R. v. B. (K.G.), [1993] 1 S.C.R. 740

Her Majesty The Queen Appellant

v.

K.G.B. Respondent

Indexed as: R. v. B. (K.G.)

File No.: 22351.

1992: October 8; 1993: February 25.

Present: Lamer C.J. and L'Heureux‑Dubé, Sopinka, Gonthier, Cory, McLachlin and Iacobucci JJ.

on appeal from the court of appeal for ontario

Criminal law ‑‑ Evidence ‑‑ Prior inconsistent statements ‑‑ Admissibility ‑‑ Witnesses' videotaped statements to police implicating accused as the murderer ‑‑ Witnesses recanting statements at trial ‑‑ Whether prior inconsistent statements admissible as evidence of the truth of their contents ‑‑ Whether common law rule as to the use of prior inconsistent statements should be changed ‑‑ Canada Evidence Act, R.S.C., 1985, c. C‑5, s. 9.

Courts ‑‑ Stare decisis ‑‑ Supreme Court of Canada ‑‑ Prior inconsistent statements admissible only to impeach witness's credibility ‑‑ Whether common law rule as to use of prior inconsistent statements should be changed ‑‑ If so, whether change to be made by Parliament rather than the courts.

The accused and three of his friends were involved in a fight with two men. In the course of the fight, one of the youths pulled a knife and stabbed one of the men in the chest and killed him. The four youths immediately fled the scene. About two weeks later, the accused's friends were interviewed separately by the police. Each was accompanied by a parent and in one case by a lawyer and each was advised of his right to counsel. It was also made clear that they were under no obligation to answer the questions and that they were not "at this time" charged with any offence. With the youths' consent the interviews were videotaped. In their statements, they told the police that the accused had made statements to them in which he acknowledged that he thought he had caused the death of the victim by the use of a knife. The accused was charged with second degree murder and tried in Youth Court. At trial, the three youths recanted their earlier statements and, during the Crown's cross‑examination pursuant to s. 9 of the Canada Evidence Act, they stated they had lied to the police to exculpate themselves from possible involvement. Although the trial judge had no doubt that the recantations were false, the witnesses' prior inconsistent statements could not be tendered as proof that the accused actually made the admissions. Under the traditional common law position, they could only be used to impeach the witnesses' credibility. In the absence of other sufficient identification evidence, the trial judge acquitted the accused and the Court of Appeal upheld the acquittal. Prior to the hearing in this Court, the three witnesses pleaded guilty to perjury as a result of their testimony at trial. In this appeal, the Crown asks this Court to reconsider the common law rule which limits the use of prior inconsistent statements to impeaching the credibility of the witness.

Held: The appeal should be allowed and a new trial ordered.

Per Lamer C.J. and Sopinka, Gonthier, McLachlin and Iacobucci JJ.: The time has come for the rule limiting the use of prior inconsistent statements to impeaching the credibility of the witness (the "orthodox rule") to be replaced by a new rule recognizing the changed means and methods of proof in modern society. The history of the orthodox rule demonstrates that it has not enjoyed consistent or unqualified support. Considering a change to the orthodox rule is not a matter better left to Parliament; the rule itself is judge‑made and lends itself to judicial reform, and it is a natural and incremental progression in the development of the law of hearsay in Canada by this Court. The guidelines which direct this Court's exercise of its jurisdiction to overrule its previous decisions adopting the orthodox rule do not suggest that it should do anything other than what it thinks best in reconsidering the orthodox rule: a reformed rule would not violate the Charter, the existing rule has been attenuated by developments in the law of hearsay and is somewhat, if not overly, technical, and reforming the rule would not directly expand the scope of criminal liability.

A reformed rule must carefully balance the accused's interests in a criminal trial with the interests of society in seeing justice done. Since the orthodox rule is an incarnation of the hearsay rule, a reformed rule must also deal with the "hearsay dangers" of admitting prior inconsistent statements for the truth of their contents ‑‑ namely, the absence of an oath or solemn affirmation when the statement was made, the inability of the trier of fact to assess the demeanour, and therefore the credibility, of the declarant when the statement was made, and the lack of contemporaneous cross‑examination by the opponent.

Following this Court's decisions in Khan and Smith, evidence of prior inconsistent statements of a witness other than an accused should be substantively admissible on a principled basis, the governing principles being the reliability of the evidence and its necessity. These criteria, however, must be adapted to the present context. As a threshold matter, the prior inconsistent statements will only be admissible if they would have been admissible as the witness's sole testimony, lest what would be excluded as the witness's primary evidence be admitted under the reformed rule simply because the witness has recanted.

The focus of the inquiry in the case of prior inconsistent statements is on the comparative reliability of the prior statement and the testimony offered at trial. Additional indicia and guarantees of reliability to those outlined in Khan and Smith must thus be secured in order to bring the prior statement to a comparable standard of reliability before such statements are admitted as substantive evidence. In this context, the absence of an oath and the inability of the trier of fact to assess the declarant's demeanour are the only "hearsay dangers" which present real concerns. The criterion of reliability will therefore be satisfied when the circumstances in which the prior statement was made provide sufficient guarantees of its trustworthiness with respect to these two dangers.

There will be sufficient circumstantial guarantees of reliability to allow the jury to make substantive use of the statement: (1) if the statement is made under oath, solemn affirmation or solemn declaration following an explicit warning to the witness as to the existence of severe criminal sanctions for the making of a false statement; (2) if the statement is videotaped in its entirety; and (3) if the opposing party, whether the Crown or the defence, has a full opportunity to cross‑examine the witness at trial respecting the statement. Alternatively, other circumstantial guarantees of reliability may suffice to render such statements substantively admissible, provided that the judge is satisfied that the circumstances provide adequate assurances of reliability in place of those which the hearsay rule traditionally requires. With the oath, solemn affirmation or solemn declaration and the warning, the first "hearsay danger" is satisfied. The witness is clearly made aware of the gravity of the situation and his duty to tell the truth. The presence of an oath, solemn affirmation or solemn declaration also increases the evidentiary value of the statement when it is admitted at trial. The trier of fact will have the opportunity to choose between two sworn statements and will not be asked to accept unsworn testimony over sworn testimony, or to render a verdict based on unsworn testimony. While it is true that the oath in itself has no power to ensure truthfulness in some witnesses, the fact that both statements were made under oath removes resort to the absence of an oath as an indicium of the alleged unreliability of the prior inconsistent statement. With a videotaped statement, the second "hearsay danger" is also satisfied. The indicia of credibility, and therefore reliability, are available to the trier of fact. Not only does the trier have access to the full range of non‑verbal indicia of credibility, but there is also a reproduction of the statement which is fully accurate, eliminating the danger of inaccurate recounting. In a very real sense, the evidence ceases to be hearsay, since the declarant is brought before the trier of fact. Finally, while a cross‑examination of the witness at trial does not satisfy entirely the absence of contemporaneous cross‑examination, given the other guarantees of trustworthiness, the third "hearsay danger" is not a sufficient reason to exclude the statement from the jury as substantive evidence. The practical difficulties in requiring contemporaneous cross‑examination tip the balance in favour of allowing cross‑examination at trial to serve as a substitute.

Unavailability is not an indispensable condition of necessity. The criterion of necessity must be given a flexible definition, capable of encompassing diverse situations. In the case of prior inconsistent statements, evidence of the same value cannot be expected from the recanting witness or other sources. Where a sufficient degree of reliability is established, the trier of fact should be allowed to weigh both statements in light of the witness's explanation for the change.

When a party gives notice that it will seek to make substantive use of a prior statement, the trial judge must on the voir dire held under s. 9 of the Canada Evidence Act satisfy himself that the indicia of reliability necessary to admit hearsay evidence of prior statements are present and genuine. If they are, he must then examine the circumstances under which the statement was obtained, to satisfy himself that the statement supported by the indicia of reliability was made voluntarily if to a person in authority, and that there are no other factors which would tend to bring the administration of justice into disrepute if the statement was admitted as substantive evidence. In most cases, as in this case, the party seeking to admit the prior inconsistent statement as substantive evidence will have to establish that these requirements have been satisfied on a balance of probabilities. The trial judge is not to decide whether the prior inconsistent statement is true, or more reliable than the present testimony, as that is a matter for the trier of fact. Once this process is complete, and all of its constituent elements satisfied, the trial judge need not issue the standard limiting instruction to the jury, but may instead tell the jury that they may take the statement as substantive evidence of its content, or, if he is sitting alone, make substantive use of the statement, giving the evidence the appropriate weight after taking into account all of the circumstances. In either case, the trial judge must direct the trier of fact to consider carefully these circumstances in assessing the credibility of the prior inconsistent statement relative to the witness's testimony at trial. Where the prior statement does not have the necessary circumstantial guarantees of reliability, and so cannot pass the threshold test on the voir dire, but the party tendering the prior statement otherwise satisfies the requirements of s. 9(1) or (2) of the Canada Evidence Act, the statement may still be tendered into evidence, but the trial judge must instruct the jury in the terms of the orthodox rule.

In this case, while the statements of the recanting witnesses were videotaped, and the accused's counsel had a full opportunity to cross‑examine the witnesses at trial, the statements were not made under oath, solemn affirmation or by a solemn declaration. Considering himself bound by the orthodox rule, the trial judge refused to consider the admissibility of the statements. Given his expressed belief that the witnesses were lying at trial, it is possible that he might have found sufficient indicia of reliability to admit the statements as substantive evidence. A new trial should be ordered at which the reformed rule relating to prior inconsistent statements will, if necessary, be applied by the trial judge, who will decide whether sufficient indicia of reliability and necessity are present in this case, and instruct the trier of fact to afford the prior statements the appropriate weight in reaching a verdict.

Per L'Heureux‑Dubé and Cory JJ.: While the rule against the substantive use of prior statements should be changed, the administering of an oath or solemn affirmation should not be regarded as an essential safeguard for ensuring the veracity of a statement. Neither the taking of the oath nor the possibility of imprisonment arising from perjury charges resulting from testimony at trial can in themselves ensure that a witness will tell the truth. It is the reliability that can be placed upon the statement that should determine its admissibility. That reliability will depend on a number of factors, among others, whether the witness giving the statement is subject to criminal prosecution for making a deliberately false statement. The less stringent requirements for prosecution for offences like obstructing justice, fabricating evidence or public mischief, which, unlike perjury, do not require corroboration, and the more frequent prosecutions for these offences would have a better deterrent effect upon those who might be prone to make false statements. A requirement that the police administer an oath to the witness cannot further deter the witness from lying and seems to be superfluous. Liability for these offences does not rest in any way on the oath and a witness cannot be found guilty of perjury arising from a sworn declaration made in the course of a criminal investigation. The absence of the oath should thus not presumptively stand in the way of the admissibility for substantive purposes of a prior inconsistent statement. Although it would be preferable to give a warning to a witness of the possibility of criminal proceedings if that person gives a false statement, it may not be essential to give such a warning in order to render the statement admissible. Canadians are well aware and indeed expect that witnesses who make deliberately false statements to the police would be subject to criminal prosecution.

A videotaped statement with its complete and comprehensive record of the questions posed, the answers given and the demeanour of the witness, will often serve as a complete answer to the issues of reliability and voluntariness of the statement. It is not essential, however, that a statement be videotaped in order to be admissible. Where a complete and comprehensive record of the statement is preserved together with satisfactory evidence of the circumstances of the interview and the demeanour of the witness all the requirements on this count will be met. If the prior statement, while not videotaped, meets all the criteria of reliability it should be admissible.

While at the time the statement was made the witness was not subjected to the rigours of cross‑examination, if the prior statement is ruled admissible then the witness will be subject to cross‑examination at trial where the trier of fact will be able to study the witness's demeanour throughout his testimony and to assess what weight, if any, should be attached to all the evidence of the witness including the prior statement. The opposing party, whether the Crown or the defence, will also be able to explore the witness's reasons for the court room recantation and the veracity of his testimony.

A prior inconsistent statement should be admitted for all purposes if upon a voir dire the trial judge is satisfied beyond a reasonable doubt that the following conditions are met: (1) the evidence contained in the prior statement is such that it would be admissible if given in court; (2) the statement has been made voluntarily by the witness and is not the result of any undue pressure, threats or inducements; (3) the statement was made in circumstances, which viewed objectively would bring home to the witness the importance of telling the truth; (4) that the statement is reliable in that it has been fully and accurately transcribed or recorded; and (5) the statement was made in circumstances that the witness would be liable to criminal prosecution for giving a deliberately false statement.

If at the conclusion of the voir dire the prior inconsistent statement is ruled admissible for all purposes then, at some time, the trial judge should advise the jury that, although the statement has been ruled admissible, it is up to them to decide what weight, if any, they should attach to it. In assessing the statement, the jury should take into account all the circumstances in which it was made and should be instructed that they may consider that the statement should be given less weight because it was not subject to cross‑examination at the time it was made and because there was not the same opportunity to assess the demeanour of the witness as there would have been had the statement been made in court.

In this case, an analysis of the conditions of admissibility indicates that it would be open to a judge conducting a voir dire at a new trial to find that the prior inconsistent statements met all the conditions for admissibility.

The Khan and Smith approach to hearsay evidence provides an alternative justification for changing the interpretation of s. 9 of the Canada Evidence Act as long as the threshold tests for reliability set out above are made a part of this approach.

Cases Cited

By Lamer C.J.

