R. v. C.L.Y., 2008 SCC 2, [2008] 1 SCR 5

                                                 SUPREME COURT OF CANADA

Between:

C.L.Y.

Appellant

v.

Her Majesty the Queen

Respondent

Coram: Bastarache, Binnie, Deschamps, Fish, Abella, Charron and Rothstein JJ.

______________________________

R. v. C.L.Y., [2008] 1 S.C.R. 5, 2008 SCC 2

C.L.Y.                                                                                                                                  Appellant

v.

Her Majesty The Queen                                                                                                 Respondent

Indexed as:  R. v. C.L.Y.

Neutral citation:  2008 SCC 2.

File No.:  31811.

2007:  November 6; 2008:  January 25.

Present:  Bastarache, Binnie, Deschamps, Fish, Abella, Charron and Rothstein JJ.

on appeal from the court of appeal for manitoba

Criminal law — Evidence — Credibility — Burden of proof — Accused convicted of two counts of sexual assault — Whether trial judge misapplied burden of proof — Whether trial judge misapprehended and mischaracterized evidence leading to convictions.

C.L.Y. was a teenager charged with three counts of sexual assault.  The charges arose out of allegations of inappropriate touching when he was babysitting a girl who was at the time between six and eight years old.  The trial judge rejected C.L.Y.’s evidence and convicted him of two counts of sexual assault.  He was acquitted of the third charge.  The convictions were upheld by the Court of Appeal.  The majority of the court rejected C.L.Y.’s argument that the trial judge, by considering the evidence of the girl first, had shifted the burden of proof.  The dissenting judge concluded that the trial judge’s acceptance of the complainant’s evidence before even considering that of C.L.Y. resulted in at least a subtle shift of the onus of proof.

Held:  The appeal should be allowed.

Per Bastarache, Abella, Charron and Rothstein JJ.:  The paramount question is whether, on the whole of the evidence, the trier of fact is left with a reasonable doubt about the guilt of the accused.  The key is whether the correct burden and standard of proof were applied, not what words were used in applying them.  What R. v. W. (D.), [1991] 1 S.C.R. 742, offered was a helpful map, not the only route.  Its purpose was to ensure that triers of fact understand that the verdict should not be based on a choice between the accused’s and Crown’s evidence, but on whether, based on the whole of the evidence, they are left with a reasonable doubt as to the accused’s guilt.  In this case, it is difficult to see how the sequence in which the trial judge set out her findings of credibility can be said to undermine her articulated and correct statement of the law, or demonstrate that she lost sight of a paramount legal principle like reasonable doubt.  Her reasons revealed that she understood that a finding that the girl was credible did not mean that the onus shifted to the accused to show that he was not guilty. [6‑8] [12]

The flaw in this case which does lead to concerns about the fairness of the trial relates to fundamental misapprehensions and mischaracterizations of the evidence leading to the convictions.  In view of the record of this case, it is clear that the trial judge’s bases for disbelieving C.L.Y. rested on misapprehensions of his evidence and played a critical role in the convictions, rendering them insupportable.  In the circumstances, the convictions should be quashed and a new trial ordered. [13] [21‑22]

Per Binnie, Deschamps and Fish JJ.:  This appeal should be allowed not only because the trial judge rejected the evidence of the accused for reasons that are not supported by the record, but also because she misapplied the burden of proof.  This is precisely the kind of case where a departure from the procedure set out in R. v. W. (D.), [1991] 1 S.C.R. 742, is fraught with particular danger.  The very purpose of adhering to that procedure is to foreclose an inadvertent shifting of the burden of proof where the complainant and the accused have both testified and the outcome of the trial turns on their credibility as witnesses.  Mere reference to this framework will not suffice in all cases and a correct statement of the burden of proof can scarcely save its evident misapplication.  In this case, the risk of an inadvertent shift in the burden of proof materialized, as the trial judge accepted the evidence of the complainant without taking into account at all the contrary evidence of the accused.  As a matter of law, the accused was presumed innocent; as a matter of fact, his fate had been sealed without any consideration at all of his evidence under oath at trial. [23] [25-27] [32]

Cases Cited

By Abella J.

