the evidence in relation to the reasonable doubt standard and acquitted the accused. The Court of Appeal upheld the acquittal. However, Roach J.A. in dissent was of the view that the trial judge had no jurisdiction at this stage of the proceedings to consider the question of reasonable doubt. Rather, that question was for the jury or for the judge qua jury only after the defence had made its election to submit or not submit additional evidence and after that evidence, if any, was in. Roach J.A. described the limited nature of the trial judge's decision at this intermediate stage in terms of a choice between "evidence which...would justify a conviction" and "no evidence" as follows at pp. 215‑16:

The question whether or not there was any evidence which, if believed, would justify a conviction, is a question of law. I have recorded sufficient of the Crown's evidence to make it perfectly clear that as a matter of law if the accused was being tried by a Judge and jury, rather than by a Judge alone, the Judge would have had to leave the case to the consideration of the jury either to acquit or to convict. He could not have instructed the jury that there was no evidence and directed them to return a verdict of acquittal.

Roach J.A.'s dissenting reasons found favour with a majority of this Court and it was unanimously agreed that a new trial should be ordered.

12. A similar question as to the scope of the trial judge's function on a motion for a directed verdict arose in the context of circumstantial evidence in R. v. Comba, supra. In that case this Court set aside a jury verdict of guilt. Sir Lyman Duff C.J. stated, in effect, that cases comprised wholly of circumstantial evidence should be measured against the rule in Hodge’s Case (1838), 1838 CanLII 1 (FOREP), 2 Lewin 227, 168 E.R. 1136. The rule states that where all the evidence is circumstantial the accused can be found guilty only if the evidence is both consistent with guilt and inconsistent with any other rational conclusion. He also remarked that this would have been an appropriate case to direct a verdict of acquittal. A line of jurisprudence subsequently developed which incorporated the rule in Hodge’s Case into the trial judge's deliberations on a motion for a directed verdict, thus creating an exception where circumstantial evidence was concerned to the narrow role of the judge at this stage of the trial.

13. The divergence of views on this point was not resolved until the mid 1970's. In John v. The Queen, 1970 CanLII 199 (SCC), [1971] S.C.R. 781, this Court stated that the rule in Hodge's Case was simply an elaboration of the reasonable doubt standard. The logical extension of this position occurred in R. v. Paul, 1975 CanLII 185 (SCC), [1977] 1 S.C.R. 181, and Lavoie v. The Queen, 1977 CanLII 212 (SCC), [1977] 1 S.C.R. 193, cases in which this Court held that the rule in Hodge’s Case and the question of reasonable doubt does not arise on a motion for a directed verdict. Although additional confusion has been generated by Ritchie J.'s seeming approval in Shephard of the notion that Sir Lyman Duff C.J. intended to create a special exception in Comba, I think his remarks must be read in light of the clear rejection only a few months later of the relevance of the Hodge's Case standard to directed verdicts in Lavoie, supra. Laskin C.J. wrote the judgment of a unanimous Court which included Ritchie J. and which reads in its entirety as follows at p. 193:

It will not be necessary to hear you, Mr. Gagnon. We are all of the opinion that there was no error in the judgment of the Court of Appeal. On the motion by the accused for a directed verdict, the rule in Hodge’s Case (1838), 1838 CanLII 1 (FOREP), 2 Lewin 227, 168 E.R. 1136, does not apply, as this Court held in R. v. Paul, 1975 CanLII 185 (SCC), [1977] 1 S.C.R. 181. The appeal is dismissed.

Accordingly, the notion of a Comba exception to general principles in terms of the standard against which to measure circumstantial evidence at this stage of a trial was finally put to rest and the restrictive role of the trial judge reaffirmed.

14. Some of the other antecedents to Shephard lend further support to the view that the basis on which a trial judge can withdraw a case from the jury is extremely narrow. As mentioned earlier, Roach J.A. of the Court of Appeal in Morabito spoke in terms of "no evidence" and several other decisions describe the directed verdict question as the "absence of evidence" question. Perhaps the strongest statement to this effect is by Pigeon J. in R. v. Kyling, 1970 CanLII 159 (SCC), [1970] S.C.R. 953. In the course of his reasons he stated at p. 955:

The first reason given by the majority for dismissing the appeal was that the decision of the trial court was not really a nonsuit, but rather an acquittal on the merits, for an insufficiency, not an absence, of evidence. It was rightly pointed out that whereas the absence of any evidence is a question of law, the insufficiency of the evidence is essentially a question of fact. The right of appeal by the prosecution, however, is only on a question of law.

And later at p. 956:

In The King v. Morabito, 1949 CanLII 1 (SCC), [1949] S.C.R. 172, this Court held that a nonsuit pronounced on a motion by the defence after the close of the case for the Crown is a decision on a question of law, namely, the absence of evidence. This rule was also held applicable to summary proceedings by the judgment in Feeley et al. v. The Queen, 1952 CanLII 11 (SCC), [1953] 1 S.C.R. 59. I have found nothing in the 1955 Criminal Code that might imply a change in the principles underlying these judgments, nor has anything to this effect been brought to our attention.

15. The language in these cases would seem to limit intervention by the trial judge to instances in which there is absolutely no evidence at all. While this would seem to offer a simple and clear‑cut standard, I believe that in this context "no evidence" is not to be taken literally as meaning a total absence of any evidence but as meaning rather no evidence capable of supporting a conviction.

16. The dissenting reasons of de Grandpré J. in Paul afford, perhaps, the clearest insight into the special meaning of "no evidence". In that case Ritchie J. for the majority ordered a new trial, thus overturning a directed verdict of acquittal at trial on a charge of knowing possession of stolen goods. He stated at p. 189 that in his view the admitted facts were such that "a properly instructed jury might reasonably have drawn the inference that they [the goods] did not get there without having been stolen by somebody". De Grandpré J. (Laskin C.J. and Spence J. concurring) disagreed. He felt that evidence on two of the essential elements of the offence in question was lacking. In the course of his reasons de Grandpré J. wrote at pp. 185‑86:

There is no question that on a motion for nonsuit the accused can only be released if there is an absence of evidence. In this connection we need only refer to the remarks of Pigeon J. in The Queen v. Kyling, 1970 CanLII 159 (SCC), [1970] S.C.R. 953, at p. 956:

As Pigeon J. points out, the applicable principles were discussed more fully in Morabito and Feeley, and I adopt the reasons expressed in those cases.

However, the expression "absence of evidence" does not mean that the motion for a nonsuit must be dismissed whenever there is an iota of evidence, no matter how inconsequential this may be. On the contrary, a careful reading of Feeley, referred to above, discloses that there were few iotas of evidence in that case. Appellants had been acquitted at first instance on a motion for nonsuit, and the Court of Appeal had set aside this acquittal and ordered a new trial. This Court unanimously reinstated the judgment of acquittal with respect to three of the four appellants. Cartwright J., as he then was, speaking for the Court, emphasized that the various aspects of the evidence "taken together, are insufficient to make out a prima facie case" (p. 64).

(Emphasis added.)

In de Grandpré J.'s view the Court of Appeal had not only applied the correct principles of law but had also correctly concluded that the evidence submitted by the Crown met the "absence of evidence" standard.

17. The division of views in Paul in a sense highlights the area in which a trial judge can legitimately impose his or her assessment of the evidence on the trier of fact. As in Comba the crucial evidence in Paul was circumstantial. The minority found that it failed to establish any direct link between the goods in question and a theft. Ritchie J. for the majority thought that the facts were such that one could infer they had been stolen. He stated at pp. 191‑92:

As I have indicated, however, I do not think that the judgment of Sir Lyman Duff in the Comba Case, supra, in which he reiterates the rule in Hodge’s Case has any application to the circumstances here disclosed. I am rather of opinion, as indicated by Mr. Justice Pigeon in R. v. Kyling, 1970 CanLII 159 (SCC), [1970] S.C.R. 953 at p. 957, that the true test to be applied before granting a nonsuit is whether the Crown evidence gives rise merely to a suspicion or whether it is such as to justify the drawing of an inference.

(Emphasis added.)

Thus, in a sense, what was a "mere suspicion" for the minority was a "valid inference" for the majority. This debate over the line between suspicion and inference is likely to appear in any case in which circumstantial evidence of an essential element of the offence is dubious. However, even where the critical link is purported to be established by direct evidence, it can, in my view, be withheld from the jury only at that same point, i.e., where the evidence approaches a similar level of extreme dubiousness or insufficiency in the sense of simply raising a suspicion or a conjecture of guilt and no more.

18. The traditional formulation of the test for a directed verdict also suggests that the "absence of evidence" standard is not meant to be interpreted in literal terms. In Feeley v. The Queen, 1952 CanLII 11 (SCC), [1953] 1 S.C.R. 59, Cartwright J. (as he then was) for the Court summarized the applicable principles as follows at pp. 60‑61:

It is common ground that had the learned Magistrate refused the motion the appellants would have had the right to call evidence for the defence if so advised and counsel for the respondent submits that the decision of this Court in The King v. Morabito, 1949 CanLII 1 (SCC), [1949] S.C.R. 172, establishes (i) that at that stage it was open to the learned Magistrate to dismiss the charge unless there was no evidence on which, had the trial been before a jury, a properly instructed jury, acting reasonably, might have convicted the accused, and (ii) that whether or not there was such evidence is "a question of law alone" within the meaning of s. 1013(4) of the Code. I agree with this submission.

Cartwright J.'s qualification of the appropriate standard in terms of whether a properly instructed jury might convict appears throughout the jurisprudence and is reflected by Ritchie J. in Shephard. Indeed, the link made by Ritchie J. in Shephard between the test for a directed verdict and that for sufficiency of evidence under s. 475 of the Criminal Code supports the notion that "no evidence" represents a degree of quality rather than an absence of any evidence, good or bad. In addition, Ritchie J. in the course of his reasons gives some indication of the point at which poor quality evidence reaches that degree of extreme insufficiency. His implicit approval of the refusal to issue a warrant for extradition in Commonwealth of Puerto Rico v. Hernandez, [1973] F.C. 1206, provides a concrete illustration of identification evidence which is so dubious that a properly instructed jury could not reasonably convict. In that case, Thurlow J. (as he then was) of the Federal Court of Appeal described the evidence in question as follows at p. 1214:

Indeed, I find it inconceivable that a person should be put on trial on such flimsy evidence as a purported identification made a year after the event by a person who did not previously know the accused and whose only opportunity to observe him was a fleeting one from a distance of some sixty feet, if indeed he ever got that close. The evidence of the witness having selected the respondent's identification card is not in itself evidence implicating the respondent, and in the circumstances as described it tends to weaken rather than to strengthen the identification made by Atilano at the hearing ... Nor was there anything else to implicate the respondent or corroborate such identification.