Applied: R. v. Smith, 1992 79 (SCC), [1992] 2 S.C.R. 915; R. v. Khan, 1990 77 (SCC), [1990] 2 S.C.R. 531; not followed: Deacon v. The King, 1947 38 (SCC), [1947] S.C.R. 531; considered: McInroy v. The Queen, 1978 175 (SCC), [1979] 1 S.C.R. 588, aff'g 1977 1937 (BC CA), [1977] 4 W.W.R. 734; R. v. Duckworth (1916), 1916 492 (ON CA), 26 C.C.C. 314; R. v. Bernard, 1988 22 (SCC), [1988] 2 S.C.R. 833; Watkins v. Olafson, 1989 36 (SCC), [1989] 2 S.C.R. 750; R. v. Salituro, 1991 17 (SCC), [1991] 3 S.C.R. 654; referred to: Attorney General v. Hitchcock (1847), 16 L.J. Ex. 259; R. v. Mannion, 1986 31 (SCC), [1986] 2 S.C.R. 272; R. v. Corbett, 1988 80 (SCC), [1988] 1 S.C.R. 670; R. v. Kuldip, 1990 64 (SCC), [1990] 3 S.C.R. 618; R. v. Stinchcombe, 1991 45 (SCC), [1991] 3 S.C.R. 326; Dubois v. The Queen, 1985 10 (SCC), [1985] 2 S.C.R. 350; State v. Saporen, 285 N.W. 898 (1939); California v. Green, 399 U.S. 149 (1970); Myers v. Director of Public Prosecutions, [1965] A.C. 1001; R. v. Chaulk, 1990 34 (SCC), [1990] 3 S.C.R. 1303; R. v. Potvin, 1989 130 (SCC), [1989] 1 S.C.R. 525; R. v. Williams (1985), 1985 113 (ON CA), 50 O.R. (2d) 321 (C.A.), leave to appeal refused, [1985] 1 S.C.R. xiv; R. v. Broyles, 1991 15 (SCC), [1991] 3 S.C.R. 595; R. v. Collins, 1987 84 (SCC), [1987] 1 S.C.R. 265; Omychund v. Barker (1744), 1 Atk. 21, 26 E.R. 15; Reference re Truscott, 1967 66 (SCC), [1967] S.C.R. 309; Ares v. Venner, 1970 5 (SCC), [1970] S.C.R. 608; Khan v. College of Physicians and Surgeons of Ontario (1992), 1992 2784 (ON CA), 9 O.R. (3d) 641; Ibrahim v. The King, [1914] A.C. 599; Prosko v. The King (1922), 1922 584 (SCC), 63 S.C.R. 226; Horvath v. The Queen, 1979 16 (SCC), [1979] 2 S.C.R. 376; Rothman v. The Queen, 1981 23 (SCC), [1981] 1 S.C.R. 640; Piché v. The Queen, 1970 182 (SCC), [1971] S.C.R. 23; Wright v. Beckett (1833), 1 M. & Rob. 414, 174 E.R. 143; Paramore v. State, 229 So.2d 855 (1969).

By Cory J.

Not followed: Deacon v. The King, 1947 38 (SCC), [1947] S.C.R. 531; considered: R. v. Khan, 1990 77 (SCC), [1990] 2 S.C.R. 531; R. v. Smith, 1992 79 (SCC), [1992] 2 S.C.R. 915; R. v. Boisjoly, 1971 153 (SCC), [1972] S.C.R. 42; referred to: R. v. Moran (1987), 1987 124 (ON CA), 36 C.C.C. (3d) 225; R. v. L.S.L. (1991), 1991 7820 (SK QB), 89 Sask. R. 267; R. v. Gravelle (1952), 1952 345 (ON CJ), 103 C.C.C. 250; R. v. MacGillivray (1971), 1971 1886 (PE SCTD), 3 Nfld. & P.E.I.R. 227; R. v. Feger (1989), 36 Q.A.C. 26; R. v. J.(J.) (1988), 1988 7099 (ON CA), 65 C.R. (3d) 371 (Ont. C.A.), leave to appeal refused, [1989] 1 S.C.R. ix; R. v. Stapleton (1982), 1982 3331 (ON CA), 66 C.C.C. (2d) 231; R. v. Howard (1972), 1971 370 (ON CA), 7 C.C.C. (2d) 211; Lessard v. La Reine, [1965] Que. Q.B. 631; R. v. Sevick (1930), 1930 478 (NS SC), 54 C.C.C. 92; R. v. Edwards (1986), 1986 2885 (SK QB), 47 Sask. R. 303; R. v. Verma (1980), 28 A.R. 233; R. v. Lindstrom (1977), 1977 3211 (NS SC), 33 N.S.R. (2d) 369; R. v. Martin (1969), 12 Crim. L.Q. 201; R. v. Snider (1953), 1953 797 (ON CJ), 17 C.R. 136; California v. Green, 399 U.S. 149 (1970); Di Carlo v. United States, 6 F.2d 364 (1925); Gibbons v. State, 286 S.E.2d 717 (1982); Vézeau v. The Queen, 1976 7 (SCC), [1977] 2 S.C.R. 277; R. v. Oakes, 1986 46 (SCC), [1986] 1 S.C.R. 103; R. v. Mack, 1988 24 (SCC), [1988] 2 S.C.R. 903; R. v. Pickett (1975), 1975 1428 (ON CA), 28 C.C.C. (2d) 297; Coulombe v. La Reine, [1976] C.A. 327.

Statutes and Regulations Cited

Act for the suppression of Voluntary and Extrajudicial Oaths, S.C. 1874, c. 37.

Act respecting Procedure in Criminal Cases, and other matters relating to Criminal Law, S.C. 1869, c. 29, s. 68.

Canada Evidence Act, R.S.C., 1985, c. C‑5, s. 9.

Civil Evidence Act 1968 (U.K.), 1968, c. 64, s. 3(1)(a).

Common Law Procedure Act, 1854 (U.K.), 17 & 18 Vict., c. 125, s. 22.

Constitution of the United States, Sixth Amendment.

Criminal Code, R.S.C., 1985, c. C‑46, ss. 131 [rep. & sub. c. 27 (1st Supp.), s. 17], 134 [idem], 137, 139(2), 140(1) [idem, s. 19].

Criminal Law Amendment Act, 1975, S.C. 1974-75-76, c. 93, s. 6.

Criminal Law Amendment Act, 1985, S.C. 1985, c. 19, s. 17.

Federal Rules of Evidence, 28 U.S.C. app., Rule 801(d)(1)(A).

Young Offenders Act, R.S.C., 1985, c. Y‑1, s. 56(2)(c), (d).

Authors Cited

Canada. Federal/Provincial Task Force on Uniform Rules of Evidence. Report of the Federal/Provincial Task Force on the Uniform Rules of Evidence. Toronto: Carswell, 1982.

Canada. Law Reform Commission. Report on Evidence. Ottawa: The Commission, 1975.

Delisle, Ronald Joseph. "Cross‑examination of Own Witness on Previous Inconsistent Statement ‑‑ s. 9(2) Canada Evidence Act" (1978‑79), 21 Crim. L.Q. 162.

Delisle, Ronald Joseph. Evidence: Principles and Problems, 2nd ed. Toronto: Carswell, 1989.

Dugdale, D. F. "Against oath‑taking", [1985] N.Z.L.J. 404.

Graham, Michael H. "Employing Inconsistent Statements for Impeachment and as Substantive Evidence: A Critical Review and Proposed Amendments of Federal Rules of Evidence 801(d)(1)(A), 613, and 607" (1977), 75 Mich. L. Rev. 1565.

Grant, Alan. "Videotaping Police Questioning: a Canadian Experiment", [1987] Crim. L.R. 375.

Heaton‑Armstrong, Anthony, and David Wolchover. "Recording Witness Statements", [1992] Crim. L.R. 160.

Laskin, Bora. "The Role and Functions of Final Appellate Courts: The Supreme Court of Canada" (1975), 53 Can. Bar Rev. 469.

McCormick, Charles Tilford. McCormick on Evidence, vol. 2, 4th ed. By John William Strong, General Editor. St. Paul, Minn.: West Publishing Co., 1992.

Miller, Joyce. "The Audio‑Visual Taping of Police Interviews With Suspects and Accused Persons by Halton Regional Police Force: An Evaluation". Ottawa: Law Reform Commission, 1988.

Morden, John Wilson. "Evidence ‑‑ Proof of Own Witness's Prior Inconsistent Statement Where "Adverse" ‑‑ Section 24, Evidence Act (Ont.)" (1962), 40 Can. Bar Rev. 96.

Morgan, Edmund M. "Hearsay Dangers and the Application of the Hearsay Concept" (1948), 62 Harv. L. Rev. 177.

Ontario. Law Reform Commission. Report on the Law of Evidence. Toronto: The Commission, 1976.

Schiff, Stanley. "The Previous Inconsistent Statement of Opponent's Witness" (1986), 36 U.T.L.J. 440.

Stuesser, Lee. "Admitting Prior Inconsistent Statements For Their Truth" (1992), 71 Can. Bar Rev. 48.

United States. Senate. Senate Report No. 93‑1277, 93rd Cong., 2d Sess. (1974), reprinted in [1974] U.S. Cong. & Adm. News 7051.

Wigmore, John Henry. Evidence in Trials at Common Law, vol. 3A. Revised by James H. Chadbourn. Boston: Little, Brown & Co., 1970.

Wigmore, John Henry. Evidence in Trials at Common Law, vol. 5. Revised by James H. Chadbourn. Boston: Little, Brown & Co., 1974.

APPEAL from a judgment of the Ontario Court of Appeal (1991), 1991 13906 (ON CA), 49 O.A.C. 30, dismissing the Crown's appeal from the accused's acquittal on a charge of second degree murder. Appeal allowed and new trial ordered.

S. Casey Hill and Scott C. Hutchison, for the appellant.

Keith E. Wright and Mary E. Misener, for the respondent.

Lamer C.J.

The judgment of Lamer C.J. and Sopinka, Gonthier, McLachlin and Iacobucci JJ. was delivered by

Lamer C.J. -- The issue in this appeal is the substantive admissibility of prior inconsistent statements by a witness other than an accused. The Crown asks this Court to reconsider the common law rule which limits the use of such statements to impeaching the credibility of the witness. In my opinion, the time has come for the orthodox rule to be replaced by a new rule recognizing the changed means and methods of proof in modern society.

I - The Facts

On April 24, 1988, Joseph Wright and his brother Steven got off a bus at an intersection in Scarborough, Ontario. The brothers crossed the street and began walking home. At about the same time, the respondent and three other young men were driving past the same intersection. An argument started between the group in the car and the two men on the street and shortly thereafter a fight occurred. The brothers were unarmed. In the course of the fight one of the four persons from the vehicle pulled a knife, slashing twice at Joseph's face and then stabbing him in the chest. The stab wound to the chest penetrated Joseph's heart and killed him. The four young men then fled the scene.

About two weeks later, the three young men involved with the respondent in the incident were interviewed separately by the police. While the appellant states that the three witnesses approached the police to make their statements, the respondent notes that two of the witnesses testified that they approached the police only after the police came to their homes in connection with the police investigation of the killing, and the third witness testified that it was his mother's idea that he give a statement to the police. Each was accompanied by a parent and in one case by a lawyer and each was advised of his right to counsel. It was also made clear that they were under no obligation to answer questions put to them by the police, and while the police told the witnesses that they were not charged with any offence, the interviewers also added the qualification "at this time" in two of the interviews. With the youths' consent the interviews were videotaped.

In their statements, the three young men told the police that the respondent had made statements to them in which he acknowledged that he thought he had, or had, caused the death of the deceased by the use of a knife. The respondent was charged with second degree murder and he entered a plea of not guilty. Following an unsuccessful attempt by the Crown to have the case transferred to adult court, the respondent's trial commenced before Judge MacDonnell in Youth Court on November 14, 1989.

When called at trial by the Crown, the three young men refused to adopt their earlier statements respecting the admissions made by the respondent. The trial judge allowed the Crown to cross-examine them on their prior statements pursuant to s. 9 of the Canada Evidence Act, R.S.C., 1985, c. C-5. They admitted they had made the statements to the police but said that they had lied to the police and that the respondent had not in fact made the incriminating statements that they had previously attributed to him. Their explanation for having lied to the police was that they did so to exculpate themselves from possible involvement. They claimed to have either forgotten what occurred when the respondent was alleged to have made his inculpatory statements, or to have not heard the respondent.

The trial judge held that the only use that could be made of the prior inconsistent statements of the three witnesses was with respect to their credibility, and that the prior inconsistent statements could not be used as evidence of the truth of the matters stated therein; that is, they could not be tendered as proof that the respondent actually made the admissions. The only other evidence of the identity of the assailant was identification evidence provided by the victim's brother, who identified the accused at trial (in a "dock" identification with little evidential value) and testified as to the appearance of the deceased's assailant. The trial judge found that the dock identification was "naked opinion given 19 months after the event", and that the brother had only a poor opportunity to observe his brother's attacker. Doubts also existed in connection with his latter evidence because of several inconsistencies between elements of his description of the attacker shortly after the incident and elements of his description at trial, including: the type of jacket and pants worn by the assailant, the clothing of the other youths, and the height and weight of the assailant relative to the other youths. At trial, the respondent argues, several other elements of the brother's description did not match the respondent's characteristics, including: the colour of the respondent's skin relative to the other youths, the fact that the assailant held the knife in his right hand while the respondent was described as left-handed by two of the recanting witnesses, whether the assailant wore jewellery, and the colour of the assailant's hair.

As a result of the doubt which existed with respect to the issue of identification, and in the absence of other admissible evidence, the trial judge acquitted the respondent.

The appellant appealed the acquittal to the Court of Appeal for Ontario, which stated that it was bound by the decisions of this Court, and accordingly dismissed the appeal.

Between the filing of the parties' facta and the hearing of this appeal, the three witnesses pleaded guilty to perjury. The Crown brought a motion on the day of oral argument in this appeal to adduce this new evidence which it asserted was relevant to two issues: (i) whether the operation of a revised rule would tend to enhance or detract from the public's perception of and respect for the administration of justice, and (ii) whether the application of the orthodox rule had caused a miscarriage of justice in this case. Counsel for the respondent consented to the admission of this fresh evidence on this basis and for this purpose alone, and not as evidence of the respondent's innocence or guilt. Counsel for the Crown agreed to this limitation.

II - Judgments Below

Youth Court

Judge MacDonnell first assessed the evidence with respect to the events which led up to Joseph Wright's death and found that the victim's brother essentially told the truth regarding the sequence of events preceding and during the fight. He rejected the evidence of the three young men who were present in the car with the respondent. Judge MacDonnell held:

... those three witnesses were obviously and deliberately untruthful in their sworn evidence before me with respect to material matters. Their account of how this fight occurred is not only at odds with the evidence of Steven Wright, Sean Dowling and in my opinion, Ruth Kazan [two independent witnesses], it is at odds with common sense.

