H.L. v. Canada (Attorney General), 2005 SCC 25, [2005] 1 SCR 401

                                                 SUPREME COURT OF CANADA

Between:

H.L.

Appellant

v.

Attorney General of Canada

Respondent

‑ and ‑

Attorney General for Saskatchewan

Intervener

Coram: McLachlin C.J. and Major, Bastarache, Binnie, LeBel, Deschamps, Fish, Abella and Charron JJ.

______________________________

H.L. v. Canada (Attorney General), [2005] 1 S.C.R. 401, 2005 SCC 25

H.L.                                                                                                                   Appellant

v.

Attorney General of Canada                                                                        Respondent

and

Attorney General for Saskatchewan                                                              Intervener

Indexed as:  H.L. v. Canada (Attorney General)

Neutral citation:  2005 SCC 25.

File No.:  29949.

Hearing:  May 13, 2004.

Present:  Iacobucci, Major, Bastarache, Binnie, LeBel, Deschamps and Fish JJ.

Rehearing:  December 13, 2004.

Judgment:  April 29, 2005.

Present:  McLachlin C.J. and Major, Bastarache, Binnie, LeBel, Deschamps, Fish, Abella and Charron JJ.

on appeal from the court of appeal for saskatchewan

Appeals — Saskatchewan Court of Appeal — Questions of fact — Applicable standard of appellate review on questions of fact in Saskatchewan — Whether Court of Appeal correct in setting aside trial judge’s pecuniary damages award for loss of past and future earnings — Whether Court of Appeal applied proper standard — The Court of Appeal Act, 2000, S.S. 2000, c. C‑42.1, s. 14.

L brought an action for sexual battery against S and the federal government for acts that had occurred 20 years earlier when L was about 14 years old.  S, who worked on a reserve for the federal government, sexually abused L on two occasions.  L left school when he was about 17 years old, without completing the eighth grade.  He was unable to retain meaningful employment between 1978‑1987.  During that time, he drank heavily, was incarcerated frequently and relied on social assistance to meet his needs.  Between 1988‑2000, he worked sporadically.  The evidence given by L and two expert witnesses satisfied the trial judge that L’s poor employment record between 1978‑1987 was attributable to his alcoholism, emotional difficulties, and criminality, which were in turn attributable to the sexual abuse perpetrated by S.  He found as well that L’s sporadic work record between 1988‑2000 was consistent with the emotional difficulties described by the experts in their assessments of the psychological effects of sexual abuse.  The trial judge maintained L’s action against S and the federal government, since he found that the criteria for the imposition of vicarious liability on the government had been met.  He awarded L non‑pecuniary damages, pecuniary damages for loss of past and future earnings and pre‑judgment interest.  With respect to L’s claim for loss of future earnings, in the absence of specific evidence in this regard, the trial judge relied inferentially on the evidence relating to L’s past earning capacity.  The Court of Appeal dismissed the federal government’s appeal as it related to vicarious liability and to the award for non‑pecuniary damages, but allowed the appeal in relation to pecuniary damages and pre‑judgment interest.  The Court of Appeal set aside the award for pecuniary damages for loss of past and future earnings on the ground that, on its assessment of the evidence, the evidence fell short of proving the loss.  Leave to this Court was granted by the Court of Appeal, pursuant to s. 37  of the Supreme Court Act, to clarify the correct standard of review applicable to the Saskatchewan Court of Appeal.

Held (Bastarache, LeBel, Deschamps and Charron JJ. dissenting in part):  The appeal should be allowed in part.  The trial judge’s award of pecuniary damages for loss of past earnings is restored, but the award must be reduced to reflect the time L spent in prison and the social assistance he received during the period covered by the award.

Per McLachlin C.J. and Major, Binnie, Fish and Abella JJ.:  In Saskatchewan, as elsewhere in Canada, a trial judge’s primary findings of fact and inferences of fact are only reviewable on appeal on a standard of palpable and overriding error.  The Court of Appeal Act, 2000, in particular s. 14, did not create for Saskatchewan an appellate court radically different, in powers and purpose, from its counterparts in the other provinces.  To the contrary, an examination of both the 2000 Act and its predecessors, their legislative history, and their judicial interpretation in this Court and by the Saskatchewan Court of Appeal itself all lead to the conclusion that the 2000 Act did not change the standard of review applicable in Saskatchewan to appellate review on questions of fact:  the appeal is a review for error, and not a review by rehearing.  Courts of appeal in Canada, absent an express legislative instruction to the contrary, cannot disregard the governing principle of appellate intervention on questions of fact.  They may make their own findings and draw their own inferences, but only where the trial judge is shown to have committed a palpable and overriding error or made findings of fact, including inferences of fact, that are clearly wrong, unreasonable, or unsupported by the evidence.  A court of appeal cannot substitute for the reasonable inference preferred by the trial judge, an equally, or even more, persuasive inference of its own.  These principles are consistent with this Court’s recent decision in Housen. [3-6] [13-16] [74] [80] [89] [110]

In this case, the Saskatchewan Court of Appeal reversed the trial judge on six points: (1) qualification of the experts, (2) causation, (3) mitigation, (4) incarceration, (5) collateral benefits, and (6) loss of future earnings.  The Court of Appeal erred in interfering with the trial judge’s findings on the first three issues because it applied the wrong standard and improperly substituted its own opinion of the facts for that of the trial judge.  The trial judge, however, made “palpable and overriding errors” on the last three issues.  His finding that S’s sexual abuse of L caused his loss of income due to imprisonment is both contrary to judicial policy and unsupported by the evidence.  L’s lack of gainful employment caused by his imprisonment resulted from his criminal conduct, not from his abuse by S or from the alcoholism.  The award for loss of past earnings should thus be reduced to reflect the time L spent in prison.  The trial judge also erred in not deducting from the same award the social assistance payments L had received during the relevant period.  The trial judge’s failure to make such deduction constitutes a severable error of principle.  Finally, the trial judge’s award for loss of future earnings must be set aside.  The finding that a person has had emotional and substance abuse problems which in the past have impacted on his earning capacity is not in itself a sufficient basis for concluding on the balance of probabilities that this state of affairs will endure indefinitely. [111] [137] [142-143] [145] [148] [152]

Per Bastarache, LeBel and Deschamps JJ. (dissenting in part):  In Saskatchewan, the nature of appellate review is by way of rehearing and not review for error.  The grammatical and ordinary sense of the words used in ss. 13 and 14 of The Court of Appeal Act, 2000, as well as the object of the Act, the object of the specific legislative provisions that form the statutory framework for appeals, and the Act’s historical foundations, clearly lead to that conclusion.  The Court of Appeal Act, 2000 is the only one among all of the statutes governing the powers of appellate courts in Canada that relieves the Court of Appeal of any obligation to adopt the view of the evidence taken by the trial judge and directs it to act on its own view of what, in its judgment, the evidence proves. [157] [243] [296]

A number of Saskatchewan Court of Appeal cases also support the conclusion that the nature of appellate review in Saskatchewan is by way of rehearing.  To the extent that there are cases from this Court and the Saskatchewan Court of Appeal that appear to conflict with this conclusion, they can be reconciled.  In particular, in Housen, the mere fact that this Court did not, on an appeal from the Saskatchewan Court of Appeal, refer to The Court of Appeal Act but instead used a statement from a different province’s Court of Appeal that is in conflict with the clear language of that Act to define the role of the appellate court in Saskatchewan, demonstrates that Housen should not be used to understand the nature of appellate review in that province.  Rather, the application of Housen as an authority should be limited to general standards of appellate review only. [259] [294-298]

Appellate review by way of rehearing is not a retrial or a de novo hearing. On an appeal by way of rehearing, the Court of Appeal is not limited to a review of the lower court’s decision and can form its own judgment on the issues and direct its attention to the merits of the case.  This does not mean, however, that the Court of Appeal can ignore the trial judge’s findings.  The special advantage of the trial judge calls for a measure of deference on the part of the Saskatchewan Court of Appeal when, pursuant to the direction in s. 14 of the Act, it is considering what the evidence proves.  Factual findings that engage the special advantage of the trial judge will be accorded some deference and the Court of Appeal will only interfere and apply its own view of the evidence if the trial judge has committed a palpable and overriding error in his or her fact finding.  Factual findings that do not engage the special advantage of the trial judge are not entitled to the same level of deference.  The Court of Appeal will only interfere and apply its own view of the evidence if the trial judge has committed a simple error in his or her fact finding.  In the case of inferences of fact, since a trial judge is in no better position than the Court of Appeal to draw inferences of fact from a base of fact properly established, the threshold that the Court of Appeal must pass before substituting its own inference of fact is reasonableness.  Nevertheless, given the respect that is to be accorded to the office of the trial judge, in the cases of inferences of fact or of findings of fact that do not engage the special advantage of the trial judge, the Court of Appeal will presuppose that the trial judge has drawn reasonable inferences of fact or made factual findings free of error. [178] [245] [253-256]

The Court of Appeal correctly applied the appropriate standard when it set aside the trial judge’s pecuniary damages award for both past and future loss of earnings, because the factual inferences on which the award was based were not reasonably supported by the evidence and were therefore not reasonable.  Even if the more stringent standard set out in Housen applied here, the Court of Appeal’s decision would still be upheld.  The trial judge’s findings were so unreasonable that they amounted to palpable error in the appreciation of the evidence and the inferences drawn.  With respect to past loss of earnings, the trial judge’s first inference that S’s sexual abuse caused L’s alcoholism was based primarily on the general expert evidence.  However, the expert witnesses in this case transcended their respective fields of expertise when they testified as to the etiology of alcoholism and the cause of L’s alcoholism in particular.  Since the expert witnesses were not properly qualified to express opinions on this subject, their evidence in this regard is entitled to no weight, and L’s testimony as to the effect of S’s sexual abuse on his alcoholism could not, on its own, provide a sufficient evidentiary basis for the trial judge’s inference that S’s sexual abuse caused L’s alcoholism.  The trial judge’s second inference that S’s sexual abuse caused L’s emotional problems which resulted in L losing employment income also lacks a sufficient evidentiary foundation.  The evidence adduced at trial only demonstrated that L did not work between 1978‑1987 and worked only sporadically between 1988‑2000.  It does not prove that L was wholly or largely unable to work because of his emotional problems.  L’s sporadic work record, in itself, is as consistent with choosing not to work as with being unable to work.  With respect to future loss of earnings, since it was not reasonable for the trial judge to conclude that L suffered a loss of employment income because of S’s sexual abuse, given the evidentiary gaps in the trial judge’s causal chain, it was likewise not reasonable for him to conclude that L will continue to suffer such a loss in the future. [306] [313-317] [323-325] [329]

Per Charron J. (dissenting in part):  There is agreement with the majority’s analysis on the governing standard of review for appeals in Saskatchewan and the Court of Appeal thus erred in finding that the standard was other than that adopted by this Court in Housen.  However, on application of the appropriate standard of review, the Court of Appeal was correct in setting aside the entire award for pecuniary damages.  There is agreement with the minority that the same error informed the trial judge’s decision to award pecuniary damages in respect of both past and future loss of earnings.  The trial judge found that there was a causal connection between the acts of sexual abuse and a lifelong inability to earn income.  The evidence did not support this finding and, consequently, the award for loss of income, past and future, is unreasonable. [347-348]

Cases Cited

By Fish J.

Applied:  Housen v. Nikolaisen, [2002] 2 S.C.R. 235, 2002 SCC 33; M.B. v. British Columbia, [2003] 2 S.C.R. 477, 2003 SCC 53; referred to:  Lensen v. Lensen, [1987] 2 S.C.R. 672; Tanel v. Rose Beverages (1964) Ltd. (1987), 57 Sask. R. 214; Athey v. Leonati, [1996] 3 S.C.R. 458; Markling v. Ewaniuk, [1968] S.C.R. 776; Kosinski v. Snaith (1983), 25 Sask. R. 73; R. v. W. (R.), [1992] 2 S.C.R. 122; Assicurazioni Generali SpA v. Arab Insurance Group, [2003] 1 W.L.R. 577; Stein v. The Ship “Kathy K”, [1976] 2 S.C.R. 802; Beaudoin‑Daigneault v. Richard, [1984] 1 S.C.R. 2; Geffen v. Goodman Estate, [1991] 2 S.C.R. 353; Toneguzzo‑Norvell (Guardian ad litem of) v. Burnaby Hospital, [1994] 1 S.C.R. 114; Hodgkinson v. Simms, [1994] 3 S.C.R. 377; Schwartz v. Canada, [1996] 1 S.C.R. 254; Harrington v. Harrington (1981), 33 O.R. (2d) 150; Pelech v. Pelech, [1987] 1 S.C.R. 801; Board of Education of the Long Lake School Division No. 30 of Saskatchewan v. Schatz (1986), 49 Sask. R. 244; Sisson v. Pak Enterprises Ltd. (1987), 64 Sask. R. 232; Rizzo & Rizzo Shoes Ltd. (Re), [1998] 1 S.C.R. 27; Knight v. Huntington (2001), 14 B.L.R. (3d) 202, 2001 SKCA 68; Bogdanoff v. Saskatchewan Government Insurance (2001), 203 Sask. R. 161, 2001 SKCA 35; Brown v. Zaitsoff Estate (2002), 217 Sask. R. 130, 2002 SKCA 18; Janiak v. Ippolito, [1985] 1 S.C.R. 146; Canada (Director of Investigation and Research) v. Southam Inc., [1997] 1 S.C.R. 748.

By Bastarache J. (dissenting in part)

Housen v. Nikolaisen, [2002] 2 S.C.R. 235, 2002 SCC 33; Fox v. Percy (2003), 214 C.L.R. 118, [2003] HCA 22; Kourtessis v. M.N.R., [1993] 2 S.C.R. 53; Farm Credit Corp. v. Valley Beef Producers Co‑operative Ltd. (2002), 223 Sask. R. 236, 2002 SKCA 100; Chieu v. Canada (Minister of Citizenship and Immigration), [2002] 1 S.C.R. 84, 2002 SCC 3; Hallberg v. Canadian National Railway Co. (1955), 16 W.W.R. 538; Taylor v. University of Saskatchewan (1955), 15 W.W.R. 459; Audergon v. La Baguette Ltd., [2002] E.W.J. No. 78 (QL), [2002] EWCA Civ 10; Gray v. Turnbull (1870), L.R. 2 Sc. & Div. 53; Bigsby v. Dickinson (1876), 4 Ch. D. 24; Coghlan v. Cumberland, [1898] 1 Ch. 704; Montgomerie & Co. v. Wallace‑James, [1904] A.C. 73; Mersey Docks and Harbour Board v. Procter, [1923] A.C. 253; Benmax v. Austin Motor Co., [1955] A.C. 370; Coventry v. Annable (1911), 19 W.L.R. 400, aff’d (1912), 46 S.C.R. 573; Greene, Swift & Co. v. Lawrence (1912), 2 W.W.R. 932; Miller v. Foley & Sons (1921), 59 D.L.R. 664; Messer v. Messer (1922), 66 D.L.R. 833; Monaghan v. Monaghan, [1931] 2 W.W.R. 1; Kowalski v. Sharpe (1953), 10 W.W.R. (N.S.) 604; Tarasoff v. Zielinsky, [1921] 2 W.W.R. 135; Matthewson v. Thompson, [1925] 2 D.L.R. 1211; French v. French, [1939] 2 W.W.R. 435; Wilson v. Erbach (1966), 56 W.W.R. 659; Tanfern Ltd. v. Cameron‑MacDonald, [2000] 1 W.L.R. 1311; Assicurazioni Generali SpA v. Arab Insurance Group, [2003] 1 W.L.R. 577; S.S. Hontestroom v. S.S. Sagaporack, [1927] A.C. 37; Whitehouse v. Jordan, [1981] 1 All E.R. 267; Laurentide Motels Ltd. v. Beauport (City), [1989] 1 S.C.R. 705; St‑Jean v. Mercier, [2002] 1 S.C.R. 491, 2002 SCC 15; Warren v. Coombes (1979), 142 C.L.R. 531; Workmen’s Compensation Board v. Greer, [1975] 1 S.C.R. 347; Toneguzzo‑Norvell (Guardian ad litem of) v. Burnaby Hospital, [1994] 1 S.C.R. 114; Lensen v. Lensen, [1987] 2 S.C.R. 672; Board of Education of the Long Lake School Division No. 30 of Saskatchewan v. Schatz (1986), 49 Sask. R. 244; Tanel v. Rose Beverages (1964) Ltd. (1987), 57 Sask. R. 214; Sisson v. Pak Enterprises Ltd. (1987), 64 Sask. R. 232; Knight v. Huntington (2001), 14 B.L.R. (3d) 202, 2001 SKCA 68; Bogdanoff v. Saskatchewan Government Insurance (2001), 203 Sask. R. 161, 2001 SKCA 35; Brown v. Zaitsoff Estate (2002), 217 Sask. R. 130, 2002 SKCA 18; Underwood v. Ocean City Realty Ltd. (1987), 12 B.C.L.R. (2d) 199; R. v. Mohan, [1994] 2 S.C.R. 9; R. v. Marquard, [1993] 4 S.C.R. 223; Parker v. Saskatchewan Hospital Assn., [2001] 7 W.W.R. 230, 2001 SKCA 60; Athey v. Leonati, [1996] 3 S.C.R. 458; State Rail Authority of New South Wales v. Wiegold (1991), 25 N.S.W.L.R. 500; M.B. v. British Columbia, [2003] 2 S.C.R. 477, 2003 SCC 53.

By Charron J. (dissenting in part)

Housen v. Nikolaisen, [2002] 2 S.C.R. 235, 2002 SCC 33.

Statutes and Regulations Cited

Alberta Rules of Court, Alta. Reg. 390/68, r. 518(c), (e).

Civil Procedure Rules 1998 (U.K.), S.I. 1998 No. 3132, Part 52 [ad. S.I. 2000 No. 221], r. 52.11(1).

Constitution Act, 1867, s. 92(13) , (14) .

Court of Appeal Act, R.S.B.C. 1996, c. 77, s. 9(2).

Court of Appeal Act, R.S.M. 1987, c. C240, s. 26(1), (2).

Court of Appeal Act, R.S.S. 1930, c. 48, s. 8.

Court of Appeal Act, R.S.S. 1978, c. C‑42, s. 8.

Court of Appeal Act, S.S. 1915, c. 9, ss. 8, 9.

Court of Appeal Act, 2000, S.S. 2000, c. C‑42.1, ss. 7(2)(a), 12, 13, 14, 16.

Courts of Justice Act, R.S.O. 1990, c. C.43, s. 134(1)(a), (4)(a).

Courts of Justice Act, R.S.Q., c. T‑16, s. 10.

Indian Act, S.C. 1951, c. 29.

Interpretation Act, 1995, S.S. 1995, c. I‑11.2, ss. 10, 35 [am. 1998, c. 47, s. 6].

Rules of the Supreme Court, 1883 (U.K.), Order 39, Order 58, rr. 1, 4.

Supreme Court Act, R.S.C. 1985, c. S‑26, s. 37 .

Supreme Court Act, R.S.P.E.I. 1987, c. 66, s. 56(1)(a), (4)(a).

Authors Cited

Andrews, N. H.  “A New System of Civil Appeals and a New Set of Problems”, [2000] Cambridge L.J. 464.

Driedger, Elmer A.  Construction of Statutes, 2nd ed.  Toronto:  Butterworths, 1983.

Great Britain.  Civil Procedure.  London:  Sweet & Maxwell, 2002.

Hohfeld, Wesley Newcomb.  Fundamental Legal Conceptions as Applied in Judicial Reasoning and Other Legal Essays, ed. by W. W. Cook.  New Haven:  Yale University Press, 1923.

Jolowicz, J. A.  “Court of Appeal or Court of Error?”, [1991] Cambridge L.J. 54.

Jolowicz, J. A.  “The New Appeal:  re‑hearing or revision or what?” (2001), 20 C.J.Q. 7.

Perell, Paul M.  “The Standard of Appellate Review and the Ironies of Housen v. Nikolaisen” (2004), 28 Advocates’ Q. 40.

Royer, Jean‑Claude.  La preuve civile, 3^e éd. Cowansville, Qué. : Yvon Blais, 2003.

Saskatchewan.  Legislative Assembly.  Debates and Proceedings (Hansard), 1st Sess., 24th Leg., June 7, 2000, pp. 1625‑26.

Sullivan, Ruth.  Sullivan and Driedger on the Construction of Statutes, 4th ed.  Markham, Ont.:  Butterworths, 2002.

 Zuckerman, Adrian A. S.  Civil Procedure.  London:  LexisNexis UK, 2003.

APPEAL from a judgment of the Saskatchewan Court of Appeal (Cameron, Vancise and Lane JJ.A.) (2002), 227 Sask. R. 165, 287 W.A.C. 165, [2003] 5 W.W.R. 421, [2002] S.J. No. 702 (QL), 2002 SKCA 131, affirming in part a decision of Klebuc J. (2001), 208 Sask. R. 183, [2001] 7 W.W.R. 722, 5 C.C.L.T. (3d) 186, [2001] S.J. No. 298 (QL), 2001 SKQB 233, and supplementary reasons (2001), 210 Sask. R. 114, [2001] 11 W.W.R. 727, [2001] S.J. No. 478 (QL), 2001 SKQB 233.  Appeal allowed in part, Bastarache, LeBel, Deschamps and Charron JJ. dissenting in part.

E. F. Anthony Merchant, Q.C., Eugene Meehan, Q.C., and Graham Neill, for the appellant.

Roslyn J. Levine, Q.C., and Mark Kindrachuk, for the respondent.

Barry J. Hornsberger, Q.C., for the intervener.

The judgment of McLachlin C.J. and Major, Binnie, Fish and Abella JJ. was delivered by

Fish J. —

I.       Introduction

1                                   This appeal turns on the applicable standard of appellate review on questions of fact in Saskatchewan, and on the application of that standard by the Court of Appeal in this case.  Our concern is with all of the facts, and nothing but the facts:  with facts proved directly and with facts inferred, but not with questions of law or questions of mixed law and fact.

2                                   Legislatures may fix by statute the powers of the appellate courts they are constitutionally authorized to create. The Legislative Assembly of Saskatchewan has done so, most recently in The Court of Appeal Act, 2000, S.S. 2000, c. C-42.1 (“2000 Act”). 

3                                   The 2000 Act did not enlarge materially the powers previously vested in the Saskatchewan Court of Appeal.  Nor did it purport to modify at all the manner in which those powers have been exercised for nearly half a century.

4                                   More particularly, the 2000 Act did not change the standard of review applicable in Saskatchewan to appellate intervention with respect to findings of fact.  The criteria that govern the exercise by the Court of Appeal of its statutory powers in this regard remain unchanged.  Like other appellate courts across the country, it may substitute its own view of the evidence and draw its own inferences of fact where the trial judge is shown to have committed a palpable and overriding error or made findings of fact that are clearly wrong, unreasonable or unsupported by the evidence.

5                                   This standard of appellate review is subject, of course, to statutory exceptions.  It does not apply where the legislature has expressly provided otherwise.  Nothing in the 2000 Act reflects any such intention or has any such effect.  It sets out the powers of the Court of Appeal in considerable detail; in other Canadian jurisdictions, equivalent powers are conferred in more general terms.  As we shall see, however, the 2000 Act neither bestows on the Court of Appeal for Saskatchewan unique powers of appellate intervention on questions of fact nor ordains their exercise in a manner that, within Canada, is exclusive to Saskatchewan. 

6                                   In my respectful view, the Court of Appeal departed from the applicable standard in this case.

7                                   I would therefore allow the appeal in part, with costs, as explained in the reasons that follow.

II.      Overview

8                                   This matter reaches us, exceptionally, with leave granted by the Court of Appeal itself, pursuant to s. 37  of the Supreme Court Act, R.S.C. 1985, c. S‑26 .  In reversing the trial judge, the Court of Appeal felt empowered by its governing statute to “rehear” the case.  Speaking for the Court of Appeal on the leave application, Bayda C.J.S. acknowledged that a very different standard — “review for error” — had been held applicable in “the recent majority decision of the Supreme Court of Canada in Housen v. Nikolaisen et al., [2002] 2 S.C.R. 235”.  “Both conclusions”, said the Chief Justice, “cannot be right” ((2003), 238 Sask. R. 167 , 2003  SKCA 78, at para. 11).  I agree, of course, and, in my respectful view, it is the standard applied by the Court of Appeal — the “rehearing” standard — that is wrong.

9                                   I shall deal later with the difference between the majority and minority reasons in Housen v. Nikolaisen, [2002] 2 S.C.R. 235, 2002 SCC 33. For present purposes, it will suffice to mention that this Court in Housen was unanimous on the issue that concerns us here: All nine Justices agreed that the standard of appellate review on questions of fact in Saskatchewan is review for error and not review by rehearing.  They agreed as well that findings of fact by the trial judge will be disturbed on appeal only for errors that can properly be characterized as palpable and overriding.

10                               It was not contended in Housen, either in the Saskatchewan Court of Appeal or in this Court, that the standard of appellate review in Saskatchewan differed significantly from the prevailing standard elsewhere in Canada.  And none of the parties found it necessary or useful to refer in their written or oral submissions in this Court to the 2000 Act or its predecessors.  This should not be thought surprising.  On second reading, the Minister of Justice assured the Legislative Assembly of Saskatchewan that Bill 80 which, upon its adoption, became the 2000 Act

doesn’t change the jurisdiction of the Court of Appeal in any way, it simply restates the historical jurisdiction of the court in a way that can be understood by users of the Act. 

(Saskatchewan Hansard, June 7, 2000, at p. 1626)

11                               Moreover, the Saskatchewan Court of Appeal, both before and after the coming into force of the 2000 Act, had consistently held that a trial judge’s findings of fact can be set aside only where palpable and overriding error is shown.  It affirmed and reiterated that principle well before this Court’s judgment in Housen, and even before Lensen v. Lensen, [1987] 2 S.C.R. 672.  Thus, for example, in Tanel v. Rose Beverages (1964) Ltd. (1987), 57 Sask. R. 214 (C.A.), Bayda C.J.S. stated that the palpable and overriding standard had been followed by the Saskatchewan Court of Appeal “for a long time and most certainly since [1960]” (p. 218).

12                               Lensen, also an appeal from Saskatchewan, was decided under the predecessor to the 2000 Act.  This Court dealt specifically in that case with the governing provision of the Saskatchewan statute, but laid down a uniform norm for appellate courts across the country.

13                               As we shall see, the decisive provisions of the 2000 Act are identical in substance to the corresponding provisions of the Act it replaced.  This underlines the present relevance of the Court of Appeal’s decisions prior to November 1, 2000, when the current Act came into effect.  And it reflects the legislative intention, mentioned earlier, not to “change the jurisdiction of the Court of Appeal in any way” (Saskatchewan Hansard, at p. 1626).

14                               Finally, I agree that the powers of the Saskatchewan Court of Appeal are set out in its constituent statute in greater detail than is the case in most other provinces.  Greater detail in an empowering statute, however, does not invariably signal a legislative intent to confer broader powers.  Often, the opposite is true.  In any event, the 2000 Act must be read in the light of this Court’s jurisprudence — and appellate decisions in Saskatchewan itself — immediately prior to its adoption. Neither the text of the Act nor its legislative history indicates a departure from the principles set out in those cases.

15                               In short, I am not at all persuaded that the 2000 Act was intended to create for Saskatchewan an appellate court radically different, in powers and purpose, from its counterparts in the other provinces.  Nothing in the record before us, in the relevant provisions of the Act, or in the Court of Appeal’s own earlier appreciation of its proper role suggests to me that it has now been invested with a general jurisdiction to “rehear” trials — that is, to apply a “rehearing” standard when it reviews judgments at trial.

16                               To a significant extent, that is what it did here.  In my respectful view, it improperly substituted its own opinion of the facts for that of the trial judge.  The court evidently viewed with skepticism the trial judge’s conclusions regarding the damages suffered by H.L. as a direct result of Mr. Starr’s proven misconduct.  Doubt as to the soundness of the trial judge’s findings of fact, however, is not a recognized ground of appellate intervention.

17                               I would therefore allow the appeal in part and restore the trial judge’s award for past loss of earnings, except where the errors imputed to him are indeed “palpable and overriding”.

III.    The Facts and Judgment at Trial

18                               H.L., a former resident of Gordon First Nation Reserve, brought an action for sexual battery against William Starr and the Government of Canada for acts that had occurred some twenty years earlier:  (2001), 208 Sask. R. 183, 2001 SKQB 233.  Mr. Starr was employed at that time by the federal Department of Indian and Northern Affairs (“Department”) as Residence Administrator of the Gordon Student Residence on Gordon First Nation Reserve.