19. Unlike my colleague, Lamer J., I do not think that the extension of the absence of evidence standard implicit in the test of whether a properly instructed jury might convict vests in the trial judge an unbounded discretion over the qualitative aspects of evidence. The use of the terms "no evidence" and "absence of evidence" at the very least imports a notion of constraint and stringency into the standard. However, I do agree with my colleague that Sir Lyman Duff C.J.'s direction in Comba with regard to the dubious nature of the evidence does not create a test of dubiousness which is separate and distinct from the test in Shephard. In my view Comba and Shephard are consistent with one another. Certainly this was the view of Ritchie J. in Shephard. He remarked on Comba at pp. 1081‑82 as follows:

It is to be observed that there was no suggestion in the Comba case that any of the evidence called by the Crown was either tainted or unreliable. It came from witnesses whose integrity was at no time put in question and who testified as to a variety of circumstances which had excited enough suspicion against the accused to occasion his arrest and trial but which taken together did not establish his guilt in accordance with the accepted standards of proof in such cases which are described by Duff C.J. ...

Ritchie J. then quotes Sir Lyman Duff C.J.'s formulation of the rule in Hodge's Case. As mentioned earlier, this causes some confusion over whether Comba imported the reasonable doubt standard into the test for a directed verdict in cases based on circumstantial evidence. In light of this Court's position in Paul and in Lavoie, it is my view that Ritchie J. in Shephard was using Comba simply to illustrate his point that poor quality evidence which merely raises a suspicion may legitimately be withheld from the jury in contra‑distinction to evidence that is poor in the sense of lacking in credibility. The question then is what content is to be given to Sir Lyman Duff C.J.'s concept of dubiousness or to Ritchie J.'s concept of a proper and reasonable basis for conviction. Can the content be articulated in a way that is useful without being overly cumbersome and inflexible?

20. In the case at bar, Hall J.A. proposed the Comba test of dubiousness as an alternative to Lord Widgery C.J.'s direction in Turnbull which had been relied on by the trial judge. Again, I do not think that the controversy in this case stems from competing judicial tests. Apart from the question of Lord Widgery C.J.'s requirement of corroboration for poor quality evidence which does not arise here, I do not think there is such inconsistency between Turnbull and Shephard that they must be viewed as imposing different standards. In fact, Turnbull makes it quite clear that what is usually required in cases which turn on eyewitness testimony is a direction to the jury as to the inherent frailty of identification evidence coupled with an instruction to consider carefully the conditions under which the identification is made. Lord Widgery C.J. elaborated on the caution to be given as follows at p. 552:

Secondly, the judge should direct the jury to examine closely the circumstances in which the identification by each witness came to be made. How long did the witness have the accused under observation? At what distance? In what light? Was the observation impeded in any way, as for example by passing traffic or a press of people? Had the witness ever seen the accused before? How often? If only occasionally, had he any special reason for remembering the accused? How long elapsed between the original observation and the subsequent identification to the police? Was there any material discrepancy between the description of the accused given to the police by the witness when first seen by them and his actual appearance?

One can deduce from this catalogue of impediments to accuracy that observations made under less than ideal conditions by witnesses can nevertheless be put to the jury. This is consistent with the Shephard formulation of the test.

21. After giving several examples of good quality evidence, Lord Widgery C.J. went on to suggest at p. 553 that where the quality is poor "as for example when it depends solely on a fleeting glance or on a longer observation made in difficult conditions", the trial judge should withdraw the case from the jury and direct a verdict of acquittal. Thus, at some point the adverse conditions for observation cannot be cured by a caution. Again this is consistent with Ritchie J.'s implicit approval of Hernandez. Unfortunately, Lord Widgery C.J. does not indicate, and it is perhaps impossible to do so, when the borderline is crossed. No one would take issue with the "fleeting glance" test; it represents the extreme of frailty which cannot be cured by a caution. The real difficulty is with "a longer observation made in difficult conditions". In this sense Turnbull is more significant for what it does not lay down. If it sets out any principle with regard to the test for a directed verdict, it is the principle that the quality of a witness' identification is directly related to the extent of the witness' opportunity for observation. However, Turnbull offers no workable criteria for determining when conditions are so difficult that an eyewitness' testimony should not be relied on or, in the language of the Canadian cases, when valid inferences become mere suspicions. The only guidance Turnbull gives is on the actual fact situation in Turnbull itself and in the accompanying appeals.

22. The fact situation in Turnbull is by and large unhelpful as the challenged evidence was recognition evidence and was supported by ample additional evidence. Accordingly, it passed Lord Widgery C.J.'s test and was found to have been properly put to the jury. In the other two appeals dealt with by the Court at the same time the convictions were quashed. However, they involved significantly worse conditions of observation than those in the present case and in this respect resemble the evidentiary basis in Hernandez. In R. v. Roberts the assailant was a stranger to the complainant, the attack was over in a few moments, the place of the attack was dark except for the flashing lights of the dance hall, the accused's subsequent conduct was consistent with his denial of guilt and there were significant contradictions in the descriptions of the identifying witnesses. In R. v. Whitby the identifications were of a man in a balaclava who had been seen for a short time during a robbery. One witness had seen him from the rear only; the other witness gave a completely inconsistent description before picking the accused out of a line‑up. There is very little one can point to in either of these cases which would redeem the obvious untrustworthiness of the identification evidence. Together with Comba and Hernandez they serve to delineate the line beyond which it would be improper in law to put the question of guilt or innocence to the jury. The case at hand requires that one delineate the opposite boundary, namely that beyond which it would be improper in law to withhold the question of culpability from the jury. I agree that within these broad parameters deference should in general be paid to the decision of the trial judge. Although the cases continually assert that whether there is an "absence of evidence" is a question of law upon which either party can appeal, this does not mean that the reviewing court should automatically intervene whenever it disagrees with the trial judge's assessment. However, unlike Lamer J., I do not believe that Wright J.'s assessment of the complainant's initial identification evidence falls within that area of deference.

23. It seems to me that Turnbull provides some useful guidance on how to delineate the line between suspicion and inference and fill the analytic gap left by Lord Widgery C.J.'s reference to "difficult conditions". One can infer from the kinds of considerations enumerated in that case as going to the quality of identification evidence that, at the very least, the trial judge should address his or her mind to the same kind of considerations in determining whether or not to direct a verdict. In addition, it would seem to follow that where the frailties in the evidence can be remedied by a caution, the judge should leave the matter to the jury. This, indeed, is the ultimate test; can the frailties be remedied by an appropriate caution? It is not, in my view, good enough for a trial judge to make a finding as to the quality of the evidence. He must go on to determine whether the adverse conditions for observation can be taken care of by an appropriate caution. If they can, then Ritchie J.'s test of sufficiency in Shephard can be met, i.e., there is "evidence upon which a reasonable jury properly instructed could return a verdict of guilty" (p. 1080).

24. I do not mean to suggest that the trial judge should in every case be held to a rigid tabulation of the factors listed in Turnbull. A mechanical test would not reflect the infinite particularity of possible fact situations. The Turnbull caution does, however, identify a number of factors which can clearly affect the quality of eyewitness evidence: length of observation, distance, light, obstructions to view, recognition, time between original observation and subsequent description to the police and discrepancies between that description and the accused's appearance. No doubt there are many more. The consistency in the witness' descriptions (relevant in this case), his or her degree of attention and awareness at the time of the crime, and the witness' reaction at subsequent encounters with the accused (also relevant in this case) are some that come readily to mind.

25. In the present case the evidence, apart from the suggestiveness of the courtroom identification, is dubious with regard to only two of the Turnbull factors, namely the poor lighting during most of the encounter and the fact that the assailant was a stranger to the complainant. These factors will be present in a large number of rape cases. I agree with Matas J.A. that we have to have regard to the realities of this particular offence. Wright J. seemed especially concerned by the lack of lighting. Admittedly in some circumstances, such as total darkness or the use of masks, a single factor by itself may require a directed verdict. However, it is clear from the Turnbull caution that imperfect or partial lighting does not always undermine accuracy. In my view, the question of whether or not it did so in this case should have been left to the jury. The complainant's first impression occurred under streetlights. The ensuing twenty minutes or so were in the dark but at close contact. The witness was fairly self‑possessed a few hours later and was able at that point to give a general but coherent description. Her subsequent descriptions including the police portrait were for the most part consistent with that first description. In addition, she visibly reacted when later shown the accused in the course of a courtroom identification.

26. In sum, except for the fact that her assailant was a stranger, the darkness behind the CN sheds, and several minor inconsistencies with regard to the length of hair and the fabric and length of the coat, the complainant's identification was detailed and coherent. It would be surprising indeed if such evidence was relegated to the category of mere suspicion and conjecture. The trial judge correctly focussed on the lighting as the crucial factor in relation to accuracy. However, I think that any weakness generated by the partially lit conditions was well within the ability of the jury to handle sensibly in combination with an appropriate caution. Any other interpretation would be inconsis­tent with the scope and purpose of Lord Widgery C.J.'s comparatively detailed direction in this regard and with the stringency of the absence of evidence standard which underlies Ritchie J.'s test in Shephard. It follows that Wright J. could not have directed himself to these considerations and in particular to the question of whether the dubiousness generated by the poor lighting could be adequately addressed in a strongly worded caution. Thus, whether one regards Wright J.'s decision as flowing from a misunderstanding of Turnbull or as a failure to apply Shephard, the conclusion seems inescapable that he committed an error of law in directing a verdict of acquittal on the basis solely of the quality of the initial identification evidence.

27. The Effect of Improper Police Procedures

28. I have considered Wright J.'s assessment of the complainant's initial identification evidence in isolation from the issue of the possible detrimental impact on it of the subsequent improper police procedures. In practice, one could consider all the factors going to quality of the evidence together and only then apply the Shephard test with regard to the capacity of a properly instructed jury to convict on it. This was the approach taken by Matas J.A. For purposes of the present discussion I think it is helpful to proceed in two stages in order to clarify the considerations which are appropriate on a motion for a directed verdict. It will also be more efficient in a case such as this to take a two‑step approach. Where the quality of a complainant's initial identification is so unreliable, e.g. the fleeting glance, that even with a strongly worded caution from the judge it could not support a verdict of guilt, there would be no need to go on to consider the impact of subsequent improper police procedures. However, where as in the present appeal, the initial identification evidence meets the test of sufficiency, the trial judge must, in my opinion, go on to the second stage and consider the impact on that evidence of the subsequent police procedures.

29. The details of the line‑up procedure employed in the present appeal were summarized by Matas J.A. and are reproduced in my colleague's reasons. Several features of it are relevant to the quality of the identification. Rather than place the accused in an identification parade the police brought the complainant to view a number of prisoners as they were brought into court. Mr. Mezzo was among them. There is no indication that this procedure was adopted because of the accused's refusal to participate in a proper parade or by urgency. The complainant knew the police had a suspect in custody. She reacted visibly when she saw Mr. Mezzo but, because her view was obstructed by the sheriff, she said that although Mr. Mezzo looked like the man she could not be certain. In addition, the police were unable to say whether Mr. Mezzo's name had been called, the charge read, or whether any of the other prisoners called at the same time looked like Mr. Mezzo. The problems in terms of evidential quality inherent in this first attempt to have the complainant identify her assailant are fairly obvious.