Judge MacDonnell next considered whether the Crown had proved that the person who used the knife was the respondent. In this regard, the judge found that the identification evidence of the victim's brother was weak and could not, standing by itself, establish beyond a reasonable doubt that the respondent was the person with the knife. Judge MacDonnell was also of the opinion that there was no circumstantial evidence supporting this identification evidence. With respect to the respondent's friends' statements to the police that the respondent admitted to using the knife in the fight with Joseph Wright, Judge MacDonnell held that he was precluded from considering these statements as evidence that the respondent made those admissions. As to their testimony given at trial, the judge was of the view that the three witnesses lied under oath. He stated that he disbelieved their testimony "based on my assessment of the entirety of the content of their testimony and, very importantly, the manner in which it was given. With respect to the latter point, in my opinion they were anything but forthright." He concluded on this point:

In my opinion, each of these three witnesses lied to me with respect to having lied to the police about what the accused said to them. I have no doubt that their recantations are false. That is, I have no doubt that on this point they were telling the police the truth as they knew it about what the accused said. That finding is not necessarily the same as a finding that the accused made the admissions, but it is tantamount to that finding.

Judge MacDonnell reluctantly followed, however, the traditional common law position that the only use that could be made of the prior inconsistent statements of a witness other than an accused which the witness does not adopt was to impeach his credibility. The judge refused to apply the reasoning of Estey J., concurring in the result in McInroy v. The Queen, 1978 175 (SCC), [1979] 1 S.C.R. 588, which would expand significantly the use to be made of unadopted prior inconsistent statements.

Court of Appeal (1991), 1991 13906 (ON CA), 49 O.A.C. 30

The Ontario Court of Appeal refused to reconsider the present validity of the common law rule as to the use of prior inconsistent statements, rejecting the appellant's submission that in the particular circumstances of this case the premises upon which the rule was founded were not applicable. The court held (at p. 32):

That prior inconsistent statements of a witness may only be used in assessing the credibility of a witness, and may not be used as evidence of the truth of the matters stated therein, is a principle of criminal law in Canada of long standing which was recognized in Deacon v. The King [1947 38 (SCC), [1947] S.C.R. 531]. In R. v. Mannion [1986 31 (SCC), [1986] 2 S.C.R. 272], the rule was restated by the Supreme Court of Canada. In that case, the court was invited to review the rule so as to make admissible as evidence of their contents such past contradictory statements by nonparty witnesses where it is shown that they had been made under oath and subject to cross-examination, but the court found it unnecessary to deal with that issue. In Corbett v. The Queen [1988 80 (SCC), [1988] 1 S.C.R. 670], the rule was again restated, and more recently in R. v. Kuldip [1990 64 (SCC), [1990] 3 S.C.R. 618], a judgment of the Supreme Court of Canada, released December 7, 1990, the principle appears to have been once again reaffirmed.

The court concluded that it was bound by the judgments of the Supreme Court of Canada and that it was inappropriate for it to review the traditional rule. Therefore, it dismissed the appeal.

III - Grounds for Appeal

The Crown now appeals to this Court on the following ground: whether the Ontario Court of Appeal erred in law when it held that the learned trial judge did not err in law when he held that the prior inconsistent statements of several witnesses could not be used as evidence of the facts alleged in the statements. In urging this Court to reconsider or refrain from reconsidering the orthodox rule, the parties also addressed the issues of stare decisis and whether the change in the law proposed by the appellant should be made by Parliament rather than this Court.

IV - Analysis

A. History and Development of the Orthodox Rule

In his reasons in McInroy, supra, Estey J. traced the origins of the orthodox rule that prior inconsistent statements are admissible only to impeach the credibility of a witness, and not as evidence of the truth of their contents. The rule was first stated by Denman C.J. in Wright v. Beckett (1833), 1 M. & Rob. 414, 174 E.R. 143, at p. 145 E.R.:

The other danger [of permitting cross-examination of a witness found to be hostile by the party which called the witness] is, that the statement, which is admissible only to contradict the witness, may be taken as substantive proof in the cause. But this danger equally arises from the contradiction of an adverse witness: it is met by the Judge pointing out the distinction to the jury, and warning them not to be misled. It is not so abstruse but that Judges may explain it, and juries perceive its reasonableness; and it is probable that they most commonly discard entirely the evidence of him who has stated falsehoods, whether sworn or unsworn.

Estey J. noted that courts continued to apply this rule after the enactment of The Common Law Procedure Act, 1854 (U.K.), 17 & 18 Vict., c. 125, s. 22, and its criminal law parallel in S.C. 1869, c. 29, s. 68, which have survived in s. 9(1) of the Canada Evidence Act. However, Estey J. also observed that the rule was not unchallenged in the 19th century. In Attorney General v. Hitchcock (1847), 16 L.J. Ex. 259, at p. 261, Pollock C.B., Parke B. concurring, said:

If you ask a witness whether he has not made a certain statement which would be material, and opposed to part of his testimony, you may then call witnesses to prove that he has made the statement, and the jury are at liberty to believe either the one account or the other.

The debate over the merits of the orthodox rule continued into the 20th century in decisions like R. v. Duckworth (1916), 1916 492 (ON CA), 26 C.C.C. 314 (Ont. S.C., App. Div.), in which the court affirmed the orthodox rule of limited admissibility by a 4 to 1 margin. For the majority, Clute J. characterized the Crown's suggestion that evidence of prior inconsistent statements by Crown witnesses who recanted should be substantively admissible as (at p. 324) a rather "startling proposition and one to which I cannot accede except upon the clearest authority. None has been cited." Clute J. further said of s. 9 of the Canada Evidence Act, R.S.C. 1906, c. 145, that (at p. 329):

This section was passed not to admit what was not evidence, but for the purpose of contradiction. It does not provide that what was not otherwise evidence becomes evidence by the section, but simply that, before the party can be contradicted, the witness must, in the opinion of the Court, prove adverse. . . .

Riddell J. concurred, relying however on an edition of Wigmore on Evidence which predated the introduction of criticism of the orthodox rule in that work. Masten J. joined the majority, although only reluctantly since he doubted the ability of the jury member to follow the judge's limiting instructions (at p. 353):

Assume that he reaches the conclusion that the testimony now being given at the trial is false, and that the earlier statement is true. The duty of the juryman then is to disregard the evidence given by the witness at the trial, as being false testimony. But, having reached that conclusion, and having struck from his mind the testimony of the witness, he must then proceed to obliterate as well the statement given by the witness on the previous occasion, which he has just concluded a moment before to be true. Is that humanly possible?

In declining to follow the orthodox rule, Meredith C.J.C.P. (at p. 359) asserted that "[c]ases decided in the dark ages of the law of evidence, when, among other strange things, a party was not at liberty to discredit his own witness, cannot be very helpful for any purpose in these days, in this Province...." Meredith C.J.C.P. continued at p. 360:

And I hold that, if he [the trial judge] had told them [the jury] that, if from anything said by the witnesses at the trial and from their demeanour, they found that their statements on the former occasion were true, then such statements would be evidence at the trial, evidence from the witnesses' own lips, as the Judge put it, as well as from that other test of truthfulness, their demeanour.

This Court first adopted the orthodox rule in Deacon v. The King, 1947 38 (SCC), [1947] S.C.R. 531, and, unlike the court in Duckworth, was unanimous on the point. However, both Kerwin J. and Rand J. exhibited a strong awareness of the arguments against the orthodox rule. Kerwin J. (for Rinfret C.J., Taschereau and Estey JJ.) held at p. 534 that the fact that a sketch made by a witness which contradicted her present testimony, was made an exhibit,

does not take the exhibit out of the category of something merely going to the credibility of the witness and raise it to the status of something that as against the accused is to be taken as evidence of the truth of the statements contained in the writing. A contrary proposition would be entirely foreign to our criminal law.

Rand J. concurred (at pp. 537-38):

That such statements generally are limited to credibility and cannot be used as evidence of the truth of the facts to which they relate, is well established.... It is quite true that it may be difficult to dissociate the matters of such statements from the facts brought before the jury by the witness and to nullify the influence they may have on the minds of the jurors in dealing with the evidence as a whole; but anything short of this would expose a person to a fabricated account of events, too dangerous to risk. But the whole field of cross-examination, in the discretion of the court, is opened and the matters of the statement can thus be brought within the test of the testimonial response of the witness. This might be taken as a reason for leaving all the facts, including the statement, to the consideration of the jury, but the long experience of the courts is against it.

The first, and only, hint of contention respecting the orthodox rule in this Court was in the reasons of Estey J. in McInroy.

In McInroy, a Crown witness had made a statement to the police in connection with the appellant McInroy's trial for murder, claiming that the appellant had admitted killing the victim and being paid to do so. At trial, the witness claimed to not remember making the statement. The Crown applied under s. 9(2) of the Canada Evidence Act to cross-examine the witness on her prior inconsistent statement, and a voir dire was held. When confronted with her written statement, the witness persisted in her claim not to remember any of the matters contained within it. In his charge to the jury, the trial judge carefully stated the orthodox rule:

... I repeat that in light of her continued inability to remember those questions and answers, they do not form part of her evidence and accordingly are not to be taken as evidence of the truth of what is contained therein, but are only to be considered by you in testing or determining her credibility as a witness.

The British Columbia Court of Appeal, 1977 1937 (BC CA), [1977] 4 W.W.R. 734, held that the trial judge had erred in permitting Crown counsel to cross-examine the witness as an adverse witness, since her credibility was not in issue and she "had testified to nothing damaging to the Crown's case" (p. 742).

The Supreme Court of Canada allowed the Crown's appeal. Martland J. (Ritchie, Spence, Pigeon, Beetz and Pratte JJ. concurring) held that the Court of Appeal had misconstrued s. 9(2) as requiring the witness to be adverse, when in fact all the trial judge was required to do under the section was to determine whether the previous statement in writing or reduced to writing was in fact inconsistent with the witness's present testimony. Martland J. was of the opinion that there was a clear inconsistency in the witness's testimony and that the trial judge had correctly exercised his discretion. Furthermore, Martland J. noted (at p. 605) that "[t]he trial judge was careful to explain ... the limited extent to which that cross-examination might be considered by the jury."

Estey J. differed in his attitude towards the orthodox rule. He wrote (at pp. 606-7):

It is in my respectful view both an error in law and an offence against common sense to instruct the jury that the witness's prior statement, particularly when given in the circumstances of this case, may be considered by the jury only on the issue as to credibility of the witness, St. Germaine, and must be disregarded on the issues of fact arising in this statement; and, more precisely, that the jury must be told that the prior statement may not be considered by them as proof or even as some evidence relating to the matters asserted in that statement.

Estey J. noted that the only basis for excluding the relevant evidence of St. Germaine's prior statement (and the admissions of the accused it reported) was the hearsay rule, but that in the case of prior inconsistent statements (at p. 614) "the speaker of the `hearsay' is indeed in the witness box available for cross-examination and subject to the scrutiny of the trier of fact". He then canvassed authorities criticizing the orthodox rule on various bases, including the difficulty of applying the distinction between permissible and impermissible uses of evidence of the prior statement, the fact that if a witness contradicts his or her evidence in chief during cross-examination, both versions may be considered to ascertain the truth, and that the circumstances at trial mean there is no real reason to classify the statement as hearsay.

He based his ultimate rejection of the orthodox rule on the narrower basis of s. 9(2), however, which he noted was enacted in 1968, after this Court's decision in Deacon, and which gave no indication that it was limited to procedural concerns only. Estey J. concluded that the proper interpretation of s. 9(2) was that it permitted the trier of fact, once the procedural requirements of the section were met, to take into appreciation all the evidence given by the witness, both in the prior statement to the police and her testimony at trial.

Estey J.'s reliance on this narrow ground to reject the orthodox rule prompted criticism from R. J. Delisle, who noted in a comment on McInroy ((1978-79), 21 Crim. L.Q. 162, at p. 166) that the departure from the orthodox rule might be welcomed, but also observed that there were problems inherent in "the piecemeal nature of the judicial amendment" which created anomalies by the different treatment of evidence admitted under s. 9(1) and (2).

This Court did not take up Estey J.'s challenge to the orthodox rule, but instead has restated the orthodox rule in several recent cases, including R. v. Mannion, 1986 31 (SCC), [1986] 2 S.C.R. 272; R. v. Corbett, 1988 80 (SCC), [1988] 1 S.C.R. 670; R. v. Kuldip, 1990 64 (SCC), [1990] 3 S.C.R. 618; R. v. Stinchcombe, 1991 45 (SCC), [1991] 3 S.C.R. 326; and R. v. Smith, 1992 79 (SCC), [1992] 2 S.C.R. 915. However, it is also important to note that in none of these cases was the orthodox rule itself directly challenged, as it has been in this appeal. Instead, those decisions used the orthodox rule as a point of comparison, or as the model for a separate rule, as must be addressed in connection with Kuldip.

Special attention must be paid to Kuldip. In that case, the respondent challenged the use of his inconsistent testimony from a prior trial to impeach his credibility at his present trial on the ground that it violated his right under s. 13 of the Canadian Charter of Rights and Freedoms, which provides:

13. A witness who testifies in any proceedings has the right not to have any incriminating evidence so given used to incriminate that witness in any other proceedings, except in a prosecution for perjury or for the giving of contradictory evidence.

I held, for the majority, that there was no violation of s. 13, since using a prior inconsistent statement from a former proceeding during cross-examination to impeach the credibility of the accused only does not thereby incriminate the accused. If the trial judge charged the jury respecting the permissible uses of the prior testimony, there would be no incrimination of the accused by the prior testimony. My holding in Kuldip should be understood not as merely restating and applying the orthodox rule, but as establishing a Charter rule recognizing the unique circumstances of compulsion surrounding such statements. In Dubois v. The Queen, 1985 10 (SCC), [1985] 2 S.C.R. 350, I wrote for the majority (at p. 358) that

the purpose of s. 13, when the section is viewed in the context of s. 11(c) and (d) [of the Charter], is to protect individuals from being indirectly compelled to incriminate themselves, to ensure that the Crown will not be able to do indirectly that which s. 11(c) prohibits.