19                               With the approval of the Department, Mr. Starr had organized various  extracurricular activities for the students and other children living on the Reserve.  It was through one of these activities, an after-school boxing club, that Mr. Starr came into contact with H.L.  H.L. was then about 14 years old.  Mr. Starr sexually abused H.L. on two occasions by subjecting him to acts of masturbation and to requests for sexual favours.

20                               H.L. testified that Mr. Starr’s assaults had a profound and enduring impact.  He felt “ashamed” and “dirty”, and was afraid to tell anyone what had happened, because he thought no one would believe him.  He “tried to find a way to get out of going to school because [he] didn’t want to be around anybody”, and “had a hard time concentrating because it was on [his] mind”. 

21                               H.L. testified that he had never even “touched” alcohol before the assaults occurred, but began consuming excessive amounts shortly thereafter, when he was 15 or 16 years old.  Alcohol provided an “escape” from his recurring thoughts about the sexual assaults.  “[M]y way of dealing with it”, he said, “was to go out and get drunk.”  That is why he “started drinking at a young age and got addicted to alcohol”. 

22                               Because he had difficulty concentrating and was by then “already into alcohol pretty bad”, H.L. left school when he was about 17 years old, without completing the eighth grade.  H.L. characterized the sexual abuse perpetrated by Mr. Starr as the most traumatic event of his life.

23                               Both H.L. and Canada called witnesses who were qualified as experts in assessing the psychological effects of sexual abuse.  Both experts had tested H.L. and interviewed him extensively.  Canada’s expert, Dr. Arnold, adverted to factors other than H.L.’s sexual abuse by Mr. Starr that had, in his view, contributed to H.L.’s addiction to alcohol.  He noted, in particular, that H.L. had grown up in a home that modelled alcohol abuse and violence.  Dr. Arnold concluded, however, that Mr. Starr’s sexual abuse of H.L. was a “specific triggering event” that led to H.L.’s abuse of  alcohol. 

24                               Asked whether H.L. would have become an alcoholic in any event, Dr. Arnold stated:  “He may have had vulnerability, but except for the exposure to the sexual abuse, may not have developed a substance abuse problem.  So I have to be careful when I say that, the risk is there, but except for that triggering event it may not have occurred.  We don’t know.”  Invited to elaborate, Dr. Arnold added:

What we have [here] is an individual who has a risk because of his upbringing, so he’s — he has a risk and a vulnerability.  If specific stressful life events come along and he’s exposed to them, such as sexual abuse, he is at more risk than someone who doesn’t have that history of vulnerability.

25                               H.L.’s expert, Mr. Stewart, testified that H.L. was primarily traumatized by the sexual abuse perpetrated by Mr. Starr, which could be linked to his withdrawal and drinking problems:

[T]hey certainly coincide with his abuse, and again research would indicate that substance abuse . . . is a direct result of being abused, so with other interviews and assessments and people that I’ve seen in therapy that have undergone sexual abuse, they find it extremely difficult to concentrate . . . .

26                               Mr. Stewart explained that some “resilient” children are able to “shrug off” sexual abuse, with the benefit of a strong home and family life and the opportunity to disclose the abuse in a safe manner.  Children who have been abused by a trusted authority figure, on the other hand, are more adversely affected.

27                               The trial judge, Klebuc J., accepted the evidence of H.L. and the experts.  He found that the sexual assaults committed by Mr. Starr caused H.L. to suffer enormous humiliation, self‑blame and loss of self‑worth, to lose interest in his education, in part due to his inability to concentrate, and to develop alcoholism.

28                               Klebuc J. recognized that H.L. had a dysfunctional home life.  He found, however, that no divisible injury could be attributed to it; nor was it a “necessary cause” of H.L.’s injuries.  There was no evidence that H.L. suffered from a “crumbling skull”, or pre-existing condition that would have led to his losses regardless of the sexual battery (see Athey v. Leonati, [1996] 3 S.C.R. 458, at paras. 34-36).  Rather, if H.L. was particularly vulnerable, this amounted to a “thin skull”, within the meaning of Athey, exonerating neither Canada nor Mr. Starr from their liability for the consequences.

29                               H.L. was unable to retain meaningful employment between 1978 and 1987 (the “first period”).  During that time, he drank heavily and was incarcerated frequently.  He relied on social assistance to meet his needs.  Between 1988 and 2000 (the “second period”), he worked sporadically. 

30                               H.L. testified that his inability to maintain steady employment was attributable to his abuse of alcohol, manifested by extensive and recurring periods of indulgence.

31                               The impact of the sexual assaults on H.L.’s ability to maintain steady employment was addressed as well by the experts.  Dr. Arnold, for example, testified to the “chain of events” set into motion by the sexual abuse perpetrated by Mr. Starr.  He stated that this “triggering event” led to alcohol and school problems, a loss of confidence in the school system, and a diminished “work ethic”, which Dr. Arnold defined as H.L.’s “ability to hold work and be able to regularly show up and those kinds of things”.  Dr. Arnold explained that sexual abuse by an authority figure, both generally and in H.L.’s specific circumstances, could lead to distrust of authority figures, including teachers, police, employers, judges, doctors, and medical care workers. 

32                               Similarly, Mr. Stewart testified that the sexual abuse would cause H.L. to have negative self-esteem, a poor self-image and a lack of confidence.  These personality traits, he added, detrimentally affect one’s ability to secure and maintain employment.

33                               The evidence given by H.L. and the experts satisfied Klebuc J. that H.L.’s poor employment record during the “first period” was attributable to his alcoholism, emotional difficulties, and criminality, which were in turn attributable to the sexual abuse perpetrated by Mr. Starr.  He found as well that H.L.’s sporadic work record during the “second period” was consistent with the emotional difficulties described by the experts in their psychological assessments.

34                               In the result, Klebuc J. maintained H.L.’s action against Mr. Starr and the Government of Canada.  He found that the criteria for the imposition of vicarious liability on Canada had been met, and awarded H.L. a total of $80,000 in non‑pecuniary damages, $296,527.09 in pecuniary damages and $30,665 in estimated pre‑judgment interest.

35                               The non‑pecuniary damages included $60,000 for the losses and injuries, including emotional distress, that H.L. had suffered — and would continue to suffer — as a consequence of Mr. Starr’s abuse, and aggravated damages of $20,000.

36                               The pecuniary damages were determined as follows.  Klebuc J. was satisfied that the appellant would have been able and willing to work, but for his emotional difficulties and resulting dependence on alcohol.  Relying on Statistics Canada data submitted on consent, Klebuc J. estimated that H.L. would have worked as a construction or agricultural labourer 25 weeks annually, during the “first period” (1978-87), earning a total of $27,150.

37                               Klebuc J. found that H.L. would have maintained full‑time employment in automotive repair during the “second period” (1988-2000).  Relying here again on Statistics Canada data, he applied the median rate of $330 per week for all persons engaged in the repair and overhaul of motor vehicles.  He discounted this amount by a 20 percent contingency factor to reflect H.L.’s vulnerability to job loss due to his limited education and cut off this branch of the award at the date of a back injury suffered by H.L.  After deducting the income actually earned by H.L., Klebuc J. estimated a residual loss in earnings of $90,187.09 for the period.

38                               Klebuc J. then considered H.L.’s claim for loss of future earnings and, in the absence of specific evidence in this regard, relied inferentially on the evidence relating to H.L.’s past earning capacity.  He estimated H.L.’s future income, but for Mr. Starr’s misconduct, at no less than $17,160 annually, and deducted H.L.’s average earnings in the past to arrive at an annual income loss of $12,533 for the remainder of H.L.’s projected working life.

IV.    The Court of Appeal

39                               The Court of Appeal dismissed Canada’s appeal as it related to vicarious liability and to the $80,000 award for non‑pecuniary damages, but allowed the appeal in relation to the pecuniary damages and pre‑judgment interest.  H.L.’s cross‑appeal was dismissed except as it related to $6,500 in damages for future care: (2002), 227 Sask. R. 165, 2002 SKCA 131. 

40                               Writing for the court, Cameron J.A. noted that the appeal and cross‑appeal were based on s. 7(2)(a) and s. 13 of the 2000 Act.  In his view, these provisions embody a legislative choice for an unlimited right of appeal, embracing every component of the decision at trial that engages s. 13 of the 2000 Act.

41                               Cameron J.A. accepted the binding authority of Lensen, which was based on s. 8 of The Court of Appeal Act, R.S.S. 1978, c. C‑42 (“1978 Act”).  Cameron J.A. acknowledged that s. 14 of the 2000 Act, which replaced s. 8, differed in syntax but not in substance from its predecessor.  He noted that Lensen had been applied by the Saskatchewan Court of Appeal on innumerable occasions to limit the broad power of appellate review under s. 14 and its predecessor on issues of credibility.  A trial judge’s assessment of credibility, he said, cannot be interfered with on appeal in the absence of palpable and overriding error.

42                               Cameron J.A. was of the opinion, however, that no such limit governs inferences of fact and questions of mixed fact and law.  This, he said, was the traditional view adopted by the Saskatchewan Court of Appeal, as evidenced by Markling v. Ewaniuk, [1968] S.C.R. 776, applied in Kosinski v. Snaith (1983), 25 Sask. R. 73 (C.A.).

43                               Cameron J.A. acknowledged that a set of uniform national standards governing appellate review has evolved in Canada for inferences of fact and questions of mixed fact and law, but considered that Housen had extended the measure of appellate deference traditionally associated with findings of credibility to other components of the decision at trial.  In his view, this trend toward increased deference required reconsideration, especially for Saskatchewan, where the right of appeal and the powers of the court to act on that right are set out in the 2000 Act.

44                               Cameron J.A. regretted that the general standard of appellate review had shifted from appeal by way of rehearing, which he viewed as traditional in Saskatchewan, to the more deferential standard of review for error.

45                               Cameron J.A. suggested that Housen underscores the divide between the current standards of judicially limited appellate review and the broad appellate power granted by the Saskatchewan legislature.

46                               On the merits of the appeal, Cameron J.A. concluded that the award for pecuniary damages lacked an evidentiary foundation and therefore could not stand.  He found the following errors in the trial judge’s awards of $117,337.09 for loss of past earning capacity and $179,190 for loss of future earning capacity:

1.    The trial judge erred in failing to consider the plaintiff’s duty to mitigate.

2.    The trial judge did not take into account the extent to which the defendant Mr. Starr’s wrongful acts contributed to the loss of earnings.  He ought to have had regard for the possibility that H.L. would have been unable to cope with his alcohol-related problems irrespective of the sexual assault by Mr. Starr.

3.    The trial judge awarded H.L. damages for loss of earning capacity while H.L. was incarcerated.  In this regard, Cameron J.A. found that the trial judge had erred in attributing the plaintiff’s criminal behaviour to the wrongdoing of Mr. Starr.

4.    The trial judge did not address the issue of whether the social assistance benefits received by H.L. constituted offsetting collateral benefits.

47                               Acting on its own view of the evidence, the Court of Appeal held that H.L. had not established that he was wholly or largely unable to work because of the sexual abuse by Mr. Starr.  In its view, the evidence simply proved that H.L. did not work during the first period (1978 to 1987) and worked only sporadically during the second period (1988 to 2000).  An inference that Mr. Starr’s abuse caused H.L.’s reduced earning capacity would require more convincing evidence than was adduced in this case.  The court found that H.L.’s sporadic work record was, in itself, as consistent with choice as with disability.

48                               Finally, the court recalled that expert witnesses can provide opinion evidence only on matters within their recognized field of expertise.  Beyond that, their opinion evidence is inadmissible and, if admitted, entitled to no weight.  According to the Court of Appeal, the two expert witnesses in this case were allowed to “roam at large” and to express opinions that they were not qualified to give.

49                               The Court of Appeal thus set aside the award of pecuniary damages on the ground that, on its assessment, the evidence fell short of proving the loss. 

50                               H.L. now appeals to this Court from the decision of the Court of Appeal.

V.     Discussion

51                               The appeal raises two main issues:

1.    What is the correct standard of review by provincial appellate courts on questions of fact, and is that standard different for the Court of Appeal for Saskatchewan?

2.    Did the Saskatchewan Court of Appeal misapply the governing standard to the trial judge’s findings of fact in this case?

A.     The Applicable Standard of Review:  Introduction

52                               Fact finding in the litigation context involves a series of cerebral operations, some simple, others complex, some sequential, others simultaneous.  The entire process is generally reserved in Canada to courts of first instance.  In the absence of a clear statutory mandate to the contrary, appellate courts do not “rehear” or “retry” cases.  They review for error.

53                               The standard of review for error has been variously described.  In recent years, the phrase “palpable and overriding error” resonates throughout the cases.  Its application to all findings of fact — findings as to “what happened” — has been universally recognized; its applicability has not been made to depend on whether the trial judge’s disputed determination relates to credibility, to “primary” facts, to “inferred” facts or to global assessments of the evidence.

54                               Nor has the standard been said to vary according to whether we are concerned with what Hohfeld long ago described as “evidential” or “constitutive” facts (see W. N. Hohfeld, Fundamental Legal Conceptions as Applied in Judicial Reasoning and Other Legal Essays (1923), at p. 32).  Nor, put differently, has the standard been said to vary according to whether our concern is with direct proof of a fact in issue, or indirect proof of facts from which a fact in issue has been inferred.

55                               “Palpable and overriding error” is at once an elegant and expressive description of the entrenched and generally applicable standard of appellate review of the findings of fact at trial.  But it should not be thought to displace alternative formulations of the governing standard.  In Housen, for example, the majority (at para. 22) and the minority (at para. 103) agreed that inferences of fact at trial may be set aside on appeal if they are “clearly wrong”.  Both expressions encapsulate the same principle: an appellate court will not interfere with the trial judge’s findings of fact unless it can plainly identify the imputed error, and that error is shown to have affected the result.

56                               In my respectful view, the test is met as well where the trial judge’s findings of fact can properly be characterized as “unreasonable” or “unsupported by the evidence”.  In R. v. W. (R.), [1992] 2 S.C.R. 122, McLachlin J. (as she then was) explained why courts of appeal must show particular deference to trial courts on issues of credibility.  At the same time, however, she noted (at pp. 131-32) that

it remains open to an appellate court to overturn a verdict based on findings of credibility where, after considering all the evidence and having due regard to the advantages afforded to the trial judge, it concludes that the verdict is unreasonable.

The statutory framework in criminal matters is, of course, different in certain respects.  But as a matter of principle, it seems to me that unreasonable findings of fact —  relating to credibility, to primary or inferred “evidential” facts, or to facts in issue — are reviewable on appeal because they are “palpably” or “clearly” wrong.  The same is true of findings that are unsupported by the evidence.  I need hardly repeat, however, that appellate intervention will only be warranted where the court can explain why or in what respect the impugned finding is unreasonable or unsupported by the evidence.  And the reviewing court must of course be persuaded that the impugned factual finding is likely to have affected the result.

57                               I find it helpful, in concluding on this point, to reproduce Professor  Zuckerman’s summary of the governing principles in England:

As a general principle, an appeal court must not interfere with findings of fact made by the lower court for the simple reason that the judge who saw and heard the witnesses is better placed to assess their reliability and draw inferences from their testimony.  An appeal court will interfere only if it concludes that no reasonable court could have reached such conclusions, or if the lower court failed to take crucial factors into consideration. . . .

                                                                   . . .

. . . It follows that, if the appeal court cannot conclude that the lower court’s inference from the primary facts was wrong, in the sense that it fell outside the range of inferences that a reasonable court could make, the appeal court should allow the lower court’s decision to stand.  The nature of the appellate evaluation of the lower court’s decision will vary in accordance with the type of judgment that the lower court was called upon to make.  But whatever the nature of the issues and however wide or narrow is the room for disagreement, the test remains the same: was the lower court’s decision wrong. . . .

A decision will be wrong if it was founded on an incorrect interpretation of statute, or if it wrongly applied a legal principle, or if it was based on a plainly erroneous factual conclusion. . . . Put another way, as long as the lower court’s conclusions represent a reasonable inference from the facts, the appeal court must not interfere with its decision.

(A. A. S. Zuckerman, Civil Procedure (2003), at pp. 765-68)

58                               Moreover, procedural changes governing civil appeals in England that took effect in May of 2000 do not appear from subsequent decisions of the Court of Appeal to have altered substantially the previous approach to appellate review:

When the Court of Appeal heard appeals on questions of fact [under the old procedure] the court was essentially conducting a review of the findings made by the judge below . . . .  Our task [under the new regime] is essentially no different from what it was — we consider the judgment testing it against the evidence available to the judge and we ask, as we used to ask, whether it was wrong.

(Assicurazioni Generali SpA v. Arab Insurance Group, [2003] 1 W.L.R. 577 (C.A.), per Ward L.J., at para. 195)

In determining whether or not the judgment appealed from was so “wrong”, whether under the new or the old regime,

the appeal court conducting a review of the trial judge’s decision will not conclude that the decision was wrong simply because it is not the decision the appeal judge would have made had he or she been called upon to make it in the court below.  Something more is required than personal unease and something less than perversity has to be established.  The best formulation for the ground in between where a range of adverbs may be used — “clearly”, “plainly”, “blatantly”, “palpably” wrong, is . . . whether that finding by the trial judge exceeded the generous ambit within which reasonable disagreement about the conclusion to be drawn from the evidence is possible.

(Assicurazioni, per Ward L.J., at para. 197)

59                               For present purposes, I find it unnecessary to consider in detail how the standard of appellate review has been applied in England either before or since the reforms that took effect in May of 2000.  I am content with two observations.

60                               First, the passages I have quoted describe the standard of appellate review in England in terms that are fully compatible with both the majority and the minority reasons in Housen.

61                               Second, on any view of the matter, English precedent provides no support for reading into Saskatchewan legislation, past or present, an appellate jurisdiction to “rehear” — in any sense of that term — determinations of fact made at trial.  The English Rules of the Supreme Court, 1883 expressly provided that “[a]ll appeals to the Court of Appeal shall be by way of rehearing.”  The governing statutes in Saskatchewan have never included equivalent or similar language.

B.      Housen v. Nikolaisen

62                               The rules governing appellate intervention in Canada on matters of fact have been set out and reaffirmed in an unbroken line of cases over nearly three decades: Stein v. The Ship “Kathy K”, [1976] 2 S.C.R. 802; Beaudoin-Daigneault v. Richard, [1984] 1 S.C.R. 2; Lensen; Geffen v. Goodman Estate, [1991] 2 S.C.R. 353; Toneguzzo-Norvell (Guardian ad litem of) v. Burnaby Hospital, [1994] 1 S.C.R. 114; Hodgkinson v. Simms, [1994] 3 S.C.R. 377; Schwartz v. Canada, [1996] 1 S.C.R. 254; and Housen.

63                               Housen, like the present case, was an appeal from Saskatchewan where the Court of Appeal had reversed the trial judge.  At issue was the trial judge’s finding that a regional municipality was liable for a portion of the damages caused to the plaintiff in a traffic accident on a rural road.  The Court was divided as to the standard of review applicable to the trial judge’s findings of negligence (a finding of mixed law and fact) and causation (a finding of fact).  In this case, we are concerned only with the standard of review on findings of fact.

64                               All nine justices agreed in Housen that an appellate court ought never to retry a case.  They agreed as well that deference is owed to all findings of fact made by the trial judge, whether those findings are based on direct evidence or on inferences drawn from facts proved directly.

65                               Speaking for the majority, Iacobucci and Major JJ. stated, for example, that “to apply a lower standard of review to inferences of fact would be to depart from established jurisprudence of this Court, and would be contrary to the principles supporting a deferential stance to matters of fact” (para. 19).  Likewise, speaking for the minority, Bastarache J. agreed that “the standard of review is identical for both findings of fact and inferences of fact” (para. 103 (emphasis added)).

66                               It was in the application of this shared view as to the governing principle that the Court divided.

67                               Speaking for the majority, Iacobucci and Major JJ. held that all findings of fact, whether based on direct or circumstantial evidence, are only reviewable on a standard of palpable and overriding error.  In their view, a panoply of policy reasons command appellate deference.  These include the need to limit the cost of litigation and to promote the autonomy of trial proceedings, two reasons that are unrelated to the superior vantage point of the trial judge in hearing viva voce evidence.

68                               Bastarache J. did not dilute, still less abandon, the principle of deference with respect to findings of fact based on inferences.  In his view, however, an inference was reviewable if it was not “reasonably . . . supported by the findings of fact that the trial judge reached”:

While the standard of review is identical for both findings of fact and inferences of fact, it is nonetheless important to draw an analytical distinction between the two.  If the reviewing court were to review only for errors of fact, then the decision of the trial judge would necessarily be upheld in every case where evidence existed to support his or her factual findings.  In my view, this Court is entitled to conclude that inferences made by the trial judge were clearly wrong, just as it is entitled to reach this conclusion in respect to findings of fact. [Emphasis added; para. 103.]

69                               As I have already mentioned, there is no meaningful difference between a standard of “clearly wrong” and a standard of “palpable and overriding error”.  As Iacobucci and Major JJ. noted in Housen, at para. 5, the New Oxford Dictionary of English (1998) defines “palpable” as “clear to the mind or plain to see” (p. 1337 (emphasis added)).  Moreover, no error could lead to a reversal unless it was “overriding” in the sense that it discredits the result.

70                               The “palpable and overriding error” standard, apart from its resonance, nevertheless helps to emphasize that one must be able to “put one’s finger on” the crucial flaw, fallacy or mistake.  In the words of Vancise J.A., “[t]he appellate court must be certain that the trial judge erred and must be able to identify with certainty the critical error” (Tanel, at p. 223, dissenting, though not on this issue).

71                               And yet, again as indicated earlier, I agree with Bastarache J. that there is no meaningful difference between

concluding that it was “unreasonable” or “palpably wrong” for a trial judge to draw an inference from the facts as found by him or her and concluding that the inference was not reasonably supported by those facts.

(Housen, at para. 104)

72                               I have not overlooked that, according to the majority in Housen, the test to be applied in reviewing inferences of fact is “not to verify that the inference can be reasonably supported by the findings of fact of the trial judge, but whether the trial judge made a palpable and overriding error in coming to a factual conclusion based on accepted facts” which, in its view, implied a stricter standard (para. 21 (emphasis in original)).  The apparent concern of the majority was that, in drawing an analytical distinction between factual findings and factual inferences, the minority position might lead appellate courts to involve themselves in reweighing the evidence (para. 22).  As well, the majority stated:

If there is no palpable and overriding error with respect to the underlying facts that the trial judge relies on to draw the inference, then it is only where the inference‑drawing process itself is palpably in error that an appellate court can interfere with the factual conclusion. [Emphasis in original; para. 23.]

73                               These passages from the majority reasons in Housen should not be taken to have decided that inferences of fact drawn by a trial judge are impervious to review though unsupported by the evidence.  Nor should they be taken to have restricted appellate scrutiny of the judge’s inferences to an examination of the primary findings upon which they are founded and the process of reasoning by which they were reached.

74                               I would explain the matter this way.  Not infrequently, different inferences may reasonably be drawn from facts found by the trial judge to have been directly proven.  Appellate scrutiny determines whether inferences drawn by the judge are “reasonably supported by the evidence”.  If they are, the reviewing court cannot reweigh the evidence by substituting, for the reasonable inference preferred by the trial judge, an equally — or even more — persuasive inference of its own.  This fundamental rule is, once again, entirely consistent with both the majority and the minority reasons in Housen.

75                               In short, appellate courts not only may — but must — set aside all palpable and overriding errors of fact shown to have been made at trial.  This applies no less to inferences than to findings of “primary” facts, or facts proved by direct evidence.

76                               Courts of appeal across Canada, despite their understandable concern over these passages, have well understood the central message of three decades of jurisprudence in this Court, culminating in Housen.  They have generally applied the palpable and overriding error standard to all findings of fact made at trial — albeit with varying degrees of enthusiasm.

C.     The Applicability of Housen in Saskatchewan

77                               We are urged to find on this appeal that the rule governing appellate intervention on matters of fact differs in Saskatchewan from the rest of Canada.

78                               Housen was an appeal from the Court of Appeal for Saskatchewan, but did not refer to the 2000 Act or its predecessors.  On the strength of this “omission”, it is now argued that the Court in Housen misapprehended the scope of appellate review in Saskatchewan.

79                               This contention rests on three propositions.  First, it is suggested that the 2000 Act, like its predecessors, vests in the Court of Appeal for Saskatchewan a broader jurisdiction than is conferred by corresponding legislation on appellate courts  elsewhere in Canada.  Second, it is argued that the Court of Appeal for Saskatchewan,  at least prior to Housen, had consistently interpreted its governing statute as granting a larger scope of review than Housen permits.  Finally, it is contended that it was the intention of the Saskatchewan legislature in 2000, when it amended The Court of Appeal Act, to clarify that appellate review in that province was to proceed by way of rehearing.

80                               None of these propositions is firmly rooted in fact or in law.  An examination of both the former and present Acts, their legislative history, and their judicial interpretation in this Court and by the Saskatchewan Court of Appeal itself all lead to the same conclusion:  appellate review in Saskatchewan has for a long time proceeded, and continues to proceed, on essentially the same basis as appellate review elsewhere in Canada.  The appeal is a review for error, and not a review by rehearing.

D.     The Saskatchewan Court of Appeal Act

81                               These are the provisions of the 2000 Act that set out the powers of the Court of Appeal for Saskatchewan:

12(1)  On an appeal, the court may:

(a) allow the appeal in whole or in part;

(b) dismiss the appeal;

(c) order a new trial;

(d) make any decision that could have been made by the court or tribunal appealed from;

(e) impose reasonable terms and conditions in a decision; and

(f) make any additional decision that it considers just.

(2)  Where the court sets aside damages assessed by a jury, the court may assess any damages that the jury could have assessed.

13  Where issues of fact have been tried, or damages have been assessed, by a trial judge without a jury, any party is entitled to move against the decision of the trial judge, by motion for a new trial or otherwise:

                                                                   . . .

(b) on the same grounds, including objections against the sufficiency of the evidence, or the view of the evidence taken by the trial judge, that are allowed in cases of trial or assessment of damages by a jury.

14  On an appeal from, or on a motion against, the decision of a trial judge or on any rehearing, the court is not obliged to grant a new trial or to adopt the view of the evidence taken by the trial judge, but the court shall act on its own view of what, in its judgment, the evidence proves, and the court may draw inferences of fact and pronounce the decision that, in its judgment, the trial judge ought to have pronounced.

Section 14 is of particular interest on this appeal.  Cameron J.A. was of the view that it frees the Court of Appeal from the view of the evidence taken by the trial judge, and entitles it to draw its own inferences of fact.

82                               While s. 14 refers to a “rehearing”, it is clear from the context of the Act that this does not confer on the Court of Appeal the power to “rehear” trials.  It simply provides that the powers available to the court on an appeal are available on the rehearing of an appeal.  The term “rehearing” is used in the 2000 Act in s. 16, which states that the court shall rehear an appeal in certain circumstances, for example, where this is made necessary by the death or resignation of two or more of the judges who heard the initial appeal.  As the then Minister of Justice of Saskatchewan explained on second reading, the 2000 Act (then Bill 80) “clarifies the procedure respecting rehearings”, which are to take place if a rehearing of the appeal is required for reasons mentioned below (Saskatchewan Hansard, at p. 1626).

83                               Though the statute uses more specific language, it is similar in effect to the corresponding statutes in other provinces and territories as regards the issue that concerns us here — authority to review primary findings of fact and inferences.

84                               Thus, for example, a review of other provincial statutes reveals that British Columbia, Alberta, Manitoba, Ontario, and Prince Edward Island all explicitly allow their courts of appeal to “draw inferences of fact”: Court of Appeal Act, R.S.B.C. 1996, c. 77, s. 9(2); Alberta Rules of Court, Alta. Reg. 390/68, r. 518(c); The Court of Appeal Act, R.S.M. 1987, c. C240, s. 26(2); Courts of Justice Act, R.S.O. 1990, c. C.43, s. 134(4)(a); Supreme Court Act, R.S.P.E.I. 1987, c. 66, s. 56(4)(a).

85                               Alberta, Manitoba, Prince Edward Island and Ontario also grant their respective courts of appeal the power to arrive at the decision the trial judge “ought” to have made: Alberta Rules of Court, r. 518(e); The Court of Appeal Act (Man.), s. 26(1) ; Supreme Court Act (P.E.I.), s. 56(1)(a); Courts of Justice Act (Ont.), s. 134(1)(a).