30. Because of the complainant's uncertainty on this first line‑up, a second courtroom viewing was arranged at which Mr. Mezzo's name was definitely called out. This time the complainant positively identified Mr. Mezzo as her assailant. Matas J.A. comments that "It is impossible to say if the complainant identified the accused on the second occasion only as the man she saw in court the first time or whether she was making a valid identification of the assailant" (p. 285). He went on to say that the problem is magnified when it is coupled with the other factors of suggestiveness and the absence of Crown evidence on the particulars of the line‑up. Matas J.A. therefore agreed with the directed verdict of acquittal.

31. Both the trial judge and Matas J.A. rely on R. v. Smierciak (1946), 1946 CanLII 331 (ON CA), 87 C.C.C. 175 in their assessment of the impact of the line‑up. The accused in that case was charged with attempting to pass a forged cheque by presenting it to a bank teller. When the teller asked for a registration card the accused searched his pockets and remarked that he must have left it in his car. He then walked out of the bank and did not return. Although the teller would not have taken any special care to observe the accused at the time of the transaction, the conditions for observation were very good. The police subsequently showed her a single photo of a suspect which she identified as the man. He was convicted. On appeal Laidlaw J.A. stated at p. 177:

In addition to such matters, and of the utmost importance, is the method used to recall or refresh the recollections of a witness who is to be relied upon to identify a person suspected of wrongdoing or who is under arrest. If a witness has no previous knowledge of the accused person so as to make him familiar with that person's appearance, the greatest care ought to be used to ensure the absolute independence and freedom of judgment of the witness. His recognition ought to proceed without suggestion, assistance or bias created directly or indirectly. Conversely, if the means employed to obtain evidence of identification involve any acts which might reasonably prejudice the accused, the value of the evidence may be partially or wholly destroyed. Anything which tends to convey to a witness that a person is suspected by the authorities, or is charged with an offence, is obviously prejudicial and wrongful. Submitting a prisoner alone for scrutiny after arrest is unfair and unjust. Likewise, permitting a witness to see a single photograph of a suspected person or of a prisoner, after arrest and before scrutiny, can have no other effect, in my opinion, than one of prejudice to such a person.

Consequently, Laidlaw J.A. quashed the appellant's conviction, stating that the police procedures had rendered the teller's evidence valueless and, in the absence of any other evidence, "it would be unfair and unsafe to convict him" (p. 180).

31. The parallel to the facts of the present case is clear. Taken at its best we have here a fairly reliable initial identification followed by two improper and suggestive line‑ups. However Laidlaw J.A.'s concern in Smierciak, supra, seems to be as much with prejudice to the accused as with the frailty of the complainant's evidence. He concludes that because of the prejudice arising from the police procedures it would be unfair and unsafe to convict. In contrast, it is the frailty of the evidence aspect that appears to have been foremost in the minds of both the trial judge and Matas J.A. in the present case. Indeed, Matas J.A. specifically states that it is not necessary to consider the issue of the exclusion of the evidence on the basis of either common law or constitutional fairness principles.

32. While suggestiveness and observational impediments may feature in both the frailty issue and the fairness issue, the two are approached from different standpoints and involve radically different policy concerns. Where the frailty or accuracy of the identification is under consideration, the standpoint is that of the complainant and of society. The underlying concerns are to redress the harm caused to her and to society by criminal behaviour and to ensure that the verdict is supported by proper evidence. Where fairness is at issue, the standpoint is that of the accused and the underlying concern is the affront to justice inherent in prejudicial and needlessly suggestive procedures. In spite of these conceptual differences, accuracy and unfairness are often treated as aspects of the same issue. This can in some instances be explained by the fact that ultimately the two standpoints converge. Obviously it would be unfair to convict an accused on frail evidence. In addition, where evidence is excluded for fairness reasons, the effect may be to render the remaining evidence frail. In the present appeal, the appellant has argued accuracy and fairness as separate points. I agree that they should be viewed as analytically distinct. The question whether evidence should be excluded is not properly part of the decision to direct a verdict except in the indirect sense mentioned above, namely where the exclusion effectively undermines the other perfectly acceptable evidence.

33. Laidlaw J.A.'s decision suggests that an improper line‑up can destroy otherwise good evidence. This principle is implicit in Matas J.A.'s conclusion in the present appeal and in the case of R. v. Faryna (1982), 1982 CanLII 3796 (MB CA), 18 Man. R. (2d) 185 (C.A.) However, a retreat from Laidlaw J.A.'s interventionist approach is displayed in the cases of R. v. Sutton, 1969 CanLII 497 (ON CA), [1970] 2 O.R. 358 (C.A.); R. v. Smith (1975), 1975 CanLII 2371 (NS CA), 12 N.S.R. (2d) 289 (C.A.); R. v. Engel (1981), 1981 CanLII 3519 (MB CA), 9 Man. R. (2d) 279 (C.A.) and R. v. Simpson (No. 2) (1981), 1981 CanLII 3284 (ON CA), 58 C.C.C. (2d) 122 (Ont. C.A.)

34. In R. v. Faryna Huband J.A., with whom O'Sullivan J.A. concurred, was of the view that the use of an improper line‑up fatally weakened the Crown's case. As in the present appeal the entire case turned on eyewitness evidence. However, unlike the present appeal, the initial identifications given by the two complainants were extremely poor, consisting in one instance of two encounters lasting a total of ten seconds and in the other instance of one encounter lasting three seconds. In addition, subsequent descriptions given by the complainants had several inconsistencies and both picked the accused out of line‑ups in which the persons in the line bore no resemblance to the accused. Huband J.A. distinguished R. v. Smith, supra, on the basis of the extraordinary frailty of the initial identifications of Mr. Faryna. He went on to describe the line‑up and its impact on the Crown's case as follows at pp. 195‑96:

She next was asked to identify the accused when he was presented to her in a police lineup, which can be best described as the situation in which the police identified their suspect to her by means of juxtaposition, and distinction from the others in the lineup. Having identified the man so indicated as her assailant, there is a reasonable apprehension that her subsequent descriptions of her assailant, both at the preliminary inquiry and at trial, were based on impressions she formed of the man in the police lineup, rather than of the man she saw for fleeting seconds at the time of the assault. This is not to impute any ill motive to Mrs. Cudmore. It simply reflects the inherent danger in identification evidence of this type and highlights the reasons why police lineups must be conducted properly if any weight is to be given to the evidence obtained as a result.

35. While Huband J.A.'s decision may have been warranted by the combination of the fleetingness of the initial encounters with the bias implicit in identification by juxtaposition, it is clear that where the reliability of the evidence rather than control of police behaviour or prejudice to the accused is the over‑riding concern, the psychological barriers generated by suggestive line‑ups do not always destroy otherwise good evidence. In R. v. Sutton, supra, Jessup J.A. ordered a new trial in part because the trial judge had failed to caution the jury with respect to the frailty generated by an improperly conducted photo line‑up. The implication is that those frailties could have been reme­died by a proper direction. Likewise in R. v. Smith and R. v. Engel, supra, the respective Courts of Appeal were of the view that a caution to the trier of fact was all that was needed to counteract the damage caused by improper line‑up procedures.

36. In Smith, the strength of the initial identification played a large part in the reviewing court's assessment of the conviction at trial. MacDonald J.A. of the Nova Scotia Court of Appeal said at pp. 297‑99:

Counsel for the appellant contends that the evidence of Mr. Walker, particularly in view of the acquittal of Pace, is too weak and uncertain to found a conviction. In support of this contention he referred to Regina v. Spatola (1970), 1970 CanLII 390 (ON CA), 10 C.R.N.S. 143, Rex v. Brown and Angus (1951), 1951 CanLII 393 (BC CA), 99 C.C.C. 141, Rex v. Goldhar, Rex v. Smokler (1941), 1941 CanLII 311 (ON CA), 76 C.C.C. 270, Regina v. Shaver (1971), 1970 CanLII 1756 (NS CA), 2 N.S.R. (2d) 225, and Regina v. Kerwin (1974), 26 C.R.N.S. 351. I subscribe to what was said in all those cases but would point out that here Mr. Walker had ample opportunity of observing the facial features of the appellant enabling him to receive an impression of a definite cast of countenance. When the distinguishing red mark is added to this, in my opinion, it cannot be said that Mr. Walker could not visualize the personal characteristics and individuality of the appellant to the extent necessary to enable him to identify him late on the day of the robbery in the lineup. This being so, I would refer again to Regina v. Spatola, supra, wherein Laskin, J.A., said in effect that identification evidence is for the jury to evaluate on proper direction.

The object of the identification parade or lineup is to make sure that the ability of the witness to recognize the suspect has been fairly and adequately tested. I think it is desirable for the police to hold an identification parade before the arrest, trial or preliminary inquiry, as the case may be. Such a parade or lineup should be carried out by placing the accused, or the suspect, with a sufficient number of people as similar physically as possible to him, leaving the witness to pick him out if he can, without assistance.

In the present case my opinion is that the construction of the lineup was unsatisfactory. The effect of such a defective lineup goes to the weight of the identification evidence which, in the present case and notwithstanding the unsatisfactory lineup was sufficient for the reasons already given to support the conviction.

37. In R. v. Engel, supra, O'Sullivan J.A., with whom Huband J.A. concurred, agreed with counsel that the line‑up was a farce. Nevertheless he felt that the conviction should be upheld in view of the fact that the trial judge sitting alone "suitably warned himself of the dangers inherent in the identification evidence and in the lack of corroboration" (p. 284).

38. These cases reflect a deference to the ability of the triers of fact to weigh poor quality evidence. This deference is implicit in Ritchie J.'s restrictive formulation of the governing test in Shephard. However, the case which best illustrates this confidence in the jury and which perhaps approximates most closely to the facts in the present appeal is R. v. Simpson (No. 2), supra.

39. The accused in Simpson was charged with attempted murder. The attempt occurred at night, albeit near a street light. The complainant spent the whole time trying to fight off her assailant. Afterwards she was able to give a description which generally corresponded to that of the accused except that she got the colour of his eyes wrong. The accused refused to participate in a line‑up. The police brought the complainant to a courtroom where he was appearing in connection with another offence. The accused entered the courtroom alone, handcuffed, and escorted by a constable. The complainant immediately began to shake and cry and was taken from the room. After she had calmed down she asked for another view in order to be sure. She was then taken back into the courtroom whereupon she positively identified the accused as her assailant.