Considered:  R. v. W. (D.), [1991] 1 S.C.R. 742; referred to:  R. v. Morin, [1988] 2 S.C.R. 345; R. v. S. (W.D.), [1994] 3 S.C.R. 521; R. v. Challice (1979), 45 C.C.C. (2d) 546; R. v. MacKenzie, [1993] 1 S.C.R. 212; R. v. Levasseur (1994), 89 C.C.C. (3d) 508, rev’d [1994] 3 S.C.R. 518; R. v. Boucher, [2005] 3 S.C.R. 499, 2005 SCC 72; R. v. Minuskin (2003), 68 O.R. (3d) 577; R. v. Chittick (2004), 228 N.S.R. (2d) 81, 2004 NSCA 135; R. v. Morrissey (1995), 97 C.C.C. (3d) 193; R. v. Lohrer, [2004] 3 S.C.R. 732, 2004 SCC 80; R. v. W. (R.), [1992] 2 S.C.R. 122; R. v. Gagnon, [2006] 1 S.C.R. 621, 2006 SCC 17.

By Fish J.

Referred to:  R. v. W. (D.), [1991] 1 S.C.R. 742; R. v. Morrissey (1995), 97 C.C.C. (3d) 193.

APPEAL from a judgment of the Manitoba Court of Appeal (Twaddle, Monnin and Freedman JJ.A.), [2006] 12 W.W.R. 385, 208 Man. R. (2d) 202, 383 W.A.C. 202,  213 C.C.C. (3d) 503, 43 C.R. (6th) 81, [2006] M.J. No. 398 (QL), 2006 CarswellMan 363, 2006 MBCA 124, affirming the accused’s conviction on two counts of sexual assault.  Appeal allowed.

Alan J. Semchuk, for the appellant.

Richard A. Saull and Brian R. Bell, for the respondent.

The judgment of Bastarache, Abella, Charron and Rothstein JJ. was delivered by

[1]                              Abella J. — C.L.Y. was a teenager charged with three counts of sexual assault.  The charges arose out of allegations of inappropriate touching when he was babysitting.  At the time of the incidents, he was between the ages of around 16 to 18. The girl was between 6 and 8 years old.  C.L.Y. denied any inappropriate touching.

[2]                              The trial judge rejected C.L.Y.’s evidence and convicted him of two counts of sexual assault (2005 CarswellMan 562).   He was sentenced to 18 months’ probation. He was acquitted of a third charge.

[3]                              The convictions were upheld by the Manitoba Court of Appeal ((2006), 208 Man. R. (2d) 202, 2006 MBCA 124).  The majority rejected C.L.Y.’s argument that the trial judge, by considering the evidence of the girl first, had shifted the burden of proof contrary to this Court’s instruction in R. v. W. (D.), [1991] 1 S.C.R. 742.  The majority’s conclusion on the issues relating to the assessment of credibility are:

In the present case, it is clear that the judge understood that her decision on credibility did not mean the accused was guilty. Her reasons disclose a full understanding of the two separate findings she was making. She did not leap from her credibility finding directly to conviction. She applied the principles regarding reasonable doubt.

. . .