Indeed, prior statements made by the accused would not have been excluded by the orthodox rule because they are admissions, and substantively admissible under the applicable hearsay exception. This reinforces my characterization of the holding in Kuldip as an independent Charter rule of evidence, rather than merely an application of the orthodox rule.

It must also be remembered that Kuldip and s. 13 of the Charter refer to a very special subset of prior inconsistent statements, in which the prior statement is made by an accused in a proceeding who testifies at a future proceeding and which, if admitted for the truth of its contents, would incriminate him in the second proceeding. Furthermore, s. 13 applies only to a witness who testifies in a "proceeding"; while this Court has yet to explore the outer boundaries of this term, cases decided to date have concentrated on judicial proceedings such as trials and preliminary inquiries: see Dubois. A police interview, even where the witness makes his or her statement under oath, may not be a "proceeding" for the purposes of s. 13. As this precise issue does not arise in this appeal, I will make no further comments on this point.

In the present appeal, and the more usual course of events, there is no question of incriminating the witness, who is not the accused, by admitting the prior statement for the truth of its contents. In the vast majority of cases where prior inconsistent statements are adduced, then, s. 13 of the Charter will not be engaged. In those cases where it is the prior inconsistent statement of the accused which is offered, s. 13 will operate to restrict its use to the purpose of impeaching credibility only.

Finally, in Smith, I explicitly refrained from assessing the impact of the principled approach to hearsay exceptions on the orthodox rule, since this supplementary ground of appeal had not been pressed before the Court, and since the conclusion I had reached in respect of hearsay evidence generally made it unnecessary to decide this ground.

B. Hearsay Rationale of the Orthodox Rule

The theoretical basis of the orthodox rule has been much debated. S. Schiff, in "The Previous Inconsistent Statement of Opponent's Witness" (1986), 36 U.T.L.J. 440, at p. 451, states that:

The orthodox doctrine limiting the trier's use of the non-party witness' statement rests on the hearsay rule. While American judges and commentators have taken this as obvious, Canadian and English judges have rarely discussed the reason for the limit.

The author cites as rare examples of Canadian judges explicitly recognizing the hearsay rule as the basis of the orthodox rule the reasons of Riddell J. in Duckworth, and Estey J. in McInroy. In Evidence: Principles and Problems (2nd ed. 1989), at p. 247, R. J. Delisle also recognized the source of the orthodox rule in the law of hearsay.

However, it is clear that commentators addressing the orthodox rule are implicitly relying on the hearsay concept when they discuss the "dangers" of admitting prior inconsistent statements for their truth, such dangers being the same as the traditional "hearsay dangers": the absence of an oath or solemn affirmation when the statement was made, the inability of the trier of fact to assess the demeanour and therefore the credibility of the declarant when the statement was made (as well as the trier's inability to ensure that the witness actually said what is claimed), and the lack of contemporaneous cross-examination by the opponent. For examples of commentators who have explicitly or implicitly justified the orthodox rule on the basis of the hearsay rule, see J. W. Morden (now Morden A.C.J. of the Ontario Court of Appeal), "Evidence -- Proof of Own Witness's Prior Inconsistent Statement Where "Adverse" -- Section 24, Evidence Act (Ont.)" (1962), 40 Can. Bar Rev. 96, at p. 103; McCormick on Evidence (4th ed. 1992), vol. 2, at pp. 117-18, and L. Stuesser, "Admitting Prior Inconsistent Statements For Their Truth" (1992), 71 Can. Bar Rev. 48, at p. 53.

Another indication of the basis of the orthodox rule in the hearsay rule is the treatment it has received from law reform bodies. The Law Reform Commission of Canada's Report on Evidence (1975) addressed the orthodox rule in that part of its Report dealing with hearsay, and the Federal/Provincial Task Force on Uniform Rules of Evidence, in its Report (1982), classified prior inconsistent statements offered for their truth as hearsay (at p. 315) "out of caution".

It is important to understand the basis and rationale of the orthodox rule in order to understand how this Court's recent decisions in R. v. Khan, 1990 77 (SCC), [1990] 2 S.C.R. 531, and Smith have affected the orthodox rule, and to determine what type of principled analysis this Court should apply to the particular problem of prior inconsistent statements.

C. Erosion of Orthodox Rule

(1) Criticism of the Rule

The orthodox rule has been almost universally criticized by academic commentators. Their criticisms can be distilled into the assertion that the hearsay dangers on which the orthodox rule is based are ill-founded or non-existent in the case of prior inconsistent statements.

Respecting the oath, commentators discount the significance of the oath in modern society. Stuesser, at pp. 53-54, is representative in arguing that "[t]he unfortunate reality in our modern society is that the power of an oath must be discounted as a means of ensuring reliability for a statement." One critic less reservedly referred to the oath as "no more than a lingering relic of primordial superstition" and "primitive mumbo-jumbo" (D. F. Dugdale, "Against oath-taking", [1985] N.Z.L.J. 404). Furthermore, commentators argue that the witness is under oath at trial when he or she is asked to adopt or deny the prior statement, thus invoking the power of the oath at the trial as it relates to the truthfulness of the prior statement. However, I note that while the witness faces the legal consequences of violating an oath or solemn affirmation at trial, in most cases there is less incentive to be truthful when the statement is made, leading to a natural preference for the testimony at trial if the alternative is unsworn or unaffirmed testimony.

Critics also claim that the lack of opportunity for the trier of fact to observe the demeanour of the witness at the time the statement was made, and thus to assess credibility based on that demeanour, is overstated in its significance. They argue that the opportunity to observe the witness as he or she denies or professes not to remember making the statement can give the trier insight into the truthfulness of the recantation, and therefore also the truthfulness of the prior statement which is denied. This does not obviate the problem of ensuring that the witness's prior statement is fully and accurately reproduced for the trier of fact. Of course, both of these criticisms of the orthodox rule are reinforced when, as in this case, the prior statement is videotaped, allowing the trier of fact to observe the witnesses' demeanour and ensuring that an accurate record of the statement is tendered as evidence.

In a study for the Law Reform Commission of Canada, Halton Regional Police Officers in Burlington videotaped all suspect interviews, where the suspect consented. Interviewing officers, court officers, Crown prosecutors and defence counsel were then surveyed to evaluate the results of this program. In an interim evaluation one year into the two-year experiment (A. Grant, "Videotaping Police Questioning: a Canadian Experiment", [1987] Crim. L.R. 375), the author stated (at p. 379) that

it certainly does not appear that people are inhibited by the camera or the general situation from making confessions and inculpatory admissions on videotape to interviewing police officers. The audio-visual process appears to record a much greater range of responses than the written statement.

The author noted that both Crown prosecutors and defence counsel responded favourably to the use of videotapes, if not for the same reasons, that in most cases no contest arose as to the introduction of the tape, and that there was no evidence that costly crews or other assistance were required to create (at p. 383) "a clear and reliable record of the interview."

The final report reached the same conclusions. In "The Audio-Visual Taping of Police Interviews With Suspects and Accused Persons by Halton Regional Police Force: An Evaluation" (1988), J. Miller summarized an evaluation paper prepared by Grant. She noted (at p. 3) that the case of Paramore v. State, 229 So.2d 855 (1969), in which the Florida Supreme Court upheld the admissibility of a taped confession, was

a milestone not only for law enforcement practices, but for the administration of justice as a whole. For the first time in the history of the criminal justice system the mysterious veil was lifted and it was now possible to present before the court the entire contents of a police interrogation in a comprehensive and accurate manner.

The conclusion remained that people were not inhibited by the camera, and that (at p. 13): "The courts's ability to asses the credibility and voluntariness of a statement was seen by the defence bar as being greatly enhanced by the more complete picture provided by videotape." Furthermore, defence counsel preferred videotaped statements to police officer's notes and memories to create an accurate record of statements. Any prejudice to the accused from being seen as he or she appeared at the time of the interview was seen by defence counsel to be outweighed by the accuracy and reliability of the record provided by the videotape. Indeed, Miller observed (at p. 17) that "it is common ground among police, Crown and defence counsel that the videotape is an accurate record of the interview as it occurred in the videotaping room." Miller concluded (at p. 23):

Aside from providing a new tool to the investigative process, the Police, Crown counsel and defence lawyers viewed the introduction of this technology as an improvement of the administration of justice. An accurate videotape record of police interviews largely eliminates courtroom conflicts over what was said and how an accused was treated. The new technology therefore, helps police in gathering evidence at the same time as it adds to the protection of the rights of the accused.

The "milestone" represented by widely available videotape technology and its introduction in the trial process, then, has gone a long way towards meeting this second hearsay danger. I also believe that demeanour evidence from sources other than a videotape might, in exceptional circumstances, also serve the same purpose to answer this criticism of the orthodox rule.

The lack of cross-examination is the most important of the hearsay dangers, but perhaps also the most overstated in the context of prior inconsistent statements. By definition, commentators argue, the maker of the statement is present in court and amenable to vigorous cross-examination respecting his or her recollection, testimonial capacity and bias at the time of the making of the prior statement. As it is argued in McCormick on Evidence, supra, at p. 120:

The witness who has told one story aforetime and another today has opened the gates to all the vistas of truth which the common law practice of cross-examination and re-examination was invented to explore. The reasons for the change of face, whether forgetfulness, carelessness, pity, terror, or greed, may be explored by the two questioners in the presence of the trier of fact, under oath, casting light on which is the true story and which the false.

Wigmore, vol. 3A (Chadbourn rev. 1970), §1018, at p. 996, stated that the only ground for excluding prior inconsistent statements as substantive evidence was the hearsay rule, and continued:

But the theory of the hearsay rule is that an extrajudicial statement is rejected because it was made out of court by an absent person not subject to cross-examination.... Here, however, by hypothesis the witness is present and subject to cross-examination. There is ample opportunity to test him as to the basis of his former statement. The whole purpose of the hearsay rule has been already satisfied.

Similarly, Morden, supra, wrote of prior inconsistent statements, at p. 103, that:

Being hearsay such a statement is technically inadmissible yet, in view of the opportunity of opposing counsel to cross-examine the witness on the statement, the main reason for the hearsay rule disappears and consequently the possibility of injustice is negatived.

Furthermore, commentators observe, the witness's recantation has accomplished all that the opponent's cross-examination could hope to: the witness now testifies under oath that the prior statement was a lie, or claims to have no recollection of the matters in the statement, thus undermining its credibility as much as cross-examination could have. As Stuesser pointed out (at p. 60), in fact "the mantle of a `hostile' cross-examiner in the case of a recanting witness is taken up by the caller of the witness".

Proponents of the orthodox rule concentrate on the importance of contemporaneous cross-examination; an oft-quoted statement is that of Stone J. in the Minnesota Supreme Court in State v. Saporen, 285 N.W. 898 (1939), at p. 901:

The chief merit of cross examination is not that at some future time it gives the party opponent the right to dissect adverse testimony. Its principal virtue is in its immediate application of the testing process. Its strokes fall while the iron is hot. False testimony is apt to harden and become unyielding to the blows of truth in proportion as the witness has opportunity for reconsideration and influence by the suggestions of others, whose interest may be, and often is, to maintain falsehood rather than truth.

Equally oft-quoted, however, is E. M. Morgan's reply, in "Hearsay Dangers and the Application of the Hearsay Concept" (1948), 62 Harv. L. Rev. 177, at p. 193:

Why does falsehood harden any more quickly or unyieldingly than truth? What has become of the idea that truth is eternal and, though crushed to earth, will rise again? Isn't the opportunity for reconsideration and for baneful influence by others even more likely to color the later testimony than the prior statement?

Critics assert that there is a greater opportunity for coercion, intimidation and fabrication in the time between the taking of the statement and the trial than when the statement was made closer to the time of the events described. Proponents of the orthodox rule point to the time between the events described and when the statement was made as presenting opportunities for identical coercion and intimidation, and assert that later coercion may be directed towards convincing the witness to tell the truth at trial. It is clear that statements are neither more nor less likely to be true based solely on a consideration of when they were made.

Furthermore, to argue about which statement is more likely to be true based on the opportunity for improper influence misconceives the nature of the problem before the court. The question is not whether it would be better to have had contemporaneous cross-examination at the time the statement was made; of course it would have, and this is a factor which must affect any consideration of the weight to be afforded the evidence by the trier of fact. Instead, the real issue to be considered is whether the absence of contemporaneous cross-examination is a sufficient reason to exclude the statement from the jury as substantive evidence. In California v. Green, 399 U.S. 149 (1970), at pp. 160-61, White J. of the United States Supreme Court wrote:

It may be true that a jury would be in a better position to evaluate the truth of the prior statement if it could somehow be whisked magically back in time to witness a gruelling cross-examination of the declarant as he first gives his statement. But the question as we see it must be not whether one can somehow imagine the jury in "a better position", but whether subsequent cross-examination at the defendant's trial will still afford the trier of fact a satisfactory basis for evaluating the truth of the prior statement.

Perhaps because it has been the most emphasized of the hearsay dangers, it is clear that the lack of contemporaneous cross-examination has also been the most criticized reason for excluding prior inconsistent statements as substantive evidence.

(2) Judicial and Legislative Reform of the Rule

The orthodox rule has attracted the attention of law reformers on the bench, in the legislatures, and at law reform commissions. All bodies which have turned their minds to the orthodox rule have recommended a departure from the strict prohibition on substantive admissibility in favour of qualified admissibility.

The Law Reform Commission of Canada, in its 1975 Report on Evidence, proposed that the hearsay rule should not exclude prior inconsistent statements by a witness, on the grounds that the witness could be cross-examined on the prior statement, that circumstances may serve to make the prior statement more reliable (since his memory may have been fresher and it may have been made before anyone had the opportunity to improperly influence the witness), the witness will have the opportunity of denying the prior statement or explaining the inconsistencies, and that the distinction between permissible and impermissible use of the evidence was a difficult one. Thus, Draft Section 28 proposed by the Commission provided: "A statement previously made by a witness is not excluded by section 27 [which makes hearsay evidence inadmissible] if the statement would be admissible if made by him while testifying as a witness."