40. At Simpson's trial the judge charged the jury as follows as related by Martin J.A. at p. 130:

You may think then that Simpson being brought into court in this fashion diminishes, if not completely does away with, the weight to be attached to Cathy's identification on that day in that court room. But you might also feel that you are entitled to overlook the unusual procedure, or at least not feel that it detracts from her identification when you consider Constable Anderson's very vivid description of how she reacted to seeing Simpson, saying "That's the man" or words to that effect. When you are looking at a man handcuffed, surrounded by police in a court room is one thing, but to react as Cathy did seems to me to be another thing, you may find, or you may not find. You may disagree‑‑you may agree. You may think then that her reactions may be sufficient, or a sufficient act of identification which eliminates the problem of the method of bringing Simpson in and the fact he was handcuffed and with the police. That is one of the areas that you are going to have to make a decision upon.

Although the accused only contested the admission of the evidence of the complainant's physical reaction to his appearance in court, I think that Martin J.A.'s rejection of that ground and approval of the trial judge's direction implied an approval of the trial judge's reliance on the ability of the jurors to balance the prejudicial impact of the in‑court identification process and the difficulties inherent in the complainant's initial observation of her assailant against the overall strengths and weaknesses of her evidence. The parallels with the circumstances in Mezzo are quite striking. In both cases the initial encounter is under difficult conditions but the complainant's subsequent descriptions are consistent and fairly specific. If anything, the line‑up process in Simpson is more suggestive than that in Mezzo. In both cases there are two viewings and in both cases the witness has a physical reaction to the sight of the suspect.

41. I think that several conclusions can be drawn from these authorities. First, the improprieties in the line‑up do not necessarily destroy otherwise good evidence. Second, consistent with Shephard, damage can often be remedied by a proper caution. Third, at this stage of the analysis the assessment of the evidential basis for conviction depends not only on separate assessments of the weaknesses in the initial observation and the weaknesses in the conduct of the line‑up but also on the impact of one or the other. As I stated at the outset, the stronger the initial identification evidence the higher the degree of subsequent impropriety it might take to undermine it.

42. It seems to me that the line‑up procedure employed in this case was not so irregular by Canadian standards as to be beyond the capabilities of the jury to weigh. Because of the failure of the Crown to record the particulars at the courtroom encounter one must assume that the viewings were tantamount to single show‑ups. However, in Marcoux v. The Queen, 1975 CanLII 10 (SCC), [1976] 1 S.C.R. 763, a unanimous decision of this Court, single show‑ups were approved by implication, at least in so far as considerations of accuracy are concerned. The accused in that case refused to take part in a line‑up. The complainant who had awakened to find a robber in his hotel room in the early morning hours subsequently identified the accused in the hallway of a police station and a few minutes later in a room occupied by Mr. Marcoux and a police constable. At trial, the defence in a motion for a directed verdict vigorously attacked the failure of the police to conduct a proper line‑up. The trial judge refused the motion. On appeal to this Court the central issues were whether line‑ups violated the privilege against self‑incrimination and whether evidence of an accused's refusal to participate in a line‑up is admissible. In the course of his reasons for the Court Dickson J. (as he then was) discussed the use of single show‑ups and in‑dock identification. Although he did not make any specific finding with regard to Canadian law, he referred briefly to American and British judicial criticism of the failure to use a proper line‑up in the following passage at pp. 773‑74:

New York cases since People v. Ballott (1967), 20 N.Y. 2d 600, have found the failure to employ a line‑up by the police, absent imperative circumstances, violative of due process‑‑see "Pre‑trial Identification", (1971) 55 Minnesota Law Review 779, at p. 791. The Court of Criminal Appeal of England has commented adversely on the absence of an identification parade and failure to explain the omission to hold one: R. v. Seiga, [1961] Crim. L.R. 541. In R. v. John (unreported, abridged in (1973) Crim. L.R. 113), evidence was given by a detective constable that the accused had refused an identification parade. On appeal from conviction, it was urged, unsuccessfully, that the trial judge had wrongly allowed one of the witnesses to identify the accused in the dock when she had not identified him before the trial. Lord Justice Cairns, delivering the judgment of the Court of Appeal, said:

This method of identification ought to have been avoided if possible. It might have been possible in this case if the Appellant had been willing to submit to an identification parade. Leaving out altogether the question of whether or not it was reasonable for him to refuse an identification parade, the fact was that the Prosecution did not have that very much better method of identification available and, that being so, it does not appear to this Court to be wrong to allow that method of identification to be used by that witness, Miriam Baker.

43. The discussion in Marcoux concerns the fairness or justice aspect of police procedures and of their impact on the accused's rights. There is no suggestion that apart from due process considerations the potential for inaccuracy generated by the station‑house show‑up rendered the evidence against Mr. Marcoux unsafe as a basis for conviction. The implication is that because Mr. Marcoux's refusal necessitated the show‑up the fairness issue does not arise and that the trial judge properly put the evidence to the jury in spite of the problems in the quality of the initial observation and the subsequent identification.

44. In the present appeal the complainant's initial identification evidence, in my view, was such as to require the trial judge to put it to the jury with an appropriate caution. The conduct of the police line‑up was clearly improper but again, on the basis of authorities such as Marcoux and Simpson, not so improper as to be beyond the power of the jury to handle given a proper direction by the trial judge. I do not think the police line‑up had the effect of destroying the complainant's initial evidence or weakening it beyond the point where it could not be taken care of by a properly worded caution. Indeed, the visible reaction of the complainant to Mr. Mezzo was, if anything, supportive of her earlier testimony and her obvious desire to be sure before she made a positive identification in the courtroom bore witness to her good faith and credibility. Matas J.A. raises the possibility that all the complainant was doing in the second courtroom encounter was recognizing the person she had seen on the first encounter. With respect, I think this is to seriously underestimate a witness who had given a perfectly sensible and commendable reason for wanting to have an unobstructed view before committing herself. But these are, in my opinion, all matters for the jury to weigh once the judge has drawn them to its attention.

45. The question remains whether the improprieties in the line‑up were such that the evidence of the complainant's identification of Mr. Mezzo as her assailant should be excluded for reasons of common law or constitutional fairness. The "imperative circumstances", namely the accused's refusal to participate in a proper line‑up, present in Marcoux and Simpson, are absent here. In addition the police by failing to record the particulars of the procedure have effectively made it impossible for the trial judge to caution the jury with respect to this portion of the evidence other than by advising them to assume the worst. In my view, these circumstances do raise a serious concern over prejudice to the accused. However, I believe that the proper context for that discussion would have been on a motion to exclude the evidence either on the basis of R. v. Wray, 1970 CanLII 2 (SCC), [1971] S.C.R. 272, and the accompanying jurisprudence or under s. 24(1) of the Charter. No such motion was made when the line‑up evidence was admitted at trial. Nor has the case been approached on that basis either by the trial judge or by the Court of Appeal.

46. In light of the standard imposed in Shephard I would agree with the majority of the Court of Appeal that the case should have been put to the jury with a proper direction. I would therefore dismiss the appeal. It will, of course, be open to the defence on the new trial, if so advised, to move to have the line‑up evidence excluded for reasons of fairness.

The judgment of Beetz, Estey, McIntyre, Chouinard and Le Dain JJ. was delivered by

47. McIntyre J.‑‑I have read the reasons for judgment prepared in this appeal by my colleagues Lamer and Wilson JJ. I agree with Lamer J. that the issue to be decided in this case, is whether the trial judge was in error in directing a verdict of acquittal, despite the fact that there was admissible evidence before the court as to the identification of the accused. With deference to Lamer J., however, I am unable to agree with his disposition of the matter. Like my colleague, Wilson J., I would dismiss the appeal and affirm the Court of Appeal's order for a new trial.

48. The facts have been outlined by Lamer J. in some detail and need no repetition in these reasons. The Crown's case rests upon the identification made by the complainant of the man who assaulted her. The evidence she gave relating to the actual assault is adequately set out by Lamer J., and the further identification evidence is set out in the passages from the judgment of Matas J.A. in the Court of Appeal, which were included in Lamer J.'s reasons.

49. The trial judge followed the judgment of Harradence J.A., speaking for the Alberta Court of Appeal in R. v. Duhamel, 1980 ABCA 247, [1981] 1 W.W.R. 22, which in turn adopted and applied the reasons of Lord Widgery C.J. in the English Court of Appeal, Criminal Division, in R. v. Turnbull, [1976] 3 All E.R. 549. This led him to a consideration of the quality of the evidence. In this process he indulged in a detailed weighing of the evidence and he concluded that a verdict should be directed. He was also moved towards his decision by what he considered to be improper police procedures taken to procure further identification evidence in the proceedings.

50. In the Court of Appeal (1983), 1983 CanLII 3697 (MB CA), 35 C.R. (3d) 272, Hall J.A., speaking for the Chief Justice and himself, acknowledged that a trial judge dealing with the question of a directed verdict should consider the quality of the evidence. He said (at p. 277):

In sum, the trial judge should, without making any finding of credibility, form a view or opinion whether, because of the dubious nature of the evidence, it would be unsafe to find the accused guilty. If he forms that opinion, he would be justified in directing a jury to return a verdict of acquittal.

He went on to conclude that in the case before the court the evidence was sufficient for the jury's consideration, after a proper direction from the judge. The trial judge was in error "not so much in adopting the wrong test but in misapplying the test in Turnbull to the facts of the case". He indicated that the preferable language to employ was that employed in R. v. Comba, 1938 CanLII 7 (SCC), [1938] S.C.R. 396. He set aside the directed verdict and ordered a new trial. Matas J.A., in dissent, set out the facts of the case clearly and completely. He reviewed many of the relevant authorities. He was concerned with the frailties of visual identification and procedures followed by the police in connection with the courtroom identification. He would have dismissed the Crown appeal.

51. 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.

52. The leading case in this Court on the subject of directed verdicts is United States of America v. Shephard, 1976 CanLII 8 (SCC), [1977] 2 S.C.R. 1067. Ritchie J. wrote the majority judgment (Martland, Judson, Ritchie, Pigeon and de Grandpré JJ.), while Spence J. dissented with the concurrence of Laskin C.J., Dickson and Beetz JJ. The United States applied for Shephard's extradition. The only evidence adduced before Hugessen J., sitting as an extradition judge, was an affidavit sworn by a co‑conspirator of Shephard. Hugessen J. considered that he had to apply the test involved in taking a case away from a jury and, saying the evidence was "so manifestly unreliable", he declined to order extradition. An appeal to the Federal Court of Appeal was dismissed and the matter came to this Court. Ritchie J. agreed with the trial judge that the test to be applied was that expressed in s. 475(1) of the Criminal Code, which is set out hereunder in its amended form:

53. (1) When all the evidence has been taken by the justice he shall

(a) if in his opinion there is sufficient evidence to put the accused on trial for the offence charged or any other indictable offence in respect of the same transaction, order the accused to stand trial; or

(b) discharge the accused, if in his opinion on the whole of the evidence no sufficient case is made out to put the accused on trial for the offence charged or any other indictable offence in respect of the same transaction.