86                               Quebec confers on its Court of Appeal “all powers necessary” to the exercise of its jurisdiction (Courts of Justice Act, R.S.Q., c. T-16, s. 10) — a general power that could hardly have been expressed in broader terms — while the Atlantic provinces, the Northwest Territories, and Nunavut grant jurisdiction consistent with dates prior to the passing of their respective Acts.

87                               In this light, I think it evident that the jurisdiction of the Saskatchewan Court of Appeal to review inferences of fact drawn by the trial judge is hardly exceptional, let alone unique.  Other provincial or territorial courts of appeal are granted similar powers, expressly or implicitly, by their governing statutes.  The 2000 Act simply sets out those powers in more detail than some. A detailed enunciation of the powers conferred does not signify a legislative intent that they be more expansively exercised.

88                               The Saskatchewan Court of Appeal is explicitly empowered to take its own view of what the evidence proves, to draw inferences of fact and to pronounce any decision that the trial judge ought to have pronounced.  I do not think it open to question that other provincial appellate courts are endowed with these very same powers.  But the scope of the powers conferred must not be confused with the manner in which they are to be exercised.  In Harrington v. Harrington (1981), 33 O.R. (2d) 150, Morden J.A., speaking for the court in a different context but on this very point, stated:

The purpose of s. 17(2) (b)(i) of the Divorce Act , which enables us to “pronounce the judgment that ought to have been pronounced” is to prescribe the general kind of disposition open to us, on allowing an appeal, as an alternative to ordering a new trial . . . and is not intended, in my view, to provide the rule governing when we will interfere with the challenged judgment, i.e., it does not set forth the standard for determining whether or not the challenged judgment should be set aside. [Emphasis in original; pp. 154-55.]

Harrington was expressly endorsed by this Court in Pelech v. Pelech, [1987] 1 S.C.R. 801, at p. 824.

89                               Neither in Saskatchewan nor elsewhere in Canada may courts of appeal, absent an express legislative instruction to the contrary, disregard the governing principle of appellate intervention on questions of fact. They may indeed make their own findings and draw their own inferences, but only where the trial judge is shown to have committed a palpable and overriding error or made findings of fact that are clearly wrong, unreasonable, or unsupported by the evidence.

90                               As I stated at the outset, the 2000 Act neither bestows on the Court of Appeal for Saskatchewan unique powers of appellate intervention on questions of fact nor ordains their exercise in a manner that, within Canada, is exclusive to Saskatchewan. 

E.      The Judicial Treatment of The Court of Appeal Act in Saskatchewan

91                               Prior to any intervention by this Court as to the appropriate standard of appellate review on questions of fact in Saskatchewan, the Court of Appeal’s own case law under the 1978 Act, the predecessor to the 2000 Act that concerns us here, was entirely consistent with the principles elaborated in Lensen and Housen.

92                               The leading case in the province was Board of Education of the Long Lake School Division No. 30 of Saskatchewan v. Schatz (1986), 49 Sask. R. 244, where the Court of Appeal found that a standard of palpable and overriding error applied to a trial judge’s findings of fact.  Sherstobitoff J.A., for himself and for Tallis J.A., provided an extensive, detailed and definitive analysis of the Court of Appeal’s decisions concerning its jurisdiction to review findings of fact.  Dealing specifically with s. 8 of the 1978 Act, Sherstobitoff J.A. stated, at p. 248:

While, on its face, s. 8 appears to confer not only the power, but a duty to “rehear” or “retry” a case, simple fairness and justice require a court of appeal to recognize that a trial judge has an immense advantage in assessing evidence and arriving at findings of fact as opposed to a court of appeal which is confined to an examination of a cold black and white record of a trial proceeding, completely devoid of the tension, emotion, colour, and atmosphere of a trial, all of which factors are immeasureably important in assisting a trial judge in arriving at his conclusions. It is for these reasons that a court of appeal must extend very substantial deference to the finding of facts of a trial judge. The issue has been considered on many occasions by the Supreme Court of Canada and its decisions bear these principles out. [Emphasis added.]

The Court of Appeal in Long Lake School Division expressly adopted the jurisprudence of this Court setting out general standards of appellate review applicable to questions of fact, in particular citing Stein v. The Ship “Kathy K”, a classic enunciation of the principle of appellate deference to the findings of fact at trial.

93                               Nor can it be contended, as Cameron J.A. suggested in the present case, that the Court of Appeal had traditionally distinguished, in considering its powers of review, between primary findings and inferences.  On the contrary, in Long Lake School Division, Sherstobitoff J.A. did not restrict his guidelines for appellate deference to matters of credibility or to the review of primary findings.  He set out instead a general guideline: “[w]here there is evidence to support a finding of fact a court of appeal should not interfere in the absence of palpable or demonstrable error” (p. 251).

94                               Similarly, in Tanel, Bayda C.J.S. set out the test for appellate review of findings of fact in these terms: “first, is there evidence to support the trial judge’s findings of fact; and second, is there an absence of palpable or demonstrable error?” (p. 218).  Later, Bayda C.J.S. referred to the “‘task of great and almost insuperable difficulty’ (per Lord Sumner [in S.S. Hontestroom v. S.S. Sagaporack, [1927] A.C. 37 (H.L.)]) that any appellant faces in trying to convince an appellate court to overturn a trial judge’s finding of fact” (p. 220).  In the same case, Vancise J.A. summarized the standard of review aptly, at p. 223:

For an appellate court to intervene in respect to findings of fact by a trial judge and to modify or substitute those findings of fact there must be palpable and overriding error.  The appellate court must be certain that the trial judge erred and must be able to identify with certainty the critical error.

95                               Yet again, and still prior to this Court’s decision in Lensen, Cameron J.A., for the court, applied a palpable and overriding error standard to the trial judge’s conclusion that the plaintiffs had not relied on the defendant’s misrepresentation, as it was “open to him on the evidence”: see Sisson v. Pak Enterprises Ltd. (1987), 64 Sask. R. 232, at p. 235.

96                               Speaking for the Court in Lensen, Dickson C.J. thus adopted the Court of Appeal’s own synthesis of its jurisprudence in Long Lake School Division, when he stated (at pp. 683-84):

It is a well‑established principle that findings of fact made at trial based on the credibility of witnesses are not to be reversed on appeal unless it can be established that the trial judge made some “palpable and overriding error which affected his assessment of the facts” . . . . While section 8 of the Saskatchewan Court of Appeal Act authorizes the Court of Appeal to “draw inferences of fact”, this task must be performed in relation to facts as found by the trial judge.  Unless the trial judge has made some “palpable and overriding error” in this regard, s. 8 should not be construed so as to modify the traditional role of the Court of Appeal with respect to those findings.

97                               In short, far from proceeding by way of rehearing, the Court of Appeal for Saskatchewan appears to have for many decades prior to both Lensen and Housen understood its legislative mandate as a power of review for error.  The court consistently and repeatedly held that it was authorized to intervene in a trial judge’s findings of fact only where palpable and overriding error was shown.

98                               No decision has been drawn to our attention where the court has asserted a power of review by rehearing.

F.      The Effect of the 2000 Amendments to The Court of Appeal Act

99                               It was argued by the Attorney General for Saskatchewan that amendments made to The Court of Appeal Act in 2000 call for a reconsideration of the principles of appellate review applicable in Saskatchewan. 

100                           Prior to those amendments, as we have just seen, a standard of palpable and overriding error had been applied with relative consistency to appellate review of findings of fact made at trial.  Neither a plain reading of the 2000 Act, nor the legislative history of the amendments, indicate a legislative intention to change that standard.

101                           As we shall see, moreover, the Court of Appeal itself did not view the 2000 Act, after its adoption, as warranting a departure from Lensen, or from its own case law prior to the decision of our Court in that case.

102                           Section 14 of the current 2000 Act is the successor to s. 8 of the 1978 Act.  The two provisions are best viewed together:

14 [Powers of court re evidence] On an appeal from, or on a motion against, the decision of a trial judge or on any rehearing, the court is not obliged to grant a new trial or to adopt the view of the evidence taken by the trial judge, but the court shall act on its own view of what, in its judgment, the evidence proves, and the court may draw inferences of fact and pronounce the decision that, in its judgment, the trial judge ought to have pronounced.

8. [Court not bound by view of evidence taken by trial judge] Upon appeal from, or motion against, the order, decision, verdict or decree of a trial judge, or on the rehearing of any cause, application or matter, it shall not be obligatory on the court to grant a new trial, or to adopt the view of the evidence taken by the trial judge, but the court shall act upon its own view of what the evidence in its judgment proves, and the court may draw inferences of fact and pronounce the verdict, decision or order that, in its judgment, the judge who tried the case ought to have pronounced.

103                            Section 14, which came into effect on November 1, 2000, is identical in substance to the former s. 8, though the drafting of the provision has been modernized.  While s. 14 refers only to a “decision” (rather than to an “order, decision, verdict or decree”), “decision” is defined in the 2000 Act to include an order, verdict or decree.  Even in this regard, s. 14 thus corresponds in substance to the former provision.

104                            In other respects, the two provisions are indistinguishable.  Section 14 of the 2000 Act merely rephrases its predecessor in plainer English.  This should cause no surprise, given the legislative history of the amendments.

105                            On second reading of the 2000 Act, Mr. Axworthy emphasized that the amendments were primarily intended to restate the historical jurisdiction of the court in modern language, and to facilitate its translation into French:

Hon. Mr. Axworthy: — Thank you, Mr. Speaker.  I rise today to move second reading of The Court of Appeal Act, 2000.  Mr. Speaker, The Court of Appeal Act was first passed when the court was created in 1915, and a number of provisions in the Act have remained unchanged since that time.  Therefore, Mr. Speaker, there’s a need to update and clarify some of these provisions.

The present section of the Act relating to jurisdiction is incomprehensible to anyone other than a legal historian.  The Bill before the House doesn’t change the jurisdiction of the Court of Appeal in any way, it simply restates the historical jurisdiction of the court in a way that can be understood by users of the Act.

The legislature will be asked to approve the re‑enactment of The Court of Appeal Act in both French and English . . . .

Mr. Speaker, the English version of this Bill required revision for clarification purposes before the French translation could go forward.  As well as adopting gender‑neutral language, this update of The Court of Appeal Act substantially improves the law by making it clear and more understandable, even to my own colleagues, Mr. Speaker.

And finally:

. . . the Bill clarifies the procedure respecting rehearings.  It states that the court shall rehear an appeal if due to death or resignation, only one judge who heard the appeal remains.  As well, if the number of judges is reduced to an even number that is deeply divided on a matter, a party to the appeal may apply for a rehearing.

(Saskatchewan Hansard, at pp. 1625-26 (emphasis added))

106                            Though of limited weight, Hansard evidence can assist in determining the background and purpose of legislation: Rizzo & Rizzo Shoes Ltd. (Re), [1998] 1 S.C.R. 27, at para. 35.  In this case, it is particularly apposite, since it was contended by the Attorney General for  Saskatchewan, an intervener in this Court, that the legislature’s purpose in revising The Court of Appeal Act was to “clarify” that the Court of Appeal was to be placed “in a position of conducting an appeal by rehearing”.

107                            Here, too, I find instructive the Saskatchewan Court of Appeal’s own interpretation of its constituent statute.  It does not appear, even prior to this Court’s judgment in Housen, to have understood the 2000 Act as an enlargement of its powers of review on questions of fact under the previous Act, as interpreted by the court itself. In Knight v. Huntington (2001), 14 B.L.R. (3d) 202, 2001 SKCA 68, at para. 28, Sherstobitoff J.A., speaking for the court, applied the palpable and overriding error standard to findings of credibility made and inferences of fact drawn by the trial judge:

In this case, much of the trial judge’s finding of fact depended primarily upon assessments of the relative credibility of the witnesses.  To that extent, his findings cannot be interfered with unless the appellants can show that there was a palpable and overriding error.  Further, to the extent that his findings depended upon drawing inferences of fact, the appellants must show that there was no evidence from which those conclusions could reasonably be drawn. [Emphasis added.]

Knight was heard by the Saskatchewan Court of Appeal in May of 2001, some six months after the 2000 Act had come into force.  R. Sullivan, Sullivan and Driedger on the Construction of Statutes (4th ed. 2002), states the “common law presumption that procedural legislation applies immediately and generally to both pending and future facts” (p. 582).  This common law rule has been codified by Saskatchewan’s Interpretation Act, 1995, S.S. 1995, c. I-11.2, s. 35 (am. S.S. 1998, c. 47, s. 6).

108                            Similarly, in Bogdanoff v. Saskatchewan Government Insurance (2001), 203 Sask. R. 161, 2001 SKCA 35, Gerwing J.A., in oral reasons for the court, applied a palpable and overriding error standard to a finding of causation made by the trial judge.  The appeal was heard more than three months after the 2000 Act came into effect.  In Brown v. Zaitsoff Estate (2002), 217 Sask. R. 130, 2002 SKCA 18, decided almost a year later, Tallis J.A. applied the same standard, again for a unanimous court.

109                            In none of these decisions was there any suggestion that the 2000 Act had enlarged the scope of appellate review of findings of fact in Saskatchewan.  Nor was the 2000 Act mentioned at all.

G.     The Standard of Appellate Review: Conclusion

110                            With respect, I do not find persuasive any of the arguments advanced in support of the contention that the rules governing appellate intervention in Saskatchewan differ from those set out in Housen.  On the contrary, I am satisfied for the reasons given that the standard of review for inferences of fact, in Saskatchewan as elsewhere in Canada, is that of palpable and overriding error and its functional equivalents, including “clearly wrong”, “unreasonable” and “not reasonably supported by the evidence”.

H.     Application of the Standard of Review

111                            The Court of Appeal reversed the trial judge on six points that are at issue in this appeal: (1) qualification of the experts, (2) causation, (3) mitigation, (4) incarceration, (5) collateral benefits, and (6) loss of future earnings.  In my respectful view, the Court of Appeal erred in interfering with the trial judge’s findings on the first three issues.  I agree, however, that the trial judge erred in awarding H.L. damages for lost earnings for the time he spent in prison, in failing to deduct the social assistance received by H.L. from the award for loss of past earnings, and in granting an award for loss of future earnings.

(1)  The Expert Evidence

112                            The trial judge based his conclusion that Mr. Starr’s sexual abuse of H.L. caused H.L.’s alcoholism on the evidence adduced before him, including that of the experts called by the parties.  The Court of Appeal, in my view, erred in substituting its own opinion of that evidence for that of the trial judge and in interfering with the judge’s conclusion on this issue.

113                            Cameron J.A. found that “the two witnesses [the experts] were pretty much allowed to roam at large, expressing all manner of opinion in relation to which they were not formally qualified” (para. 255).  Specifically, Cameron J.A. felt that the experts should not have been allowed to speak to the cause of H.L.’s alcoholism (para. 256).  He then concluded that, in the absence of this expert evidence, there was no basis for an inference that Mr. Starr’s abuse had caused H.L.’s alcoholism and consequent loss (para. 258).

114                            I am unable to share this view.  Both experts testified that H.L.’s sexual abuse by Mr. Starr had caused his alcoholism.  Both were qualified to speak to the  long-term psychological effects of that sexual abuse.  Both had extensive clinical and professional experience in that area, and both had tested H.L. and interviewed him extensively.  Characterizing the testimony of the experts as evidence concerning the etiology of alcoholism in general ignores its real content and true import:  rather than appreciating the experts’ testimony for its relevance, purpose and significance as evidence of the effects of Mr. Starr’s tortious conduct on H.L. himself, Cameron J.A. misapprehended it as testimony about the causes of alcoholism generally.

115                            Both experts were psychologists with extensive knowledge and experience concerning sexual abuse.  They were qualified to speak to the effects of such abuse, including substance abuse.  Both testified that Mr. Starr’s abuse bore a causal relationship to H.L.’s substance abuse; the difference in their respective opinions related only to the extent of that causal relationship in the circumstances of this case. 

116                            Moreover, the difference in their opinions had no bearing on the liability of Mr. Starr or Canada for the damages found by the trial judge to have been suffered by H.L.: Dr. Arnold stated that H.L.’s family life enhanced his vulnerability to alcoholism, but nonetheless described the abuse as the “specific triggering event”, without which H.L.’s pre-existing vulnerability may not have caused him harm.  In Dr. Arnold’s opinion, we just “don’t know” what would have happened to H.L. had he not suffered abuse at the hands of Mr. Starr, because, in fact, he did.

117                            With respect, it is neither accurate nor helpful to say that the trial judge allowed the experts to “roam at large”.  On the contrary, they were “reigned in” by the trial judge upon proper objections by counsel, for example on the issue of H.L.’s “earning capacity”.

118                            To sum up, then, both experts testified on direct examination that the abuse H.L. experienced bore a causal relationship to his substance abuse.  The respondent’s position on that issue is therefore unacceptable: In effect, the respondent seeks to disavow in this Court the evidence he himself adduced at trial on the ground that his own witness, Dr. Arnold, was not qualified to answer the questions he himself put to the witness without objection by opposing counsel. 

119                            Dr. Arnold’s answers were indeed detrimental to the respondent’s case.  But it is too late in the day to contend that Dr. Arnold was not qualified to speak to the relationship between the sexual abuse inflicted on H.L. and his ensuing problems — the very issue upon which he was examined deliberately by counsel who called him.

120                            I would therefore allow the appeal on this branch of the matter, since the trial judge did not err in qualifying the witnesses, in making findings on their relative credibility or in relying on their expert opinions.

(2)  Causation

121                            In my respectful view, the Court of Appeal erred in setting aside the trial judge’s findings on causation.

122                            Causation is a factual inference: Housen, at paras. 70 and 75 of the majority reasons and paras. 111 and 159 of the minority reasons.

123                            This Court explained the test for causation in Athey, at paras. 13-19:

Causation is established where the plaintiff proves to the civil standard on a balance of probabilities that the defendant caused or contributed to the injury: Snell v. Farrell, [1990] 2 S.C.R. 311; McGhee v. National Coal Board, [1972] 3 All E.R. 1008 (H.L.).

The general, but not conclusive, test for causation is the “but for” test, which requires the plaintiff to show that the injury would not have occurred but for the negligence of the defendant: Horsley v. MacLaren, [1972] S.C.R. 441.

The “but for” test is unworkable in some circumstances, so the courts have recognized that causation is established where the defendant’s negligence “materially contributed” to the occurrence of the injury.

                                                                   . . .

The law does not excuse a defendant from liability merely because other causal factors for which he is not responsible also helped produce the harm . . . . It is sufficient if the defendant’s negligence was a cause of the harm. . . . [Emphasis in original.]

124                            The causal question at issue here is whether Mr. Starr’s sexual abuse of H.L. reduced  H.L.’s employment income during the first and second periods.  The trial judge answered that question in the affirmative (para. 65).  He drew this inference  from both experts’ opinions that H.L.’s alcoholism was caused by the sexual abuse; from their opinions that sexual abuse results in a loss of self‑esteem and self‑confidence which, in turn, affect employability or “work ethic”; and from the evidence of H.L. himself.

125                            The trial judge based his assessment of damages on the finding that the sexual abuse by Mr. Starr caused H.L.’s  emotional difficulties and alcoholism, which in turn caused his inability to secure and maintain full‑time employment.  Ultimately, then, the question is whether this was a reasonable inference on the facts as found by the trial judge.

126                            The experts in this case gave their evidence regarding (1) the link between H.L.’s sexual abuse and his emotional problems and alcoholism, and (2) the link between H.L.’s low self‑esteem and self-confidence and his reduced employability.  The opinion of an expert was not necessary to make the link between H.L.’s alcoholism and his reduced ability to sustain remunerative employment.  That link, which might appear to be a matter of common experience to many, was nonetheless provided by H.L. himself.

127                            The first inference drawn by the trial judge was that the sexual abuse caused H.L.’s emotional problems and alcoholism.  Both experts testified that sexual abuse would cause a victim to have a negative self‑image and a lack of self-confidence.  As we saw earlier, H.L. also testified that the abuse made him feel humiliated and ashamed, caused him to lose concentration, and led to his withdrawal from schooling at an early stage.

128                            Both experts also identified the abuse as having triggered H.L.’s excessive drinking and addiction to alcohol.  As I have already suggested, on either expert’s testimony, the test in Athey was met.

129                            The second inference drawn by the trial judge was that H.L.’s emotional problems and alcohol abuse reduced his capacity to secure and retain employment.  On this point, Dr. Arnold, the defendant Canada’s expert, testified on cross‑examination by counsel for H.L. that sexual abuse would affect a victim’s “work ethic”:

Q:   Would you think it likely that [sexual abuse by someone associated with the school system] would have affected [a victim’s] work ethic?

A:   Work ethic as in — perhaps to define that, I think what you’re saying is his ability to hold work and be able to regularly show up and those kinds of things?

Q:   Yes.

A:   Yes, and I would refer to the chain of events I just referred to.  You have an event, then — sorry, an event — I better be clear here — event of abuse, you have alcohol and, yes, indeed that chain of events would logically go there and —

No objection was taken to this testimony. 

130                            Mr. Stewart was asked on direct examination by counsel for H.L. whether self‑esteem and self‑confidence affected employability, and answered that he was “sure it would, yes”.  This testimony was allowed by the court, despite the objection taken by counsel for Canada, on the basis that it was within the realm of Mr. Stewart’s (and Dr. Arnold’s) expertise.

131                            Canada’s earlier objection to questions put in chief to Mr. Stewart regarding H.L.’s earning capacity had been sustained on the basis that Mr. Stewart was not a vocational expert.  Without endorsing that finding, I find it sufficient to mention that both experts were allowed to express their opinions whether the emotional problems caused by Mr. Starr’s abuse affected H.L.’s ability to find and keep a job, but not whether they reduced his earning capacity when he did secure employment.

132                            In addition, as already mentioned, there was an evidentiary basis for the trial judge’s finding that alcoholism had affected H.L.’s earning capacity.  H.L. himself testified that he was unable to sustain employment for more than five or six months due to his drinking problem, and that his lack of education, criminal record and alcoholism deterred employers from hiring him.  This was, of course, a matter within H.L.’s personal experience, and the trial judge was entitled to give it appropriate weight. 

133                            There was thus sufficient evidence on the record to support the trial judge’s findings that the sexual abuse of H.L. by Mr. Starr caused emotional problems and alcoholism, which in turn hindered H.L.’s efforts to hold down a job.  On this evidence, a reasonable trier of fact could draw a causal inference.  The trial judge therefore committed no reviewable error in awarding damages for loss of past earnings, and the Court of Appeal erred in setting aside that award.

(3)  Loss of Past Earnings: Mitigation

134                             The onus rests on the defendant to prove that the plaintiff failed to mitigate his loss: Janiak v. Ippolito, [1985] 1 S.C.R. 146, at p. 163.  Here, the trial judge concluded that the Crown led no evidence on the issue of mitigation.  The Court of Appeal pointed to H.L.’s failure to upgrade his education and training as well as his failure to enter rehabilitation as evidence that he failed to mitigate his loss (para. 232).

135                            H.L. testified that he failed to upgrade his education because he had a poor memory and dropped out of an auto mechanics course after two months.  This is consistent with the trial judge’s finding that H.L.’s alcoholism, poor self‑image and lack of confidence affected his ability to learn a trade and his ability to find and keep a job.  This does not point to a failure to mitigate.  And though the record is essentially silent regarding H.L.’s efforts at rehabilitation, it appears from his evidence at trial that he was at least then making an effort to abstain from any further consumption of alcohol.

136                            Since the evidence as to H.L.’s mitigation of his damages was inconclusive at best, Canada’s burden had not been discharged.  The Court of Appeal therefore erred in reversing the trial judge’s finding on this issue.

(4)  Loss of Past Earnings: Incarceration

137                            In calculating H.L.’s loss of past earnings, the trial judge did not reduce the damages awarded to reflect the time H.L. spent in prison.  The Court of Appeal intervened in this respect — quite properly, in my view.  As Cameron J.A. noted, to compensate an individual for loss of earnings arising from criminal conduct undermines the very purpose of our criminal justice system (paras. 240‑41); an award of this type, if available in any circumstances, must be justified by exceptional considerations of a compelling nature and supported by clear and cogent evidence of causation.

138                            The trial judge inferred that H.L.’s alcohol abuse, which was caused by the sexual abuse, “led to [his] numerous convictions on alcohol and theft related offences” (para. 29).  As already noted, the inference that sexual abuse caused H.L.’s alcoholism is supported by the evidence.  It is the relationship between H.L.’s alcoholism and his loss of earnings due to imprisonment that is the focus of my concern here: The question before the trial judge was not whether H.L. had committed certain crimes while drunk, but whether his ensuing incarceration was caused by his addiction to alcohol.

139                            In examination‑in‑chief by H.L.’s counsel, Mr. Stewart testified that there is a relationship between sexual abuse and criminal conduct, in “that a number of individuals — in fact a wide number of individuals, I don’t have the exact number, who have been either physically or sexually abused in childhood, a great proportion of those end up being abusers themselves once they reach adulthood”.

140                            In cross‑examination, Mr. Stewart explained that his statement concerned the likelihood that a child who is sexually abused will become an abuser as an adult.  None of H.L.’s periods of incarceration related to charges of sexual abuse.

141                            The expert evidence did not disclose a more general link between sexual abuse and criminality.  Nor did the materials before the trial judge entitle him to conclude that those suffering from alcoholism were more inclined to commit crimes.

142                            In any event, the chain of causation linking H.L.’s sexual abuse to his loss of income while incarcerated was interrupted by his intervening criminal conduct.  During these periods, his lack of gainful employment was caused by his imprisonment, not by his alcoholism; and his imprisonment resulted from his criminal conduct, not from his abuse by Mr. Starr nor from the alcoholism which it was found to have induced.

143                            Thus, on any view of the matter, the trial judge’s finding that Mr. Starr’s sexual abuse of H.L. caused his loss of income due to imprisonment is both contrary to judicial policy and unsupported by the evidence. 

144                            I would therefore dismiss H.L.’s appeal under this head.

(5)  Loss of Past Earnings: Social Assistance

145                            The Court of Appeal found, again correctly in my view, that the trial judge had erred in not deducting from H.L.’s award for loss of past earnings the social assistance payments he had received during the relevant period.

146                            Klebuc J. found that H.L. “generally relied on social assistance to meet his needs” during the first period for which he assessed damages for lost earnings, but did not account for those or any other social assistance payments in fixing his award (para. 64). With respect to the second period for which Klebuc J. assessed damages for lost earnings, he did, however, deduct the income earned by H.L.

147                            This Court recently had occasion to consider whether social assistance payments are to be deducted from damage awards for lost earnings in M.B. v. British Columbia, [2003] 2 S.C.R. 477, 2003 SCC 53.  In that case, McLachlin C.J. affirmed the “common sense proposition that social assistance benefits are a form of wage replacement” and deductible at common law to avoid double recovery (para. 28).

148                            Klebuc J. did not have the benefit of this Court’s decision in M.B.  His understandable — but nonetheless erroneous — failure to deduct social assistance benefits constitutes a severable error of principle: see Canada (Director of Investigation and Research) v. Southam Inc., [1997] 1 S.C.R. 748, and Housen.

149                            Unfortunately, the amount of social assistance received by H.L. during the first and second periods is not available on the record.  In the absence of agreement between the parties, this calculation must therefore be left to the trial court for proof and determination.

(6)  Loss of Future Earnings

150                            Finally, the trial judge awarded H.L. $179,190 for loss of future earnings.  The Court of Appeal set this award aside on the basis of what it found to be factual errors by the trial judge.  With respect, I do not share the Court of Appeal’s findings of factual error, but I do agree that the trial judge’s disposition on this branch of the award lacked an evidentiary basis — quite unlike his award for loss of past earnings, which was supported by the evidence of H.L. and the expert witnesses called by H.L. and Canada. 

151                            In quantifying the damages for loss of future earnings, the trial judge acknowledged explicitly that the parties had presented no evidence regarding H.L.’s future earning capacity (para. 70).

152                            The finding that a person has had emotional and substance abuse problems which in the past have impacted on his earning capacity is not in itself a sufficient basis for concluding on the balance of probabilities that this state of affairs will endure indefinitely.  To assume, without additional evidence, that H.L. will continue to suffer from substance abuse and emotional problems, will not upgrade his education or enter into rehabilitation, and will continue to have a reduced earning capacity, would be to do him an unnecessary and unwarranted disservice — particularly in the light of his own evidence that he had already at the time of trial taken steps to end his addiction to alcohol.