The Ontario Law Reform Commission, in its Report on the Law of Evidence (1976), took a different approach, maintaining the orthodox rule where a prior inconsistent statement was made by the calling party's witness, but allowing evidence of the prior inconsistent statements of the opposing party's witness to be admitted for their truth once proof is given that the witness in fact made the statement.

The Federal/Provincial Task Force recommended that prior inconsistent statements be substantively admissible if they were made under oath and subject to cross-examination at the time they were made.

Most jurisdictions in the United States have abandoned the orthodox rule through both judicial and legislative action. Federal Rules of Evidence Rule 801(d)(1)(A) provides:

Rule 801. ...

(d) Statements which are not hearsay. -- A statement is not hearsay if --

(1) Prior statement by witness. -- The declarant testifies at the trial or hearing and is subject to cross-examination concerning the statement, and the statement is (A) inconsistent with the declarant's testimony, and was given under oath subject to the penalty of perjury at a trial, hearing, or other proceeding, or in a deposition....

This expanded rule of admissibility is now in force in 21 states. Four states and the District of Columbia adhere to the orthodox rule, 8 states follow modified rules different from the Federal Rule, and 17 states allow prior inconsistent statements to be admitted for their truth with no restrictions. Some states require additional indicia of reliability, including: personal knowledge of the facts contained in the statement, a demonstration of the reliability of the evidence, a requirement that the statement have been made voluntarily and not as the result of coercion or merely in response to allegations by the investigator, or contemporaneous recording, transcription or videotaping of the statement. It can readily be seen that one of these additional requirements (personal knowledge) would be required in any event for non-hearsay evidence from a witness to be admissible, mirroring the requirement in the Law Reform Commission of Canada's Report that to be substantively admissible, the prior statement must be one which would have been admissible if made by the declarant while testifying as a witness.

In England and Scotland, the orthodox rule applies to criminal proceedings, although, as in Canada and the United States, there has been widespread criticism of the rule in that context. The English Civil Evidence Act 1968 (U.K.), 1968, c. 64, renders the prior inconsistent statement of a witness admissible for its truth in civil proceedings:

3. -- (1)Where in any civil proceedings --

(a)a previous inconsistent or contradictory statement made by a person called as a witness in those proceedings is proved by virtue of section 3, 4 or 5 of the Criminal Procedure Act 1865

that statement shall by virtue of this subsection be admissible as evidence of any fact stated therein of which direct oral evidence by him would be admissible.

Thus, it can be seen that the majority of law reformers, legislators and the judiciary have heeded the almost unanimous call to reform the orthodox rule. In all reformed rules, the hearsay dangers are addressed in one combination or another, supporting the characterization of the orthodox rule as an incarnation of the hearsay rule, and presenting this Court with a variety of alternatives from which to choose. The variety of approaches chosen after careful consideration of the issue by these various bodies also indicates that we must carefully balance the interests of the accused in a criminal trial with the interests of society in seeing justice done when deciding what guarantees of reliability will suffice to render prior inconsistent statements admissible for the truth of their contents. Any reformed rule must address the hearsay dangers, yet not be so restrictive towards that end so as to be of limited utility in the majority of cases.

(3) Developments in the Law of Hearsay

Finally, it is clear that the orthodox rule, in so far as it is based on the hearsay rule, has been undermined by the decisions of this Court in Khan and Smith. In Smith, I stated that the decision in Khan "should be understood as the triumph of a principled analysis over a set of ossified judicially created categories" (p. 930), and that that decision (at p. 933)

signalled an end to the old categorical approach to the admission of hearsay evidence. Hearsay evidence is now admissible on a principled basis, the governing principles being the reliability of the evidence, and its necessity.

I will return to Smith and the principled approach to the hearsay rule as it applies in the particular case of prior inconsistent statements, but it is important to note that any erosion of the categorical approach to the hearsay rule must influence the Court's consideration of the orthodox rule as one instance of that rule.

D. Stare Decisis

Before considering a new rule pertaining to the admissibility of prior inconsistent statements for the truth of their contents, it is necessary to address the arguments of the respondent, strenuously advanced before this Court, that (i) this Court should not overrule Deacon, and (ii) that any change to the orthodox rule is for Parliament to make and not this Court. The arguments are related, and to some extent undermine one another, since the first identifies the source of the orthodox rule as it applies in Canada in the common law of evidence as developed by this Court, even while the second denies that the common law is the appropriate vehicle to change the orthodox rule. I will deal with the second argument first, since the issue of overruling Deacon does not arise if this Court decides to wait for Parliament to reform the orthodox rule.

The rules of evidence are primarily judge-made, and the Law Reform Commission of Canada called the hearsay rule "the most characteristic ... rule of our system of evidence" (p. 69). In Smith, I affirmed the statement of Lord Donovan, dissenting in Myers v. Director of Public Prosecutions, [1965] A.C. 1001 (H.L.), at p. 1047, that "[t]he common law is moulded by the judges and it is still their province to adapt it from time to time so as to make it serve the interests of those it binds" (p. 931). Significantly, Myers was a decision in which a majority of the House of Lords refused to create a new hearsay exception relating to business records, leading to a legislative response overcoming the decision.

This duty of the courts to review common law rules has a pragmatic basis: courts are best situated to assess the operation and possible deficiencies of common law rules in practical situations. Finally, if any other authority for the jurisdiction and, in some cases, the responsibility of the courts to reform the common law were needed, the late Chief Justice Laskin wrote, in "The Role and Functions of Final Appellate Courts: The Supreme Court of Canada" (1975), 53 Can. Bar Rev. 469, at pp. 478-79:

When everything considered relevant has been weighed and an overruling decision commends itself to a judge, he ought not at that stage to stay his opinion and call upon the legislature to implement it. This is particularly true in respect of those areas of the law which are judge-made, and to a degree true in respect of those areas where legislation is involved which is susceptible of a number of meanings. A final court must accept a superintending responsibility for what it or its predecessors have wrought, especially when it knows how little time legislatures today have (and also, perhaps, little inclination) to intrude into fields of law fashioned by the courts alone, although legislatures may, of course, under the prodding of law reform agencies and of other public influences, from time to time do so.

While this Court has limited changes to the common law to those which are "slow and incremental" rather than "major and far-reaching" (McLachlin J. in Watkins v. Olafson, 1989 36 (SCC), [1989] 2 S.C.R. 750, at p. 760), and only "those incremental changes which are necessary to keep the common law in step with the dynamic and evolving fabric of our society" (Iacobucci J. in R. v. Salituro, 1991 17 (SCC), [1991] 3 S.C.R. 654, at p. 670), it is important to assess the change in the correct frame of reference.

In the context of any particular impugned rule standing on its own, no change could ever satisfy the above criteria. Viewed this way, changing the orthodox rule will mean eliminating the orthodox rule because of its absolute prohibition on the substantive use of prior inconsistent statements. The respondent correctly submits that this change is not slow and incremental.

However, if the change to the orthodox rule is viewed within the larger context of the trend within evidence law towards greater admissibility and a correspondingly increased emphasis on the weight to be accorded admissible evidence, the change is clearly incremental as it renders one particular type of evidence admissible for a certain purpose subject to certain conditions. Even if the frame of reference is narrowed to the adoption, by this Court, of a principled approach to hearsay evidence in Khan, and Smith, the change to a reformed rule is but an application of that principled approach to a particular type of otherwise excluded hearsay evidence, which is appropriate and "necessary to keep the common law in step with the dynamic and evolving fabric of our society" for all the same reasons that the adoption of the principled approach was necessary in Khan and Smith. As I wrote in Smith (at p. 932):

What is important, in my view, is the departure signalled by Khan from a view of hearsay characterized by a general prohibition on the reception of such evidence, subject to a limited number of defined categorical exceptions, and a movement towards an approach governed by the principles which underlie the rule and its exceptions alike. The movement towards a flexible approach was motivated by the realization that, as a general rule, reliable evidence ought not to be excluded simply because it cannot be tested by cross-examination.

Notably, in Salituro, at p. 666, Iacobucci J. cited the decision of McLachlin J. in Khan as an example of this Court's "willingness to adapt and develop common law rules to reflect changing circumstances in society at large".

Therefore, I do not believe that considering a change to a reformed prior inconsistent statements rule is a matter better left to Parliament; the rule itself is judge-made and lends itself to judicial reform, and it is a natural and incremental progression in the development of the law of hearsay in Canada by this Court.

Having decided that it is the province and duty of this Court to review the orthodox rule, the question then becomes whether this Court's decision in Deacon stands in the way of implementing the rule we think best. I do not think that it does.

In Salituro, at p. 665, Iacobucci J. stated that "[t]his Court is now willing, where there are compelling reasons for doing so, to overturn its own previous decisions." However, there are guidelines which direct our exercise of this jurisdiction to overrule previous decisions. In R. v. Chaulk, 1990 34 (SCC), [1990] 3 S.C.R. 1303, at p. 1353, I adopted the considerations listed by Dickson C.J. (in dissent) in R. v. Bernard, 1988 22 (SCC), [1988] 2 S.C.R. 833. Those guidelines were:

(1)whether the rule or principle under consideration must be varied in order to avoid a Charter breach;

(2)whether the rule or principle under consideration has been attenuated or undermined by other decisions of this or other appellate courts;

(3)whether the rule or principle under consideration has created uncertainty or has become "unduly and unnecessarily complex and technical;" and

(4)whether the proposed change in the rule or principle is one which broadens the scope of criminal liability, or is otherwise unfavourable to the position of the accused.

While it is not necessary to vary the rule to avoid a Charter breach, the respondent submitted that any change would in fact result in a Charter violation, since a change to the orthodox rule would be inconsistent with any right the accused may have under the Charter to "confrontation," analogous to the U.S. constitutional right under the Sixth Amendment (which provides that "[i]n all criminal prosecutions, the accused shall enjoy the right to . . . be confronted with the witnesses against him . . . "). The respondent relied heavily on R. v. Potvin, 1989 130 (SCC), [1989] 1 S.C.R. 525, in which Wilson J. held that the introduction of preliminary inquiry evidence at trial pursuant to s. 715 of the Criminal Code (formerly s. 643) did not violate the accused's rights under ss. 7 or 11(d) of the Charter, as long as the accused had an adequate opportunity to cross-examine the witness (at p. 543):

It is, in my view, basic to our system of justice that the accused have had a full opportunity to cross-examine the witness when the previous testimony was taken if a transcript of such testimony is to be introduced as evidence in a criminal trial for the purpose of convicting the accused.

With respect to s. 11(d) of the Charter, Wilson J. held (at p. 546): "...this claim must also fail if his constitutional right to have had a full opportunity to cross-examine the witness on the earlier occasion was respected".

However, it is clear that Wilson J.'s emphasis on the importance of cross-examination at the time the evidence was given was necessitated by the present unavailability of the witness, as required by the list of circumstances in s. 715(1)(a) to (d): for the purposes of that provision, the witness must be proven to have refused to testify, or be dead, insane, too ill to travel, or absent from Canada. Cross-examination at the time the evidence was given is clearly the only kind of cross-examination possible where a witness is unavailable at trial, so Wilson J.'s words should not be read in an overly restrictive fashion. In the case of prior inconsistent statements, by definition the witness is presently available for cross-examination, and cross-examination at trial represents an adequate safeguard of the accused's ss. 7 and 11(d) rights as a substitute for cross-examination at the time the prior inconsistent statement was made. I have discussed the arguments respecting the value of contemporaneous cross-examination, and will return to them below. In addition, Wilson J. also expressed the more lenient proposition in Potvin (at p. 544) that "It is, as I have said, a principle of fundamental justice that the accused have had a full opportunity to cross-examine the adverse witness", indicating that her references to contemporaneity were perhaps not intended to define the content of the right to cross-examine. This more lenient requirement is clearly satisfied in the case of prior inconsistent statements.

I note also that the effect of this Court's holding in Khan, in which the evidence of a child complainant was admitted through her mother's testimony, was to deprive the accused of any opportunity to cross-examine the hearsay declarant, either at the time the statement was made or at trial. Similarly, in Smith, the hearsay declarant was dead, Smith being tried for her murder. Indeed, such is the case with all of the hearsay exceptions in which necessity is created by the unavailability of the hearsay declarant, either under the old categorical approach or under the new principled approach. In neither Khan nor Smith was it suggested that the inability of the accused to cross-examine the hearsay declarant at any time infringed any of the accused's Charter rights. In this case, the accused has the added protection of full cross-examination at trial.

Finally, the operation of a modified rule to admit the accused's self-incriminating prior inconsistent statements from a previous proceeding in violation of s. 13 is foreclosed by this Court's decision in Kuldip, which recognized that this provision of the Charter required a form of the orthodox rule to be applied to the accused's prior statements, even though such statements, being admissions, would not be excluded as substantive evidence by the hearsay rule.

The other guidelines also suggest that this Court should not decline to overrule Deacon if it thinks best to do so. While the orthodox rule itself has not been attenuated by this Court, the traditional hearsay rules from which it developed and which supply its rationale certainly have been in Khan and Smith. Successful attacks on the categorical approach to the hearsay rule also weaken the underpinnings of the orthodox rule. It is true that this Court has recently restated the orthodox rule, but it has not done so in the context of reconsidering the orthodox rule itself.

Part of the rationale for the orthodox rule has similarly been attenuated by changes in the methods of proof and demonstration in the modern trial process. When hearsay evidence is tendered, the trier of fact is being asked to act on second-hand testimony: the trier never actually hears or sees the hearsay declarant make the statement, and so has no basis on which to evaluate the statement for him or herself. Rather, the trier of fact must rely on the account of the statement given by the testifying witness, with all of the inherent weaknesses of recollection and perception that human witnesses possess. All of this has changed with the advent of videotaping. In a manner not possible when Wright v. Beckett, Duckworth, or Deacon were decided, prior statements can be placed before the trier of fact in their entirety and in a form which ensures their integrity. This change has significantly undermined the "presence" rationale for the orthodox rule.