He said, at p. 1080:

I agree that the duty imposed upon a "justice" under s. 475(1) is the same as that which governs a trial judge sitting with a jury in deciding whether the evidence is "sufficient" to justify him in withdrawing the case from the jury and this is to be determined according to whether or not there is any evidence upon which a reasonable jury properly instructed could return a verdict of guilty. The "justice", in accordance with this principle, is, in my opinion, required to commit an accused person for trial in any case in which there is admissible evidence which could, if it were believed, result in a conviction.

Later, after referring to the trial judge's comments that the evidence was "manifestly unreliable", he added, at p. 1087:

With the greatest respect I cannot accept the proposition that a trial judge is ever entitled to take a case from the jury and direct an acquittal on the ground that, in his opinion, the evidence is "manifestly unreliable". If this were the law it would deprive the members of the jury of their function to act as the sole judges of the truth or falsity of the evidence and would thus, in my opinion, be contrary to the accepted role of the jury in our legal sytem.

53. Ritchie J. has stated what has been called the `classic statement' of the rule that has been the law of Canada since before the Criminal Code was adopted. In this, he has adopted as the measure of sufficiency the concept of a prima facie case, that is, a case containing evidence on all essential points of a charge which, if believed by the trier of fact and unanswered, would warrant a conviction.

54. Ritchie J. based his formulation of the test to be applied by a trial judge on a line of authorities going back as far as 1788. In R. v. Atwood and Robbins (1788), 1 Leach 464, 168 E.R. 334, it was recognized, at pp. 334‑35, in the words of Buller J.:

The distinction between the competency and the credit of a witness has been long settled. If a question be made respecting his competency, the decision of that question is the exclusive province of the Judge; but if the ground of the objection go to his credit only, his testimony must be received and left with the Jury, under such directions and observations from the Court as the circumstances of the case may require, to say whether they think it sufficiently credible to guide their decision on the case.

This is consistent with the well‑known expression of the rule in Metropolitan Railway Co. v. Jackson (1877), 3 App. Cas. 193, where Lord Cairns L.C. said, at p. 197:

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 negligence may be reasonably inferred; the jurors have to say whether, from those facts, when submitted to them, negligence ought to be inferred. It is, in my opinion, of the greatest importance in the administration of justice that these separate functions should be maintained, and should be maintained distinct.

These words, though spoken in a civil case, were held to be applicable in a criminal case in this Court in R. v. Morabito, 1949 CanLII 1 (SCC), [1949] S.C.R. 172 at p. 174.

55. In other early cases the rule was stated and applied: see R. v. Dowsey (1866), 6 N.S.R. 93. This was a charge of murder committed on a ship at sea. The only issue on the appeal was whether the case should have been left to the jury. At page 119, Dodd J. said:

The facts of the case were for the jury, and if there was any legal evidence upon which they could assume the guilt of the prisoner, it was within their province to do so....If there was any legal evidence against a party charged with a criminal offence, the case cannot be withdrawn from the jury.

In R. v. Lloyd (1890), 19 O.R. 352, in the Ontario Supreme Court where evidence given by a rape victim was weak, MacMahon J. said, at p. 357:

So in the case in hand we may consider the evidence of the prosecutrix weak; we may, in view of the admissions made during the trial, consider some of the statements made against the prisoners Herman and George as improbable; but the probable truth or falsity of the statements was for the consideration of the jury, and although we may think they did not exercise a sound judgment in reaching their verdict, we cannot say there was no evidence to be submitted to them, and therefore the learned Judge was not bound to withdraw the case from the jury.

In this Court in Girvin v. The King (1911), 1911 CanLII 42 (SCC), 45 S.C.R. 167, the prima facie test was applied, at p. 169, in the words of Sir Charles Fitzpatrick C.J.:

I have always understood the rule to be that the Crown, in a criminal case, is not required to do more than produce evidence which, if unanswered, and believed, is sufficient to raise a prima facie case upon which the jury might be justified in finding a verdict.

56. It will be seen from the foregoing that the rule enunciated by Ritchie J. in Shephard has a long and respectable ancestry, firmly rooted in the concept that the functions of judge and jury are separate and distinct. The confusion which has arisen in this matter stems largely from the words of Sir Lyman Duff C.J. in R. v. Comba, supra, where, after expressing the view, at p. 397, that the rule in Hodge’s Case relating to the treatment of circumstantial evidence applied as part of the law of Canada, he said:

We have no doubt that the facts adduced have not the degree of probative force that is required in order to satisfy the test formulated by this rule; which is one that courts of justice in Canada are governed by and are bound to apply.

We agree with the majority in the Court of Appeal, whose reasons for their judgment we find convincing and conclusive, that the learned trial judge ought, on the application made by counsel for the prisoner at the close of the evidence for the Crown, to have told the jury that, in view of the dubious nature of the evidence, it would be unsafe to find the prisoner guilty, and to have directed them to return a verdict of acquittal accordingly. It is not, and could not, with any plausibility, be suggested that the case for the Crown was in any way strengthened or improved by the evidence put before the jury on behalf of the defence.

The appeal is dismissed.

The full significance of these words cannot be understood without a careful consideration of what happened at the trial and on the appeal.

57. The case concerned the murder of a young girl. Comba was charged and convicted. The case against him was entirely circumstantial and, while facts were revealed which could have been consistent with guilt, nothing was shown which was not equally consistent with innocence. The appeal to the Court of Appeal was heard by five judges (1938 CanLII 14 (ON CA), [1938] O.R. 200). Middleton J.A. wrote the principal judgment for the majority, with shorter concurring judgments by Masten, Fisher and Henderson JJ.A. Latchford C.J.A. dissented. The appeal was allowed. The conviction was quashed and the accused was acquitted. Middleton J.A. examined very closely the evidence led at trial. Before doing so, he referred to the rule in Hodge’s Case and noted that it had been approved in this Court in Fraser v. The King, 1936 CanLII 42 (SCC), [1936] S.C.R. 1. He referred to Rinfret J.'s quotation of the rule in that case from Wills, On Circumstantial Evidence, in these terms:

In order to justify the inference of guilt, the inculpatory facts must be incompatible with the innocence of the accused and incapable of any other reasonable hypothesis than that of guilt.

He added that it was not enough that the facts are consistent with guilt. Middleton J.A. then proceeded to examine closely the nine distinct points relied upon by the Crown to support its case, and he found in each case that the evidence was not inconsistent with Comba's innocence. He referred again to Fraser v. The King, 1936 CanLII 25 (SCC), [1936] S.C.R. 296, and quoted the words of Rinfret J. (as he then was), speaking for the Court in the Fraser case, at pp. 300‑01:

It would appear, therefore, that, when the evidence in a criminal case is purely circumstantial and at the same time equally consistent with the innocence as with the guilt of the accused, the Court of Criminal Appeal in England regards that evidence as insufficient to justify the jury in convicting, holds the verdict unsatisfactory and quashes the conviction, on the ground that it cannot be supported, having regard to the evidence.

Middleton J.A. then stated (at p. 212):

Upon this principle of law it is evident to the majority of the Court that this verdict must be set aside and the verdict of acquittal recorded.

The other judges of the majority agreed with Middleton J.A. Masten J.A. said, at p. 213:

I agree in the conclusion that the conviction must be quashed on the ground that the Crown's evidence is not incompatible with the innocence of the accused.

Fisher J.A. concluded his concurring reasons with the words (at p. 214):

My conclusion is that the evidence is as consistent with the accused's innocence as with his guilt, and that the jury was not justified in making the finding it did.

Henderson J.A. entirely concurred with Middleton J.A. and said, at p. 215:

The chain of events is consistent with his innocence and inconsistent with his guilt.

It is evident then that all four judges of the majority considered and applied the rule in Hodge's Case to the evidence which had been led at trial. They all found that the evidence was as consistent with innocence as with guilt. Thus, it will be seen that Comba, as decided in the Ontario Court of Appeal, is a ‘no evidence’ case and not inconsistent with the line of authority which culminates with Ritchie J. in Shephard. It is in this context that one must consider the words of Sir Lyman Duff C.J. in Comba in the Supreme Court of Canada. At page 397, he restated the test from Hodge’s Case in these words:

It is admitted by the Crown, as the fact is, that the verdict rests solely upon a basis of circumstantial evidence. In such cases, by the long settled rule of the common law, which is the rule of law in Canada, the jury, before finding a prisoner guilty upon such evidence, must be satisfied not only that the circumstances are consistent with a conclusion that the criminal act was committed by the accused, but also that the facts are such as to be inconsistent with any other rational conclusion than that the accused is the guilty person.

58. He then concluded that the facts did not have the probative force required to satisfy the rule. He expressed agreement with the majority of the Court of Appeal, whose reasons he found convincing and conclusive, that the trial judge should have told the jury that "in view of the dubious nature of the evidence, it would be unsafe to find the prisoner guilty, and to have directed them to return a verdict of acquittal accordingly". These are the words that from time to time have caused uncertainties. It has been suggested that the use of the word dubious' has given the trial judge a power or discretion to weigh and consider the quality of the evidence and to remove it from the jury's consideration. In my view, no such result occurred. The word dubious', as used here, must be considered in the context of the whole case and, particularly, the judgment of the majority of the Ontario Court of Appeal which Sir Lyman Duff C.J. specifically approved. The evidence was dubious only because in the application of the test in Hodge's Case it cast the balance neither way. It led to no result and left no evidence upon which a conclusion could be based. It is evident then that like the Court of Appeal, this Court treated Comba as a `no evidence' case. It is then consistent with past authority. Accordingly, there is no authority from this Court which is inconsistent with Shephard.

59. I am aware, of course, that in R. v. Knox, 1967 CanLII 692 (BC CA), [1968] 2 C.C.C. 348, and R. v. Sawrenko (1971), 1971 CanLII 1218 (YK CA), 4 C.C.C. (2d) 338, the British Columbia Court of Appeal and the Court of Appeal for the Yukon Territory held that, when faced with a motion for a directed verdict in a case dependent on circumstantial evidence, it was the duty of the trial judge to decide whether the evidence satisfied the rule in Hodge’s Case. Only if it did, was the case to go to the jury for determination of guilt or innocence. In this sense a weighing process or an assessment of quality was contemplated on the part of the judge. Those cases, however, were effectively overruled in this Court in R. v. Paul, 1975 CanLII 185 (SCC), [1977] 1 S.C.R. 181, and Lavoie v. The Queen, 1977 CanLII 212 (SCC), [1977] 1 S.C.R. 193, and even in circumstantial cases the law now is that any determination as to compliance with the rule in Hodge’s Case would be left to the jury.