VI.    Disposition

153                            For all of these reasons, I would allow the appeal in part, with costs. 

154                            I would confirm the trial judge’s award of pecuniary damages for loss of past earnings, but order that they be reduced to reflect the time the appellant spent in prison and the social assistance he received during the period covered by the award.  In the absence of an agreement between the parties as to the amounts involved, they are to be fixed on an application by either party to the trial court. 

155                            Finally, I would dismiss the appeal with respect to the trial judge’s award of damages for loss of future earnings. 

The reasons of Bastarache, LeBel and Deschamps JJ. were delivered by

Bastarache J. (dissenting in part) —

I.       Overview

156                            Appeals are creatures of statute; therefore, legislative — not judicial — policy choice must be considered paramount.  Moreover, because appeals in civil cases are founded on provincial legislation which may vary from one province to another, the rights of appeal and the powers of an appellate court to act on those rights will not necessarily be uniform across the country.

157                            Among all of the statutes governing the powers of appellate courts in Canada, Saskatchewan’s Court of Appeal Act, 2000, S.S. 2000, c. C-42.1, is the only one that relieves the Court of Appeal of any obligation to adopt the view of the evidence taken by the trial judge and directs it to act on its own view of what, in its judgment, the evidence proves.  This must “mean something”.  In my view, it means that in Saskatchewan, the nature of appellate review is by way of rehearing and not review for error.

158                            In this appeal, we are particularly concerned with the conditions under which, in the context of an appeal by way of rehearing, the Court of Appeal will overrule a trial judge’s factual inference.  I contend that the court will overrule such an inference when it is not reasonable.  While it can therefore be said that the standard of review in Saskatchewan for factual inferences is reasonableness, as I will demonstrate more fully in these reasons, it is awkward to speak in terms of a “standard of review” in that regard, given the fact that, in Saskatchewan, the Court of Appeal is not limited to a “review” of the lower court’s decision but is, instead, directed to take its own view of the evidence.  Nevertheless, for the purposes of my analysis in this context and to promote clarity, I will accept the use of “standard of review” language and agree that the standard applicable to factual inferences is indeed reasonableness.

159                            On the facts of this case, I am of the view that the Court of Appeal did not misapply this standard when it set aside the trial judge’s award of pecuniary damages. On the contrary, it correctly interfered in this regard because the factual inferences on which the damages award was based were unreasonable, as they were unsupported by the evidence.  As will be further demonstrated, even if the more stringent standard adopted in Housen v. Nikolaisen, [2002] 2 S.C.R. 235, 2002 SCC 33, applied here, I would still uphold the decision of the Court of Appeal.

II.      Facts

160                            The following facts, as found by the trial judge, are not in dispute.

161                            H.L. is a status Indian within the meaning of the Indian Act, S.C. 1951, c. 29, and is a member of the Gordon First Nation Reserve.  When he was six months old, his father died, leaving his mother as the sole caregiver for 10 children of whom he was the youngest.  His mother subsequently entered into a relationship with S.W. This relationship was punctuated with frequent physical abuse of H.L.’s mother by S.W. and excessive use of alcohol by both of them.  During the first 12 years of his life, H.L.’s mother frequently moved her family between the Gordon First Nation Reserve and the Moscowegan First Nation Reserve of which S.W. was a member.  These relocations were often precipitated by acts of violence on the part of S.W.

162                            When H.L. resided at the Gordon First Nation Reserve, he attended a public school in Punnichy.  At no time did he attend Gordon’s Day School or reside at the Gordon Student Residence (formerly known as the Gordon Indian Residential School).   However, in 1974 or 1975, he joined a boxing club on the Reserve that was operated by the Department of Indian and Northern Affairs and administered by William Starr.  Starr was also the administrator of the Student Residence.  During this period of time, Starr sexually assaulted the appellant by subjecting him to two acts of masturbation.

163                            H.L. brought an action against Starr and the Government of Canada for damages suffered as a consequence of the abuse.

III.    Judicial History

A.     Saskatchewan Court of Queen’s Bench

164                            The trial judge, Klebuc J., found that the injuries and losses complained of by H.L. were attributable to Starr’s assaults. Specifically, he stated that:

[H.]L. unquestionably suffered enormous humiliation, self‑blame and loss of self‑worth as a consequence of Starr’s sexual abuse and such emotional problems in turn caused him to lose interest in pursuing an education, due in part to his inability to concentrate.  Immediately after the second assault, he commenced excessive alcohol consumption which in turn led to numerous convictions on alcohol and theft related offences, including convictions between 1978 and 2000 for driving while disqualified and driving while impaired.  These difficulties, as well as his difficulty with being “emotionally close” with women, in my view are attributable to Starr’s sexual abuse of him.  To the extent his dysfunctional family or [S.]W.’s misconduct may be viewed as a cause, I am of the opinion that Starr’s abuse is such an extraordinary occurrence that it constitutes a novus actus interveniens which severed any chain of causation that may have existed between the aforementioned causes and the damages ultimately experienced by [H.]L.

(H.L. v. Canada (Attorney General) (2001), 208 Sask. R. 183, 2001 SKQB 233, at para. 29)

Consequently, Klebuc J. granted the appellant judgment against Starr, as well as the Government of Canada, since he found that the criteria for the imposition of vicarious liability on the Government of Canada had been met.

165                            As for H.L.’s entitlement to damages, the trial judge concluded that H.L. was entitled to non-pecuniary damages of $60,000 for the emotional distress he suffered and will continue to suffer as a consequence of Starr’s abuse and aggravated damages of $20,000 for the humiliation and indignation he suffered as a result of Starr’s conduct. Klebuc J. also concluded that H.L. was willing and able to work but for his emotional and alcohol-related problems, which were attributable to Starr’s sexual abuse.  Therefore, the trial judge awarded the appellant $117,337.09 for the past loss of income earning capacity and $179,190 for future loss.  This latter amount was based solely on the evidence relating to H.L.’s past earning capacity.  Finally, Klebuc J. awarded H.L. punitive damages against Starr in the amount of $20,000.

166                            In supplemental reasons, Klebuc J. held that H.L. was entitled to claim pre-judgment interest against each defendant from the date he served his statement of claim: see H.L. v. Canada (Attorney General) (2001), 210 Sask. R. 114, 2001 SKQB 233.

B.      Saskatchewan Court of Appeal

167                            The Attorney General of Canada appealed to the Court of Appeal on the ground that the trial judge erred in holding the Government of Canada vicariously liable for Starr’s acts.  The Attorney General of Canada also made the following alternative submissions: (i) the award of damages for emotional distress was excessive; (ii) the award of damages for loss of earning capacity, past and future, was ill-founded; and (iii) the award of pre-judgment interest was contrary to law.  H.L. cross-appealed, taking issue with the trial judge’s assessment of damages and claiming that, together with pre-judgment interest, he was entitled to damages in the amount of $527,000.

168                            Cameron J.A., writing for the Court of Appeal, began his reasons for judgment with a review of the statutory framework for appeals and their adjudication in the province of Saskatchewan, and he came to the following conclusion:

On appeal from a decision of a judge of the Court of Queen’s Bench sitting without a jury, taken pursuant to sections 7(2)(a) and 13 of the Court of Appeal Act, 2000, it is the duty of the court acting under section 14 of the Act to rehear the case in the context of the grounds of appeal and make up its own mind, not disregarding the judgment appealed from, and giving special weight to that judgment in cases where the credibility of witnesses comes into question, but with full liberty to draw its own inferences from the facts proved or admitted, and to decide accordingly . . . .

(H.L. v. Canada (Attorney General) (2002), 227 Sask. R. 165, 2002 SKCA 131, at para. 77) (“H.L. (C.A.)”)

In coming to this conclusion, Cameron J.A. was cognizant of the divide that is setting in between the adjudicative framework suggested by the general standards of appellate review and that provided by The Court of Appeal Act, 2000; however, he maintained that while, in other provinces, appeal may be by way of review for error, in Saskatchewan, appeals have traditionally been and today still are by way of rehearing.

169                            Turning to the grounds of appeal advanced by the parties, Cameron J.A. dismissed the Attorney General of Canada’s appeal as it related to the trial judge’s conclusion that the Government of Canada was vicariously liable, entitling H.L. to $80,000 in non-pecuniary damages.  However, Cameron J.A. allowed the Attorney General of Canada’s appeal in relation to the trial judge’s awards of pecuniary damages for past and future loss of earning capacity and pre-judgment interest.  As to the pecuniary damages award, Cameron J.A., taking his own view of the evidence, concluded that the basic evidentiary foundation for the award was lacking.  In addition to this fundamental error, Cameron J.A. also found that the trial judge erred in four respects in his calculation of the award: (i) the trial judge failed to consider the plaintiff’s duty to mitigate; (ii) he unreasonably concluded that the plaintiff did not have a “crumbling skull” and therefore attributed too much to Starr’s wrongful acts in his assessment of pecuniary damages; (iii) he did not reduce the damages award to reflect the time H.L. was incarcerated; and (iv) he failed to account for the social assistance payments H.L. received during the relevant period.

170                            As for H.L.’s cross-appeal, Cameron J.A. dismissed it except as it related to H.L.’s claim of damages for the cost of future care.  The court allowed H.L.’s appeal in this regard, and awarded him $6,500.

171                            H.L. applied to the Court of Appeal pursuant to s. 37  of the Supreme Court Act, R.S.C. 1985, c. S-26 , for leave to appeal to this Court on the following grounds:

(1)               What is the correct standard of review of the appellate court of a province, and is that standard different for the appellate court of Saskatchewan?

(2)               Did the Saskatchewan Court of Appeal misapply that standard regarding:

a)       expert witnesses;

b)       pecuniary damages?

172                            In his reasons for judgment on the application, Bayda C.J.S. noted that the scope of the Court of Appeal’s powers was uncertain at present and that this controversy must be resolved: (2003), 238 Sask. R. 167, 2003 SKCA 78.  Therefore, he granted H.L. leave to appeal to this Court on the grounds stipulated.  He also granted the Attorney General of Canada leave to cross-appeal on the ground that the court erred in its determination that the Government of Canada was vicariously liable for Starr’s acts; however, the Attorney General of Canada discontinued the cross-appeal, and it was not argued before us.

IV.    Analysis

A.  The Nature and Standard of Appellate Review in Saskatchewan for Questions of Fact

(1)  Introduction

173                            Before beginning my analysis regarding my view of the applicable standard of appellate review in Saskatchewan for questions of fact, it is necessary to clarify what I respectfully perceive to be some confusion unfortunately apparent regarding the meaning of the term “appeal by way of rehearing”.

174                            Because the word “rehearing” can be used in a number of different senses, to avoid confusion three situations need to be identified and explained: (1) appeal by way of review (for error); (2) appeal by way of rehearing; and (3) a rehearing which is a new trial or occasionally a new appeal, also known as a de novo hearing:  see A. A. S. Zuckerman, Civil Procedure (2003), at pp. 761-62.  On an appeal by way of review, the appeal court’s duty is limited to a review of the lower court’s decision, and it may only interfere in limited circumstances identified by reference to the standard of review applicable to the particular type of question before the court (i.e., questions of fact, law or mixed fact and law): Zuckerman, at p. 762.  In general, in Canada appeals are conducted by way of review: see, e.g., Housen.

175                            In contrast, on an appeal by way of rehearing, the court is not limited to a scrutiny of the lower court’s decision but is expected to form its own judgment on the issues: Zuckerman, at p. 769.  In the case at bar, the Saskatchewan Court of Appeal held that this is the type of appeal that is available for civil matters tried by a judge alone in that province.  In its reasons for judgment, the Court of Appeal described the difference between an appeal by way of rehearing and an appeal by way of review for error as follows:

Rehearing is oriented to the decision upon the merits of the case.  Review for error is oriented to the process by which the decision is made.

(H.L. (C.A.), at para. 86)

176                            Finally, an appeal by way of rehearing must be distinguished from the last category of appeal types — an appeal by way of a hearing de novo.  As recently noted by the Australian High Court, an appeal by way of rehearing does not involve a completely fresh hearing by the appellate court of all the evidence: see Fox v. Percy (2003), 214 C.L.R. 118, [2003] HCA 22, at para. 22.  Instead, the court “proceeds on the basis of the record and any fresh evidence that, exceptionally, it admits”.

177                            It is especially important not to conflate the concept of an appeal by way of rehearing with an actual rehearing or a “retrial” (a.k.a. an appeal by way of a hearing de novo); however, with respect, it appears to me that in certain passages in his reasons for judgment in this case, my colleague Fish J. may have done so.  For instance, at para. 15 of his reasons, Fish J. states that “[n]othing in the record before us, in the relevant provisions of the Act, nor in the Court of Appeal’s own earlier appreciation of its proper role suggests to me that it has now been invested with a general jurisdiction to ‘rehear’ trials — that is, to apply a ‘rehearing’ standard when it reviews judgments at trial.”  As explained above, it is my view that there is a significant difference between “rehearing” trials (i.e., conducting a de novo hearing) and applying a “rehearing” standard when reviewing judgments at trial (i.e., conducting an appeal by way of “rehearing”).

178                            Similarly, at para. 52 of his reasons, Fish J. notes that “[i]n the absence of a clear statutory mandate to the contrary, appellate courts do not ‘rehear’ or ‘retry’ cases.” As briefly noted above, in the case at bar, the Court of Appeal concluded that, in Saskatchewan, on an appeal from a decision of a trial judge without a jury, the appeal is by way of rehearing, the Court of Appeal being directed to “make up its own mind, not disregarding the judgment appealed from, and giving special weight to that judgment in cases where the credibility of witnesses comes into question, but with full liberty to draw its own inferences from the facts proved or admitted, and to decide accordingly”:  H.L. (C.A.), at para. 77 (emphasis added).  In my opinion, in H.L. (C.A.), it is clear that when the Court of Appeal asserted that appeals in Saskatchewan are heard by way of rehearing, it was not claiming it has the power to conduct retrials or de novo hearings; rather, it was saying that it was not limited to a review of the lower court’s decision but could instead direct its attention to the merits of the case (para. 86).  I would immediately note that this language can be somewhat confusing because, as I shall explain later, a Court of Appeal will only interfere where it finds that the trial judge committed some error.  There is always a degree of deference to trial judges in an appeal by way of rehearing.

179                            With this semantic issue hopefully clarified, I will proceed with my analysis of the applicable standard of appellate review in Saskatchewan for questions of fact.  I will begin with a review of the provisions of The Court of Appeal Act, 2000 at issue in this appeal — namely ss. 7(2)(a) and 13, which pertain to the right of appeal, and ss. 12 and 14, which pertain to the powers of the Court of Appeal to act on that right — and I will apply the modern interpretation rule set out by E. A. Driedger in Construction of Statutes (2nd ed. 1983), at p. 87, to ss. 13 and 14 in particular in order to determine if they vest the Saskatchewan Court of Appeal with the jurisdiction to conduct appeals by way of rehearing or by way of review for error.  After identifying the nature of appellate review in Saskatchewan, I will consider the effect of judicial policy concerns in relation to the court’s exercise of its review powers in certain circumstances.  I will then offer my conclusion regarding the standard of appellate review in Saskatchewan for questions of fact, and I will endeavour to reconcile past jurisprudence with this conclusion.

(2)  Statutory Framework

(a)  Background

180                            Before commencing my analysis of the appropriate interpretation of the statutory provisions at issue in this appeal, it is necessary to make note of two background points that will influence my reasoning in this regard.

181                            First, as noted by La Forest J. in Kourtessis v. M.N.R., [1993] 2 S.C.R. 53, at pp. 69-70:

Appeals are solely creatures of statute;  see R. v. Meltzer, [1989] 1 S.C.R. 1764, at p. 1773.  There is no inherent jurisdiction in any appeal court.  Nowadays, however, this basic proposition tends at times to be forgotten.  Appeals to appellate courts and to the Supreme Court of Canada have become so established and routine that there is a widespread expectation that there must be some way to appeal the decision of a court of first instance.  But it remains true that there is no right of appeal on any matter unless provided for by the relevant legislature.

(See also Fox v. Percy, at para. 20.)

Because appeals are creatures of statute, legislative — not judicial — policy choice must be considered paramount: see, e.g., Farm Credit Corp. v. Valley Beef Producers Co-operative Ltd. (2002), 223 Sask. R. 236, 2002 SKCA 100, at para. 34.  Moreover, because appeals in civil cases are founded on provincial legislation, which may vary from one province to another, it must be accepted that the rights of appeal and the powers of the court to act on those rights will not necessarily be uniform across the country.  Thus, when considering the appropriate interpretation of statutory appeal provisions, such as those at issue in this case, it is necessary to have regard for such statutory variations and differences in appeal traditions as may exist between provinces: Valley Beef Producers Co-operative, at para. 36.

182                            Second, s. 10 of The Interpretation Act, 1995, S.S. 1995, c. I-11.2, reads as follows:

Every enactment shall be interpreted as being remedial and shall be given the fair, large and liberal construction and interpretation that best ensure the attainment of its objects.

As noted by the Court of Appeal in Valley Beef Producers Co-operative, s. 10 of The Interpretation Act tells us that the provisions of The Court of Appeal Act, 2000, including those pertaining to both the right of appeal and the powers of the court, must be “construed and interpreted liberally to the end of fulfilling their legislative objectives or, to put it another way, to the ultimate end of implementing the legislative policy they reflect”:  Valley Beef Producers Co-operative, at para. 43; see also H.L. (C.A.), at para. 14.

(b)  Statutory Provisions at Issue

183                            The following provisions of The Court of Appeal Act, 2000 are at issue in this appeal:

7 . . .

(2)  Subject to subsection (3) and section 8, an appeal lies to the court from a decision:

(a) of the Court of Queen’s Bench or a judge of that court;

12(1)   On an appeal, the court may:

(a)   allow the appeal in whole or in part;

(b)   dismiss the appeal;

(c)   order a new trial;

(d)  make any decision that could have been made by the court or   tribunal appealed from;

(e)   impose reasonable terms and conditions in a decision; and

(f)    make any additional decision that it considers just.

(2)  Where the court sets aside damages assessed by a jury, the court may assess any damages that the jury could have assessed.

13  Where issues of fact have been tried, or damages have been assessed, by a trial judge without a jury, any party is entitled to move against the decision of the trial judge, by motion for a new trial or otherwise:

(a)  within the same time that is allowed in cases of trial or assessment of damages by a jury; and

(b)  on the same grounds, including objections against the sufficiency of the evidence, or the view of the evidence taken by the trial judge, that are allowed in cases of trial or assessment of damages by a jury.

14  On an appeal from, or on a motion against, the decision of a trial judge or on any rehearing, the court is not obliged to grant a new trial or to adopt the view of the evidence taken by the trial judge, but the court shall act on its own view of what, in its judgment, the evidence proves, and the court may draw inferences of fact and pronounce the decision that, in its judgment, the trial judge ought to have pronounced.

184                            Sections 7(2)(a) and 13 pertain to the right of appeal and ss. 12 and 14 pertain to the powers of the court.  In the course of my analysis of the statutory provisions at issue, I will focus on s. 13 of the Act, given that it specifically pertains to the right of appeal when issues of fact have been tried by a judge alone and that the particular issue in this appeal is the standard of appellate review for questions of fact, and s. 14, given that it provides the remedy associated with the right conferred by s. 13.

(c)  Did the Province Act Within Its Authority When It Enacted The Court of Appeal Act, 2000?

185                            Before commencing the substantive portion of my analysis of the appropriate interpretation of the statutory provisions at issue in this appeal, as a preliminary point, it is important to note that the Saskatchewan legislature acted within its authority when it enacted The Court of Appeal Act, 2000. Specifically, the constitutional authority for this Act is founded on the exclusive provincial jurisdiction over property and civil rights and the administration of justice: see Constitution Act, 1867, ss. 92(13)  and 92(14) .  Although I made note of this point earlier, in my view it is also important to reiterate here that, because appeal rights and powers for civil matters are generally a matter of provincial jurisdiction, the different common law jurisdictions across Canada need not have the same nature or standards of appellate review.  As noted by the Attorney General of Canada in his written submissions, just as it is with all of the heads of provincial power under s. 92  of the Constitution Act, 1867 , the exercise of the power over property and civil rights and the administration of justice is destined to result in different approaches to similar issues. One need only look to the various provincial statutes pertaining  to limitation of actions, contributory negligence, juries and no-fault accident insurance schemes as examples of this. Therefore, in my view it is clear that inter-provincial variation in the nature and standards of appellate review is both possible and acceptable in our federal system. 

(d)  Statutory Interpretation

186                            A determination of the standard of appellate review in Saskatchewan for questions of fact turns on the interpretation given to the provisions of The Court of Appeal Act, 2000 quoted above.  On numerous occasions, this Court has confirmed that the preferred approach to statutory interpretation is that set out by Driedger, at p. 87:

Today there is only one principle or approach, namely, the words of an Act are to be read in their entire context and in their grammatical and ordinary sense harmoniously with the scheme of the Act, the object of the Act, and the intention of Parliament.

187                            Despite this Court’s adherence to this approach to statutory interpretation, as noted by this Court in Chieu v. Canada (Minister of Citizenship and Immigration), [2002] 1 S.C.R. 84, 2002 SCC 3, at para. 28, the interpretive factors enumerated by Driedger need not be applied in a formulaic fashion, particularly because they are closely related and interdependent.

188                            As explained earlier, in the course of my analysis of the appropriate interpretation of the statutory provisions at issue in this appeal, I will focus on ss. 13 and 14 the Act.  I will first consider the grammatical and ordinary sense of the words used in these two sections. I will then proceed to read these sections in their broader context.  This inquiry will include an examination of (i) the object of the Act, (ii) the object of the specific legislative provisions that form the statutory framework for the business of appeal, and (iii) the historical foundations of the Act.

(i)   Grammatical and Ordinary Sense

1.    Section 13

189                            Section 13 augments the right of appeal conferred by s. 7(2)(a), “[w]here issues of fact have been tried, or damages have been assessed, by a trial judge without a jury.”  In particular, s. 13(b) sets out the grounds upon which a party can object to the decision of the trial judge.  On an ordinary and grammatical reading of this paragraph, it is clear that it sets out two distinct grounds.  First, s. 13(b) incorporates by reference the same grounds of objection that are allowed in cases of trial or assessment of damages by a jury, including the sufficiency of the evidence. Second, para. (b) expands the scope of an appeal from a decision of a judge alone beyond the scope of an application for a new trial following a trial by jury by entitling a party to object to the view of the evidence taken by the trial judge.  The fact that s. 13(b) provides a party with two discrete grounds for objection is supported by the legislature’s use of the word “or” between “the sufficiency of the evidence” and “the view of the evidence taken by the trial judge”.  As noted by the Court of Appeal, because the two grounds for objection are expressed in the alternative, given the presumption against tautology, they are presumed not to be saying the same thing: H.L. (C.A.), at para. 22; see also R. Sullivan, Sullivan and Driedger on the Construction of Statutes (4th ed. 2002), at pp. 158-62.

2.    Section 14

190                            On an ordinary reading of s. 14, it is clear that it relieves the court of any obligation “to adopt the view of the evidence taken by the trial judge” and directs the court in imperative terms to “act on its own view of what, in its judgment, the evidence proves”.   The section then goes on to empower the court in permissive terms to “draw inferences of fact” and to “pronounce the decision that, in its judgment, the trial judge ought to have pronounced”:  see also H.L. (C.A.), at paras. 28 and 63.

191                            For reasons that will be more fully explained below, I agree with the Court of Appeal that the nature of powers conferred on the court by s. 14, in light of the right of appeal established by s. 13, are associated with appeal by way of rehearing and not retrial or review for error as generally understood.  Not only do I agree with the Court of Appeal’s understanding of the nature of the powers conferred on the court by s. 14, I also respectfully disagree with Fish J.’s reading of this section in two respects.

192                            First, at para. 82 of his reasons, Fish J. focuses on the use of the word “rehearing” in s. 14 and concludes that, given the context of the Act and especially s. 16(1), it is clear that this does not confer on the Court of Appeal the power to “rehear” trials; it simply provides that the powers available to the court on an appeal are available on the rehearing of an appeal, which would occur in the event of the resignation of two or more judges who heard the initial appeal, for example.

193                            As a preliminary point, and with respect, I wish to re-emphasize that, contrary to Fish J.’s assertion in para. 82 and elsewhere, the Court of Appeal did not claim it had the power to “rehear” trials; it claimed it had the power to conduct an appeal by way of rehearing rather than review for error.  Semantic issues aside, I agree with Fish J. that s. 14 does provide that the powers available to the court on an appeal are available on the rehearing of an appeal.  However, I do not agree that the use of the word “rehearing” in s. 14 assists in determining the nature of appellate review in Saskatchewan. In my respectful view, the use of the word “rehearing” in s. 14 is a “red herring”, so to speak, in that it is not relevant to an inquiry into the nature of appellate review in Saskatchewan.  In order to determine the nature of appellate review in Saskatchewan (i.e., whether appeals are conducted by way of rehearing or review for error), one must examine the powers conferred on the court by s. 14 from a functional perspective.  As I will explain later, when viewed functionally, it is clear that the powers conferred by s. 14 vest the court with the power to conduct an appeal by way of rehearing.

194                            I also respectfully take issue with Fish J.’s use of other provincial statutes to read down The Court of Appeal Act, 2000.  For instance, at para. 87, Fish J. states:

. . . I think it evident that the jurisdiction of the Saskatchewan Court of Appeal to review inferences of fact drawn by the trial judge is hardly exceptional, let alone unique.  Other provincial or territorial courts of appeal are granted similar powers, expressly or implicitly, by their governing statutes.  The 2000 Act simply sets out those powers in more detail than some.

195                            As noted previously, interprovincial variation in the nature and standards of appellate review is acceptable in our federal system, and I agree with the Court of Appeal’s reasoning in this regard:

The provinces, of course, constitute discrete jurisdictions for the purpose at hand.  Hence, the nature of appeal may differ from one jurisdiction to the next.  So, too, may the right of appeal, which may be more or less limited, and the powers of the appeal courts, which may be more or less extensive.  It is well to bear this in mind so as not to inadvertently import something from another jurisdiction which, however apt in that jurisdiction, may be inapt in this one.

 (H.L. (C.A.), at para. 31)

196                            On an ordinary reading of s. 14 of the Act, this section frees the Court of Appeal from the view of the evidence taken by the trial judge and empowers it to draw its own inferences of fact.  In my view, the powers conferred on the court by s. 14 are associated with an appeal by way of rehearing, and this makes the Saskatchewan Act unique.  In fact, these powers were once described by Gordon J.A. in Hallberg v.  Canadian National Railway Co. (1955), 16 W.W.R. 538 (Sask. C.A.), at p. 544, as “the widest powers given an appellate court in Canada”.  Moreover, I agree with the intervener Attorney General for Saskatchewan that a review of the statutes governing the powers of other appellate courts in Canada today confirms that no other jurisdiction in Canada has a provision equivalent to s. 14. For instance, while British Columbia, Alberta, Manitoba, Ontario and Prince Edward Island do allow their courts of appeal to draw inferences of fact, except for the British Columbia and Alberta courts of appeal, the circumstances in which they are permitted to do so are limited, and, more importantly, only the Saskatchewan legislation relieves the Court of Appeal of any obligation to adopt the view of the evidence taken by the trial judge and directs it to act on its own view of what, in its judgment, the evidence proves: see Court of Appeal Act, R.S.B.C. 1996, c. 77, s. 9(2); Alberta Rules of Court, Alta. Reg. 390/68, r. 518(c); Courts of Justice Act, R.S.O. 1990, c. C.43, s. 134(4)(a); The Court of Appeal Act, R.S.M. 1987, c. C240, s. 26(2); Supreme Court Act, R.S.P.E.I. 1987, c. 66, s. 56(4)(a).  This does not mean that the Court of Appeal can ignore the findings of the trial judge; I will deal with this issue later.

197                            Because the unique nature of appellate review in Saskatchewan is apparent on an ordinary reading of s. 14, and interprovincial variation in the nature and standards of appellate review is acceptable in our federal system, it is inappropriate to rely upon other provincial statutes to read down The Court of Appeal Act, 2000.

(ii)   Broader Context

198                            I will now proceed to examine ss. 13 and 14 of The Court of Appeal Act, 2000 in their broader context.  In order to do so, I will explore the following contextual factors: (i) the object of the Act, (ii) the object of the specific legislative provisions that form the statutory framework for appeals, and (iii) the historical foundations of the Act and ss. 13 and 14 in particular.  I will conclude that this contextual examination confirms that the nature of appellate review in Saskatchewan is by way of rehearing and not review for error.