Respecting the complexity or technicality of the orthodox rule's distinction between permissible and impermissible uses of prior inconsistent statements, I wrote in Kuldip, at p. 635, that "While such a distinction may be somewhat troublesome to the jury, it is my view that with the benefit of clear instructions from the trial judge the jury will not be unduly burdened with this distinction." The distinction of the orthodox rule itself is not overly complex, but once viewed as part of the categorical approach to hearsay evidence, it is clear that it is one of the more artificial and technical of the rules of evidence from the point of view of a juror. It would be far better, in my view, to replace the blanket exclusion of reliable evidence for a particular purpose with a general rule of admissibility for all purposes when certain criteria are met.

Finally, there is the matter of whether reforming the orthodox rule would be more favourable to the Crown than to the accused, or would otherwise broaden the scope of criminal liability. In R. v. Williams (1985), 1985 113 (ON CA), 50 O.R. (2d) 321, leave to appeal to S.C.C. refused, [1985] 1 S.C.R. xiv, Martin J.A. for the Ontario Court of Appeal stated, at p. 341, that "[g]enerally speaking, the accused, rather than the Crown, is the beneficiary of the rule that prior inconsistent statements of a witness (unless adopted by him in his evidence) are not evidence of the facts contained in them." This is because the Crown bears the burden of leading evidence, much of which will be based on police interviews with potential witnesses, leading to a greater likelihood that a Crown witness will recant and allow the Crown to take advantage of a reformed rule. The accused and his counsel will not often have the same access to witnesses to collect prior statements, and may not in fact lead evidence at all.

However, I do not think that it can be conclusively stated that a reformed rule would necessarily broaden the scope of criminal liability in the sense intended by Dickson C.J. in Bernard. In Bernard, at pp. 860-61, Dickson C.J. wrote of the fourth guideline:

Respect for the principle of certainty and the institutional limits imposed upon the law-making function of the courts should constrain the Court from overruling a prior decision where the effect would be to expand criminal liability. It is not for the courts to create new offences, or to broaden the net of liability, particularly as changes in the law through judicial decision operate retrospectively. The same argument does not apply, however, where the result of overruling a prior decision is to establish a rule favourable to the accused.

Thus, it can be seen that Dickson C.J. was speaking of a more causally direct expansion of criminal liability; for example, in Bernard by reconsidering the effect of drunkenness on criminal liability, and in Chaulk by reconsidering the meaning of insanity. Merely admitting substantive evidence which will more often perhaps, incriminate the accused than it will exculpate the accused does not, in my view, expand criminal liability in the sense intended by Dickson C.J. in Bernard. To enable the Crown to secure more convictions with a new evidentiary rule expanding the scope of admissibility is not to expand the scope of criminal liability; it is, rather, to find more criminals liable. It cannot be said that the accused has any "certainty" interest in the orthodox rule, unless one considers the situation where the accused convinces a Crown witness to recant, secure in the knowledge that the prior statement will not be substantively admissible. This is not a result that the value of certainty is intended to promote. In addition, the courts are not institutionally limited from adjusting common law rules, as I have stated above, the same way they are from making wholesale changes to the elements of substantive offences. While it may be true that prior inconsistent statements will more often be tendered by the Crown, this is not to say that such statements will always be believed by the trier of fact in preference to the witnesses' present testimony.

Therefore, the guidelines established in Bernard do not suggest that this Court should do anything other than what it thinks best in reconsidering the orthodox rule: a reformed rule would not violate the Charter, the existing rule has been attenuated by developments in the law of hearsay and is somewhat, if not overly, technical, and reforming the rule would not directly expand the scope of criminal liability. All that remains is to consider what this Court ought to do.

V - The New Admissibility Rule

A. Requirements and the Hearsay Dangers

I am of the view that evidence of prior inconsistent statements of a witness other than an accused should be substantively admissible on a principled basis, following this Court's decisions in Khan and Smith. However, it is clear that the factors identified in those cases -- reliability and necessity -- must be adapted and refined in this particular context, given the particular problems raised by the nature of such statements. Furthermore, there must be a voir dire before such statements are put before the jury as substantive evidence, in which the trial judge satisfies him or herself that the statement was made in circumstances which do not negate its reliability.

As a threshold matter, before discussing the specific requirements of the reformed rule, I would adopt the requirement embodied in the provision proposed by the Law Reform Commission of Canada, and in the English Civil Evidence Act 1968, that prior inconsistent statements will only be admissible if they would have been admissible as the witness's sole testimony. That is, if the witness could not have made the statement at trial during his or her examination-in-chief or cross-examination, for whatever reason, it cannot be made admissible through the back door, as it were, under the reformed prior inconsistent statement rule.

There are two situations which provide examples of this requirement. First, it may be that the content of the prior inconsistent statement would not normally be admissible because it is hearsay. If the witness makes a prior statement relating direct evidence about a material fact ("I saw Y fire the gun"), there will be no barrier to admitting the prior statement. Such direct evidence would have been subject to no exclusionary rule of evidence.

However, if the witness's prior statement merely repeated the direct evidence of another person ("X said he saw Y fire the gun"), such a statement, even when made with circumstantial guarantees of reliability, will not be substantively admissible for the truth of the evidence of that other person: because the statement is naked hearsay if offered to prove the fact that Y fired the gun, it would not have been admissible to prove that fact (absent a hearsay exception) as direct evidence from the testifying witness, and it does not become admissible evidence to prove that fact through the operation of the reformed rule. The only use that can be made of such prior statements is as proof that the statement was made to the witness, but not, obviously, as proof that Y fired the gun. This is no more than the hearsay rule applied to the prior statement as if it was evidence tendered at trial.

The final possibility in this category is the present case ("Y told me he fired the gun"), in which the statement relates hearsay which is admissible under an established hearsay exception: while the statement repeats the evidence of another person, as in the second example, the hearsay exception applicable to the reported statement because it is an admission applies to the prior statement, again, as if the prior statement itself had been the witness's sole testimony. Because of the circumstantial guarantees of reliability attaching to admissions (on the assumption that an accused would not falsely incriminate himself), the hearsay statement may be admitted to prove the fact that Y fired the gun. But it is crucial that the matters in the prior statement would have been admissible if offered as the witness's sole testimony.

A second category of inadmissible evidence must also be discussed. In some cases, the prior statement may be that of a state agent who repeats an admission by the accused. Imagine the case of a police agent (perhaps a friend of the accused who is conscripted by the police) who is placed in the cell with the accused, and to whom the accused makes an admission. The police agent repeats the admission for the police, but recants at trial. Again, for the prior statement to be substantively admissible under the reformed rule, the admission must have been obtained in such a way that it would have been admissible as the witness's sole evidence. This Court has recently discussed the law applicable to such statements in R. v. Broyles, 1991 15 (SCC), [1991] 3 S.C.R. 595. For the Court, Iacobucci J. held that the right to silence under s. 7 of the Charter was engaged only when the accused was detained, and that the right to silence would be violated when a person acting as an agent of the state, but who is not obviously a state agent in the eyes of the accused, elicits a statement from the accused. Evidence so obtained in violation of the accused's right to silence would then be subject to exclusion under s. 24(2) of the Charter, and with respect to the first of the factors enumerated in R. v. Collins, 1987 84 (SCC), [1987] 1 S.C.R. 265, Iacobucci J. held (at p. 618) that "the fact that evidence is obtained by conscripting the accused against himself or herself will generally be sufficient to render the trial unfair". The other factors (seriousness of the violation, and the effect of excluding the evidence) would also have to be considered, but the first factor may well be determinative.

The distinctions established in Broyles apply equally here. If the accused was not in detention when he or she made the admission, the prior statement of the state agent reporting the admission will be admissible under the reformed rule, whether it was elicited or not. If the accused was in detention when he or she made the admission, the prior statement of the state agent reporting the admission will be admissible only if the admission was made without elicitation by the state agent. In circumstances where detention and elicitation are combined in circumstances where the evidence should be excluded under s. 24(2), the state agent would not have been able to repeat the admission in his testimony at trial, so substantive evidence of that admission cannot be put before the jury through the operation of the reformed rule.

These are but two examples of circumstances in which the admissibility of the prior statement must be examined, lest what would be excluded as the witness's primary evidence be admitted under the reformed rule simply because the witness has recanted.

I turn now to a consideration of the circumstantial guarantees of reliability required in the reformed rule.

(1) Reliability

The reliability of prior inconsistent statements is clearly a key concern for law reformers and courts which have reformed the orthodox rule, and, as I have outlined, this concern is centred on the hearsay dangers: the absence of an oath, presence, and contemporaneous cross-examination. The reliability concern is sharpened in the case of prior inconsistent statements because the trier of fact is asked to choose between two statements from the same witness, as opposed to other forms of hearsay in which only one account from the declarant is tendered. In other words, the focus of the inquiry in the case of prior inconsistent statements is on the comparative reliability of the prior statement and the testimony offered at trial, and so additional indicia and guarantees of reliability to those outlined in Khan and Smith must be secured in order to bring the prior statement to a comparable standard of reliability before such statements are admitted as substantive evidence.

In my opinion, and as my discussion of these dangers above indicates, only the first two of these dangers present real concerns in this context, and if these two dangers are addressed, a sufficient degree of reliability has been established to allow the trier of fact to weigh the statement against evidence tendered at trial by the same witness. The ultimate reliability of the statement and the weight to be attached to it remain, as with all evidence, determinations for the trier of fact. What the reliability component of the principled approach to hearsay exceptions addresses is a threshold of reliability, rather than ultimate or certain reliability.

The history of the common law exceptions to the hearsay rule suggests that for a hearsay statement to be received, there must be some other fact or circumstance which compensates for, or stands in the stead of the oath, presence and cross-examination. Where the safeguards associated with non-hearsay evidence are absent, there must be some substitute factor to demonstrate sufficient reliability to make it safe to admit the evidence.

I turn now to a consideration of what "substitute" indicia of trustworthiness might suffice to permit reception of prior inconsistent statements, bearing in mind that the question of reliability is a matter for the trial judge, to be decided on the particular circumstances of the case.

(i) The Oath

It is undeniable that the significance of the oath has drastically changed since its introduction. Originally the oath was grounded upon a belief that divine retribution would visit those who lied under oath. Accordingly, witnesses were required to believe in this retribution if they were to be properly sworn and their evidence admissible. In Omychund v. Barker (1744), 1 Atk. 21, 26 E.R. 15, Willes L.C.J. stated, at p. 31 E.R.:

Though I have shewn that an Infidel in general cannot be excluded from being a witness, and though I am of the opinion that infidels who believe a God, and future rewards and punishments in the other world, may be witnesses; yet I am as clearly of opinion, that if they do not believe a God, or future rewards and punishments, they ought not to be admitted as witnesses.

Similarly, Lord Hardwicke L.C. referred (at p. 32 E.R.) to the oath as "an appeal to the Supreme Being, as thinking him the rewarder of truth, and avenger of falsehood". The difference between sworn and unsworn evidence was therefore crucial, and the absence of an oath when the prior statement was made was an important obstacle to admitting prior inconsistent statements for their truth.

We no longer require this belief in divine retribution; in Reference re Truscott, 1967 66 (SCC), [1967] S.C.R. 309, at p. 368, this Court stated in the context of child witnesses that the witness need only understand "the moral obligation of telling the truth". In this sense the oath can be said to have a changed significance, and if critics of the oath suggest only that its original supernatural force has disappeared, I agree with that observation.

It is also true, as Cory J. notes, that sanctions attach to statements not made under oath. A witness who tells one story to the police and another at trial is currently exposed to prosecution under ss. 139 (obstructing justice) and 140 (public mischief) of the Criminal Code, R.S.C., 1985, c. C-46. Furthermore, with the Court's decision in this case, prior statements which satisfy the criteria of admissibility will be used as substantive evidence in a subsequent trial; as a result, a witness who makes a false statement will also be liable to prosecution under s. 137 (fabricating evidence), once he or she is informed that the statement can, and indeed will, be used at trial if he or she recants. Finally, it may well be that, in light of this decision, Parliament will wish to extend other criminal sanctions (such as the offence of perjury, for example) to one who lies under oath, solemn affirmation or solemn declaration in the course of a criminal investigation.

However, there remain compelling reasons to prefer statements made under oath, solemn affirmation or solemn declaration. While the oath will not motivate all witnesses to tell the truth (as is indicated by the witnesses' perjury in this case), its administration may serve to impress on more honest witnesses the seriousness and significance of their statements, especially where they incriminate another person in a criminal investigation.

In addition to this positive effect on the declarant, the presence of an oath, solemn affirmation or solemn declaration will increase the evidentiary value of the statement when it is admitted at trial. First, it will mean that the trier of fact will not be asked to accept unsworn testimony over sworn testimony; instead, the trier will have the opportunity to choose between two sworn statements, and the trier's ultimate decision will not be made on the basis of unsworn or unaffirmed testimony. Similarly, should the prior statement be decisive, there is no danger of the accused being convicted solely on the basis of unsworn testimony.

Second, the presence of the oath during the making of the prior statement eliminates the explanation offered by many recanting witnesses, including one of the witnesses in this case: when confronted with the prior inconsistent statement, witnesses explain that it was not made under oath, and assert that the oath they took at trial persuaded them to tell the truth. This naturally privileges the trial testimony in the mind of the trier of fact. If both statements were made under oath, such an explanation can no longer be employed. Furthermore, since both statements cannot be true, the trier of fact has an indication of the low regard in which the witness holds the oath. Therefore, while it is true that the oath in itself has no power to ensure truthfulness in some witnesses, the fact that both statements were made under oath removes resort to the absence of an oath as an indicium of the alleged unreliability of the prior inconsistent statement.