In the case at bar, the judge provided a clear and concise analysis for her finding of credibility with respect to the accused. The accused complains that the analysis is flawed and should be disregarded because of the judge’s reliance on the complainant’s excellent recall of events to discredit his credibility. That in my view goes not to the lack or sufficiency of reasons being provided but more to the reasoning process itself. It is not, in my view, for an appellate court to question the thought or reasoning process that a trial judge utilizes in arriving at a decision unless it is clearly in error. Such is not the case here. [paras. 14 and 16]

[4]                              Twaddle J.A. dissented. He concluded that despite the trial judge’s articulated assurance that she was aware of the principles in W. (D.), there was a “subtle shift” of the onus of proof:

[T]he acceptance of the complainant’s evidence before even considering that of the accused must, as I see it, result in at least a subtle shift of the onus of proof. Having accepted the complainant’s evidence, the trial judge will inevitably find the accused’s evidence less believable than it would have been before the complainant’s evidence was accepted. [para. 32] 

He also found that it was unreasonable for the trial judge to “disbelieve the accused’s denials of guilt solely on the ground of his having too good a recollection of events” (para. 37).  He would have overturned the convictions and directed verdicts of acquittal.

[5]                              While I disagree with Twaddle J.A.’s view that the trial judge’s failure to follow the steps articulated in W. (D.) resulted in “a subtle shift of the onus of proof”, I am of the view that the way the trial judge approached the evidence and credibility of C.L.Y. was highly problematic.  Like Twaddle J.A., therefore, I would set aside the convictions.  Unlike him, however, I would order a new trial rather than directing the entry of acquittals.

[6]                              To start, I see no error in connection with the onus of proof.  This Court has consistently warned that verdicts of guilt should not be based on “whether [triers of fact] believe the defence evidence or the Crown’s evidence” (W. (D.), at p. 757).  Rather, the paramount question remains whether, on the whole of the evidence, the trier of fact is left with a reasonable doubt about the guilt of the accused (R. v. Morin, [1988] 2 S.C.R. 345, at p. 361).   The following suggested steps in W. (D.) are intended to ensure that the trier of fact remains focused on the principle of reasonable doubt:

First, if you believe the evidence of the accused, obviously you must acquit.

Second, if you do not believe the testimony of the accused but you are left in reasonable doubt by it, you must acquit.

Third, even if you are not left in doubt by the evidence of the accused, you must ask yourself whether, on the basis of the evidence which you do accept, you are convinced beyond a reasonable doubt by that evidence of the guilt of the accused. [p. 758]

[7]                              However, and significantly, this Court has also consistently confirmed that these steps need not be religiously followed or articulated.  Cory J. made this very clear in W. (D.):

. . . the failure to use such language is not fatal if the charge, when read as a whole, makes it clear that the jury could not have been under any misapprehension as to the correct burden and standard of proof to apply. [p. 758]

He reiterated this caveat in R. v. S. (W.D.), [1994] 3 S.C.R. 521, when he said, at p. 533,  that the W. (D.) procedure was not meant to be followed “word for word as some magic incantation”.  The key is whether the correct burden and standard of proof were applied, not what words were used in applying them.

[8]                              It is noteworthy that in W. (D.) itself, despite the trial judge’s error in instructing the jury that they were engaged in a credibility contest, the conviction was upheld.  This of course does not give trial judges licence to wrongly analyse credibility issues, but it does serve to remind that what W. (D.) offered was a helpful map, not the only route.  Its purpose was to ensure that triers of fact — judges or juries — understand that the verdict should not be based on a choice between the accused’s and Crown’s evidence, but on whether, based on the whole of the evidence, they are left with a reasonable doubt as to the accused’s guilt (R. v. Challice (1979), 45 C.C.C. (2d) 546 (Ont. C.A.), at pp. 556-57; R. v. MacKenzie, [1993] 1 S.C.R. 212, at pp. 219 and 240).  As Fish J.A. noted in dissent in R. v. Levasseur (1994), 89 C.C.C. (3d) 508 (Que. C.A.), at p. 532, in language approved by this Court ([1994] 3 S.C.R. 518):

The trial judge must make it indisputably clear to the jury that reaching a verdict is not simply a question of choosing the more believable of the two competing stories . . . .

To protect the innocent from conviction, we require proof beyond a reasonable doubt.  The application of this standard to questions of credibility is an entrenched part of our law.  The direction most consonant with this principle is a clear and specific instruction, where credibility is an important issue, that the jury must apply to it the test of reasonable doubt.