1.    The Object of The Court of Appeal Act, 2000

199                            I agree with the Court of Appeal that the principal object of The Court of Appeal Act, 2000, “aside from continuing the Court of Appeal for Saskatchewan, is to confer rights of appeal, as in sections 7 and 13, and to empower the court to act on those rights, as in sections 12 and 14”: H.L. (C.A.), at para. 11.  I also agree that rights of appeal are substantive rights of major importance to persons who find themselves before the courts and tribunals, and that the fullness of a right to appeal depends on the fullness of the powers of the court to act on it (para. 13).  Thus, in this context, in determining the scope of powers conferred on the Court of Appeal by the Act, it is necessary to keep in mind the object of the right of the appeal:  Valley Beef Producers Co-operative, at para. 45.

200                            With this in mind, I will now turn to an examination of the object of the specific legislative provisions that form the statutory framework for appeals in Saskatchewan.

2.    The Object of the Specific Legislative Provisions That Form the Statutory Framework for Appeals in Saskatchewan

a.  Section 7(2)(a)

201                            As noted by the Court of Appeal in this case, the right of appeal conferred by s. 7(2)(a) is expressed to be subject to ss. 7(3) and 8; however, in this instance, because neither of these sections apply, s. 7(2)(a) confers an unlimited right of appeal upon a party proceeding in the Court of Queen’s Bench: H.L. (C.A.), at para. 15; Valley Beef Producers Co-operative, at para. 49.

202                            In Valley Beef Producers Co-operative, the Court of Appeal noted that although they are highly trained and competent, judges of the Court of Queen’s Bench may on occasion fail in relation to one or more components of judicial decision making, or fail on the whole to pronounce such judgment or make such order as the  dispute requires: Valley Beef Producers Co-operative, at para. 50.  In this context, the legislature created the right of appeal found in s. 7(2)(a), the object of which is “[t]o provide parties to proceedings in the Court of Queen’s Bench with the most comprehensive and effective means of redress possible in relation to such failures.”

b.    Section 13

203                            As explained previously, s. 13 augments the right of appeal conferred by s. 7(2)(a), “[w]here issues of fact have been tried, or damages have been assessed, by a trial judge without a jury.”  In particular, s. 13(b) sets out two distinct grounds upon which a party can object to the decision of the trial judge: (1) the same grounds of objection that are allowed in cases of trial or assessment of damages by a jury, including the sufficiency of evidence; and (2) the view of the evidence taken by the trial judge.

204                            I agree with the Court of Appeal in Valley Beef Producers Co-operative, at para. 63, that “the object of [this] section may be seen to lie in expanding the scope of the grounds upon which a party is entitled to object in relation to issues of fact tried by judge alone”.  This strongly suggests that decisions of judges are not to be treated as the equivalent of jury verdicts in terms of the nature and standard of appellate review.

205                            In sum, I agree with the Court of Appeal that in cases such as the one at bar, ss. 7(2)(a) and 13 provide parties with a facially unlimited right of appeal, which has first and foremost to do with relief from error: Valley Beef Producers Co-operative, at para. 65.

c.    Section 12(1)

206                            As noted by the Court of Appeal in Valley Beef Producers Co-operative, “[t]he legislature, in empowering the Court of Appeal for Saskatchewan as it did in s. 12(1) could hardly have expressed itself in broader terms”:  para. 70.   In the case at bar, the Court of Appeal noted in particular the scope of clauses (d) and (f), which empower the court, in turn, to “make any decision that could have been made by the court . . . appealed from” and “make any additional decision that it considers just”.  Given the nature of all judicial decision making, the Court of Appeal concluded, and I agree, that the exercise of these particular powers “entails ascertaining the material facts by one method or another, identifying the governing law, and applying the law to the facts as in the judgment of the court seems right”: H.L. (C.A.), at para. 27. Besides their breadth, the powers conferred on the court by s. 12(1) are also generally remedial in nature, in that their object is to empower the court “to redress error or deficiency in relation to the resolution of the controversy in the first instance with a view to setting matters right”:  Valley Beef Producers Co-operative, at para. 70.

d.    Section 14

207                            As I have mentioned previously, s. 14 frees the Court of Appeal from the view of the evidence taken by the trial judge and directs it to act on its own view of what the evidence proves.  In the course of so doing, the court may draw inferences of fact and pronounce the decision that, in its judgment, the trial judge ought to have pronounced.  In light of the clear conferral of these broad powers, I am of the view that the object of s. 14 is to relieve the Court of Appeal from the strictures pertaining to a motion for a new trial following a jury verdict: Valley Beef Producers Co-operative, at para. 78.  As noted by the Court of Appeal in the case at bar, a party can object to a jury verdict on the grounds of misdirection, the improper reception or rejection of evidence, unfairness in the proceedings and insufficiency of the evidence relative to the verdict: H.L. (C.A.), at para. 19.  However, it may not object to the view of the evidence taken by the jury.  As explained by Culliton J.A. (as he then was) in Taylor v. University of Saskatchewan (1955), 15 W.W.R. 459 (Sask. C.A.), at p. 463, when objecting to a jury verdict, “[t]he issue . . . is not whether the court agrees with the finding of the jury, but whether the jury, if acting judicially, might properly reach the decision which it did.”

e.  Conclusion

208                            After examining the object of the specific legislative provisions that form the statutory framework for appeals in Saskatchewan, in my view it is clear that the legislature intended to provide parties to proceedings in the Court of Queen’s Bench with the most comprehensive and effective means to address error in any component of a trial decision, including the view of the evidence taken by the trial judge.  As will become clearer after an examination of the historical foundations of the Act, this particular type of appellate review is consistent with an appeal by way of rehearing — not merely by way of review for error.  For instance, in the context of a case involving a trial judge’s exercise of discretion, Jonathan Parker L.J. noted that “a decision by the appeal court to proceed by way of rehearing frees it from such constraints [involved in an appeal by way of review] and allows it to exercise the discretion afresh in circumstances where it would have been unable to do so had the appeal proceeded in the normal way, by way of review”: Audergon v. La Baguette Ltd., [2002] E.W.J. No. 78 (QL), [2002] EWCA Civ 10, at para. 85.  Nevertheless, the powers granted must be exercised in a manner consistent with applicable and proper judicial policy.  This is addressed later.

3.    Historical Foundations

209                            In my opinion, an examination of the historical foundations of The Court of Appeal Act, 2000 and ss. 13 and 14 in particular confirms that the nature of appellate review in Saskatchewan is by way of rehearing, not review for error.

a.    Historical Foundations of the Act

210                            In Valley Beef Producers Co-operative, the Court of Appeal had occasion to describe the historical foundations of The Court of Appeal Act, 2000:

The Court of Appeal Act, 2000 is the latest in a series of such enactments, the original of which was enacted in 1915, when the Court of Appeal for Saskatchewan was created [see The Court of Appeal Act, S.S. 1915, c. 9].

                                                                   . . .

The original was founded in turn and in significant part on the Judicature Act, S.S. 1909, c. 52 (ss. 24 to 29); The Supreme Court of Judicature Act, 1873 (36 and 37 Vict., c. 66, ss. 4, 18 and 19) as amended from time to time to January 1, 1889; and the Rules of The Supreme Court, 1883, Order 58.  The Supreme Court of Judicature Act, 1873 created the Court of Appeal in England and provided generally for its jurisdiction and powers. Order 58 of the Rules of the Supreme Court, which had the force of law, clothed the court with more specific powers in relation to appeal.  The Saskatchewan Court of Appeal was created and empowered along these lines, and the Court of Appeal Act, 2000 still reflects these historical foundations, as did its predecessors.  Indeed, these foundations constitute an important part of the external context in which the Act was passed and serve as guiding lights, as it were, when it comes to understanding several of its provisions, especially those concerning the right of appeal and the powers of the court. [paras. 37-38]

b.    Historical Foundations of Sections 13 and 14

211                            Order 58 of the English Rules of the Supreme Court, 1883 contained two rules of particular significance, namely, rr. 1 and 4. Rule 1 stated that “[a]ll appeals to the Court of Appeal shall be by way of rehearing”, and r. 4 provided that, among other things, the court “shall have the power to draw inferences of fact and to give any judgment and make any order which ought to have been made, and to make such further or other order as the case may require”.  Order 58 applied only to appeals, which, as noted previously, are creatures of statute; applications for a new trial following a trial by jury were governed by different rules, including the rules found in Order 39:  see Valley Beef Producers Co-operative, at para. 40.

212                            As noted by the Court of Appeal in the case at bar, the powers conferred on the court by r. 4 were picked up, first, by the Supreme Court of Saskatchewan en banc and then, later, by s. 9 of The Court of Appeal Act, S.S. 1915, c. 9.  In fact, s. 9 was made even more explicit than r. 4 of Order 58, in that, in addition to empowering the Court of Appeal to draw inferences of fact and pronounce the decision that, in its judgment, ought to have been pronounced, it also stated that “it shall not be obligatory on the court to grant a new trial, or to adopt the view of the evidence taken by the trial judge, but the court shall act upon its own view of what the evidence in its judgment proves . . . .”

213                            As noted by the Court of Appeal in the case at bar, rr. 1 and 4 of Order 58 and s. 9 of The Court of Appeal Act of 1915 were enacted in the midst of controversy regarding the right of appeal as it pertained to issues of fact tried by a judge alone and the extent of the powers of the Court of Appeal to act on that right.  The court explained that “[a]t the heart of the matter lay the question of whether a decision of a judge without a jury should be treated as the equivalent of a jury verdict, especially for the purpose of an appeal engaging issues of fact” (para. 36).  Specifically, some appellate judges, most notably Lord Chelmsford in Gray v. Turnbull (1870), L.R. 2  Sc. & Div. 53 (H.L.), so regretted that trial judges’ findings of fact seemed subject to appeal, especially when made in the face of conflicting evidence, that they placed a heavy burden of persuasion on an appellant, which virtually foreclosed appeal on the ground of the view of the evidence taken by the trial judge.  Others, such as James L.J. in Bigsby v. Dickinson (1876), 4 Ch. D. 24 (C.A.), leaned against this and adopted a more generous approach: see H.L. (C.A.), at para. 37.

214                            In the two sections that follow, I will briefly review how this controversy was resolved first in England, and second in Saskatchewan, in order to better place what are now ss. 13 and 14 of The Court of Appeal Act, 2000 in their proper historical context.

c.    Resolution of Controversy in England

215                            In England, this controversy was largely laid to rest by the adoption of Order 58, which provided for appeal by way of rehearing and expressly empowered the Court of Appeal to draw inferences of fact and give any judgment which ought to have been given, and by the subsequent decision of the Court of Appeal in Coghlan v. Cumberland, [1898] 1 Ch. 704, which proved to be a seminal case in both England and Saskatchewan.

216                            In Coghlan v. Cumberland, Lindley M.R. discussed the nature of appellate review of a decision of a judge alone as follows:

The case was not tried with a jury, and the appeal from the judge is not governed by the rules applicable to new trials after a trial and verdict by a jury.  Even where, as in this case, the appeal turns on a question of fact, the Court of Appeal has to bear in mind that its duty is to rehear the case, and the Court must reconsider the materials before the judge with such other materials as it may have decided to admit.  The Court must then make up its own mind, not disregarding the judgment appealed from, but carefully weighing and considering it; and not shrinking from overruling it if on full consideration the Court comes to the conclusion that the judgment is wrong. [pp. 704‑5]

Although a court is under a duty to make up its own mind,  Lindley M.R. noted that it must nevertheless be cognizant of the inherent difficulty of doing so in respect of findings of fact that rest on the trial judge’s assessments of credibility:

When, as often happens, much turns on the relative credibility of witnesses who have been examined and cross‑examined before the judge, the Court is sensible of the great advantage he has had in seeing and hearing them. It is often very difficult to estimate correctly the relative credibility of witnesses from written depositions; and when the question arises which witness is to be believed rather than another, and that question turns on manner and demeanour, the Court of Appeal always is, and must be, guided by the impression made on the judge who saw the witnesses.  But there may obviously be other circumstances, quite apart from manner and demeanour, which may shew whether a statement is credible or not; and these circumstances may warrant the Court in differing from the judge, even on a question of fact turning on the credibility of witnesses whom the Court has not seen. [p. 705]

217                            Similarly, in Montgomerie & Co. v. Wallace‑James, [1904] A.C. 73 (H.L.), Earl of Halsbury L.C. stated that, if called upon to take up an issue of fact on appeal, an appellate tribunal must do so to the best of its ability, including its ability to draw inferences:

My Lords, I think this appeal should be allowed. It is simply a question of fact, and doubtless, where a question of fact has been decided by a tribunal which has seen and heard the witnesses, the greatest weight ought to be attached to the finding of such a tribunal.  It has had the opportunity of observing the demeanour of the witnesses and judging of their veracity and accuracy in a way that no appellate tribunal can have.  But where no question arises as to truthfulness, and where the question is as to the proper inferences to be drawn from truthful evidence, then the original tribunal is in no better position to decide than the judges of an Appellate Court. [Emphasis added; p. 75.]

218                            Despite the clear language of Order 58 that an appeal of a decision of a judge alone shall be by way of rehearing, some appellate judges continued to espouse the view that, for the purposes of appeal, such a decision was to be treated as the equivalent of a jury verdict.  This prompted the House of Lords to revisit this issue in both Mersey Docks and Harbour Board v. Procter, [1923] A.C. 253, and Benmax v. Austin Motor Co., [1955] A.C. 370. 

219                            In Mersey Docks and Harbour Board v. Procter, Viscount Cave L.C. adopted the principles articulated by Lindley M.R. in Coghlan v. Cumberland and by Earl of Halsbury L.C. in Montgomerie & Co. v. Wallace‑James in his discussion of the duty of a court hearing an appeal from the decision of a judge alone:

My Lords, it was contended on behalf of the appellants that the finding of Branson J., being a finding of a trial judge on a question of fact, should not have been disturbed by the Court of Appeal.  In my opinion there is no ground for such a contention.  The duty of a Court hearing an appeal from the decision of a judge without a jury was clearly defined by Sir Nathaniel Lindley M.R. in Coghlan v. Cumberland, and by Lord Halsbury in Montgomerie & Co. v. Wallace‑James, and is no longer in doubt.  The procedure on an appeal from a judge sitting without a jury is not governed by the rules applicable to a motion for a new trial after a verdict of a jury. In such a case it is the duty of the Court of Appeal to make up its own mind, not disregarding the judgment appealed from and giving special weight to that judgment in cases where the credibility of witnesses comes into question, but with full liberty to draw its own inference from the facts proved or admitted, and to decide accordingly. . . . The material facts, so far as they are known, are undisputed; and the Court of Appeal was at liberty, and indeed was bound, to draw its own inference from them. [Footnotes omitted; pp. 258-59.]

220                            Likewise, in Benmax v. Austin Motor Co., the House of Lords upheld the decision of the Court of Appeal on the ground that, while the ability of the Court of Appeal to overrule a trial judge’s decision that is based on assessments of credibility may be limited, its ability to address the trial judge’s inference from the evidence as a whole is not.  Specifically, Lord Morton, who delivered the judgment of the House, stated that

in the present case it would appear that the learned judge did not doubt the credibility of any witness, and formed his views by inference from the evidence as a whole.  The Court of Appeal formed the opposite view by the same method and I agree with that court. [p. 374]

221                            Viscount Simonds, who delivered concurring but more extensive reasons in Benmax v. Austin Motor Co., addressed what he perceived to be a source of confusion: the distinction between “the finding of a specific fact and a finding of fact which is really an inference from facts specifically found, or, as it has sometimes been said, between the perception and evaluation of facts” (p. 373).  As for inferences from facts, Viscount Simonds stated that “an appellate court should form an independent opinion, though it will naturally attach importance to the judgment of the trial judge” (p. 374).

222                            These decisions effectively settled the controversy regarding the nature of appellate review in England for questions of fact; that is, until May 2, 2000, when Parliament introduced a new system of civil appeals: H.L. (C.A.), at para. 48. As will be discussed below, this new appeal regime appears to have changed the nature of appellate review in England.

d.    Resolution of Controversy in Saskatchewan

223                            In order to explain how the controversy was laid to rest in Saskatchewan, it is necessary to begin with the years preceding the enactment of The Court of Appeal Act of 1915.

224                            In Coventry v. Annable (1911), 19 W.L.R. 400, the Supreme Court of Saskatchewan en banc heard an appeal from a decision of a judge alone on the question, among others, of whether the judge was wrong in finding no fraud on the part of the defendant.  Wetmore C.J. adopted and applied Lindley M.R.’s statement of principle in Coghlan v. Cumberland, finding fraud and deciding the case accordingly. The appeal to the Supreme Court of Canada was dismissed, with three of the six judges filing individual reasons.  Anglin J., in particular, agreed with Wetmore C.J. and adopted the essence of the principle put forward by Lindley M.R. in Coghlan v. Cumberland that is quoted above: see Annable v. Coventry (1912), 46 S.C.R. 573, at p. 587.

225                            Anglin J. also referred to the decision in Coghlan v. Cumberland in Greene, Swift & Co. v. Lawrence (1912), 2 W.W.R. 932 (S.C.C.), at p. 944, adding:

However loath we may be to reverse the decision of a trial judge on the question of fact, “it is our duty to do so if the evidence coerces our judgment so to do.”  The Gairloch, 1899, 2 Ir. 1, 13; Coghlan v. Cumberland, 1898, 1 Ch. 704, 67 L.J. Ch. 402.

226                            In 1918, The Court of Appeal Act of 1915 was proclaimed and thereafter appeals went to the Court of Appeal for Saskatchewan where they were governed by ss. 8 and 9 (now ss. 13 and 14), “the purpose of which lay in putting to rest the controversy in Saskatchewan along the lines it had been put to rest in England, though along even more explicit and decisive lines”:  H.L. (C.A.), at para. 56. In paragraphs 57 to 59 of its reasons in the case at bar, the Court of Appeal succinctly sets out how, since The Court of Appeal Act of 1915 was enacted, it has adopted the “rehearing” approach to appellate review that was first laid out in Coghlan v. Cumberland and approved by Anglin J. in Annable v. Coventry and Greene, Swift & Co. v. Lawrence: see,  e.g.,  Miller v. Foley & Sons (1921), 59 D.L.R. 664; Messer v. Messer (1922), 66 D.L.R. 833; Monaghan v. Monaghan, [1931] 2 W.W.R. 1; Kowalski v. Sharpe (1953), 10 W.W.R. (N.S.) 604; Tarasoff v. Zielinsky, [1921] 2 W.W.R. 135; Matthewson v. Thompson, [1925] 2 D.L.R. 1211; French v. French, [1939] 2 W.W.R. 435, at p. 443; and Wilson v. Erbach (1966), 56 W.W.R. 659, at p. 666.

227                            In conclusion, the court notes that “the controversy that had prevailed in Saskatchewan prior to the enactment of the Court of Appeal Act of 1915 was laid to rest here following the enactment of sections 8 and 9 of that Act”:  H.L. (C.A.), at para. 60.  In my view, this brief discussion of the historical foundations of The Court of Appeal Act, 2000 and ss. 13 and 14 in particular confirms that in Saskatchewan, for the purpose of appeal, a decision of a judge alone is not to be taken as the equivalent of a jury verdict, and that the nature of appellate review of such a decision is by way of rehearing.

e.    Legislative History

228                            Although the section numbering may have changed and the language may have been modernized over time, the Acts’ historical foundations remain relevant today because the Saskatchewan legislature has faithfully adhered to the content of what are now ss. 13 and 14 throughout the years: H.L. (C.A.), at para. 61.  Therefore, I share the view of the Court of Appeal that the 2000 legislative amendments to The Court of Appeal Act did not have any substantive effect on the nature of appellate review in Saskatchewan and the scope of the powers of the Court of Appeal; rather, these amendments to The Court of Appeal Act

served to maintain and augment at least 85 years of appellate practice in Saskatchewan, where appeal in relation to an unlimited right of appeal from a decision of a trial judge without a jury has traditionally been by way of “rehearing”, with the Court of Appeal being directed to take its own view of the evidence and being empowered to draw inferences of fact and pronounce the decision that ought to have been pronounced by the trial judge. [para. 62]

229                            In contrast, a major English legislative initiative in 2000 appears to have had a much more marked effect on the nature of appellate review in that country.  On May 2, 2000, a new system of civil appeals took effect in England, and the new rules appear mostly in Part 52 of the Civil Procedure Rules 1998, S.I. 1998 No. 3132: see Great Britain, Civil Procedure (2002), vol. 1, at pp. 1182ff.  These new rules seek to restrict resort to the appeal process to cases that really justify its use and to ensure that appeals, when they are warranted, are conducted in an efficient and effective manner: Zuckerman, at p. 719.  In Tanfern Ltd. v. Cameron-MacDonald, [2000] 1 W.L.R. 1311 (C.A.), at para. 50, Brooke L.J. described the changes imposed by this new appeal regime as “the most significant changes in the arrangements for appeals in civil proceedings in this country for over 125 years”.  These changes were welcomed by N. H. Andrews in “A New System of Civil Appeals and a New Set of Problems”, [2000] Cambridge L.J. 464, at p. 465, since they would “reduce the delay, expense, and uncertainty of civil proceedings” and “increase the incentive for litigants to ‘get it right first time round’”.  Nevertheless, Andrews also noted that “the same changes will reduce the chances of rectifying defective decisions”, and that “[t]his is the price paid for achieving the impressive benefits of the new system of appeals.”

230                            It is of particular relevance to the issues that arise in this case that under this new appeal regime in England, as a general rule, appeals to the Court of Appeal are no longer by way of rehearing but, rather, are limited to a review of the lower court’s decision: see Civil Procedure Rules, r. 52.11(1).  This obvious change in terminology from “rehearing” to “review” would suggest, at first blush at least, that the English Parliament intended that the nature of appellate review in that country be changed, specifically from a more robust appeal by way of rehearing to a more limited review of the lower court’s decision.  However, it appears that there is currently some controversy in England regarding the difference between appeal by way of rehearing and appeal by way of review, and what effect the terminology change in the new rules had on the nature of appellate review in that country.

231                            For instance, in Assicurazioni Generali SpA v. Arab Insurance Group, [2003] 1 W.L.R. 577 (C.A.), Ward L.J., in a separate opinion, stated that prior to the reform of civil appeals in 2000, “interlocutory appeals in the Court of Appeal were treated as reviews of the lower court’s decision”, even though they were “nominally by way of rehearing” (para. 194).  Therefore, despite the change in language from “rehearing” to “review”, Ward L.J. concluded that an appellate court’s task is essentially no different from what it was before the new rules came into effect; that is, “[t]he Court of Appeal can only interfere if the decision of the lower court was wrong and in deciding whether or not findings of fact were wrong, we take a retrospective look at the case and do not decide it afresh untrammelled by the judge’s conclusion” (para. 195).  Similarly, in his reasons for judgment on behalf of the court in Assicurazioni, Clarke L.J. acknowledged that there is plainly force in the submission that the nature of appellate review changed with the change in language, but he nonetheless concluded that although the previous rule expressly referred to a rehearing, “the exercise upon which the court was engaged was essentially one of review” (para. 13).

232                            Conversely, Jolowicz argues that under the former Rules of the Supreme Court appeals to the Court of Appeal were by way of rehearing, and this “meant that appellate judges were most unlikely to interfere with the trial judge’s findings of fact in so far as they depended on his assessment of the credibility of witnesses, but it did not mean that they were judges only of law”:  J. A. Jolowicz, “The New Appeal: re-hearing or revision or what?” (2001), 20 C.J.Q. 7, at p. 7.  Specifically, Jolowicz argues that the provisions in the former Rules of the Supreme Court were enough to ensure that the English Court of Appeal was indeed a “court of appeal”, before which issues are to be decided afresh in fact and in law, and not what is called elsewhere a “[c]ourt of cassation”, before which only the conformity of the lower court judgment to the rules of law is argued (pp. 7-8).

233                            Similarly, Lord Sumner stated in S.S. Hontestroom v. S.S. Sagaporack, [1927] A.C. 37 (H.L.), at p. 47, that “[o]f course, there is jurisdiction to retry the case on the shorthand note, including in such retrial the appreciation of the relative values of the witnesses, for the appeal is made a rehearing by rules which have the force of statute”:  see also J. A. Jolowicz, “Court of Appeal or Court of Error?”, [1991] Cambridge L.J. 54.  Nonetheless, like Jolowicz, Lord Sumner also noted that “not to have seen the witnesses puts appellate judges in a permanent position of disadvantage as against the trial judge”; therefore, he opined that “[i]f [the trial judge’s] estimate of the man forms any substantial part of his reasons for his judgment the trial judge’s conclusions of fact should, as I understand the decisions, be let alone” (p. 47).

234                            Not only does Jolowicz argue that, under the former rules, appeals to the Court of Appeal were by way of rehearing, in the sense that the Court of Appeal was empowered to “retry the case on the shorthand note”, he is also of the view that because an appeal court conducting a “review” of the decision of the lower court under the new rules may take account of the evidence given at trial and may exercise all the powers conferred on it by the rules (including the power to draw inferences of fact), the new “review” provided for by Part 52 of the Civil Procedure Rules “differs little, if at all, from the procedure formerly used in the Court of Appeal”:  see Jolowicz, “The New Appeal: re-hearing or revision or what?”, p. 11.

235                            Of course, it is not this Court’s place to resolve this controversy in English law.  Whatever the change in language in the new English appeal rules may mean, as noted above, it is clear that over the years Saskatchewan has not substantively changed the language in its Court of Appeal Act.  Therefore, this supports the argument that the nature of appellate review in Saskatchewan has not deviated from its historical roots.  In other words, in Saskatchewan, appeals were and continue to be by way of rehearing.

236                            Moreover, if one accepts that an appeal by way of rehearing is different from a review of the lower court decision and that the change in language in the new English appeal rules from “rehearing” to “review” signified a shift to a more restricted form of appellate review in that country (as a plain reading of the new rule would seem to suggest at least), then it can be argued that the Saskatchewan legislature was not willing to pay the price to which Andrews refers — i.e., unlike the English Parliament, the Saskatchewan legislature was not willing to reduce the chances of rectifying defective decisions in order to reduce the delay, expense and uncertainty of civil proceedings.  It saw no need for it.  Moreover, because an appeal is a statutory creature, legislative policy choices in this area must be seen to be paramount.  It appears that, in enacting The Court of Appeal Act, 2000, the Saskatchewan legislature re-affirmed its policy choice to have its Court of Appeal proceed with appeals by way of rehearing.  If the legislature is now concerned or becomes concerned in the future about the nature of appellate review in Saskatchewan, it is open to the legislature to amend The Court of Appeal Act, 2000: see Chieu v. Canada (Minister of Citizenship and Immigration), at para. 66.  For now, the statute is clear: in Saskatchewan, the nature of appellate review is by way of rehearing.

237                            However, even if one subscribes to the view professed by the Lord Justices in Assicurazioni that appellate review in England was and still is by way of review, unlike my colleague Fish J., I contend that the English understanding of appeal by way of “review” is different from the Canadian understanding of this concept, which was recently articulated by this Court in Housen.  Moreover, the English understanding of appeal by way of review is actually more closely in line with the nature of appellate review in Saskatchewan, which, as I understand it, is by way of rehearing.

238                            For example, Zuckerman notes that one of the general principles underlying an appeal by way of review in England is that an appeal court should not interfere with findings of fact made by the lower court because the judge who saw and heard the witnesses is better placed to assess their reliability and draw inferences from their testimony. Zuckerman states that this principle led to a distinction in the English case law “between conclusions concerning primary facts which followed entirely from the assessment of the reliability of witnesses, and conclusions based on a combination of testimonial assessment and analysis of documents and surrounding circumstances, with which the appeal court would more readily interfere” (p. 766). Zuckerman explains that the reason for this distinction is that “an appeal court is just as well placed as the trial judge to determine the proper inferences to be drawn from circumstantial or documentary evidence”:  see also Whitehouse v. Jordan, [1981] 1 All E.R. 267 (H.L.), per Lord Fraser.