60. The significance of all this is now somewhat limited because of this Court's decision in R. v. Cooper, 1977 CanLII 11 (SCC), [1978] 1 S.C.R. 860, which has placed the rule in Hodge’s Case into the general consideration of reasonable doubt. In any event on the law as it stood when Comba was decided, it is clear that Sir Lyman Duff C.J. was treating circumstantial evidence, which did not comply with the rule in Hodge's Case, as no evidence at all going to show guilt. In this, he was not disturbing the long line of authority against directed verdicts where there was some evidence of guilt.

61. Reliance was placed in argument upon the judgment of the Court of Appeal of Alberta in R. v. Duhamel, supra, and the English case of R. v. Turnbull, supra, upon which it is largely based. In Turnbull, Lord Widgery C.J., for the Court of Appeal, Criminal Division (Roskill, Lawton L.JJ., Cusack and May JJ.), stressed the dangers of injustice that can arise and in some cases have arisen because of errors made in visual identification of persons charged with crime. At the same time, he was at pains to make it clear that visual identification could not be eliminated as admissible evidence in criminal proceedings. He dwelt at length on the nature and scope of the directions which a trial judge should give to a jury which had to deal with evidence of visual identification, and he said that failure to give an adequate direction could result in convictions being quashed. In all, he was referring to what he described as the quality of the evidence. He said, at p. 552:

If the quality is good and remains good at the close of the accused's case, the danger of a mistaken identification is lessened; but the poorer the quality, the greater the danger.

It is impossible to disagree with Lord Widgery C.J. when he speaks of the danger of error in visual identification. Nobody could disagree with his assertion of the need for a careful and complete direction to the jury with regard to their treatment of such evidence. When, however, he introduces the suggestion that the trial judge should consider the quality of the evidence and, where he finds it wanting, take the case from the jury, he enters more controversial ground and authorizes the trial judge to encroach upon the jury's territory. Such a step blurs or even obliterates the clear line separating the functions of judge and jury. Questions of credibility and the weight that should be given to evidence are peculiarly the province of the jury. The term ‘quality’, as applied by Lord Widgery C.J., is really nothing more than a synonym for ‘weight’. To consider it, the trial judge exceeds his function. I am fully aware that untruthfulness on the part of a witness is not the only factor which may render his evidence weak or even useless. The most honest witness may be mistaken or, by external interference or distraction or by lack of opportunity to observe, the value of his evidence may be diminished or even eliminated so that it should receive little if any weight. All these considerations affect the weight of the evidence and the jury, twelve indifferent persons who must be unanimous to convict the accused, may with a careful direction from the trial judge take account of all these relevant factors and reach a conclusion with at least as much reliability as a single judge.

62. I do not adopt this position "out of deference to the verdict of a jury". I do so because it merely accords with the duty the law has imposed upon juries. 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.

63. Turning to the case at bar, it is acknowledged that there was direct evidence of identification. It should have been left to the jury, with a proper caution in the terms suggested in Turnbull. I would therefore dismiss the appeal and confirm the direction for a new trial. I agree with Hall J.A. that any question regarding the allegedly improper police procedures in connection with the later identification in the courtroom should be left for the new trial.

The reasons of Lamer and La Forest JJ. were delivered by

64. Lamer J. (dissenting)‑‑The issue in this appeal can be simply stated as follows: Did the trial judge err in law in directing a verdict of acquittal despite the fact that there was some evidence as to the identification of the accused?

65. The appellant was tried for rape before a jury in the Court of Queen's Bench of Manitoba. The accused at trial and throughout does not contest the fact that the victim was raped. He takes issue solely with her identification of him as her assailant. She was the only witness as regards that issue. At the close of the Crown's case, Wright J. granted a motion for a directed verdict of acquittal. The Court of Appeal, one judge dissenting, allowed the Crown's appeal and ordered a new trial: (1983), 1983 CanLII 3697 (MB CA), 35 C.R. (3d) 272. The accused comes to this Court as of right under s. 618(2)(a) of the Criminal Code.

The Facts

66. On June 13, 1982, at approximately 12:45 a.m., the complainant was walking down Water Avenue in Winnipeg. She was attacked by a man who grabbed her around the neck and pulled her behind the C.N.R. sheds on Water Avenue, where he proceeded to sexually assault her. The complainant managed to escape and ran to the street where she was rescued by a passing motorist. According to the complainant, the entire incident took about twenty minutes. The complainant stated that before being dragged behind the shed she had an opportunity to see the man's face when he was beside her and that she had a front view of his face long enough to see him. She admitted it was dark behind the shed area but said that she was still able to see her attacker's face.

67. The most comprehensive summary of the identification evidence is to be found in Matas J.A.'s reasons in the Court of Appeal at pp. 278‑81:

The complainant described her assailant to the police on three occasions. The first description was given in the early morning of 13th June, a short time after the attack. Later in the day, at about 2:00 p.m., the complainant gave a description to a different police officer. And on 17th June, she described her assailant to a police artist in order to have a sketch prepared.

At the first interview, according to the police officer, the attacker was described as male, white, 26 years old, with a chubby build, squinty eyes, a large mouth and dark short hair. (The police officer noted this description of the assailant's hair because the complainant said that the hair came to the top of the assailant's ears.) The assailant was said to be wearing a blue coat, blue jeans and suede shoes.

At the second interview, the police officer was told that the assailant was male, white, 26 years of age, 5 feet 9 inches to 5 feet 10 inches, with a heavy build, chunky and muscular, squinty eyes, that he had dark hair just below the ears, and that he was wearing blue jeans, dark‑coloured shoes, a light‑coloured T‑shirt, and a blue three quarter‑length smock‑type jacket with large patch pockets on each side.

The complainant did not say anything about eye colour or skin tone or colour. At the trial, she admitted she would not have been able to give that information because the attacker was always in a dark area and it was difficult to tell colours.

The description given by the complainant to the police artist at the opening of their session was that the assailant was male, Caucasian (with ‘some Indian mix’), 27 years of age, 5 feet 10 inches, over 200 pounds and muscular, and that he was wearing a three‑quarter‑length blue jacket, a beige‑coloured T‑shirt, jeans and "hush‑puppy"‑style shoes.This was the first time actual weight was mentioned and was the first reference to Indian background. Further discussion between the artist and the complainant resulted in the sketch which was filed in court as Ex. 1. The complainant said she thought that the sketch resembled her attacker but she had a feeling something was missing. At the trial, the artist said Mr. Mezzo did not appear to have an Indian background.

Mr. Mezzo was arrested on 27th June, 1982, about two weeks after the attack on the complainant in this case. The complainant and two police officers testified about the in‑court or docket identification of Mr. Mezzo which took place on 28th June and 30th June, 1982.

It was not suggested to the complainant that she take part in an identification line‑up.

Constable Steinthorson telephoned the complainant on 27th June to ask her to come to the Public Safety Building the next day for the courtroom identification. The complainant met Constable Steinthorson at 9:00 a.m. the next morning. Constable Steinthorson suspected Mr. Mezzo was her assailant. He knew Mr. Mezzo well enough to know what he looked like and to identify him. He told the complainant that they would be doing a courtroom identification; she was to accompany him to the courtroom, where a number of persons would appear. The complainant expected to see her attacker, but she did not know when. She also knew he was in jail.

At about 10:00 a.m. the complainant went into the courtroom. She sat on the left side of the public gallery, in the third row, with a rape crisis worker and Constable Steinthorson. The prisoners came into the courtroom from a door on the left. At about 10:45 a.m., when Mr. Mezzo, the fifth prisoner, came in and sat down, the police officer saw the complainant react visibly by shaking and moving from side to side. Mr. Mezzo was about 20 to 25 feet away from the complainant. He was in the courtroom about 90 seconds. Shortly after Mr. Mezzo's appearance, the complainant left the courtroom at the suggestion of the police officer.

The complainant told the police officer that Mr. Mezzo looked like her assailant but she was not sure. Nevertheless, Constable Steinthorson felt that the complainant had identified Mr. Mezzo as her attacker.

The complainant explained to the court that she had not been sure of the identification because her view of Mr. Mezzo was obstructed by a sheriff who was standing beside Mr. Mezzo. She said she did not get a good look at Mr. Mezzo for that reason.

The police were unable to say whether Mr. Mezzo's name was called, or whether the offence he was charged with was mentioned. No evidence was available about who the previous prisoners were or what they looked like. Constable Steinthorson agreed on cross‑examination that: `So as far as we know, the only individual who came forward that morning who was of the rough description that the complainant had originally given was Michael Mezzo.'

Later, the complainant received a call from Constable Legge, who said he wanted her to come in and identify some people. Constable Legge had become involved in the case on 27th June. In his direct examination he explained his purpose in this way:

"I indicated to her at that time that in the course of our investigation we had a suspect in the matter of the attack upon herself and we wished her to accompany myself to a courtroom on that day and to view all the persons in the courtroom that day to see if there is anything she could tell me about anyone in the courtroom."

On 30th June, Constable Legge and the complainant seated themselves in the courtroom in the first row, but this time on the right‑hand side. They were 30 to 35 feet away from the prisoner. Although Constable Legge said he had asked the Crown attorney to arrange not to call Mr. Mezzo's name, the name was called. By coincidence, Mr. Mezzo was again the fifth person to come out of the one door into the courtroom. The complainant positively identified Mr. Mezzo as her assailant.

At the preliminary hearing, the complainant identified Mr. Mezzo as her assailant. The only people in the courtroom on that occasion were the judge, the Crown attorney, Mr. Mezzo's lawyer and Mr. Mezzo.

At the trial, the complainant identified a photograph of Mr. Mezzo, taken in June 1982, and identified the accused in the box. She described her assailant as 5 feet 8 inches to 5 feet 9 inches, 185 to 200 pounds, husky, 26 years of age, with dark black hair, ear‑length, an oval face, squinty eyes, a long and pudgy nose, a full mouth and lips, and dark tone skin similar to native in racial origin. She said he had been wearing a beige T‑shirt, blue jeans, dark suede shoes, and a navy blue jacket, three‑quarter length, like a hip jacket, of heavy‑weight nylon.

The description of the jacket corresponded to the jacket seized from Mr. Mezzo and filed in court as an exhibit. It was different from her earlier descriptions. And the description of Mr. Mezzo at the trial was more in accord with his actual appearance in June 1982 than the first descriptions given by the complainant to the police.

The Law

68. This Court's latest pronouncement as regards the test to be applied by a judge when deciding a motion for a directed verdict is to be found in the reasons of Ritchie J., writing for the majority in United States of America v. Shephard, 1976 CanLII 8 (SCC), [1977] 2 S.C.R. 1067. It was there reiterated that it was the same test as that to be applied by a Justice of the Peace when deciding whether to commit an accused to trial. That test is set out in s. 475 of the Criminal Code as being whether "in his opinion the evidence is" (or is not) "sufficient to put the accused on trial". Before this Court's decision in Shephard, that test had been interpreted in three different ways: the probable guilt test, the prima facie case test where the judge cannot weigh the evidence at all to determine sufficiency, and the prima facie test where the judge does weigh the evidence.