[9]                              This was most recently reinforced by Deschamps J., writing for the majority in R. v. Boucher, [2005] 3 S.C.R. 499, 2005 SCC 72:

The approach set out in W. (D.) is not a sacrosanct formula that serves as a straitjacket for trial courts.  Trial judges deliver oral judgments every day and often limit their reasons to the essential points.  It would be wrong to require them to explain in detail the process they followed to reach a verdict.  They need only give reasons that the parties can understand and that permit appellate review: R. v. Sheppard, [2002] 1 S.C.R. 869, 2002 SCC 26, and R. v. Burns, [1994] 1 S.C.R. 656. [para. 29]

Charron J., dissenting in part, nonetheless echoed this flexibility:

I agree with Deschamps J. that a ritual incantation is not required in every case. The trial judge did not have to repeat the formula set out in W. (D.) to demonstrate that she had relied on correct legal principles in assessing the accused’s credibility.  Moreover, she is presumed to know those principles. Thus, when a trial judge states that he or she “rejects” an accused’s testimony, it can generally be concluded that the testimony failed to raise a reasonable doubt in the judge’s mind.  [para. 59]

(See also R. v. Minuskin (2003), 68 O.R. (3d) 577 (C.A.), at para. 22, per Rosenberg J.A.; R. v. Chittick (2004), 228 N.S.R. (2d) 81, 2004 NSCA 135, at para. 21, per Cromwell J.A.)

[10]                          Here we are dealing with a trial judge sitting without a jury.  She is presumed to know a principle as elementary as the presumption of innocence.  Her own words confirm her knowledge and appreciation of the correct approach.  She properly articulated the test in W. (D.) and explained:

. . . I am acutely aware of the test in W. (D.) and the necessity of not simply preferring one testimony over another. I am also aware of the recent Manitoba Court of Appeal decisions on assessing credibility in these cases, yet it is difficult to start, other than with an assessment of the complainant’s credibility, as that is the cornerstone of the Crown’s case. I will, however, be guided by the concerns expressed by the Court of Appeal. [para. 4]

[11]                          Unlike my colleague Justice Fish, I respectfully see no blueprint for error in the trial judge’s failure to observe W. (D.) as a catechism.  Among several useful observations in R. v. Morrissey (1995), 97 C.C.C. (3d) 193 (Ont. C.A.), at pp. 203-5, cautioning appellate judges not to dissect, parse, or microscopically examine the reasons of a trial judge, the following passages by Doherty J.A. are particularly apt reminders in this case:

A trial judge’s reasons cannot be read or analyzed as if they were an instruction to a jury.  Instructions provide a road-map to direct lay jurors on their journey toward a verdict.  Reasons for judgment are given after a trial judge has reached the end of that journey and explain why he or she arrived at a particular conclusion.  They are not intended to be, and should not be read as a verbalization of the entire process engaged in by the trial judge in reaching a verdict. [p. 204]

                                                                            . . .

. . . Where a phrase in a trial judge’s reasons is open to two interpretations, the one which is consistent with the trial judge’s presumed knowledge of the applicable law must be preferred over one which suggests an erroneous application of the law: R. v. Smith (1989), 95 A.R. 304 (C.A.) at pp. 312-13; affirmed [1990] 1 S.C.R. 991, 109 A.R. 160, 111 N.R. 144. [pp. 203-4]

[12]                          The trial judge’s reasons reveal that she understood that a finding that the girl was credible did not mean that the onus shifted to the accused to show that he was not guilty.  I find it difficult to see how the sequence in which the trial judge set out her findings of credibility can be said to undermine her articulated and correct statement of the law, or demonstrate that she lost sight of a paramount legal principle like reasonable doubt.

[13]                          The flaw in this case which does lead me to have concerns about the fairness of the trial relates to fundamental misapprehensions and mischaracterizations of the evidence leading to the convictions.