239                            Similarly, in Assicurazioni, although Clarke and Ward L.JJ. both stated that, despite the change in language from “rehearing” to “review”, the Court of Appeal’s task has always been and still is to review the lower court’s judgment for error, it is clear from their judgments that they accept that, in the course of such a review, the greater the advantage the trial judge has over the appellate court (e.g., with respect to credibility assessments), the more reluctant the appellate court should be to interfere.  However, when the relative advantage of the trial judge is not engaged, such as when the issue is with respect to the drawing of inferences, then the appellate court may more readily interfere.  This proposition was specifically recognized by Ward L.J. as follows:

Where the primary facts are not challenged and the judgment is made from the inferences drawn by the judge from the evidence before him, then the Court of Appeal, which has the power to draw any inference of fact it considers to be justified, may more readily interfere with an evaluation of those facts. [para. 197]

240                            From my perspective, it appears that the type of “review” to which Zuckerman and the Lord Justices in Assicurazioni are referring is not the same as the general “review for error” concept subscribed to in Canada. In Housen, all nine justices agreed in principle that, in the course of a review for error, the standard of review should be identical for both findings of fact and inferences of fact, although, as will be explained below, the majority and minority disagreed on the articulation of the standard of review for the latter.  In his reasons in this case, Fish J. confirms that the same standard of review should apply to findings of fact as well as to inferences of fact: see, e.g., at paras. 52-55.  In contrast, as explained above, in England there is authority for the proposition that, in the course of an appeal by way of review in that country, there is a distinction to be made between findings of fact that engage the special advantage of the trial judge (e.g., those that involve assessments of credibility) and inferences of fact that do not.  The appellate court will more readily interfere in the latter case, and, in my view, this implies that, contrary to the Canadian position, the same standard of review cannot be applied to both circumstances.

241                            Not only is the type of appeal by way of “review” described by Zuckerman and the Lord Justices in Assicurazioni different from the Canadian understanding of this concept, I also suggest that in some respects it is actually more in line with the nature of appellate review in Saskatchewan.  In these reasons, I will go on to explain that in Saskatchewan, where the nature of appellate review is by way of rehearing, when the trial judge’s factual findings engage the special advantage he or she has over an appellate tribunal, the Court of Appeal will only interfere and apply its own view of the evidence if the trial judge has committed a palpable and overriding error in his or her fact finding.  In contrast, because the trial judge is in no better position than the Court of Appeal to draw inferences of fact from a base of fact properly established, I will contend that the Court of Appeal will more readily interfere when inferences are at issue.  Specifically, it is my view that the Saskatchewan Court of Appeal will overrule a trial judge’s inference of fact and draw its own when it concludes that the inference is not reasonable.  It is apparent that this approach to factual findings and inferences of fact, which I will explain more fully below, is quite similar to that followed in England (where appeals are now by way of “review”), in that both approaches differentiate between the two and grant more deference to the trial judge in the former case, when his or her special advantage is engaged.

242                            All of this merely demonstrates that terminology (i.e., “rehearing” versus “review”) can be misleading.  Therefore, in these circumstances, and particularly because appeals are statutory creations, it is best to focus upon the statute that sets out the appellate court’s jurisdiction and powers, in order to determine the nature of appellate review.  As I have explained in these reasons, the statute in this case is clear: in Saskatchewan, the nature of appellate review is by way of rehearing.

(iii)  Conclusion Regarding the Nature of Appellate Review in Saskatchewan

243                            After examining the grammatical and ordinary sense of the words used in ss. 13 and 14 of The Court of Appeal Act, 2000, as well as the object of the Act, the object of the specific legislative provisions that form the statutory framework for appeals, and the Act’s historical foundations, to me it is clear that the nature of appellate review in Saskatchewan is by way of rehearing, with the Court of Appeal being directed to take its own view of the evidence and being empowered to draw inferences of fact and pronounce the decision that ought to have been pronounced by the trial judge.  In light of this conclusion, one question remains: although the Court of Appeal is not constrained by the view of the evidence taken by the trial judge, in what circumstances will the Court of Appeal apply its own view of the evidence and, if necessary, pronounce the decision that ought to have been pronounced?  In particular, in this appeal, we are concerned with when the Court of Appeal will do so in relation to questions of fact.  The issue then is to determine what judicial policy mandates in the particular context of The Court of Appeal Act, 2000.

(3)  Judicial Policy Concerns

244                            Contrary to the Saskatchewan Court of Appeal, I believe that the direction contained in s. 14 of the Act to the Court of Appeal to take its own view of what the evidence proves is subject to the judicial policy concern that trial judges enjoy a special advantage over an appellate court, in that they hear the testimony of witnesses viva voce and are exposed to the case as a whole.  This is also the view of most authors, for instance J.-C. Royer, La preuve civile (3rd ed. 2003), at p. 324.  The trial judge’s special advantage has been recognized by this Court on a number of occasions: see, e.g., Laurentide Motels Ltd. v. Beauport (City), [1989] 1 S.C.R. 705, at p. 794, per L’Heureux-Dubé J.; St-Jean v. Mercier, [2002] 1 S.C.R. 491, 2002 SCC 15, at para. 36.  Of particular relevance to this appeal, the trial judge’s special advantage was recently described by the High Court of Australia in the context of an appeal by way of rehearing as follows:

On the one hand, the appellate court is obliged to “give the judgment which in its opinion ought to have been given in the first instance”.  On the other, it must, of necessity, observe the “natural limitations” that exist in the case of any appellate court proceeding wholly or substantially on the record.  These limitations include the disadvantage that the appellate court has when compared with the trial judge in respect of the evaluation of witnesses’ credibility and of the “feeling” of a case which an appellate court, reading the transcript, cannot always fully share.  Furthermore, the appellate court does not typically get taken to, or read, all of the evidence taken at the trial.  Commonly, the trial judge therefore has advantages that derive from the obligation at trial to receive and consider the entirety of the evidence and the opportunity, normally over a longer interval, to reflect upon that evidence and to draw conclusions from it, viewed as a whole.

(Fox v. Percy, at para. 23 (footnotes omitted))

245                            The special advantage of the trial judge calls for a measure of deference on the part of the Saskatchewan Court of Appeal when, pursuant to the direction in s. 14 of the Act, it is considering what the evidence proves: see Valley Beef Producers Co-operative, at para. 87.  Specifically, when the trial judge’s decision is based upon issues that engage this special advantage (most notably, factual findings based on credibility assessments), the Court of Appeal should make due allowance in this respect: see Fox v. Percy, at para. 25.  Nevertheless, it must be kept in mind that the Court of Appeal is charged with the statutory mandate to conduct appeals by way of rehearing, and, as noted by the High Court of Australia:

[T]he mere fact that a trial judge necessarily reached a conclusion favouring the witnesses of one party over those of another does not, and cannot, prevent the performance by a court of appeal of the functions imposed on it by statute.  In particular cases incontrovertible facts or uncontested testimony will demonstrate that the trial judge’s conclusions are erroneous, even when they appear to be, or are stated to be, based on credibility findings. [para. 28]

246                            Furthermore, although it is my view that the Court of Appeal should accord some deference to decisions that are based upon issues that engage the special advantage of the trial judge, the same deferential stance does not extend to drawing inferences of fact or to evaluating a body of fact against a legal standard: see H.L. (C.A.), at para. 68. As noted by the Court of Appeal in the case at bar and confirmed in several other cases, the Court of Appeal “is in as good a position as the trial judge to draw inferences of fact from a base of fact proven or admitted”: see Montgomerie & Co. v. Wallace-James, at p. 75; Mersey Docks and Harbour Board v. Procter, at pp. 258-59; Warren v. Coombes (1979), 142 C.L.R. 531 (H.C. Austl.), at p. 551, cited with approval in Fox v. Percy, at para. 25.  This point was, in fact, confirmed by this Court in Workmen’s Compensation Board v. Greer, [1975] 1 S.C.R. 347, at pp. 357-58, where, after quoting from the aforementioned House of Lords’ decision in Montgomerie & Co. v. Wallace-James, it stated that

the practice of this Court, which reflects a reluctance to interfere with concurrent findings of fact in two provincial courts, does not apply with the same force to inferences drawn from conflicting professional opinions as it does to findings based on direct factual evidence.

In Toneguzzo-Norvell (Guardian ad litem of) v. Burnaby Hospital, [1994] 1 S.C.R. 114, at p. 122, McLachlin J. (as she then was) agreed with this Court’s decision in Greer, but also elaborated on it as follows:

I agree that the principle of non‑intervention of a Court of Appeal in a trial judge’s findings of facts does not apply with the same force to inferences drawn from conflicting testimony of expert witnesses where the credibility of these witnesses is not in issue.  This does not however change the fact that the weight to be assigned to the various pieces of evidence is under our trial system essentially the province of the trier of fact, in this case the trial judge.

247                            My conclusion that the same appellate deference does not extend to the drawing of inferences is strengthened by the fact that s. 14 of the Act expressly empowers the Court of Appeal to draw inferences of fact and pronounce the decision that, in its judgment, ought to have been pronounced, a power that necessarily entails drawing evaluative inferences: H.L. (C.A.), at para. 68.

(4)  When Faced With a Question of Fact, in What Circumstances Will the Court of Appeal Apply Its Own View of the Evidence and, if Necessary, Pronounce the Decision that Ought to Have Been Pronounced?

248                            In light of my consideration of the impact of judicial policy concerns regarding the special advantage of the trial judge, especially with regard to factual findings based on assessments of credibility, I will now address the question: when faced with a question of fact, in what circumstances will the Court of Appeal apply its own view of the evidence and, if necessary, pronounce the decision that ought to have been pronounced?

249                            In general, I agree with the Court of Appeal’s statement:

On appeal from a decision of a judge of the Court of Queen’s Bench sitting without a jury, taken pursuant to sections 7(2)(a) and 13 of the Court of Appeal Act, 2000, it is the duty of the court acting under section 14 of the Act to rehear the case in the context of the grounds of appeal and make up its own mind, not disregarding the judgment appealed from, and giving special weight to that judgment in cases where the credibility of witnesses comes into question, but with full liberty to draw its own inferences from the facts proved or admitted, and to decide accordingly . . . .

(H.L. (C.A.), at para. 77)

250                            To this general statement, I would add the following two points.  First, I would note that deference should not only be accorded to trial decisions where the credibility of witnesses comes into question, but also to all such decisions where the special advantage of the trial judge is engaged.  This would include cases where the trial judge’s conclusion on an issue is dependent on his or her holistic assessment of the evidence presented at trial (all of which may not be available to the Court of Appeal), also described as his or her “feeling” of the case: see Fox v. Percy, at para. 23. Second, as I mentioned above, an appropriate interpretation of The Court of Appeal Act, 2000 indicates that the primary function of the Court of Appeal is to correct error or deficiency in the particular case.  In light of its role as a “court of error”, I would emphasize that “[i]f, making proper allowance for the advantages of the trial judge, [the Court of Appeal concludes] that an error has been shown, [it is] authorised, and obliged, to discharge [its] appellate duties in accordance with the statute”:  Fox v. Percy, at para. 27.  Here again I must note that the language used is awkward because of the word “error”.  But in the context I have discussed, one will understand that even though the palpable error threshold is not applicable, the Court of Appeal will only substitute its view of the facts if it finds some error in the reasoning of the trial judge.

251                            In my view, contrary to the submissions of the Attorney General for Saskatchewan, this general statement of when the Saskatchewan Court of Appeal will interfere with the decision of a trial judge does not imply that factual findings that do not engage the special advantage of the trial judge and inferences of fact are to be reviewed on the basis of the correctness standard. As a preliminary point, in Saskatchewan, the nature of appellate review is, as earlier demonstrated, by way of rehearing, not review for error; therefore, the notion of “reviewing” a decision on any standard, let alone the correctness standard, is not applicable in these circumstances.  Moreover, the correctness standard implies that the reviewing court will accord no deference to the decision of the lower court, but it is clear from the general statement of the Court of Appeal in H.L. (C.A.) that I quoted above that even though it is the duty of the court acting under s. 14 of the Act to rehear the case in the context of the grounds of appeal and make up its own mind, it will not disregard the judgment appealed from. The office of a trial judge is deserving of respect, and the decisions of such judges will be presupposed (and not presumed) to be free from error: see Valley Beef Producers Co-operative, at paras. 117 and 120.

252                            Turning now to specifically address the circumstances in which the Court of Appeal, when faced with questions of fact, will apply its own view of the evidence and, if necessary, pronounce the decision that ought to have been pronounced, I will distinguish between three types of questions of fact: (i) factual findings that engage the special advantage of the trial judge; (ii) factual findings that do not; and (iii) inferences based on findings of fact.  I distinguish between factual findings that engage the special advantage of the trial judge and factual findings that do not because, as noted above, more deference is called for with regard to the former.  I distinguish between factual findings generally and inferences of fact because there is an analytical difference between the two: see Housen, at para. 103 (for the minority).  Inferences involve logical deductions that rely upon findings of fact in order to come to either legal or factual conclusions.  In this case, we are only concerned with factual inferences.

(a)  Factual Findings That Engage the Special Advantage of the Trial Judge

253                            As noted previously, factual findings that engage the special advantage of the trial judge will be accorded some deference by the Court of Appeal.  To the extent that “standards of review” language is useful in the context of appellate review by way of rehearing, if only for clarity’s sake, it can be argued that, although the Court of Appeal is not constrained by the view of the evidence taken by the trial judge, when the trial judge’s factual findings engage the special advantage he or she has over an appellate tribunal, the Court of Appeal will only interfere and apply its own view of the evidence if the trial judge has committed a palpable and overriding error in his or her fact finding: H.L. (C.A.), at para. 77.  That being said, it must be borne in mind that the primary function of the Saskatchewan Court of Appeal is to correct “error or deficiency” in the particular case, and it is not relieved of this function “by the fact that the trial judge has, expressly or implicitly, reached a conclusion influenced by an opinion concerning the credibility of witnesses”:  Fox v. Percy, at para. 29.  In such a case, “making all due allowances for the advantages available to the trial judge”, the Court of Appeal must not shrink from applying its own view of the evidence if it concludes that the trial judge’s view is tainted by a palpable and overriding error.

(b)  Factual Findings That Do Not Engage the Special Advantage of the Trial Judge

254                            Factual findings that do not engage the special advantage of the trial judge are not entitled to the same level of deference as those that do.  Therefore, while the Court of Appeal will presuppose that such factual findings are free from error, given the respect that is to be accorded to the office of a trial judge, if the Court of Appeal concludes that some error has indeed been made, it will apply its own view of the evidence: see Valley Beef Producers Co-operative, at para. 117.  As above, to the extent that “standards of review” language is helpful, it can be argued that, although the Court of Appeal is not constrained by the view of the evidence taken by the trial judge, it will only interfere and apply its own view of the evidence if the trial judge has committed a simple error in his or her fact finding.

255                            The distinction between simple error and palpable and overriding error should be noted.  Because the trial judge is typically in no better position than the Court of Appeal to make factual findings that are not dependent on hearing the testimony of witnesses viva voce or being exposed to the case as a whole, no greater measure of deference is called for than presupposing (though not presuming) that the trial judge’s fact finding is free from error: see Valley Beef Producers Co-operative, at para. 117; H.L. (C.A.), at para. 73. If a simple error in this type of fact finding is detected — it need not be palpable and overriding —, the Court of Appeal must interfere and apply its own view of the evidence.  This approach is supported by my position that an appropriate interpretation of The Court of Appeal Act, 2000 indicates that the primary function of the Court of Appeal is to correct error or deficiency in the particular case.

(c)  Inferences of Fact

256                            In the case at bar, we are particularly concerned with conditions under which, in the context of an appeal by way of rehearing, the Court of Appeal will overrule a trial judge’s inference of fact and instead draw its own.  Once again, I reiterate that a trial judge is in no better position than the Court of Appeal to draw inferences of fact from a base of fact properly established, either in the first instance or on appeal: see Valley Beef Producers Co-operative, at para. 86; H.L. (C.A.), at para. 68.  Furthermore, it must be remembered that the primary function of the Court of Appeal is to correct error or deficiency in the particular case.  Because inferences are matters of logic and their accuracy can never be conclusively established, it is awkward to speak in terms of “error” with respect to these inferences; instead, I favour the use of the concept of reasonableness.  Because s. 14 of the Act states in part that “the court shall act on its own view of what, in its judgment, the evidence proves, and the court may draw inferences of fact . . . .”, it can be argued that the Court of Appeal “may” draw its own inference of fact when, after considering what, in its own view, the evidence proves and, comparing that with the inference of fact made by the trial judge, it concludes that the trial judge’s inference of fact was not reasonable.  In other words, in the context of an appeal by way of rehearing, the threshold that the Court of Appeal must pass before substituting its own inference of fact is reasonableness.  Nevertheless, as was the case with findings of fact that do not engage the special advantage of the trial judge, out of respect for the office of the trial judge, the Court of Appeal will presuppose that he or she has drawn reasonable inferences of fact: see Valley Beef Producers Co-operative, at para. 120.  The Court of Appeal in Valley Beef Producers Co-operative aptly summarized this concept as follows:

Should the court, having determined what inferences may properly be drawn, conclude that those of the chambers judge were reasonable, as presupposed, it will not interfere.  Should it reach the opposite conclusion, or be left in serious doubt, it will take its own view of what the evidence warrants in the way of inference and establish the secondary facts as it sees fit in the exercise of its faculty of judgment. [para. 120]

(5)  Previous Jurisprudence Regarding Standards of Appellate Review

257                            In light of my conclusion regarding the nature of appellate review in Saskatchewan and the circumstances in which the Court of Appeal will apply its own view of the evidence and, if necessary, pronounce the decision that ought to have been pronounced, in this final section, I will endeavour to reconcile past jurisprudence on this topic with my conclusion.

258                            Earlier, I explained why the nature of appellate review in Saskatchewan is by way of rehearing and not review for error, such that the Court of Appeal is relieved of any obligation to adopt the view of the evidence taken by the trial judge and is directed instead to act on its own view of what, in its judgment, the evidence proves.  Moreover, I noted that the Saskatchewan Act, which gives the Court of Appeal its mandate to conduct appeals in this manner, is unique.  In this context, I contend that when examining previous jurisprudence on appellate review with a view to reconciling it with my conclusion about the nature and (so-called) standards of appellate review in Saskatchewan, the focus should be on Saskatchewan-specific cases, and in particular those from this Court and the Saskatchewan Court of Appeal.  This is particularly so when the general standards of appellate review seem to be drifting away from appeal by way of rehearing and toward appeal by way of review for error: H.L. (C.A.), at para. 86.  Additionally, as noted by the Court of Appeal, these general standards “have evolved not so much on the basis of statutory provision as on the basis of judicial policy concerns — concerns having to do with finality in litigation, with cost and delay occasioned by appeal, with taxing scarce judicial resources, and so on” (para. 81).  However, because appeals are creatures of statute, these judicial policy concerns must be assessed in light of the statutory mandate conferred on the Court of Appeal.  In this case, the Saskatchewan legislature, through The Court of Appeal Act, 2000, has instructed its Court of Appeal to conduct appeals by way of rehearing rather than review for error, with the overall aim of redressing error and setting matters right: Valley Beef Producers Co-operative, at para. 70.  In this light, the judicial policy concerns noted above lose their relevance and should not be used to limit the facially unlimited right of appeal and the broad powers of the appellate court to act on that right that are provided by statute in the province of Saskatchewan. Instead, other policies apply.

259                            Examining the judicial interpretation by this Court and the Saskatchewan Court of Appeal of the nature and (so-called) standards of appellate review in Saskatchewan, in contrast with Fish J.’s assertion at para. 98 that “[n]o decision has been drawn to our attention where the [Saskatchewan Court of Appeal] has asserted a power of review by rehearing”, I have found that there are a number of Saskatchewan Court of Appeal cases that support my conclusion that the nature of appellate review in Saskatchewan is by way of rehearing and not review for error.  Moreover, to the extent that there are cases from this Court and the Saskatchewan Court of Appeal that appear to conflict with my conclusion, in my view, they can be reconciled in one of three ways.  First, a number of these cases hold that findings of fact made at trial based on the credibility of witnesses are not to be reversed on appeal unless the trial judge made some palpable and overriding error which affected his or her assessment of the facts.  This specific holding is not at odds with my conclusion regarding the nature of appellate review in Saskatchewan and the circumstances in which the Saskatchewan Court of Appeal, when faced with factual findings that engage the special advantage of the trial judge, will interfere and apply its own view of the evidence.  Therefore, these cases can be reconciled by restricting their impact to their specific rationes decidendi. Second, some of these cases do not refer to the Saskatchewan Act at all, despite the fact that it is the only statute in Canada that frees the Court of Appeal from the view of the evidence taken by the trial judge and directs it to take its own view of the evidence.  Therefore, I contend that these cases are better understood as applying to appellate review in general, the nature of which has drifted toward appeal by way of review, rather than appellate review in Saskatchewan under The Court of Appeal Act, 2000, which continues to be by way of rehearing.  Finally, some cases, or at least aspects of them, may simply require reconsideration in light of the above analysis regarding the nature of appellate review in Saskatchewan.  This Court is not bound by the particular decisions of lower courts, particularly when there is conflict within the case law.

(a)  Saskatchewan Court of Appeal Cases in Support of My Conclusion That the Nature of Appellate Review in Saskatchewan Is by Way of Rehearing

260                            At para. 97 of his reasons, Fish J. states that “the Court of Appeal for Saskatchewan appears to have for many decades prior to both Lensen and Housen understood its legislative mandate as a power of review for error”.  With respect, I must disagree with this statement because, as noted previously, since The Court of Appeal Act of 1915 was enacted, the Saskatchewan Court of Appeal has adopted the rehearing approach to appellate review that was first laid out in Coghlan v. Cumberland and approved by Anglin J. in Annable v. Coventry and Greene, Swift & Co. v. Lawrence: see, e.g., Miller v. Foley & Sons; Messer v. Messer; Monaghan v. Monaghan; Kowalski v. Sharpe; Tarasoff v. Zielinsky; Matthewson v. Thompson; French v. French, at p. 443; and Wilson v. Erbach, at p. 666.

261                            For example, in Miller v. Foley & Sons, the appeal turned entirely upon a finding of fact by the trial judge, and the Court of Appeal held that “[t]he rule no doubt is that, in cases of this kind, it is the duty of this Court to re-hear the case and to over-rule it when on full consideration the Court is convinced that the judgment appealed from is wrong” (p. 665).  Similarly, in Monaghan v. Monaghan, the Court of Appeal confirmed that, pursuant to (then) s. 8 of The Court of Appeal Act, R.S.S. 1930, c. 48, it was duty-bound to reconsider the evidence:

We are dealing with an appeal from a decision of a Judge sitting without a jury, and it is therefore our duty to reconsider the evidence.  This is especially true in this Court, where we are bound to give due effect to sec. 8 of The Court of Appeal Act, R.S.S., 1930, ch. 48.  This case does not turn upon the relative credibility of witnesses, and the question of which witness is to be believed rather than another does not arise; if it did, different considerations would, of course, apply. [p. 5]

262                            Moreover, in Tarasoff v. Zielinsky, at p. 138, the Court of Appeal confirmed that it was in as good a position as the trial judge to draw inferences of fact, which, in this case, was that the deceased was killed by the defendant’s bull. Likewise, in French v. French, at p. 443, the Court of Appeal noted that it is required to act on its own view of what the evidence proves when it cannot agree with the inferences the trial judge drew from the facts:

In the result, I can only say, with great deference, that I cannot agree with the inferences which [the trial judge] draws from the facts and, such being the case, it is my duty to act upon my own view of what the evidence proves:  The Court of Appeal Act, R.S.S. 1930, ch. 48, sec. 8; Lysnar v. National Bank of N.Z., [1935] 1 W.W.R. 625.

(See also Wilson v. Erbach, at p. 666.)

263                            Not only did the Saskatchewan Court of Appeal adopt the rehearing approach to appellate review in cases decided long before the decisions of this Court in Lensen v. Lensen, [1987] 2 S.C.R. 672, and Housen (discussed below), following these two decisions, the Court of Appeal also appears to have maintained its adherence to this form of appellate review.

264                            For instance, in Valley Beef Producers Co-operative, the Court of Appeal took occasion to examine in detail the provisions of The Court of Appeal Act, 2000, given its newness at the time and the concern among members of the bar of the Province that the role of the court is on the wane, in that it will not or cannot address appeals as robustly as it once did.  The court explained this concern in the following manner:

The concern, which seems to have been mounting over the last several years, stems from what are nowadays referred to in increasingly comprehensive terms as the “standards of appellate review”, a continually evolving set of uniform, national standards governing the determination of issues of fact and questions of law raised before the court on appeal.  It behoves us to recognize this concern and to address it. [para. 32]

265                            In its interpretation of The Court of Appeal Act, 2000, the court noted that the right of appeal conferred by s. 7(2)(a), and augmented by s. 13 when unlimited, is “a right of the person which has first and foremost to do with relief from error or deficiency, if such be found, in the decision affecting that person” (para. 65).  With regard to s. 12(1), the court stated that the legislature “could hardly have expressed itself in broader terms, evincing an intention to fully empower the court to take such steps in respect of the decision under appeal as occasion requires” (para. 70).  Although full consideration of ss. 13 and 14 was not needed to decide the case, which essentially involved the trial judge’s identification and application of the governing law, the court nevertheless went on to state that “[s]ection 14 is concerned with issues of fact, as is s. 13, but it extends beyond this, inasmuch as it goes on to empower the court to pronounce judgment, which entails the application of law to fact once properly established” (para. 73).  The court concluded that “[a]ll of this, of course, dovetails with the notion of appeal by way of re-hearing.”

266                            As for the central issue in the appeal — in what circumstances will the Court of Appeal exercise its remedial powers conferred on it by s. 12(1) and take issue with a chambers judge’s decision on the grounds of (i) the judge’s identification of the governing law and (ii) the judge’s application of that law to the facts of the case — the court held that the court will exercise these powers “if satisfied of material error of law or material deficiency in respect of the decision under appeal” (para. 92).  Moreover, although the court was of the view that appeal by way of rehearing remains the case in Saskatchewan, to the extent the basis for the exercise of the powers conferred on the court is expressed in terms of “standards of review”, it went on to offer its opinion as to the standards in Saskatchewan (paras. 95-96).

267                            First, with regard to questions of law, which, in the court’s view, include both the judge’s identification of the law as well as the judge’s application of the law to the facts, the court stated that the standard in Saskatchewan was correctness (paras. 96 and 111).  On this point, the court acknowledged that its decision was in contrast with the majority opinion in Housen; it attempted to justify this conflict in a number of ways, including by noting that the decision in Housen did not address the Saskatchewan Act (para. 108).  Second, with regard to questions of fact, the court held that, for so-called primary and secondary facts (i.e., what I term “factual findings” and “inferences of fact”) that are not dependent on hearing viva voce evidence or assessments of credibility, “no greater measure of deference is called for than that pertaining to the office of the judge, who being a superior court judge may be presupposed to have made reasonable findings of fact” (para. 117).  In contrast, for findings of fact that do arise out of viva voce evidence and assessments of credibility, the court held that “again the court will address the issue on the standard of ‘reasonableness’, determined on the same basis and to the same extent as otherwise, but accompanied by an appreciably higher measure of deference by reason of the judge’s advantage, denied this court, of having seen and heard the witnesses” (para. 119).

268                            While it is apparent that there is some divergence between my conclusion and that of the Court of Appeal in Valley Beef Producers Co-operative regarding the circumstances in which the Court of Appeal will interfere with the trial judge’s decision and instead apply its own view of the evidence, this is of secondary importance.  If necessary, the conflict can be reconciled by this Court.  What is of primary importance is the fact that in Valley Beef Producers Co-operative, a decision that was rendered after The Court of Appeal Act, 2000 was enacted and post-Housen, the Court of Appeal confirmed that in Saskatchewan, at least, “appeal by way of re-hearing remains the case” (para. 95).

269                            This understanding of the nature of appellate review in Saskatchewan was repeated by the Court of Appeal in the case at bar: see H.L. (C.A.), at para. 77.                                             

(b)  Reconciliation of Cases That Appear to Conflict With Conclusion That the Nature of Appellate Review in Saskatchewan Is by Way of Rehearing

270                            To the extent that there are cases from this Court and the Saskatchewan Court of Appeal that appear to conflict with my conclusion regarding the nature and (so-called) standards of review in Saskatchewan, in my view, they can be reconciled in one of three ways: (i) they can be limited to their specific rationes decidendi; (ii) they can be distinguished because they do not refer to The Court of Appeal Act, 2000 or its predecessors or (iii) they may simply require reconsideration in light of the above analysis regarding the nature of appellate review in Saskatchewan.  In this section, I will discuss in turn each of the cases cited by Fish J. that appear to conflict with my understanding that the nature of appellate review in Saskatchewan is by way of rehearing.