69. The Shephard decision has put to rest the debate between probable guilt as against the prima facie test, and comes down on the side of the proponents of the prima facie test. But while there was agreement as regards the exclusion of the probable guilt test, the Court divided as to the formulation of the prima facie test and as to whether and to what extent the evidence could be weighed.

70. Ritchie J. stated that the test laid down in R. v. Comba, 1938 CanLII 7 (SCC), [1938] S.C.R. 396, was limited to cases where the evidence was circumstantial and the test he laid down for the majority as to the "sufficiency" of the evidence is as follows, at p. 1080:

... whether or not there is any evidence upon which a reasonable jury properly instructed could return a verdict of guilty. The `justice', in accordance with this principle, is, in my opinion, required to commit an accused person for trial in any case in which there is admissible evidence which could, if it were believed, result in a conviction.

71. In Shephard, an extradition case, the evidence was that of an accomplice who gave affidavit evidence in which he acknowledged that, if he co‑operated and gave successful evidence against Shephard to the satisfaction of the American prosecuting authorities, all charges against him would be dropped. It was common ground, that the test to be applied by a judge in extradition matters is the same as that of a judge sitting at a preliminary inquiry or a judge deciding a motion for a directed verdict. I might mention that some cases have suggested that an extradition judge's discretion may be more restrictive, but this has no bearing on this case and I need not enter into that issue.

72. In Shephard, Ritchie J. said that (at p. 1084) "Credibility in my view must always be a question for the jury" and not the judge, and (at p. 1087) "I cannot accept the proposition that a trial judge is ever entitled to take a case from the jury and direct an acquittal on the ground that, in his opinion, the evidence is `manifestly unreliable'."

73. The minority said the test was that which had been set out in R. v. Comba, supra. Spence J. writing for the minority said at p. 1076:

I am of the opinion that the trial judge, upon considering whether he should direct a verdict at the close of the Crown's case, has a discretion to direct a verdict not confined only to those cases where the evidence was wholly circumstantial. Rather, his discretion is, in the words of Chief Justice Duff, ‘whether the evidence is of such a dubious nature as to be dangerous’. In Puerto Rico v. Hernandez, supra, the evidence was dubious and dangerous to use as a basis for conviction because the witness purported to make observations which the learned trial judge believed were impossible. In the present case, the evidence is dangerous and dubious because it was given by a witness who was quite evidently acting in hope of a reward which had been promised to him in detail. I use Thurlow J.'s words in Hernandez, ‘I find it inconceivable that a person should be put on trial on such flimsy evidence ...’.

74. At first blush, Ritchie J.'s unqualified reference to "credibility" and "unreliability" could support the proposition that the weight to be given to a witness' testimony is the exclusive realm of the jury. But this interpretation does not survive a closer analysis of his reasons in Shephard.

75. The probative value, the weight to be given to a witness' testimony, is resultant of an assessment of (1) the witness' sincerity, honesty, and, (2) the witness' capacity to be accurate. When Ritchie J. in Shephard, refers to a witness' credibility, or to the reliability of the evidence, I take it, given the issue in that case, that he was referring to the witness' honesty, not to the witness' capacity to be accurate or not to be honestly mistaken.

76. For this interpretation I find support in Ritchie J.'s comments on the Federal Court of Appeal's decision in the case of Commonwealth of Puerto Rico v. Hernandez, [1973] F.C. 1206. In Hernandez the evidence was identification and what was then at issue was not the witness' honesty but the witness' accuracy. Ritchie J. said at pp. 1087‑88:

I also differ from the learned Chief Justice of the Federal Court in that I consider there to be a clear distinction in principle between this case and that of Hernandez. In the latter case Mr. Justice Thurlow characterized the evidence which the extradition judge had found to be insufficient to justify the issuing of a warrant for extradition, in the following paragraph:

After having read and re‑read the evidence of the witnesses, Atilano and Ortiz and after hearing the very able argument of counsel for the Commonwealth of Puerto Rico I am not persuaded that the evidence was reasonably sufficient to warrant putting the respondent on trial. Indeed, I find it inconceivable that a person should be put on trial on such flimsy evidence as a purported identification made a year after the event by a person who did not previously know the accused and whose only opportunity to observe him was a fleeting one from a distance of some sixty feet, if indeed he ever got that close. The evidence of the witness having selected the respondent's identification card is not in itself evidence implicating the respondent, and in the circumstances as described it tends to weaken rather than strengthen the identification made by Atilano at the hearing ... Nor was there anything else to implicate the respondent or corroborate such identification.

It thus appears to me clear that in that case it was held that there was not enough evidence to support a prima facie case against the accused and in fact that it was `inconceivable that a person should be put on trial on such flimsy evidence'. It was upon this ground that the extradition judge in that case based his opinion as to insufficiency.

Ritchie J. thus distinguishes Shephard from Hernandez by specifically referring to the witness' capacity to be accurate.

77. When commenting on Comba, Ritchie J. said at pp. 1081‑82:

It is to be observed that there was no suggestion in the Comba case that any of the evidence called by the Crown was either tainted or unreliable. It came from witnesses whose integrity was at no time put in question and who testified as to a variety of circumstances which had excited enough suspicion against the accused to occasion his arrest and trial but which taken together did not establish his guilt in accordance with the accepted standards of proof in such cases which are described by Duff C.J. in the following passage:

It is admitted by the Crown, as the fact is, that the verdict rests solely upon a basis of circumstantial evidence. In such cases, by the long settled rule of the common law, which is the rule of law in Canada, the jury, before finding a prisoner guilty upon such evidence, must be satisfied not only that the circumstances are consistent with a conclusion that the criminal act was committed by the accused, but also that the facts are such as to be inconsistent with any other rational conclusion than that the accused is the guilty person.

(Emphasis added.)

Here again Ritchie J. focuses on the distinction between accuracy and honesty, this time by highlighting in the Comba test the fact that the honesty of the witnesses, their "integrity", to use his words, was not in issue and therefore was not being weighed by the judge.

78. The distinction between a witness' truthfulness and a witness' accuracy has sometimes been made by referring to the former as the witness' credibility, and to the latter as the quality of the witness' testimony. (See amongst others, R. v. Turnbull, [1976] 3 All E.R. 549 (C.A.); R. v. Sutton, 1969 CanLII 497 (ON CA), [1970] 2 O.R. 358 (C.A.); R. v. Spatola, 1970 CanLII 390 (ON CA), [1970] 4 C.C.C. 241 (Ont. C.A.); R. v. Duhamel, 1980 ABCA 247, [1981] 1 W.W.R. 22 (Alta. C.A.); R. v. Atfield (1983), 1983 ABCA 44, 42 A.R. 294 (C.A.))

79. My reading of the majority reasons in Shephard is that when assessing the sufficiency of the evidence, that is when determining whether "there is any evidence upon which a reasonable jury properly instructed could return a verdict of guilty", the judge cannot assess the witness' honesty but may assess the witness' capacity to be accurate and his or her resulting reliability. If I am wrong in reading that distinction in Ritchie J.'s reasons, then I propose that this Court find that to be the proper test. This is not a proper case for reconsidering Shephard with a view to adopting the views of the minority; we have not been invited to do so, nor do the issues in this case raise the matter squarely.

80. "Quality" of evidence is truly the realm of the trial judge. I could not put matters better than by quoting in support of this proposition what Lord Devlin said in Trial by Jury (1966), at pp. 62‑63, speaking of the trial judge's function:

... if he is not satisfied that there is evidence on which reasonable men could act, he must not let the case go to the jury. Again, so stated, the rule might seem to be a commonplace. If there is not evidence upon which reasonable men could act, it must be immaterial whether the case is left to the jury or not since their verdict could only be negative. So the rule would be otiose were it not for the fact that judges and juries can take a different view of what reasonable men should think. When there is such a difference, it is not, as might be supposed, the view of the jury who ex officio are the reasonable men that prevails, but the view of the judge. This understood, the rule can be restated in its true terms: the verdict cannot be supported unless a reasonable lawyer thinks that there is evidence upon which reasonable men could act. There can be no doubt about the difference between the reasonable lawyer and the reasonable man. When twelve men unanimously return a certain verdict, they must believe that there is evidence to support it. When the judge rejects their verdict, he is not really telling them that not a single one among them is a reasonable man. What he is telling them is that there is a certain minimum of evidence which the law requires and that as a trained lawyer he is a better judge than they are of what that minimum should be. What the minimum should be is not a question of law but a question for lawyers.

(Emphasis added.)

81. When applied to circumstantial evidence this would mean that the question as to whether and to what extent the evidence is dubious will be determined assuming the witnesses are being truthful. The question would be: "Would it be unsafe to convict?" This would involve, given the particular nature of circumstantial evidence, asking whether innocence can be logically and reasonably excluded. Morris J. Fish, in "Committal for Trial: ‘Some’ Evidence is Not ‘Sufficient’ " (1979), 39 R. du B. 607, at p. 618, suggests the following test, which I would adopt:

If guilt and innocence are equally consistent with the facts, there is no evidence. If guilt is more probable than innocence, there is some evidence, but unless innocence can be logically and reasonably excluded, the evidence is insufficient and the case should end with a discharge, non‑suit or directed verdict. Finally, if innocence can be logically and reasonably excluded, the evidence is sufficient and may be considered by the trier of fact. Conviction will not automatically ensue. The trier of fact, it need hardly be repeated, will only convict if ‘convinced beyond a reasonable doubt that the guilt of the accused is the only reasonable explanation of the facts established by the evidence’.

It has somewhere been said that semantics teaches us to take our exercise in other ways than by jumping to conclusions. In this vein, there is a purpose to the foregoing discussion. A justice or judge should not jump to the conclusion that there is evidence for a jury to consider whenever the facts are consistent with guilt. He is bound to consider whether guilt can (not, of course, will) properly and reasonably be inferred from the evidence. And he must bear in mind that reasonableness is a question for him, not the jury:...

82. When applied to testimonial evidence this approach involves an assessment by the judge of the circumstances under which the witness purports to be, not truthful (as in Shephard), but accurate (as in Hernandez and in this case). The reliability of the witness' accuracy may be undermined by the circumstances under which the witness was called upon to register what he later on relates in the belief that it was and still is accurate. This might be, when considering the accuracy of identification testimony, the fact that the witness had only "a fleeting glance", or had a "longer observation" but one "made in difficult circumstances"; or again the result of improper police identification proceedings that would tend to induce the witness into mistakingly identifying someone.