[14]                          Three witnesses testified at trial — the girl, who was by then nine years old, her mother, and C.L.Y.  The mother was not present at either of the alleged incidents. In the words of the trial judge, “Other than the acts complained of, the other evidence from all three is surprisingly consistent” (para. 3).

[15]                          In reaching the conclusion that C.L.Y.’s evidence did not raise a reasonable doubt in her mind, the trial judge relied primarily on what she said were two features of his evidence: that he was reluctant to admit to any tickling at all, a reluctance she found did not “ring true”, and that he remembered a “surprising amount of detail” surrounding the events.   I set out her reasoning on these issues in full:

He denied anything inappropriate with [the complainant].  He had babysat her on more than a few occasions and had seen her last some seven or eight months before the complaint was lodged, yet he was able to provide a surprising amount of detail for incidents in which, according to him, nothing happened.

At the first house, some four years earlier, he testified that the first night they read a book on the floor, that her mother had gone to a party, and that they were reading a Leapfrog book that spelled words. [This related to the third charge of sexual assault for which C.L.Y. was acquitted.]

There was a similar amount of detail provided respecting the second incident, that her mother had to work late, that when they got to her house he asked her if she had any homework, that they then did a number of spelling or number games while sitting on the bed beside each other, that the mother’s boyfriend came home about five o’clock when it was expected he, [the accused], would be there until nine. Again, why would you remember all this detail if nothing happened?

On both the first two occasions, [the accused] was very loathe to admit to the possibility that there might have been any tickling at all, even of an innocuous variety. That does not ring true. There was similarly an incredible amount of detail provided for the third incident, yet it was some seven or eight months before the police came to see him. He even offered that [the girl] changed into jeans from shorts when she came home, although he had no reason to remember that.

In short, respecting [the accused’s] testimony, I do not believe it, nor does it raise a reasonable doubt in my mind. The entirety of the circumstances persuades me of the guilt of the accused beyond a reasonable doubt on the last two offences.  [Emphasis added; paras. 9-13.]

[16]                          There is no support in the transcript for the trial judge’s conclusion that C.L.Y. “was very loathe to admit to the possibility that there might have been any tickling at all, even of an innocuous variety”.  C.L.Y. readily acknowledged that he may have tickled the girl, but insisted that he had never tickled her inappropriately. His testimony, under cross-examination, was as follows:

Q:   Okay. Now, do you think tickling someone when you’re babysitting, do you think there’s something wrong with that?

A:    I don’t believe so, sir.

Q:   Well, wouldn’t that be something that you might have engaged with her, tickling her?

A:    Yes, but not in the area that she specifies.

Q:   Okay. So if I asked you to think back, for example, when you were reading her the book, if there was something funny or something going on in the story, did you tickle her then around the waist?

A:    I don’t remember doing that on that particular occasion of reading that particular book to her, no, sir.

Q:   Do you recall occasions when you tickled her?

A:    Not there, sir, no.

Q:   Well, anywhere. Do you recall tickling her in her waist?

A:    I don’t recall tickling her at that residence, sir.

Q:   Well, or [sic] any other occasion, do you recall tickling her?

A:    No, sir.

Q:   All right. So there was no tickling that took place then, --

A:    No, sir.

Q:   -- as far as you can recall?

A:    No, sir.

Q:   Even though there’s nothing wrong with tickling, you just said that’s not something that happened?

A:    No, sir.

Q:   All right. So, for example, when she was playing on the Mo-Mo-Mi-Mi computer, if I would suggest that, you know, she got an answer right or an answer wrong, you might have started tickling her in response to that? That -- nothing like that happened?

A:    I may have done that on that occasion.

Q:   Well, so there may have -- there may have been tickling by you?

A:    Yes. Oh, I guess I misunderstood your last question with regards -- I didn’t know that you meant all three residences.