(i)     Board of Education of the Long Lake School Division No. 30 of  Saskatchewan v. Schatz

271                            Prior to Lensen (discussed below), the leading case in Saskatchewan regarding appellate review was Board of Education of the Long Lake School Division No. 30 of Saskatchewan v. Schatz (1986), 49 Sask. R. 244.  In this case, Sherstobitoff J.A., for himself and for Tallis J.A., held that “palpable and overriding error” is the proper standard to apply to intervention by a Court of Appeal in respect of findings of fact by a trial judge. Although Sherstobitoff J.A. acknowledged that a “rehearing” is similar to what s. 8 (now s. 14) of The Court of Appeal Act requires of the court, he also stated:

While, on its face, s. 8 appears to confer not only the power, but a duty to “rehear” or “retry” a case, simple fairness and justice require a court of appeal to recognize that a trial judge has an immense advantage in assessing evidence and arriving at findings of fact as opposed to a court of appeal which is confined to an examination of a cold black and white record of a trial proceeding, completely devoid of the tension, emotion, colour, and atmosphere of a trial, all of which factors are immeasurably important in assisting a trial judge in arriving at his conclusions.  It is for these reasons that a court of appeal must extend very substantial deference to the finding of facts of a trial judge.  The issue has been considered on many occasions by the Supreme Court of Canada and its decisions bear these principles out. [p. 248]

272                            After canvassing the case law to date regarding an appellate court’s jurisdiction to review findings of fact, Sherstobitoff J.A. summarized the generally accepted principles that emerged from the authorities as follows:

1.  This court is obliged under s. 8 of the Court of Appeal Act to review the impugned findings of fact of a trial judge and if it finds error, to make its own findings of fact where possible in substitution for those of the trial judge.

2.  In a review of the findings of fact of a trial judge, the proper starting point is deference to a trial judge’s findings in recognition of the great advantage which a trial judge has in assessing the evidence.

3.  With respect to determinations of credibility and findings of fact based thereon, a court of appeal should not intervene unless it is certain that the trial judge erred and can identify the critical error.

4.  Where the issue is the trial judge’s view of the evidence as a whole, a court of appeal should not interfere unless there has been a palpable or overriding error.  It must be shown that the trial judge has failed to use or has palpably misused his advantage.

5.  Where there is evidence to support a finding of fact a court of appeal should not interfere in the absence of palpable or demonstrable error. [Emphasis in original; p. 251.]

273                            Although this does not dispose of all the points of conflict between the decision in Long Lake School Division and my conclusion in these reasons, it must be noted that in his reasons Sherstobitoff J.A. accepted that a “rehearing” is similar to what s. 8 of The Court of Appeal Act requires of the Court of Appeal, and he acknowledged that “[t]his court is obliged under s. 8 of the Court of Appeal Act to review the impugned findings of fact of a trial judge and if it finds error, to make its own findings of fact where possible in substitution for those of the trial judge” (p. 251).  These comments regarding the court’s power to conduct an appeal by way of (or at least similar to) rehearing are in contrast with Fish J.’s assertion at para. 98 of his reasons that “[n]o decision has been drawn to our attention where the court has asserted a power of review by rehearing.”

274                            Furthermore, and contrary to Fish J.’s assertion in his reasons at para. 93, it is arguable that Sherstobitoff J.A.’s comments regarding the need for appellate deference to factual findings ought to be restricted to matters that engage the special advantage of the trial judge, which include but are not limited to assessments of credibility.  For instance, although Sherstobitoff J.A. refers simply to a “finding of fact” when he states in his fifth numbered point quoted above that “[w]here there is evidence to support a finding of fact a court of appeal should not interfere in the absence of palpable or demonstrable error”, in the three previous points, he makes reference to the special advantage of the trial judge and circumstances that engage it, namely, determinations of credibility and viewing the evidence as a whole.  Thus, it can be argued that, in light of his discussion of the special advantage of the trial judge immediately preceding it, Sherstobitoff J.A.’s reference to a “finding of fact” in his fifth point (and the appellate deference it is owed) ought to be read as pertaining to only those findings that engage the special advantage of the trial judge.  This argument is strengthened by Sherstobitoff J.A.’s conclusion on this point in the appeal that “[t]he matter reduces itself to whether the trial judge erred in accepting the evidence of Schatz as to intent.  This is a matter of credibility where an appeal court should not interfere” (p. 252).  Reading down Sherstobitoff J.A.’s comments regarding the need for appellate deference to factual findings in this manner would accord with my conclusion that the powers conferred by s. 14 (then s. 8) of the Act are subject to the judicial policy concern that trial judges enjoy a special advantage over an appellate court, such that when the trial judge’s factual findings engage this special advantage, the Court of Appeal will only interfere and apply its own view of the evidence if the trial judge has committed a palpable and overriding error in his or her fact finding.

275                            To the extent that it remains unclear whether Sherstobitoff J.A.’s comments regarding the need for appellate deference to factual findings can be restricted to matters that engage the special advantage of the trial judge or whether they are to apply to all factual findings and inferences, I would simply say that the reasoning in this case may be in need of reconsideration in light of my analysis regarding the nature and (so-called) standards of appellate review in Saskatchewan.  In H.L. (C.A.), the Court of Appeal acknowledged that this might be the case and stated that:

In Board of Education of Long Lake School Division No. 30 v. Schatz et al., we made some effort to reconcile the general standards of review as they were then emerging with what was then section 8, and is now section 14, of the Act.  However, we did not then appreciate where the continuing evolution of these standards would ultimately lead.  The appeal in that case centred on a finding of fact based on the trial judge’s assessment of the credibility and reliability of the testimony of one of the parties, whose intention was in issue.  That, of course, made it unnecessary, strictly speaking, to do more than accommodate the test of “palpable and overriding error” to what is now section 14. In any event, given the way all of this has played out, we may have to reconsider some of what we had to say in Schatz in light of what a closer examination of sections 13 and 14 reveals. [para. 91]

276                            Moreover, it must be noted that Wakeling J.A. dissented in Long Lake School Division.   His comments support my position that the court’s decision in this case may be in need of reconsideration.  In particular, Wakeling J.A. noted that the appellate process is based upon the generally accepted view that three or more heads are better than one and that this was probably the reason for the broad language employed in s. 8 of The Court of Appeal Act (1978): Long Lake School Division, at p. 254. In this regard, Wakeling J.A. stated:

There can be little doubt as to the extent of the legal mandate that has been given this court, and if review on appeal is to be as meaningful as the legislators clearly intended, I see no reason for limiting this right of appellate review to the point where the conclusion reached by the trial judge is completely unsupportable, or misses the mark so widely as to call into question whether the mark was ever a point of aim.  I make this point for I am of the opinion that it is possible to so focus on the restrictions on appellate review that one useful aspect of the appeal process relating to factual considerations becomes largely emasculated.  The party suffering from a judicial error of opinion has only the slightest opportunity for correction, even though such party suffers an injustice which in my view must be as hard to accept as that arising from an error of law. [pp. 254-55]

277                            Wakeling J.A. concluded his reasons in this case by stating that it would be a strange anomaly if

in a province which apparently grants this court the broadest powers of review of any appellate court in Canada (Hallberg v. Canadian National Railway Company (1955), 16 W.W.R. (N.S.) 539, at 544), we are nonetheless as restrictive as any such court in our ability to review evidence given at trial (where such evidence is not subject to degrees of credibility) and to differ from a trial judge’s conclusions of fact drawn from such evidence. [p. 259]

(ii)   Tanel v. Rose Beverages (1964) Ltd. and Sisson v. Pak Enterprises Ltd.

278                            In Tanel v. Rose Beverages (1964) Ltd. (1987), 57 Sask. R. 214 (C.A.), both Bayda C.J.S., for himself and Wakeling J.A., and Vancise J.A. in dissent followed and applied the principles regarding the review of a trial judge’s findings of fact that were set out by Sherstobitoff J.A. in Long Lake School Division. Cameron J.A., for himself, Gerwing and Sherstobitoff JJ.A., in Sisson v. Pak Enterprises Ltd. (1987), 64 Sask. R. 232 (C.A.), did likewise.  Therefore, I do not consider it necessary to analyse these cases separately.  To the extent that the decision in Long Lake School Division can be reconciled with my conclusion regarding the nature and (so-called) standards of appellate review in Saskatchewan, so too then can the decisions in these two cases. If it needs to be reconsidered, so do these other cases.

(iii)  Lensen

279                            In Lensen, this Court addressed whether it is proper for the Court of Appeal to reverse a finding of fact made by a judge at first instance.  In his reasons for the Court, Dickson C.J. made explicit reference to s. 8 of The Court of Appeal Act (1978), and noted that “[d]espite its apparently broad language, s. 8 has been given a relatively narrow interpretation”:  p. 682, citing the Court of Appeal’s decision in Long Lake School Division in support. In this regard, Dickson C.J. stated:

It is a well-established principle that findings of fact made at trial based on the credibility of witnesses are not to be reversed on appeal unless it can be established that the trial judge made some “palpable and overriding error which affected his assessment of the facts”: . . . .  While section 8 of the Saskatchewan Court of Appeal Act authorizes the Court of Appeal to “draw inferences of fact”, this task must be performed in relation to facts as found by the trial judge.  Unless the trial judge has made some “palpable and overriding error” in this regard, s. 8 should not be construed so as to modify the traditional role of the Court of Appeal with respect to those findings. [Emphasis added; pp. 683-84.]

280                            In terms of reconciling the decision of this Court in Lensen with my conclusion regarding the nature and (so-called) standards of appellate review in Saskatchewan, two points must be made.  First, to the extent that the decision in Lensen relies upon the Court of Appeal’s decision in Long Lake School Division, I simply refer to my earlier comments regarding that case. Second, the specific issue in Lensen was “whether the trial judge found as a fact that no oral agreement existed between the parties and, if so, whether the Court of Appeal erred by substituting its version of the facts” (p. 679).  Dickson C.J. held that it was clear that the trial judge had made a finding that there was no agreement as alleged between the parties.  Moreover, he held that “the trial judge was entitled to believe the defendant father’s evidence and the evidence of his witnesses and reject the son’s testimony and the testimony of his witnesses as to the existence of an oral contract between the parties”: p. 684.  Thus, Dickson C.J. concluded that because the trial judge did not make any palpable and overriding error in his assessment of the facts, the Court of Appeal exceeded its role as set out in s. 8 of the Act when it substituted its version of the facts for that given by the trial judge.

281                            From this brief review, it is clear that the factual finding at issue in Lensen (i.e., whether an oral contract existed between the parties) was dependent on the trial judge’s assessments of credibility; therefore, I conclude that this Court’s decision in Lensen must be limited to its specific ratio decidendi, which, in my view, is the following: findings of fact made at trial based on the credibility of witnesses are not to be reversed on appeal unless it can be established that the trial judge made some “palpable and overriding error which affected his assessment of the facts” (p. 684).  If this Court’s decision in Lensen is limited in this way, it accords with my conclusion that the powers conferred by s. 14 (then s. 8) of the Act are subject to the judicial policy concern that trial judges enjoy a special advantage over an appellate court, such that when the trial judge’s factual findings engage this special advantage (e.g., when based on credibility), the Court of Appeal will only interfere and apply its own view of the evidence if the trial judge has committed a palpable and overriding error in his or her fact finding.

282                            Nevertheless, I do acknowledge that in the excerpt from Lensen quoted above, Dickson C.J. states that while s. 8 of the Act authorizes the Court of Appeal to draw inferences of fact, this task (inference drawing) must be performed in relation to facts as found by the trial judge, and, unless the trial judge has made some palpable and overriding error in this regard, s. 8 should not be construed so as to modify the traditional role of the Court of Appeal with respect to those findings.  I also acknowledge that, at first blush, this statement may appear to suggest that a palpable and overriding error standard should also apply to inferences.  However, a closer reading of this statement reveals that Dickson C.J. is merely restating the principle he articulated immediately prior to this statement.  In other words, he is stating that while the Act permits the Court of Appeal to draw inferences and inferences are logical conclusions based on established facts, this inference-drawing power does not change the fact that the court may only intervene and reverse factual findings based on the credibility of witnesses when the trial judge has made some palpable and overriding error which affected his assessment of the facts.  As noted above, when the decision in Lensen is limited to this specific ratio decidendi, it is consistent with my conclusion that the powers conferred by s. 14 (then s. 8) of the Act are subject to the judicial policy concern that trial judges enjoy a special advantage over an appellate court, such that when the trial judge’s factual findings engage this special advantage (e.g., when based on credibility), the Court of Appeal will only interfere and apply its own view of the evidence if the trial judge has committed a palpable and overriding error in his or her fact finding.

(iv)  Bogdanoff v. Saskatchewan Government Insurance; Knight v. Huntington; and Brown v. Zaitsoff Estate

283                            The Court of Appeal’s recent decisions in Knight v. Huntington (2001), 14 B.L.R. (3d) 202, 2001 SKCA 68, Bogdanoff v. Saskatchewan Government Insurance (2001), 203 Sask. R. 161, 2001 SKCA 35, and Brown v. Zaitsoff Estate (2002), 217 Sask. R. 130, 2002 SKCA 18, all deal with the issue of the conditions under which the Court of Appeal can interfere with a trial judge’s decision on a question of fact.  Because all of these cases explicitly rely upon this Court’s decision in Lensen, comments regarding that case, as well as my comments regarding Long Lake School Division, the decision upon which Lensen relies, should be kept in mind when reviewing them.

284                            Moreover, in none of these cases was there any mention of The Court of Appeal Act, 2000.  As I will more fully explain with regard to this Court’s decision in Housen, this omission is indicative of a need to reconsider the precedential value of these three cases, given the uniqueness of the Saskatchewan Act.

285                            Finally, to the extent that the Court of Appeal in these three cases applied the palpable and overriding error standard to findings of fact in which the special advantage of the trial judge is not engaged or to inferences of fact, the court’s reasoning in this regard may be in need of reconsideration in light of my analysis regarding the nature and (so-called) standards of appellate review in Saskatchewan.  For example (and as noted by Fish J. at para. 108 of his reasons), in Bogdanoff v. Saskatchewan Government Insurance, Gerwing J.A., in oral reasons for the court, applied a palpable and overriding error standard to a finding of causation. In Brown v. Zaitsoff Estate, Tallis J.A. applied the same standard to a finding that the respondent had not exerted undue influence on his mother.  Similarly, in Knight v. Huntington, Sherstobitoff J.A. stated that “to the extent that [the trial judge’s] findings depended upon drawing inferences of fact, the appellants must show that there was no evidence from which those conclusions could reasonably be drawn” (para. 28).   In contrast, as I mentioned earlier, in Saskatchewan, where the nature of appellate review is by way of rehearing, the Court of Appeal will interfere with findings of fact that do not engage the special advantage of the trial judge when there is a simple error, and for inferences of fact, the threshold for appellate intervention is reasonableness.

(v)  Housen v. Nikolaisen

286                            The appeal in Housen arose out of a motor vehicle accident which occurred on a rural road in the Municipality of Shellbrook, Saskatchewan.  At trial, the judge found that the driver of the vehicle, Douglas Nikolaisen, was negligent in travelling the rural road at an excessive rate of speed and in operating his vehicle while impaired. The trial judge also found the respondent, the Municipality of Shellbrook, to be at fault for breaching its duty to keep the road in a reasonable state of repair as required by s. 192 of The Rural Municipality Act, 1989, S.S. 1989-90, c. R-26.1.  The Court of Appeal overturned the trial judge’s finding that the municipality was negligent. At issue in the appeal to this Court was whether the Court of Appeal had sufficient grounds to intervene in the decision of the lower court.  In Housen, the Court was divided as to the proper standard of review for questions of mixed law and fact, as well as for inferences drawn from findings of fact.  In the case at bar, we are concerned only with the latter issue; that is, when the Saskatchewan Court of Appeal will interfere with a trial judge’s decision regarding inferences drawn from findings of fact.

287                            In his reasons, Fish J. states that the majority and the minority in Housen shared the same view as to the standard of review for questions of fact, which include findings of fact and inferences of fact.  In his opinion, “[i]t was in the application of this shared view as to the governing principle that the Court divided”: para. 66 (emphasis in original).  In this regard, Fish J. states that, “according to the majority in Housen, the test to be applied in reviewing inferences of fact is ‘not to verify that the inference can be reasonably supported by the findings of fact of the trial judge, but whether the trial judge made a palpable and overriding error in coming to a factual conclusion based on accepted facts’ which, in its view, implied a stricter standard” (para. 72, citing Housen, at para. 21 (emphasis in Housen)).  Fish J. notes that the majority was concerned that drawing an analytical distinction between factual findings and factual inferences, as advocated by the minority, might lead appellate courts to involve themselves in unjustified reweighing of evidence.  He also notes that the majority stated that “[i]f there is no palpable and overriding error with respect to the underlying facts that the trial judge relies on to draw the inference, then it is only where the inference-drawing process itself is palpably in error that an appellate court can interfere with the factual conclusion” (para. 72, citing Housen, at para. 23 (emphasis in Housen)).

288                            Fish J. then goes on to clarify these passages from the majority reasons in Housen.  Specifically, he states that they cannot be taken to have restricted appellate scrutiny of the judge’s inferences to an examination of the primary findings upon which they are founded and the process of reasoning by which they were reached. Instead, Fish J. asserts that “[a]ppellate scrutiny determines whether inferences drawn by the [trial] judge are ‘reasonably supported by the evidence’” (para. 74).  This statement is in direct conflict with the majority’s statement in Housen that “the standard of review is not to verify that the inference can be reasonably supported by the findings of fact of the trial judge” (para. 21 (emphasis in original)).  Moreover, to the extent that Fish J.’s clarification of the majority reasons in Housen can be understood to mean that the views of the majority and minority as to the standard of review for inferences of fact can be reconciled, I respectfully disagree.  In my view, there was a major rift between the views of the majority and the minority on this issue.

289                            For instance, although I agreed in my reasons for the minority in Housen that the standard of review is identical for both findings of fact and inferences of fact, I stated that “it is nonetheless important to draw an analytical distinction between the two” (para. 103).  I was concerned that “[i]f the reviewing court were to review only for errors of fact, then the decision of the trial judge would necessarily be upheld in every case where evidence existed to support his or her factual findings.” Thus, it was my view that, “this Court is entitled to conclude that inferences made by the trial judge were clearly wrong, just as it is entitled to reach this conclusion in respect to findings of fact”.  Thus, when reviewing the making of an inference, I stated that “the appeal court will verify whether it can reasonably be supported by the findings of fact that the trial judge reached and whether the judge proceeded on proper legal principles”.  The fact that the trial judge referred to some evidence to support his or her finding on an issue will not insulate the finding from review (para. 169).

290                            Iacobucci and Major JJ. for the majority in Housen disagreed with my articulation of the standard of review for inferences of fact in two respects.  First, they stated that “the standard of review is not to verify that the inference can be reasonably supported by the findings of fact of the trial judge, but whether the trial judge made a palpable and overriding error in coming to a factual conclusion based on accepted facts, which implies a stricter standard” (para. 21 (emphasis in original)).  With respect, I still do not understand how an appellate court can reasonably make such a distinction, or how it can be said that the legislature would have wanted that a court clearly mandated to correct errors be so limited. I cannot see on what basis such a judicial policy can be justified.  Second, in their view, drawing an analytical distinction between factual findings and factual inferences could lead appellate courts to engage in unjustified reweighing of the evidence.  Iacobucci and Major JJ. stated that although they agreed that it is open to an appellate court to find that an inference of fact made by the trial judge is clearly wrong, they added the caution that “where evidence exists to support this inference, an appellate court will be hard pressed to find a palpable and overriding error” (para. 22).  This caution was based on the idea that “where evidence exists which supports [a factual conclusion], interference with this conclusion entails interference with the weight assigned by the trial judge to the pieces of evidence”. Because the majority felt that it is not the role of appellate courts to second-guess the weight assigned to various items of evidence, given what it perceived to be the advantageous position of the trial judge in this respect, it stated that “[i]f there is no palpable and overriding error with respect to the underlying facts that the trial judge relies on to draw the inference, then it is only where the inference-drawing process itself is palpably in error that an appellate court can interfere with the factual conclusion” (para. 23 (emphasis in original)).

291                            In response to the majority’s criticism of my approach to the standard of review for inferences of fact, I would now say that we all agree that an appellate court must verify whether the making of an inference can be reasonably supported by the trial judge’s findings of fact.  It is the question of the standard of review that has to be addressed in more clear terms.  In my view, there is no difference between concluding that it was “unreasonable” or “palpably wrong” for a trial judge to draw an inference from the facts as found by him or her, and concluding that the inference was not reasonably supported by those facts.  The distinction is merely semantic.  To transport the analysis to the inference-drawing process is, in my view, of no assistance.

292                            In light of the above discussion and in contrast with Fish J.’s assertion that it was in the application of a shared view as to the standard of review that the Court divided, I contend that the majority and the minority in Housen articulated different standards of review for inferences of fact.  The majority was of the view that where evidence exists to support the trial judge’s inference of fact, an appellate court will be hard pressed to find a palpable and overriding error.  The error would have to be in the process.  In contrast, the minority argued that it is not enough for an appellate court to verify that some evidence “exists” to support the trial judge’s inference; rather, the appellate court must verify whether the inference can be reasonably supported by the findings of fact that the trial judge reached and whether the judge proceeded on proper legal principles.  The standard articulated by the minority “entitles the appellate court to assess whether or not it was clearly wrong for the trial judge to rely on some evidence when other evidence points overwhelmingly to the opposite conclusion” (Housen, at para. 169). P. M. Perell, in his article entitled “The Standard of Appellate Review and the Ironies of Housen v. Nikolaisen” (2004), 28 Advocates’ Q. 40, at p. 52, notes that “this is a fundamental principle of Aristotelian logic, the law of contradiction that a proposition and its negation cannot both be true at the same time and in the same respect”.

293                            My contention that the majority and the minority in Housen articulated different standards of review for factual inferences is supported by the fact that, on the issue of causation, the majority and minority came to opposite conclusions.  For instance, the majority held that the trial judge’s finding of fact that a hidden hazard existed at the curve where the accident occurred should not be interfered with, and this finding formed part of the basis of her findings concerning causation.  Thus, because her finding regarding the existence of a hidden hazard had a basis in the evidence, the majority held that “her conclusions on causation grounded in part on the hidden hazard finding also had a basis in the evidence” and should not have been interfered with by the Court of Appeal (paras. 73 and 75).  In contrast, the minority held that in coming to her conclusion on causation, the trial judge erred by: (a) misapprehending the evidence; (b) ignoring relevant evidence; and (c) drawing erroneous conclusions: see paras. 158-69; see also Perell, at p. 48.  The minority emphasized that even though the trial judge referred to some evidence to support her findings on causation, this did not insulate her findings from review.

294                            The differences between the majority and minority opinions in Housen aside, in my opinion this case can be reconciled with my conclusion regarding the nature and (so-called) standards of appellate review in Saskatchewan by limiting its application as an authority to general standards of appellate review only.

295                            Housen was an appeal from the Court of Appeal for Saskatchewan, but it did not refer to Saskatchewan’s unique Court of Appeal Act (the modernized version of this Act, The Court of Appeal Act, 2000, came into force after the Court of Appeal’s decision in Housen was released and while the appeal to this Court was pending). The Court of Appeal in H.L. (C.A.) surmised that this Court’s failure to refer to the former Act, and especially s. 8 (now s. 14), may have been due to the fact that s. 8 “had ceased over time to attract much attention, with the exception of the limited attention given to it by this court in Board of Education of Long Lake School Division No. 30 v. Schatz et al. . . ., and by the Supreme Court of Canada in Lensen v. Lensen”: para. 90.  In addition to this, the court stated that “the general standards of review have tended to evolve outside the statutory framework regarding appeal”.

296                            Regardless of the reason why, in my view the mere fact that this Court did not refer to The Court of Appeal Act is indicative of a need to limit its precedential value to general standards of review only. As I noted earlier, Saskatchewan’s legislative scheme for appeals is, as far as I am aware, the only one among all of the statutes governing the powers of appellate courts in Canada that frees the Court of Appeal from the view of the evidence taken by the trial judge and directs it to take its own view of the evidence.  I conclude that this makes the nature of appellate review in Saskatchewan unique.

297                            With respect, one would reasonably think that such a unique legislative mandate is deserving of mention.  However, because this mandate was not brought to the Court’s attention by the parties, the majority in Housen states that the role of the appellate court was aptly defined in Underwood v. Ocean City Realty Ltd. (1987), 12 B.C.L.R. (2d) 199, at p. 204, a British Columbia Court of Appeal decision in which the court stated:

The appellate court must not retry a case and must not substitute its views for the views of the trial judge according to what the appellate court thinks the evidence establishes on its view of the balance of probabilities.

This statement from the British Columbia Court of Appeal is in direct contrast with the legislative direction to the court in s. 8 of The Court of Appeal Act (now s. 14 of The Court of Appeal Act, 2000) to “act upon its own view of what the evidence in its judgment proves”.

298                            In my respectful view, the Court’s failure to mention The Court of Appeal Act at all and its use of a statement from a different province’s Court of Appeal that is in conflict with the clear language of the Act demonstrates that the decision in Housen should not be used to understand the nature and (so-called) standards of review in Saskatchewan.  Rather, this decision if left untouched should be limited in its application to general standards of appellate review.  Thus, I agree with the Court of Appeal that “the majority judgment in Housen v. Nikolaisen represents the culmination of the evolution of the general standards of appellate review and serves to supply an adjudicative framework that differs in material respects from that laid down by the Court of Appeal Act, 2000 and, in particular, section 14 of the Act”:  H.L. (C.A.), at para. 92.

299                            Before leaving my discussion of Housen, there is one further point to consider.  I said earlier that a trial judge is in no better position than the Court of Appeal to draw inferences of fact from a base of facts properly established.  I concluded that, in Saskatchewan, where the nature of appellate review is by way of rehearing, the threshold that the Court of Appeal must pass before substituting its own inference of fact is reasonableness.  I used the term reasonableness because it is awkward to speak of a “correct” inference.

300                            In contrast, I acknowledged that factual findings that engage the special advantage of the trial judge should be accorded some deference by the Court of Appeal.  Therefore, I concluded that the Court of Appeal will only interfere with this type of factual finding and apply its own view of the evidence if the trial judge has committed a palpable and overriding error.

301                            Implicit in my analysis is the idea that, in the context of an appeal by way of rehearing, a palpable and overriding error standard, as it applies to factual findings engaging the special advantage of the trial judge, is more deferential than the reasonableness standard, as it applies to inferences drawn from those facts.  It has not escaped me that some might find this reasoning to be potentially inconsistent with my reasons in Housen.

302                            As discussed above, in Housen, I stated that “[t]he Court [of Appeal] will not overturn a factual finding unless it is palpably and overridingly, or clearly wrong”:  para. 102.  I also stated that “[i]n reviewing the making of an inference, the appeal court will verify whether it can reasonably be supported by the findings of fact that the trial judge reached and whether the judge proceeded on proper legal principles”: para. 103.  I then concluded that “the standard of review is identical for both findings of fact and inferences of fact”, noting that “there is no difference between concluding that it was ‘unreasonable’ or ‘palpably wrong’ for a trial judge to draw an inference from the facts as found by him or her and concluding that the inference was not reasonably supported by those facts” (paras. 103-4).

303                            Because I conclude that there is no difference between a palpably wrong inference and an inference that is not reasonably supported by the facts, my attempt to distinguish the so-called standard of review in Saskatchewan for factual findings engaging the special advantage of the trial judge (i.e., palpable and overriding error) from that for factual inferences (i.e., reasonableness) may be seen to conflict with my conclusion.

304                            The conflict is only apparent. Specifically, in my opinion, it is necessary to consider foremost that the nature of appellate review in Saskatchewan is very different than the general nature of appellate review in the rest of Canada; that is, in Saskatchewan, appeal is conducted by way of rehearing, while in the rest of Canada, it appears that an appellate court reviews for error.  Thus, in Saskatchewan, the Court of Appeal is not actually reviewing the inferences drawn by the trial judge to see if they are reasonably supported by the facts; rather, pursuant to the legislative direction in s. 14 of The Court of Appeal Act, 2000, the court is making its own inferences and comparing them with those drawn by the trial judge.  The Court of Appeal will only interfere and substitute its own inferences if, on the basis of this comparison, it concludes that those inferences drawn by the trial judge were not reasonable.  Therefore, in this light, it would be inappropriate to compare my comments in Housen about the general standard of review for factual findings and inferences of fact with those I made in the context of appellate review under The Court of Appeal Act, 2000 in Saskatchewan.  Although this may be a fine distinction, I would add that the appearance of this conflict is, in my view, indicative of the problem with using “standards of review” language in the context of appellate review by way of rehearing.  While expressing the circumstances in which the Saskatchewan Court of Appeal will apply its own view of the evidence and, if necessary, pronounce the decision that ought to have been pronounced in language that does not connote the concept of appeal by way of review for error would most likely avoid conflicts such as the one discussed in this section, for purposes of clarity, it may still be necessary (for the purposes of this appeal, at least) to use “standards of review” language, subject, perhaps, to the proviso that such language is not meant to invite comparison to actual standards of review employed in appeals by way of review for error.