The presence of an oath, solemn affirmation or solemn declaration will have yet another positive effect on the declarant's truthfulness and the administration of justice. A sworn prior statement will be highly persuasive evidence in any prosecution against the declarant related to false testimony (whether in the statement or at trial), and the knowledge that this evidence exists for this purpose should weigh heavily on the mind of one who considers lying in a statement, or recanting his or her prior statement to lie at trial.

Of course, the incentives provided by the declarant's exposure to prosecution under ss. 137, 139 and 140 in relation to the first statement, and his or her fear of a perjury prosecution in relation to testimony given at trial, will only be effective if these sanctions are made known to the declarant. For this reason, the witness should be warned by the person taking the statement that the statement may be used as evidence at a subsequent trial if the witness recants (thereby engaging s. 137), and also that severe criminal sanctions will accompany the making of a false statement. This warning should refer specifically to ss. 137, 139 and 140 of the Criminal Code, and repeat the elements of and sanctions for those offences. As does the formal swearing of the witness in the trial process, this warning and the administration of the oath should serve to bring home to the witness the gravity of the situation and his duty to tell the truth.

Therefore, the best indicium of reliability on the principled approach of Smith in the case of prior inconsistent statements is that the statement, to be substantively admissible, has been made (i) under oath, solemn affirmation, or solemn declaration, and (ii) following the administration of an explicit warning to the witness of his or her amenability to prosecution if it is discovered that he or she has lied. This indicium satisfies the first hearsay danger entirely: in no case will the trier of fact be asked to accept unsworn testimony over sworn testimony, verdicts will not be based on unsworn testimony, and the circumstances which promote truthful trial testimony will have been recreated as fully as is possible.

Were the oath an absolute requirement for a finding of reliability, the only prior inconsistent statements which could be received would be statements made in circumstances where the person receiving the statement is authorized to administer the warning and the oath or affirmation. Thus, statements made to family members or friends would generally not comply, unless the witness then repeats the statement for appropriately authorized persons. In the case of police interviews, this would likely present no real difficulty, since each police station will usually have a Justice of the Peace present or readily available for interim release hearings. The Justice could then administer the warning before the statement is made, and the oath after the statement is made. Similarly, police officers on duty as the Officer in Charge could be made commissioners for the taking of oaths in the province, and administer the warning and oath in the Justice's absence.

However, I do not wish to create technical categorical requirements duplicating those of the old approach to hearsay evidence. It follows from Smith that there may be situations in which the trial judge concludes that an appropriate substitute for the oath is established and that notwithstanding the absence of an oath the statement is reliable. Other circumstances may serve to impress upon the witness the importance of telling the truth, and in so doing provide a high degree of reliability to the statement. While these occasions may not be frequent, I do not foreclose the possibility that they might arise under the principled approach to hearsay evidence.

(ii) Presence

Proponents of the orthodox rule emphasize the many verbal and non-verbal cues which triers of fact rely upon in order to assess credibility. When the witness is on the stand, the trier can observe the witness's reaction to questions, hesitation, degree of commitment to the statement being made, etc. Most importantly, and subsuming all of these factors, the trier can assess the relationship between the interviewer and the witness to observe the extent to which the testimony of the witness is the product of the investigator's questioning. Such subtle observations and cues cannot be gleaned from a transcript, read in court in counsel's monotone, where the atmosphere of the exchange is entirely lost.

All of these indicia of credibility, and therefore reliability, are available to the trier of fact when the witness's prior statement is videotaped. During the course of the hearing, counsel for the appellant screened a brief excerpt from the videotape of one of the interviews. In the main portion of the television screen is a medium-length shot of the witness facing the camera and seated across a table from the interviewing officer, showing the physical relationship between the two people. In one upper corner is a close-up of the witness's face as he or she speaks, capturing nuances of expression lost in the main view. Along the bottom of the screen is a line showing the date and a time counter, with the seconds ticking off, ensuring that the continuity and integrity of the record is maintained. The audio-visual medium captures other elements of the statement lost in a transcript, such as actions or distinctive motions which the witness demonstrates (as in this case), or answers given by nodding or shaking the head. In other words, the experience of being in the room with the witness and the interviewing officer is recreated as fully as possible for the viewer. Not only does the trier of fact have access to the full range of non-verbal indicia of credibility, but there is also a reproduction of the statement which is fully accurate, eliminating the danger of inaccurate recounting which motivates the rule against hearsay evidence. In a very real sense, the evidence ceases to be hearsay in this important respect, since the hearsay declarant is brought before the trier of fact.

Of course, the police would not resort to this precaution in every case; it may well be reserved for cases such as this, where a major crime such as murder is being investigated, the testimony of the witnesses is important to the Crown's case, and the character of the witnesses suggests that such precautions would be advisable. It is quite possible that such equipment would be available to police of given forces at a central location, and that such crucial though unstable witnesses will be taken to such locations to make their statements, or, where the statements have already been made, to repeat them in a form which may be substantively admissible should the witness recant.

In addition to an oath or solemn affirmation and warning, then, a complete videotape record of the type described above, or one which duplicates the experience of observing a witness in the courtroom to the same extent, is another important indicium of reliability which will satisfy the principled basis for the admission of hearsay evidence.

Again, it may be possible that the testimony of an independent third party who observes the making of the statement in its entirety could, in exceptional circumstances, also provide the requisite reliability with respect to demeanour evidence. I would only note at this point that there are many persons who could serve this function: police stations will have justices of the peace present or available, the witness may have his or her own lawyer present, and ss. 56(2)(c) and 56(2)(d) of the Young Offenders Act, R.S.C., 1985, c. Y-1, provide that a young person making a statement has a right of access to counsel, parents, or adult relatives. It will be a matter for the trial judge to determine whether or not a sufficient substitute for a videotape record has been provided to allow the trier of fact access to sufficient demeanour evidence to make the statement admissible.

(iii) Cross-examination

The final hearsay danger is the lack of contemporaneous cross-examination when the statement is made. The appellant is correct to concede that this is the most important of the hearsay dangers. However, in the case of prior inconsistent statements, it is also the most easily remedied by the opportunity to cross-examine at trial. This is a feature of prior inconsistent statements that conclusively distinguishes them from other forms of hearsay. As the United States Supreme Court noted in California v. Green, supra, at p. 159:

... the inability to cross-examine the witness at the time he made his prior statement cannot easily be shown to be of crucial significance as long as the defendant is assured of full and effective cross-examination at the time of trial. The most successful cross-examination at the time the prior statement was made could hardly hope to accomplish more than has already been accomplished by the fact that the witness is now telling a different, inconsistent story, and -- in this case -- one that is favorable to the defendant.

Furthermore, unlike the oath and presence, it is the hearsay danger which is impossible to address outside of judicial or quasi-judicial processes. Whereas the police can easily administer a warning and oath, and videotape a statement in the course of a witness interview, it would restrict the operation of a reformed rule to judicial or quasi-judicial proceedings to require contemporaneous cross-examination, and thereby severely restrict the impact of a reformed rule. Consider the facts of the present case: when the three witnesses were interviewed by the police, no one had yet been charged with an offence. Who could have cross-examined the witnesses at that point? How could cross-examination have been effective before the case to be met was known? These and other practical difficulties in requiring contemporaneous cross-examination tip the balance in favour of allowing cross-examination at trial to serve as a substitute. Again, we must remember that the question is not whether it would have been preferable to have had the benefit of contemporaneous cross-examination, but whether the absence of such cross-examination is a sufficient reason to keep the statement from the jury as substantive evidence. Given the other guarantees of trustworthiness, I do not think that it should be allowed to be a barrier to substantive admissibility. Of course, it will be an important consideration for the trier of fact in deciding what weight to attach to the prior inconsistent statement, and it is likely that opposing counsel will stress the absence of such cross-examination to the trier of fact.

Therefore, the requirement of reliability will be satisfied when the circumstances in which the prior statement was made provide sufficient guarantees of its trustworthiness with respect to the two hearsay dangers a reformed rule can realistically address: if (i) the statement is made under oath or solemn affirmation following a warning as to the existence of sanctions and the significance of the oath or affirmation, (ii) the statement is videotaped in its entirety, and (iii) the opposing party, whether the Crown or the defence, has a full opportunity to cross-examine the witness respecting the statement, there will be sufficient circumstantial guarantees of reliability to allow the jury to make substantive use of the statement. Alternatively, other circumstantial guarantees of reliability may suffice to render such statements substantively admissible, provided that the judge is satisfied that the circumstances provide adequate assurances of reliability in place of those which the hearsay rule traditionally requires.

(2) Necessity

Prior inconsistent statements present vexing problems for the necessity criterion. The necessity criterion has usually been satisfied by the unavailable witness: in Khan, the child declarant who could not be sworn, and in Smith, the dead declarant. By definition, the declarant in the case of prior inconsistent statements is available at trial; it is his or her prior statement that is unavailable because of the recantation.

However, it is important to remember that the necessity criterion "must be given a flexible definition, capable of encompassing diverse situations" (Smith, at pp. 933-34). Wigmore, vol. 5 (Chadbourn rev. 1974), § 1421, at p. 253, referred to two classes of necessity:

(1) The person whose assertion is offered may now be dead, or out of the jurisdiction, or insane, or otherwise unavailable for the purpose of testing. This is the commoner and more palpable reason....

(2) The assertion may be such that we cannot expect, again, or at this time, to get evidence of the same value from the same or other sources.... The necessity is not so great; perhaps hardly a necessity, only an expediency or convenience, can be predicated. But the principle is the same. [Emphasis in original.]

As an example of the second type of necessity, many established hearsay exceptions do not rely on the unavailability of the witness. Some examples include admissions, present sense impressions and business records. This is because there are very high circumstantial guarantees of reliability attached to such statements, offsetting that fact that only expediency or convenience militate in favour of admitting the evidence.

Indeed, in shaping the law of hearsay in Canada, this Court has not treated necessity in the sense of unavailability as the sine qua non of admissibility. In Ares v. Venner, 1970 5 (SCC), [1970] S.C.R. 608, for example, nurses' records were admitted as evidence at a medical negligence trial. The nurses, though present in court through the trial, were not called as witnesses. The Alberta Court of Appeal ordered a new trial, on the basis that Wigmore's necessity ground for the admission of hearsay was not satisfied since the nurses were available to testify. This Court allowed an appeal from this decision, holding at p. 626 that:

Hospital records, including nurses' notes, made contemporaneously by someone having a personal knowledge of the matters then being recorded and under a duty to make the entry or record should be received in evidence as prima facie proof of the facts stated therein.

The Court made no reference to the present availability of the nurses as it related to the admissibility of the hearsay evidence, except to note their presence in court meant that the defendant could have challenged the accuracy of the notes if he had so wished. Similarly, the maker of the prior inconsistent statement is present in court as a witness to be examined and cross-examined as to the accuracy of the recording of the statement. Ares v. Venner stands as an example of a judicially-created hearsay exception which did not require unavailability. While the decisions in Khan and Smith established that Canadian courts will no longer carve out categorical "exceptions", the new approach shares the same principled basis as the existing exceptions.

I note also that some lower courts interpreting this Court's decision in Khan have also given the necessity criterion a flexible definition. In Khan v. College of Physicians and Surgeons of Ontario (1992), 1992 2784 (ON CA), 9 O.R. (3d) 641, the Ontario Court of Appeal returned to the incident which led to this Court's decision in R. v. Khan. The disciplinary hearing took place some four years after the initial incident, at which time the child complainant was old enough to testify. However, the committee also heard evidence from the complainant's mother and other persons to whom the complainant had described the assault. The Ontario Court of Appeal held that this was not an error, since the interval between the time of the assault and the time of the disciplinary hearing meant that the complainant's ability to recall and recount the assault had deteriorated, compared to her statements immediately after the incident. Thus, the court held that the testimony of certain of the witnesses to whom the complainant described the assault was admissible under the Khan test, since it was "reasonably necessary" in the circumstances, and (at p. 655) "[a] rule which would automatically exclude the out-of-court statement where the child testifies is inconsistent with Ares v. Venner, supra, the authority relied on in Khan."

The precise limits of the necessity criterion remain to be established in the context of specific cases. It may be that in some circumstances, the availability of the witness will mean that hearsay evidence of that witness's prior consistent (the kind of statement at issue in Khan) statements will not be admissible. However, I am not prepared, at this point, to adhere to a strict interpretation that makes unavailability an indispensable condition of necessity.

In the case of prior inconsistent statements, it is patent that we cannot expect to get evidence of the same value from the recanting witness or other sources: as counsel for the appellant claimed, the recanting witness holds the prior statement, and thus the relevant evidence, "hostage." The different "value" of the evidence is found in the fact that something has radically changed between the time when the statement was made and the trial and, assuming that there is a sufficient degree of reliability established under the first criterion, the trier of fact should be allowed to weigh both statements in light of the witness's explanation of the change.

B. The Voir Dire

Pursuant to the circumstantial guarantees of reliability described above, prior statements may be used as substantive evidence of their contents by the jury. The two-stage process by which this may be done must now be described. After the calling party invokes s. 9 of the Canada Evidence Act, and fulfils its requirements in the voir dire held under that section, the party must then state its intention in tendering the statement. If the party indicates that it wishes to use the statement only to impeach the credibility of the witness, that is the end of the matter as regards the reformed rule: the trial proceeds as it did under the orthodox rule, with the judge instructing the jury accordingly. If, however, the party gives notice that it will seek to make substantive use of the statement, the trial judge must continue the voir dire to satisfy him or herself on the appropriate measure (which I will discuss below) that these indicia of reliability, or acceptable substitutes, are present: the oath, affirmation, or solemn declaration will be proved, the person who administered the oath, affirmation, or solemn declaration will testify that he or she also administered the warning (or perhaps this could be incorporated into the oath, affirmation, or solemn declaration), and the videotape will be tendered into evidence, its authenticity sworn to, and, if the trial judge wishes, screened to ensure its veracity and integrity.

With respect to the burden of proof in the voir dire, ordinarily the trial judge should be satisfied that these indicia of reliability are established on the balance of probabilities, the normal burden resting upon a party seeking to admit evidence. This is no more than a corollary of the requirement that the prior statements must relate evidence which would have been admissible as the witness's sole testimony had he or she not recanted.