Q:   Okay. I was under the impression that you don’t recall any incident at any house where you were tickling her. Am I wrong about that?

A:    I may have tickled her with regards to the Mi-Mi-Mo-Mo residence, but again, not in the area she specified.

Q:   Where did you tickle her on that occasion?

A:    If I had tickled her, which I do not remember tickling her, I would have tickled her somewhere or around her ribcage area.

Q:   Though I suppose if you don’t remember, it could have been you were tickling her on her thighs?

A:    No, sir, I wouldn’t have done that.

Q:   Is there something wrong with tickling her on her thighs?

A:    Possibly.

Q:   Well, what? What would be wrong with tickling a girl on her -- a young girl on her thighs?

A:    That is fairly close to her private area, sir. [Emphasis added; A.R., at pp. 112-13.]

As this passage demonstrates, what C.L.Y. was “loathe to admit” was that he had touched the girl inappropriately, not that he had tickled her.  This, it seems to me, is a critical error in the trial judge’s appreciation of the evidence.

[17]                          As to the trial judge’s conclusion that C.L.Y. was able “to provide a surprising amount of detail”, the record shows that he was, in fact, often unable to recall many details.  On at least 20 occasions, C.L.Y. testified that he was uncertain or simply unable to remember certain details surrounding the events in question.  Again, a review of the transcript reveals that the details that he was able to recall were anodyne and explicable, not by any means “surprising”.

[18]                          With respect, where, as here, a trial judge attributes an inappropriate capacity for recollection of detail to a witness who in fact revealed no such capacity, there remain serious concerns about the reliability of the credibility finding.

[19]                          This Court has said a material misapprehension of the evidence may justify appellate intervention.  In R. v. Lohrer, [2004] 3 S.C.R. 732, 2004 SCC 80, at para. 1, Binnie J. cited, with approval, the following analysis by Doherty J.A. in Morrissey:

Where a trial judge is mistaken as to the substance of material parts of the evidence and those errors play an essential part in the reasoning process resulting in a conviction, then, in my view, the accused’s conviction is not based exclusively on the evidence and is not a “true” verdict. . . .  If an appellant can demonstrate that the conviction depends on a misapprehension of the evidence then, in my view, it must follow that the appellant has not received a fair trial, and was the victim of a miscarriage of justice. This is so even if the evidence, as actually adduced at trial, was capable of supporting a conviction. [Emphasis added; p. 221.]

As Binnie J. pointed out, this represents a stringent standard, and the misapprehension must “play an essential part not just in the narrative of the judgment but ‘in the reasoning process resulting in a conviction’” (para. 2).

[20]                          In my view, the threshold has been met in this case.

[21]                          In reaching this conclusion, I am keenly aware that a trial judge occupies a singular perch in assessing credibility (R. v. W. (R.), [1992] 2 S.C.R. 122, at p. 131; R. v. Gagnon, [2006] 1 S.C.R. 621, 2006 SCC 17, at paras. 20-21). However, based on the record in this case, it is clear that the trial judge’s bases for disbelieving C.L.Y. rested on misapprehensions of his evidence and played a critical role in the convictions, rendering them insupportable.

[22]                          In the circumstances, the appeal is allowed, the convictions quashed, and a new trial ordered.

The reasons of Binnie, Deschamps and Fish JJ. were delivered by

[23]                          I would allow this appeal on two grounds.  The first is that the trial judge rejected the evidence of the appellant for reasons that are not supported by the record. I agree with Justice Abella in this regard and have nothing to add. In my respectful view, however, the appeal should be allowed on a second ground as well.  Reading the trial judge’s reasons as a whole, as we must, I agree with Twaddle J.A., dissenting in the Court of Appeal,  that the trial judge misapplied the burden of proof in this case.