(6)  Conclusion

305                            The Court of Appeal Act, 2000 is unique in Canada, and its provisions must “mean something”.  However, in his reasons in this case, Fish J. concludes that he is “not at all persuaded that the 2000 Act was intended to create for Saskatchewan an appellate court radically different, in powers and purpose, from its counterparts in the other provinces”: para. 15.  Based on my examination of the grammatical and ordinary sense of the words used in ss. 13 and 14 of The Court of Appeal Act, 2000, as well as the object of the Act, the object of the specific legislative provisions that form the statutory framework for the business of appeal, and the Act’s historical foundations, I respectfully disagree.  In my view, The Court of Appeal Act, 2000 mandates that the nature of appellate review in Saskatchewan is by way of rehearing, with the Court of Appeal being directed to take its own view of the evidence and being empowered to draw inferences of fact and pronounce the decision that ought to have been pronounced by the trial judge.

B.      Application of the Standard of Review

306                            As will be explained more fully below, the trial judge’s determination of H.L.’s entitlement to pecuniary damages was based on a series of factual inferences. In Saskatchewan, the (so-called) standard of review for such inferences is reasonableness.  In my view, the Court of Appeal correctly applied this standard when it set aside the trial judge’s pecuniary damages award, because the factual inferences on which the award was based were not reasonably supported by the evidence and were therefore not reasonable.  As mentioned earlier, I would find the reasons of the trial judge reviewable on the general standard set out in Housen as well.  The findings were so unreasonable that they amounted to palpable error in the appreciation of the evidence and the inferences drawn.

307                            While this conclusion is sufficient to dispose of this appeal, in my view, it is also necessary to comment upon the trial judge’s assessment of damages for past loss of earning capacity, since I agree with the Court of Appeal that the trial judge erred in his assessment in four respects.  First, he failed to consider the plaintiff’s duty to mitigate.  Second, he unreasonably concluded that the plaintiff did not have a “crumbling skull” and therefore attributed too much to Starr’s wrongful acts in his assessment of pecuniary damages.  Third, he did not reduce the damages award to reflect the time H.L. was incarcerated.  Fourth, he erred in not accounting for the social assistance payments H.L. received during the relevant period.

308                            As it is the more fundamental error, I will commence my reasons in this section with a discussion of why the basic evidentiary foundation for the pecuniary damages award for past and future loss of earnings is lacking.

(1)  No Evidentiary Foundation for Award of Pecuniary Damages for Past and Future Loss of Earnings

309                            In order to find that the Government of Canada is liable to H.L. for pecuniary damages for past and future loss of earnings, one must determine whether H.L. suffered and will suffer a loss of employment income because of the two acts of sexual abuse to which he was subjected by Starr.  At trial, Klebuc J. came to an affirmative conclusion on this issue.  In support of this conclusion, he drew two factual inferences: (i) Starr’s sexual abuse of H.L. caused H.L.’s emotional and alcohol-related problems; and (ii) these problems caused H.L.’s past loss of earnings and will cause H.L. to lose earnings in the future.  In particular, with regard to the second inference, the trial judge stated:

In my view, [H.]L.’s sporadic work record is consistent with the emotional difficulties described by Arnold and Stewart in their psychological assessments. However, his limited work history demonstrated to my satisfaction his willingness and ability to work as a construction worker and farm labourer but for his problem with alcohol. 

                                                                   . . .

[H.]L.’s continuous participation in the repair of motor vehicles for reward satisfied me that he had the aptitude and interest necessary to secure and maintain full-time employment as a mechanic but for his emotional difficulties and consequential difficulty with alcohol. [paras. 65-66]

310                            The Court of Appeal took its own view of what the evidence, in its judgment, proves, and, on this basis, it concluded that not only does the evidence fail to prove that the plaintiff was wholly or largely unable to work because of Starr’s sexual abuse, it even fails to prove the basic premise that the plaintiff was wholly or largely unable to work.  Although the court noted that the drawing of such an inference was not inconceivable, in its view, this inference was not readily drawn and would require more and better evidence than was adduced in this case.

311                            In my view, the Court of Appeal correctly followed the statutory direction in s. 14 of the Act when it took its own view of what the evidence proves.  Moreover, for the reasons that follow, I contend that the Court of Appeal was correct to interfere and overturn the trial judge’s inference that H.L. suffered and will suffer loss of employment income because of Starr’s two acts of sexual abuse, since this inference was not reasonably supported by the evidence and, therefore, not reasonable.

312                            I will begin with a discussion of why the trial judge’s conclusion regarding the liability of the Government of Canada and Starr for past loss of earnings was not reasonable.

(a)  Liability for Past Loss of Earnings

313                            As noted above, at trial, Klebuc J. drew two factual inferences in support of his conclusion that the Government of Canada and Starr were liable to H.L. for his past loss of earnings: (i) Starr’s sexual abuse of H.L. caused H.L.’s emotional and alcohol-related problems; and (ii) these problems caused H.L.’s past loss of earnings.  In my view, this causal chain must fail.  First, the inference that Starr’s sexual abuse caused H.L.’s alcoholism lacks a sufficient evidentiary basis.  That being so, there is no need to consider whether H.L.’s alcoholism caused him to lose employment income.  Second, the evidence adduced at trial cannot reasonably support the inference that H.L.’s emotional problems caused him to lose employment income.  Therefore, it is not necessary to comment upon whether Starr’s sexual abuse caused H.L.’s emotional problems.  Given these evidentiary gaps in the trial judge’s chain of reasoning, it was not reasonable for him to conclude that H.L. suffered a loss of employment income because of the two acts of sexual abuse to which he was subjected by Starr, and the Court of Appeal was correct to interfere and set aside the award of pecuniary damages for past loss of earnings.

(i)   The Inference That Starr’s Sexual Abuse Caused H.L.’s Alcoholism

314                            The trial judge’s inference that Starr’s sexual abuse caused H.L.’s alcoholism primarily derives from the opinions of Dr. Arnold and, especially, Mr. Stewart (paras. 26-27 and 29).  Both of these witnesses expressed a variety of opinions, including those on alcoholism and the cause of H.L.’s alcoholism in particular.  I recognize, like the Court of Appeal, that alcoholism is now increasingly being recognized as a disease, the etiology of which is likely to be outside the experience and knowledge of a judge or jury.  Therefore, expert opinion evidence was necessary to enable the trial judge to appreciate this issue: see R. v. Mohan, [1994] 2 S.C.R. 9, at p. 23.  However, while expert opinions on the etiology of alcoholism and the cause of H.L.’s alcoholism were necessary, Dr. Arnold and Mr. Stewart were not qualified to express opinions on this subject.  Relying on improperly qualified expert opinions led to error.

315                            As noted in Mohan, in order to be admissible, expert evidence must be provided by a properly qualified expert, that is, “a witness who is shown to have acquired special or peculiar knowledge through study or experience in respect of the matters on which he or she undertakes to testify” (pp. 20 and 25).  Dr. Arnold, a psychologist with a doctorate in psychology who had taught at the University of Saskatchewan, worked with persons subjected to sexual abuse and then went into private practice.  He was formally qualified by the trial judge “as an expert in . . . the assessment, testing and treatment of victims of sexual abuse”.  Mr. Stewart, a family consultant and therapist with a masters degree in psychology, who was experienced in working with sexually abused children, was formally qualified by the trial judge “as a clinical therapist qualified to provide expert testimony in relation to the assessment . . . and treatment of adult and child victims of sexual abuse and assault with expertise in providing therapy and conducting personality assessments in general and in particular in relation to First Nations individuals”.  I agree with the Court of Appeal that “[t]hese are such open-ended designations as to invite all manner of opinion, including opinion that may be unreliable and that transcends the witnesses’ field of expertise” (para. 255).  Moreover, I agree that Dr. Arnold and Mr. Stewart in fact transcended their respective fields of expertise when they testified as to the etiology of alcoholism and the cause of H.L.’s alcoholism in particular. Opinion evidence on these issues ought to have come from a witness who has acquired special knowledge relating to alcoholism.

316                            Since Dr. Arnold and Mr. Stewart were not properly qualified to express opinions on the etiology of alcoholism and the cause of H.L.’s alcoholism in particular, their evidence in this regard is entitled to no weight.  Accordingly, because the trial judge based his conclusion that Starr’s sexual abuse caused H.L.’s alcoholism primarily on this general expert evidence, it lacks a sufficient evidentiary foundation and was properly overturned by the Court of Appeal.  Since there is not a sufficient evidentiary basis for the inference that Starr’s sexual abuse caused H.L.’s alcoholism, there is no need to consider whether H.L.’s alcoholism caused him to lose employment income.

317                            I do note that, in addition to Dr. Arnold and Mr. Stewart, H.L. himself testified as to the effect of Starr’s sexual abuse on his alcoholism.  Nevertheless, I maintain that since the etiology of alcoholism is not understood by the average person, expert opinion evidence was necessary to enable the trial judge to appreciate this issue.  Therefore, H.L.’s testimony in this regard could not, on its own, provide a sufficient evidentiary basis for the trial judge’s inference that Starr’s sexual abuse caused H.L.’s alcoholism.

318                            Before examining the trial judge’s factual inference that H.L.’s emotional problems caused his past loss of earnings, it is necessary to make two further points. First, I echo the Court of Appeal’s call for rigour and discipline at the qualification stage, so that an expert witness can be formally qualified, if so qualified at all, relative to the opinion or opinions that he or she is expected to express.  As this Court stated in R. v. Marquard, [1993] 4 S.C.R. 223, at p. 243, “[t]he proper practice is for counsel presenting an expert witness to qualify the expert in all the areas in which the expert is to give opinion evidence.”  The need to carefully qualify expert witnesses before they testify was also aptly noted by the Saskatchewan Court of Appeal in Parker v. Saskatchewan Hospital Assn., [2001] 7 W.W.R. 230, 2001 SKCA 60, at para. 112, as follows:

A rigorous approach at this stage can avoid difficulty, especially the difficulty posed by the potential reception of opinion evidence that transcends the scope of expertise of the witness.  Strictly speaking such evidence is not admissible, and its admission can be troublesome.

319                            Second, I do note that it is for opposing counsel to object if an expert witness goes beyond the proper limits of his or her expertise, and, in this case, not only did the Attorney General of Canada fail to object to the expert testimony on the etiology of alcoholism and the cause of H.L.’s alcoholism, but his own witness, Dr. Arnold, was one of the experts who testified with regard to these issues.  Nevertheless, in Marquard, this Court stated that

[i]n the absence of objection, a technical failure to qualify a witness who clearly has expertise in the area will not mean that the witness’s evidence should be struck.  However, if the witness is not shown to have possessed expertise to testify in the area, his or her evidence must be disregarded and the jury so instructed. [p. 244]

320                            Unlike the situation in Marquard, in this case, the expert witnesses did not possess expertise sufficient to permit them to testify in relation to an area in which they were not formally qualified to give expert opinion.  Therefore, the fact that counsel for the Attorney General of Canada failed to object to Mr. Stewart’s testimony in relation to the etiology of alcoholism and the cause of H.L.’s alcoholism and failed to steer his own witness, Dr. Arnold, away from this area does not change the fact that this evidence is entitled to no weight, since it was beyond the experts’ expertise, both in fact and as it was formally recognized by the trial judge.

(ii)   The Inference That H.L.’s Emotional Problems Caused His Past Loss of Earnings

321                            Since there is no evidentiary basis for the inference that Starr’s sexual abuse caused H.L.’s alcoholism, the conclusions relating to H.L.’s alcoholism that make up the trial judge’s causal chain are no longer a consideration.  All that remains in support of the trial judge’s overall conclusion that H.L. suffered a loss of employment income because of Starr’s sexual abuse is the following chain of reasoning: Starr’s sexual abuse caused H.L.’s emotional problems, and these problems caused H.L. to lose employment income.

322                            At trial, among other emotional problems, Klebuc J. found that H.L. suffered from self-blame and a loss of self-worth (para. 29), and both experts testified that, in general, self-confidence and self-esteem affect work ethic and employability. In particular, Dr. Arnold testified on cross-examination that if a child is sexually abused by an individual from a school, the abuse would more likely cause the child to lose confidence in the school system:

Q:   Would — if the sexual perpetrator was from a school and just not speaking about [H.L.], but generally, if a sexual perpetrator were from a school, would it be more likely to cause the student to lose confidence in the school system, the school leaders?

A:   That would be my opinion.

Dr. Arnold also opined that such an act of abuse, followed by alcohol problems, would logically impact upon the individual’s ability to hold work:

Q:   Would you think it likely that this would have affected his work ethic?

A:   Work ethic as in — perhaps to define that, I think what you’re saying is his ability to hold work and be able to regularly show up and those kinds of things?

Q:   Yes.

A:   Yes, and I would refer to the chain of events I just referred to. You have an event, then — sorry, an event — I better be clear here — event of abuse, you have alcohol and, yes, indeed that chain of events would logically go there and —

Similarly, Mr. Stewart was asked on direct examination whether self-esteem and self-confidence affected employability, and he responded in the affirmative.

323                            In my view, the testimony of the experts in this regard is of a general nature and not probative of whether H.L., in particular, was wholly or largely unable to work because of his emotional problems.  Moreover, I agree with the Court of Appeal that the evidence adduced at trial only demonstrated that H.L. did not work during the first period and worked only sporadically during the second.  It does not prove that H.L. was wholly or largely unable to work because of his emotional problems.

324                            For instance, H.L.’s sporadic work record, in itself, is as consistent with choosing not to work as with being unable to work.  In fact, certain pieces of evidence militate in favour of the former rather than the latter.  For example, H.L. quit the only permanent job he ever had, working on a poultry farm outside Regina, in order to move to his then spouse’s reserve, where she could pursue her career as a home care worker. H.L. was also enrolled in an auto mechanics course that was being held on the Muskowekwan Reserve where he was then living; however, he quit the course because he was not being paid to attend.

325                            Therefore, because the evidence adduced at trial does not prove that H.L. was wholly or largely unable to work because of his emotional problems, the trial judge’s inference that H.L.’s emotional problems caused him to lose employment income lacks a sufficient evidentiary foundation and was properly overturned by the Court of Appeal.  Since there is not a sufficient evidentiary basis for the inference that H.L.’s emotional problems caused him to lose employment income, it is not necessary to comment upon whether Starr’s sexual abuse caused H.L.’s emotional problems.  The causal chain is already broken.  Furthermore, because this was the only remaining chain of reasoning that could support the trial judge’s overall conclusion that H.L. suffered a loss of employment income because of Starr’s sexual abuse, it is clear that the trial judge’s conclusion on this issue cannot stand and was indeed correctly overturned by the Court of Appeal.

326                            In sum, the Court of Appeal did not err when it set aside the trial judge’s award of pecuniary damages for past loss of earnings.  As discussed above, the (so-called) standard of review in Saskatchewan for factual inferences is reasonableness. Since the damages award was based on an unreasonable chain of reasoning (i.e., one that was not supported by the evidence), the Court of Appeal was correct to intervene and overrule it.

(b)  Liability for Future Loss of Earnings

327                            Turning now to the Government of Canada’s liability for future loss of earnings, it must be noted that the trial judge acknowledged that the parties presented no evidence regarding H.L.’s future earning capacity and that this necessitated a determination based solely on the data applied in assessing his loss of past earning capacity (para. 70).  The trial judge ultimately calculated the discounted present lump sum required to fund H.L.’s loss of future earning capacity to be $179,190.

328                            The Court of Appeal set aside this damages award, because it was of the view that the evidence fell short of proving the loss, and that the award suffered from the same errors as the award for past loss of earnings (para. 258).

329                            I agree that the award for future loss of earnings must be set aside, because, like the award for past loss of earnings, it lacks a sufficient evidentiary foundation.  In other words, since it was not reasonable for the trial judge to conclude that H.L. suffered a loss of employment income because of Starr’s sexual abuse, given the evidentiary gaps in the trial judge’s causal chain, it was likewise not reasonable for him to conclude that H.L. will continue to suffer such a loss in the future.

(2)  Trial Judge’s Assessment of Damages for Past Loss of Earnings

330                            Given my conclusion that the pecuniary damages award for past and future loss of earnings lacks an evidentiary foundation, it is not necessary to review the trial judge’s assessment of damages.  However, in order to provide further guidance in this area of the law, I will briefly comment upon his assessment, especially because I agree with the Court of Appeal that the trial judge erred in four respects.  First, he failed to consider the plaintiff’s duty to mitigate.  Second, he unreasonably concluded that the plaintiff did not have a “crumbling skull” and therefore attributed too much to Starr’s wrongful acts in his assessment of pecuniary damages.  Third, he did not reduce the damages award to reflect the time H.L. was incarcerated.  Fourth, he erred in not accounting for the social assistance payments H.L. received during the relevant period.

(a)  Duty to Mitigate

331                            I agree with the Court of Appeal that the trial judge erred in failing to consider the plaintiff’s duty to mitigate.  As noted by the court, the Attorney General of Canada, who bore the onus of proof in respect of this issue, adduced evidence that the plaintiff had been offered opportunities to upgrade his education and training and to enter into rehabilitative treatment in relation to his excessive consumption of alcohol; yet, he declined or failed to pursue these offers (para. 232).  Therefore, to the extent that the trial judge associated the plaintiff’s loss of earnings with his lack of education or training or alcohol problems, I agree with the Court of Appeal that the trial judge ought to have turned his mind to the plaintiff’s duty to mitigate, and he erred in not taking up this issue.

332                            I must also respectfully take issue with Fish J.’s reasons on this issue.  At para. 134 of his reasons, Fish J. states that “the trial judge concluded that the Crown led no evidence on the issue of mitigation”; however, from my reading of the trial judge’s reasons for judgment, it is clear that, rather than come to any such conclusion, the trial judge simply did not consider this issue at all and therein lies the error.

333                            Furthermore, at para. 135, Fish J. notes that H.L. testified that he failed to upgrade his education because he had a poor memory.  Instead of pointing to a failure to mitigate, in my colleague’s view, this testimony is consistent with the trial judge’s finding that H.L.’s alcoholism, poor self-image and lack of confidence affected his ability to learn a trade and his ability to find and keep a job.  I have already discussed the evidentiary gaps in the trial judge’s reasoning in this respect.

334                            Fish J. is also of the view that while “the record is essentially silent regarding H.L.’s efforts at rehabilitation, it appears from his evidence at trial that he was at least then making an effort to abstain from any further consumption of alcohol” (para. 135).  With respect I do not agree that the record is essentially silent regarding H.L.’s efforts at rehabilitation.  For example, at trial, the Attorney General of Canada filed as an exhibit H.L.’s admission reports from the Regina Provincial Correctional Centre (“P.C.C.”).  H.L. was incarcerated at the Regina P.C.C. for a number of alcohol and theft-related offences.  These reports indicate that upon his various admissions to the P.C.C., H.L. expressed no interest in counselling, education or treatment programs relating to alcoholism.  Therefore, while H.L. may have been making an effort to abstain from any further consumption of alcohol at the time of the trial, this does not change the fact that the trial judge ought to have considered the evidence regarding H.L.’s efforts at rehabilitation in the past, especially since H.L. was claiming damages for past loss of earnings.  He erred in not doing so.

(b)  Crumbling Skull

335                            Assuming for the purposes of my analysis at this juncture that Starr’s sexual abuse was at least a cause of H.L.’s loss of earnings (although it is my contention that there is no evidentiary basis for such a conclusion), I find that the trial judge also erred in his assessment of pecuniary damages by attributing too much to Starr’s two acts of sexual abuse.  This error came as a result of the trial judge’s unreasonable inference that H.L. did not have a “crumbling skull”.

336                            The “crumbling skull” rule was described by Major J. in Athey v. Leonati, [1996] 3 S.C.R. 458, at para. 35, as follows:

The so‑called “crumbling skull” rule simply recognizes that the pre‑existing condition was inherent in the plaintiff’s “original position”. The defendant need not put the plaintiff in a position better than his or her original position.  The defendant is liable for the injuries caused, even if they are extreme, but need not compensate the plaintiff for any debilitating effects of the pre‑existing condition which the plaintiff would have experienced anyway.  The defendant is liable for the additional damage but not the pre‑existing damage . . . .  Likewise, if there is a measurable risk that the pre‑existing condition would have detrimentally affected the plaintiff in the future, regardless of the defendant’s negligence, then this can be taken into account in reducing the overall award . . . .  This is consistent with the general rule that the plaintiff must be returned to the position he would have been in, with all of its attendant risks and shortcomings, and not a better position. [Emphasis in original; citations omitted.]

The trial judge in the case at bar concluded that there was no evidence that H.L. then had a crumbling skull.  In his view, if H.L. was predisposed to emotional injury as a consequence of S.W.’s abuse or the dysfunctional family environment he grew up in, H.L. had a “thin skull”, meaning that the Government of Canada would remain responsible for the injuries triggered by Starr’s abuse (para. 28).

337                            I agree with the Attorney General of Canada that the evidence regarding H.L.’s upbringing in a home where there was alcohol abuse and family violence suggests it is quite likely that H.L.’s earning capacity would have been substantially the same even if the sexual assaults had not occurred.  Contained in his family history are a number of disturbing events that undergird this conclusion.  For instance, at trial, evidence was adduced to show that on one occasion, S.W. assaulted H.L.’s mother, brandished a rifle and threatened to kill her, which prompted H.L. and his older brother to crawl out of their bedroom window in mortal fear of being shot. They hid under the front steps while their mother fled. In my view, this startling event, combined with H.L.’s troubled home life in general, likely would have detrimentally affected H.L. in the future, regardless of Starr’s abuse.

338                            Moreover, both expert witnesses offered testimony that gives support to this conclusion.  Dr. Arnold’s opinion was that H.L.’s alcoholism was largely attributable to causes other than Starr’s sexual abuse, such as H.L.’s exposure to alcohol abuse and family violence. Similarly, on cross-examination, Mr. Stewart admitted that H.L.’s early home life and S.W.’s abuse contributed in some way to his current problems.  I agree with the Court of Appeal that since the trial judge relied on the evidence of these experts for the purpose of connecting the plaintiff’s problem with alcohol to Starr’s wrongful acts, albeit incorrectly in my view, he also should have had regard to it for the purpose of assessing the extent of H.L.’s loss of earnings that was attributable to these acts, in order to ensure that H.L. was not placed in a better position, through an award of damages, than he would have been had it not been for Starr’s sexual abuse.

339                            Therefore, in light of the evidence regarding H.L.’s upbringing in a home where there was alcohol abuse and family violence, as well as the experts’ acknowledgement that H.L.’s formative life experience contributed at least in some way to his current problems, it was unreasonable for the trial judge to infer that H.L. did not have a crumbling skull, and that this error led him to attribute too much of H.L.’s loss of earning capacity to Starr’s two acts of sexual abuse.  Instead, as recommended by the Court of Appeal, the trial judge ought to have engaged in a retrospective contingency assessment, in order to consider potential deductions to the pecuniary damages award in light of possible contributing factors other than Starr’s sexual abuse (para. 237).

(c)  Time Incarcerated

340                            At trial, Klebuc J. inferred that H.L.’s numerous convictions on alcohol and theft-related offences were attributable to Starr’s sexual abuse (para. 29).  Accordingly, in assessing H.L.’s loss of earnings, the trial judge included periods of time in which he was incarcerated.  I agree with the Court of Appeal that the trial judge erred in attributing H.L.’s criminal behaviour and its earning-capacity consequence to Starr’s wrongful acts, because there was no evidence to support this causal inference.

341                            The evidence adduced at trial regarding the relationship between sexual abuse and criminality focused on the risk that child abuse victims may go on to become abusers themselves.  For instance, during direct examination, Mr. Stewart offered the following testimony on this issue:

Q:   Would the sexual abuse have impacted upon — you mentioned substance abuse. What about criminality?

A:   It certainly could. Criminality in the sense that a number of individuals — in fact a wide number of individuals, I don’t have the exact number, who have been either physically or sexually abused in childhood, a great proportion of those end up being abusers themselves once they reach adulthood. So in that sense, yes, it’s possible.

On cross-examination, Mr. Stewart was asked to clarify his opinion regarding the relationship between sexual abuse and criminal behaviour.  In response, he stated:

The only criminal behaviour that’s really I think associated with that [sexual abuse], strongly associated, is the tendency to possibly abuse other people, other kids maybe, but not between robbing a bank and being sexually abused.  I don’t think that there is that strong of a relationship, no.

342                            In contrast, H.L.’s criminal record demonstrates that none of the crimes for which he was incarcerated related to the abuse of other people.  Instead, as noted above, they were, for the most part, alcohol and theft related.

343                            Because the expert evidence adduced at trial only described the risk that child abuse victims may go on to become abusers themselves, I find no evidence to support the trial judge’s inference that H.L.’s numerous convictions on alcohol and theft-related offences were attributable to Starr’s sexual abuse.  Therefore, the trial judge erred in not reducing the damages award to reflect the time H.L. was incarcerated.

344                            Moreover, I agree with the Court of Appeal that to award a plaintiff damages for loss of earning capacity while incarcerated would undermine the principles of our criminal justice system (paras. 240-41).  As noted by the Court of Appeal, the risk was well described by Samuels J.A. in State Rail Authority of New South Wales v. Wiegold (1991), 25 N.S.W.L.R. 500 (C.A.), at p. 514, as follows:

If the plaintiff has been convicted and sentenced for a crime, it means that the criminal law has taken him to be responsible for his actions, and has imposed an appropriate penalty.  He or she should therefore bear the consequences of the punishment, both direct and indirect.  If the law of negligence were to say, in effect, that the offender was not responsible for his actions and should be compensated by the tortfeasor, it would set the determination of the criminal court at nought.  It would generate the sort of clash between civil and criminal law that is apt to bring the law into disrepute.

(d)  Receipt of Social Assistance

345                            The trial judge noted that from January 1, 1980 to December 31, 1987, H.L. generally relied on social assistance to meet his needs; however, in calculating the award for past loss of earning capacity for this period, he did not account for H.L.’s receipt of social assistance payments (para. 64).  The Court of Appeal found that Klebuc J. erred in not addressing the issue of whether the social assistance benefits received by H.L. should be deducted from the damages award for loss of past earnings, but it refrained from commenting any more upon the matter, since a recent decision from the British Columbia Court of Appeal directly on point was currently under appeal to this Court.  That decision, M.B. v. British Columbia, [2003] 2 S.C.R. 477, 2003 SCC 53, has now been rendered. In M.B., this Court held that social assistance is a form of income replacement and is therefore deductible at common law.  In light of this decision, it is clear that the trial judge committed a mixed error of law and fact when he failed to account for the social assistance payments H.L. received during the relevant period.

V.     Disposition

346                            For all these reasons, I would dismiss the appeal.

The following are the reasons delivered by

347                        Charron J. (dissenting in part) — I have had the benefit of reading the reasons of my colleagues Justice Fish and Justice Bastarache.  I agree with the analysis of Fish J. on the governing standard of review for appeals in the province of Saskatchewan and conclude that the Court of Appeal erred in finding that the standard was other than that adopted by this Court in Housen v. Nikolaisen, [2002] 2 S.C.R. 235, 2002 SCC 33.  As Fish J., I would nonetheless conclude that the Court of Appeal was correct in setting aside the trial judge’s award of pecuniary damages for future loss of earnings.  I also agree with Fish J. that the trial judge erred in awarding pecuniary damages during the period of time that the appellant was incarcerated and that he further erred in failing to account for H.L.’s receipt of social assistance payments.   However, I respectfully depart from Fish J.’s reasoning in the following respect.

348                        I would conclude, on application of the appropriate standard of review, that the Court of Appeal was correct in setting aside the entire award for pecuniary damages. In my view, the same error informed the trial judge’s decision to award pecuniary damages in respect of both past and future loss of earnings.  In effect, the trial judge found that there was a causal connection between the acts of sexual abuse and a lifelong inability to earn income.  In my respectful view, the evidence did not support this finding and, consequently, the award for loss of income, past and future, is unreasonable. On this point, I am essentially in agreement with the reasons of my colleague Bastarache J.  Consequently, I would dismiss the appeal.

Appeal allowed in part, with costs, Bastarache, LeBel, Deschamps and Charron JJ. dissenting in part.

Solicitors for the appellant:  Merchant Law Group, Regina.

Solicitor for the respondent:  Attorney General of Canada, Toronto.

Solicitor for the intervener:  Attorney General for Saskatchewan, Regina.