83. A final observation. The only difference in my view between the majority and the minority in Shephard is whether the judge can assess the truthfulness of witnesses when assessing whether there is sufficient evidence. With respect, I fail to see any essential difference between the test expressed in Comba, and that expressed by the majority in Shephard. The Comba formulation focuses on the nature of the evidence (dubious) and the resulting unsafe conviction, while the Shephard test focuses on the unreasonableness of the jury in founding a guilty verdict on the evidence of a given case. A reasonable jury that was properly instructed (the Shephard test), would never return a verdict of guilty "where the only evidence is of such dubious nature" that the verdict of guilty would be "unsafe" (the Comba test). I personally prefer the Shephard formulation of the test. It is expressed in terms that apply to all kinds of evidence while the Comba formulation more easily befits the kind of assessment one would make when considering circumstantial evidence. What really divided the Court in Shephard when assessing the quality of the evidence was the fact that the minority was of the view that the judge could, whatever the formulation of the test, consider the affiant's honesty while the majority found that he could not. Now to come back to this case.

The Trial Judge

84. It is the appellant's position that the trial judge in the case under appeal applied the test laid down by the majority in Shephard and that he did not, nor did he need to on the facts of this case (as would have been permissible under certain circumstances had the opinion of the minority in Shephard prevailed in any way), consider the witness' underlying credibility. He addressed the problem, says the appellant, by considering the quality of the identification evidence, not by weighing the witness' truthfulness.

85. The trial judge's decision to grant the motion for a directed verdict of acquittal rested on the finding that the identification evidence was of too poor a quality to be relied upon. This finding as regards the quality of that evidence is the resultant of two other findings of fact: the difficult conditions under which the victim made her observations as regards the identity of her aggressor and the improper police identification procedures. He assumed she was being truthful and withdrew the case because it was too dangerous to rest a conviction on her testimony, honest as it might well have been. This I conclude from a general reading of his reasons given in Court to the parties in the absence of the jury, and more particularly the following passage:

I am influenced significantly as to the law by the unanimous judgment of the Alberta Court of Appeal in R. v. Duhamel, 1980 ABCA 247, 1981, 1 W.W.R., 22 wherein Mr. Justice Harradence quotes approvingly and at length from the judgment of the Court of Appeal in England in R. v. Turnbull, 1976, 3 All England Reports, 549. In particular I refer to pages 28 and 29 of the Duhamel judgement wherein Mr. Justice Harradence sets out the principles expressed in Turnbull relative to the potential for miscarriage of justice based on visual identification. Towards the end of the statement of those principles the following declaration is made and I would say is made in no uncertain terms:

"When, in the judgment of the trial judge, the quality of the identifying evidence is poor, as for example when it depends solely on a fleeting glance or on a longer observation made in difficult conditions, the situation is very different. The judge should then withdraw the case from the jury and direct an acquittal unless there is other evidence which goes to support the correctness of that identification."

I miss out a line which I do not think is relevant and continue:

"For example, X sees the accused snatch a woman's handbag; he gets only a fleeting glance of the thief's face as he runs off but he does see him entering a nearby house. Later he picks out the accused on an identity parade. If there was no more evidence than this, the poor quality of the identification would require the judge to withdraw the case from the jury;"

Assistance in determining the measurement of poor quality evidence as identified by the court in Turnbull can be found in the paragraph immediately preceding the portion that I have just quoted. This paragraph reads:

"All these matters go to the quality of the identification evidence. If the quality is good and remains good at the close of the accused's case, the danger of a mistaken identification is lessened: but the poorer the quality, the greater the danger. In our judgment, when the quality is good, as for example when the identification is made after a long period of observation, or in satisfactory conditions by a relative, a neighbour, a close friend, a workmate and the like, the jury can safely be left to assess the value of the identifying evidence even though there is no other evidence to support it; provided always, however, that an adequate warning has been given about the special need for caution.Were the courts to adjudge otherwise, affronts to justice would frequently occur."

(Emphasis added.)

86. It is clear from this passage that the judge was addressing his mind to an assessment of the quality of the identification evidence. With "quality" in mind, he then considered the circumstances of the observation and the police identification procedures. As regards the conditions under which the observations were made, he said:

Now, it appears to me that it follows from this statement that where the evidence of identification is dependent solely on a witness who had no previous knowledge of an accused person and whose opportunity for observation has been in difficult circumstances, that evidence, in the view of the English Court of Appeal, cannot be safely left with the jury.

In the present case I am satisfied there is no evidence going to the correctness of the identification other than that emanating from the complainant. The complainant had only a very limited and restrictive opportunity to observe her assailant. Although she may have been in his presence for as long as 20 minutes, the circumstances were such that her observations of the assailant were made in very difficult conditions. The place where she was first apprehended provided minimum lighting and the place behind the CNR shed where the attack proceeded and where it is evident most of the 20 minutes, if it was that, was spent, was very dark indeed. The complainant's description of her assailant, although matching the accused in many respects, is, nevertheless, one that could match any number of persons and contains no special identifying characteristics. I have concluded that these facts fall squarely within the directions in Turnbull, approved by the Alberta Court of Appeal, and even where there has been a proper or adequate identity parade, require that the case be taken from the jury.

As regards the police procedure, Wright J. said:

In the present case there is the additional factor that I do not believe the procedure followed by the police in the subsequent identification of the accused by the complainant at the police station was proper and that fact has further damaged the complainant's identification evidence.

(Emphasis added.)

He reviewed the authorities on point, made reference to R. v. Faryna (1982), 1982 CanLII 3796 (MB CA), 18 Man. R. (2d) 185 (C.A.), and read out the following passage in R. v. Smierciak (1946), 1946 CanLII 331 (ON CA), 87 C.C.C. 175 (Ont. C.A.) at p. 177, to which the court in Faryna referred:

... of the utmost importance, is the method used to recall or refresh the recollections of a witness who is to be relied upon to identify a person suspected of wrongdoing or who is under arrest. If a witness has no previous knowledge of the accused person so as to make him familiar with that person's appearance, the greatest care ought to be used to ensure the absolute independence and freedom of judgment of the witness. His recognition ought to proceed without suggestion, assistance or bias created directly or indirectly. Conversely, if the means employed to obtain evidence of identification involve any acts which might reasonably prejudice the accused, the value of the evidence may be partially or wholly destroyed. Anything which tends to convey to a witness that a person is suspected by the authorities, or is charged with an offence, is obviously prejudicial and wrongful. Submitting a prisoner alone for scrutiny after arrest is unfair and unjust.

(Emphasis added.)

Wright J. then concluded on this latter point as follows:

So it is clear I think from the comments that I have made, the observations I have expressed, that I am unable to find that the identification procedure followed reasonably meets the standards expressed in the law that the Manitoba Court of Appeal has approved. So this then becomes an added reason for concluding, as I have, that the motion of the defence should be accepted in favour of a directed verdict of acquittal and that is what I now propose to do. So we will call the jury in.

(Emphasis added.)

When the jury returned, Wright J. gave them only a summary of the law in support of his reasons for his decision, as is apparent from his first remarks to them:

I am going to read you a small portion of the law that is applicable in order that you will understand what is taking place. The law, you will recall, I am the one who decides as to its correctness or not and which I am applying at this time is as follows:

"When in the judgment of the trial judge the quality of the identifying evidence is poor as, for example, when it depends solely on a fleeting glance or on a longer observation made in difficult conditions, the judge should then withdraw the case from the jury and direct an acquittal unless there is other evidence which goes to support the correctness of the identification."

That statement of the law is made in the context of the issue of identification. And as you all know that really is the issue that we have narrowed this case down to.

I have reviewed the evidence carefully that has been introduced by the Crown and you have heard the Crown indicate that the Crown's evidence was complete earlier. I have concluded as a matter of law that the circumstances are such as to fall squarely within the framework of that statement that I just quoted to you. That means that at this point I have to withdraw the case from you and direct this verdict of acquittal.

(Emphasis added.)

87. I mention the difference between his remarks to the jury and his reasons given previously in their absence, because the former appear, when read out of context, to be elevating the "fleeting glance" or "difficult conditions" situations to the level of a test, whilst they are but relevant factual elements upon which a conclusion as regards the quality of the evidence will be based.

The Court of Appeal

88. In his dissenting reasons Matas J.A. agreed with the trial judge and would have directed a verdict of acquittal. He did express some concern as regards a possible misconstruction by the trial judge of the applicable test at pp. 282‑83:

In my respectful opinion, Wright J. takes the applicable principles expressed in Turnbull too far if he meant to say that a judge should take a case from a jury whenever identification of an accused is dependent solely on an identifying witness who has no previous knowledge of the accused and whose opportunity for observation has been in difficult circumstances. Many crimes of violence, for example, rape or robbery, are committed in the absence of a witness other than the victim and in the absence of independent corroborating evidence. And it would be rare if the circumstances of the offence would not be difficult. Depending on the circumstances of the case, it may be that the only just course would be to direct a verdict of acquittal, as I think it was here. But the decision to direct a verdict of acquittal must depend on the circumstances of the case.

(Emphasis added.)

89. Of course he is right, given the hypothesis he postulates. The majority shared this concern. But the essence of their decision was different and is found at p. 277 where Hall J.A., writing for himself and Monnin C.J., stated that the error of law committed by the trial judge was the misapplication of the proper test to the facts of the case.

It is very easy to attribute to evidence the quality of being dubious, and therefore a view to that effect should be arrived at with restraint and not lightly if we are to have due regard for the process of trial by jury. In my view, the evidence of identification in the present case was sufficient for the jury to have passed on it after proper directions and instructions by the presiding judge. In that respect, the error of law of the trial judge was not so much in adopting the wrong test but in misapplying the test in Turnbull, supra, to the facts of the case. The preferable language to employ is that set out in Comba, supra, and adopted in Duhamel, supra.

90. It appears to me, when reading all of Wright J.'s reasons, that he applied the proper test, found the quality of the evidence such that it would be unsafe to rest a conviction upon it, and directed a verdict of acquittal. The manner in which he referred to "the fleeting glance" and "difficult conditions" situations offers some footing for the conclusion that he turned the examples into a test. But the context in which these references were made was one where he specifically referred in general terms to the "quality" of the identification as being the purpose of the scrutiny. Added to that is the fact that it is obviously the "quality test" that was on his mind when he addressed the question whether the improper police procedure had "partially or wholly destroyed" the "value of the evidence". The majority of the Court of Appeal found that it would not have been unsafe to convict, while Matas J.A. shared the trial judge's view that it would. This did not in my view raise a question of law alone in the Court of Appeal, and the Crown's appeal should have been dismissed.

91. I would therefore allow this appeal and restore the acquittal.

Appeal dismissed, Lamer and La Forest JJ. dissenting.

Solicitors for the appellant: Wolch, Pinx, Tapper, Scurfield & Wyant, Winnipeg.

Solicitor for the respondent: The Department of the Attorney General, Winnipeg.