II

[24]                          Unlike Justice Abella, I find little comfort in the trial judge’s allusion to the analytical framework recommended by R. v. W. (D.), [1991] 1 S.C.R. 742.  And I find less comfort still in the explanation she gave for departing from that framework here. The trial judge felt bound to assess the credibility of the complainant before considering the evidence of the accused because, in her words, the complainant’s credibility was “the cornerstone of the Crown’s case” (2005 CarswellMan 562, at para. 4). 

[25]                          It seems to me, on the contrary, that this is precisely the kind of case where a departure from W. (D.) is fraught with particular danger.  The very purpose of adhering to the procedure set out in W. (D.) is to foreclose an inadvertent shifting of the burden of proof where the complainant and the accused have both testified and the outcome of the trial turns on their credibility as witnesses.

[26]                          That was the case here.  And the risk of an inadvertent shift in the burden of proof materialized, as the reasons of the trial judge make plain.  The trial judge accepted the evidence of the complainant without taking into account at all the contrary evidence of the appellant.  Before even considering the appellant’s evidence, the trial judge had concluded not only that the complainant’s evidence was “credible”, but that it was in fact true: “I believe the complainant”, she stated (at para. 6).

[27]                          This conclusion, premature at best, amounted to a finding that the appellant was guilty as charged. As a matter of law, the appellant was presumed innocent; as a matter of fact, his fate had been sealed without any consideration at all of his evidence under oath at trial.

[28]                          Only then did the trial judge turn to the evidence of the accused, and she did so in these terms:

In assessing [the accused’s] testimony, I am aware that I have already found [the complainant] to be credible, but I will explain why I disbelieve [the accused] and why his evidence does not raise a reasonable doubt in my mind. [para. 8]

[29]                          I accept the proposition from R. v. Morrisey (1995), 97 C.C.C. (3d) 193 (Ont. C.A.), adopted by my colleague that a trial judge’s reasons “should not be read as a verbalization of the entire process engaged in by the trial judge in reaching a verdict” (p. 204).  That is a very different thing, however, from disregarding what a trial judge carefully and explicitly describes as the pathway to her decision.  Her statement quoted above could not be understood by the convicted accused, and should not be misunderstood by us, as anything other than that the trial judge had accepted the credibility of the complainant before “assessing” (her word) the evidence of the accused.

[30]                          Unfortunately, the appellant’s presumption of innocence had by that point been displaced by a presumption —  indeed, a finding —  of guilt. The trial judge could hardly believe both the appellant and the complainant.  Before even considering the appellant’s evidence, she had already concluded that she believed the complainant.  In effect, the trial judge had thus decided to convict the appellant unless his evidence persuaded her to do otherwise.

III

[31]                          This brings me back to the first and additional ground for allowing the appeal.  In order to lend credit to the appellant’s evidence, the trial judge would have had to abandon her conclusion, already reached, that the complainant’s evidence was not only credible, but true. As Justice Abella demonstrates, the trial judge instead rejected the appellant’s evidence for reasons that are in fact groundless. I feel bound to say that the trial judge would likely have guarded herself against invoking those reasons had she followed the analytical framework set out in W. (D.).

[32]                          Adherence to that framework, though generally appropriate, is by no means mandatory.  Likewise, mere reference to it will not suffice in all cases: A correct statement of the burden of proof can scarcely save its evident misapplication, as in this case.

[33]                          In short, judges may know the law, yet err in its application; they may know the facts, yet make findings of credibility unsupported by the record.  What matters in either instance is the substance and not the form of the decision.  For the reasons given, I have concluded, with respect, that the judgment at trial is flawed in substance:  The trial judge erred in law by shifting the burden of proof, and she erred as well in rejecting the appellant’s evidence on grounds unsupported by the record.

[34]                          I would therefore allow the appeal, set aside the appellant’s convictions and order a new trial.

Appeal allowed.

Solicitors for the appellant:  Johnston & Company, Dauphin, Manitoba.

Solicitor for the respondent:  Attorney General of Manitoba, Winnipeg.