A different situation might exist where the prior statement reports an admission made by the accused. If the statement is not made to a person in authority no special burden is required, since the ordinary burden for the admission of evidence would have applied to the witness's testimony at trial had he or she not recanted. However, if the prior statement reports an admission of the accused made to a person in authority, the higher burden associated with the law relating to confessions may well apply. Such incidents will be rare, since persons in authority who receive statements in the course of their duties from accused persons will not often recant. Additionally, if an agent of the state elicits a statement from a detained accused, the case law developed under the Charter in this respect would have to be considered with respect to the burden during the voir dire.

As neither of these issues arise in this case (since the recanting witnesses were clearly neither persons in authority in relation to the accused, nor were they agents of the state when the accused made his admissions to them, nor was the accused detained), I would leave those rare and theoretical situations to be addressed when and if they arise.

However, I would incorporate another aspect of the rule relating to confessions in the voir dire. Even where there has been a warning and oath administered, and the statement videotaped, or sufficient substitutes established, the trial judge will still have the discretion to refuse to allow the jury to make substantive use of the statement. Prior statements share many characteristics with confessions, especially where police investigators are involved. Proponents of the orthodox rule voice the concern that malign influences on the witness by police may precede the making of the statement and shape its content, in the same way that confessions may be suspect if coerced by police investigators. That is, it still may be the case that the oath and videotape, and the acknowledgement of the warning, were made under circumstances that make them suspect. For this reason, the test developed by this Court for the admission of confessions is well-suited to making a threshold determination of whether the circumstances under which the statement was made undermine the veracity of the indicia of reliability.

The classic statement of the first part of the confession rule appears in Ibrahim v. The King, [1914] A.C. 599 (P.C.), at p. 609:

It has long been established as a positive rule of English criminal law, that no statement by an accused is admissible in evidence against him unless it is shewn by the prosecution to have been a voluntary statement, in the sense that it has not been obtained from him either by fear of prejudice or hope of advantage exercised or held out by a person in authority.

Ibrahim was first adopted by this Court in Prosko v. The King (1922), 1922 584 (SCC), 63 S.C.R. 226, and was extended in decisions such as Horvath v. The Queen, 1979 16 (SCC), [1979] 2 S.C.R. 376, in which Beetz J. wrote (at pp. 424-25):

Furthermore, the principle which inspires the rule remains a positive one; it is the principle of voluntariness. The principle always governs and may justify an extension of the rule to situations where involuntariness has been caused otherwise than by promises, threats, hope or fear, if it is felt that other causes are as coercive as promises or threats, hope or fear and serious enough to bring the principle into play.

I would apply this test to prior statements. The trial judge must satisfy him or herself (again, in the majority of cases on the balance of probabilities) on the voir dire that the statement was not the product of coercion of any form, whether it involves threats, promises, excessively leading questions by the investigator or other person in a position of authority, or other forms of investigatory misconduct.

I would add another element to the trial judge's inquiry to address situations where the first factor might be satisfied, but there are other aspects of police conduct which militate against rewarding that conduct by admitting the evidence. In Rothman v. The Queen, 1981 23 (SCC), [1981] 1 S.C.R. 640, at p. 696, I wrote that even if the Ibrahim test was satisfied to make a confession admissible, such a confession "shall nevertheless be excluded if its use in the proceedings would, as a result of what was said or done by any person in authority in eliciting the statement, bring the administration of justice into disrepute."

It must be stressed that the trial judge is not making a determination on the voir dire as to the ultimate reliability and credibility of the statement. As I have indicated, that is a matter for the trier of fact. The trial judge need not be satisfied that the prior statement was true and should be believed in preference to the witness's current testimony. This distinction is also derived from the law relating to confessions. In Piché v. The Queen, 1970 182 (SCC), [1971] S.C.R. 23, at pp. 25-26, Cartwright C.J. noted that:

The main reason assigned for the rule that an involuntary confession is to be excluded is the danger that it may be untrue but, as has been recently reasserted by this Court in DeClerq v. The Queen [1968 24 (SCC), [1968] S.C.R. 902], the answer to the question whether such a confession should be admitted depends on whether or not it was voluntary not on whether or not it was true.

Similarly, in Rothman, I wrote, at p. 691, that

a statement before being left to the trier of fact for consideration of its probative value should be the object of a voir dire in order to determine, not whether the statement is or is not reliable, but whether the authorities have done or said anything that could have induced the accused to make a statement which was or might be untrue. It is of the utmost importance to keep in mind that the inquiry is not concerned with reliability but with the authorities' conduct as regards reliability.

Thus, to summarize the discussion of the voir dire: in the part of the voir dire addressing the new rule, the trial judge must first satisfy him or herself that the indicia of reliability necessary to admit hearsay evidence of prior statements -- a warning, oath, solemn affirmation, or solemn declaration, and videotape record, or sufficient substitutes -- are present and genuine. If they are, he or she must then examine the circumstances under which the statement was obtained, to satisfy him or herself that the statement supported by the indicia of reliability was made voluntarily if to a person in authority, and that there are no other factors which would tend to bring the administration of justice into disrepute if the statement was admitted as substantive evidence. In most cases, as in this case, the party seeking to admit the prior inconsistent statements as substantive evidence will have to establish that these requirements have been satisfied on the balance of probabilities. The trial judge is not to decide whether the prior inconsistent statement is true, or more reliable than the present testimony, as that is a matter for the trier of fact. Once this process is complete, and all of its constituent elements satisfied, the trial judge need not issue the standard limiting instruction to the jury, but may instead tell the jury that they may take the statement as substantive evidence of its contents, or, if he or she is sitting alone, make substantive use of the statement, giving the evidence the appropriate weight after taking into account all of the circumstances. In either case, the judge must direct the trier of fact to consider carefully these circumstances in assessing the credibility of the prior inconsistent statement relative to the witness's testimony at trial. For example, where appropriate the trial judge might make specific reference to the significance of the demeanour of the witness at all relevant times (which could include when making the statement, when recanting at trial, and/or when presenting conflicting testimony at trial), the reasons offered by the witness for his or her recantation, any motivation and/or opportunity the witness had to fabricate his or her evidence when making the previous statement or when testifying at trial, the events leading up to the making of the first statement and the nature of the interview at which the statement was made (including the use of leading questions, and the existence of pre-statement interviews or coaching), corroboration of the facts in the statement by other evidence, and the extent to which the nature of the witness's recantation limits the effectiveness of cross-examination on the previous statement. There may be other factors the trier of fact should consider, and the trial judge should impress upon the trier of fact the importance of carefully assessing all such matters in determining the weight to be afforded prior inconsistent statements as substantive evidence.

Where the prior statement does not have the necessary circumstantial guarantees of reliability, and so cannot pass the threshold test on the voir dire, but the party tendering the prior statement otherwise satisfies the requirements of s. 9(1) or (2) of the Canada Evidence Act, the statement may still be tendered into evidence, but the trial judge must instruct the jury in the terms of the orthodox rule.

VI - Disposition of This Appeal

In this case, while the statements of the recanting witnesses were videotaped, and counsel for the respondent had a full opportunity to cross-examine the witnesses at trial, the statements were not made under oath, solemn affirmation or by a solemn declaration. However, it may well be that sufficient substitutes for the oath and warning as indicia of reliability exist in this case. As this would be an issue of central importance at any new trial ordered for the respondent, I will express no opinion as to whether the new test is satisfied in this case beyond saying that it is at least possible. A trial judge might, having heard the evidence respecting the circumstances under which these statements were made, be satisfied on the balance of probabilities that the statements are sufficiently reliable to be substantively admissible.

In this case, the trial judge (sitting without a jury) refused to even consider the admissibility of the statements, considering himself bound by the orthodox rule. With this decision, the rule respecting prior inconsistent statements has been reformed. Given the trial judge's expressed belief that the witnesses were lying at trial, it is possible that he might have found sufficient indicia of reliability to admit the statements as substantive evidence. I would therefore allow the appeal and order a new trial at which the reformed rule relating to prior inconsistent statements will, if necessary, be applied by the trial judge, who will decide whether sufficient indicia of reliability and necessity are present in this case, and instruct the trier of fact to afford the prior statements the appropriate weight in reaching a verdict.

Cory J.

The reasons of L'Heureux-Dubé and Cory JJ. were delivered by

Cory J. -- At issue on this appeal is the interpretation that should be given to the Canada Evidence Act, R.S.C., 1985, c. C-5, s. 9. Until now the courts have interpreted this section to permit the use of prior contradictory statements solely for the purpose of attacking the credibility of the witness. In this case the prior statements of the witnesses were videotaped by the police during the course of their investigation of the death by stabbing of Joseph Wright. The videotaped statements of the witnesses implicated the respondent K.G.B. as the person responsible for the stabbing. At the respondent's trial for murder, the three witnesses recanted their earlier statements. The trial judge found that they were lying and that their earlier statements were true. However, he noted that he was bound by prior authority and was unable to accept the prior statements as evidence and therefore could not use them to establish the identity of the killer. The trial judge was proven to be correct in his assessment of the witnesses. Subsequently, they were charged with perjury as a result of their testimony at the trial and all three entered pleas of guilty to that charge.

I have read with interest the excellent reasons of the Chief Justice. I am in agreement with him that the rule against the substantive use of prior statements must be changed; that this Court is the appropriate institution to undertake the change; and that this change can be effected without infringing the Charter rights of an accused person. Further, I agree that a new trial must be ordered so that the substantive admissibility of the statements in this case can be considered. However, with respect, I differ from him with regard to the nature of the new rule.

The Chief Justice takes the position that as a general rule, in order to be admissible, a prior inconsistent statement must have been made under oath; the witness advised that he or she may be subject to some form of criminal prosecution (other than perjury) for falsehood; and the statement must be videotaped. Alternatively, he has determined that, exceptionally, substitutes for these requirements may be found. This involves an application of R. v. Khan, 1990 77 (SCC), [1990] 2 S.C.R. 531, and R. v. Smith, 1992 79 (SCC), [1992] 2 S.C.R. 915, which allow for hearsay to be admitted if its introduction is necessary and the evidence is reliable. With the greatest respect I find the first component of the rule requiring videotaping, a mandatory warning as to criminal liability for falsehood, and the administration of the oath to be too restrictive. Secondly, while I agree that Khan and Smith provide an alternative justification for changing the interpretation of s. 9 of the Canada Evidence Act set out in Deacon v. The King, 1947 38 (SCC), [1947] S.C.R. 531, I do not think that their unmodified application to prior inconsistent statements would adequately protect the interests of the accused from the potential dangers that surround the introduction of statements made out of court.

I. Factual Background

In order to demonstrate the unfortunate result of the present rule, it is necessary for me to set out in some detail the factual background of this case. At 2:30 a.m. on Sunday, April 24, 1988, Joseph Wright and his brother Steven, got off a bus at the corner of Morningside and Sheppard Avenues in Scarborough, Ontario. A car containing four young men, pulled up beside them and words were exchanged. The people in the car got out and a fight ensued. Joseph Wright was slashed twice in the face and stabbed in the heart. This last wound caused his death. When Joseph Wright fell to the ground, the four young men got back into the car and drove away. The only issue at trial was which of the occupants of the car had stabbed Joseph Wright. The four occupants were the respondent K.G.B., P.L. then 16 years of age, P.M. 17 years of age and M.T. 16 years of age. Steven Wright testified that only one occupant of the car had a knife. That person was black. He saw him attacking his brother with a knife. He stated that, after stabbing his brother, this same person had attempted to stab him as well. The fact that Wright testified that the man he saw attacking his brother with the knife was black is of significance only because it eliminates as a suspect M.T., who is white.

There were differences between the description of the knife wielding attacker which Wright gave to the police following the incident and the description contained in his testimony at trial. As well his description differed from the manner in which P.L., P.M. and M.T. described the respondent. These differences related to the relative darkness of the attacker's complexion and the height and hair colour of K.G.B. For what it was worth, when he gave his evidence at the trial, Steven Wright identified the accused as the man who had stabbed his brother.

A. The Evidence of M.T. and the Circumstances of his Prior Statement

M.T. learned that the police wished to speak to him. Some two weeks after the incident, on May 9, 1988, he, in the company of his father, went to the police station. When they arrived the two investigating officers told them that while M.T. was not charged with any offence, that he could have a lawyer present during the interview. This offer was declined. At one point the interview was stopped in order to permit M.T.'s father who was present at all times to call and consult a lawyer. Counsel advised him that since his son was not under arrest, there was no need for legal assistance. The police asked M.T. whether he would like another relative present during the interview instead of his father. He replied that he did not. He was later allowed to consult in private with his father. The officers explained that they were investigating the death of Joseph Wright and that the statement was being videotaped. One of the officers asked M.T. on a number of occasions whether he was telling the truth. On every occasion he replied that he was.

During the course of the videotaped interview, M.T. stated that all four occupants were at the respondent's house the Sunday evening following the incident. The respondent, he said, then told him "I shouldn't a used the knife ... I pulled it out just to scare him thinking he would run away". The witness then demonstrated to the officers the slashing motion that the respondent used while he spoke these words. On the Monday afternoon following the killing, the four occupants of the car again met, this time in P.L.'s home. At this time, the respondent again said that he should not have used the knife. The respondent was also reported to have said that he threw the knife out. M.T. told the officers that the respondent was the only one of the four occupants of the car who had a knife.

At trial, the testimony of M.T. was radically different. He stated that while there was a conversation at the respondent's home after the stabbing, he could not hear what was being said by the respondent. He admitted to making the earlier videotaped statements to the police but said that he was lying at the time because he was "really scared". He testified that he was telling the truth in the witness box because he was under oath.

B. Prior Statement and Evidence at Trial of P.M.

The police came to P.M.'s home to speak to him. As a consequence of this visit he went to the police station on May 6, accompanied by his mother. He was told that he was not going to be charged with any offence at that time. He was advised that he could have a lawyer or any adult relative present during the course of the interview. He asked to have his mother with him during the interview and this request was, of course, granted. He was told that the interview would be videotaped. During the interview he told the officers that the respondent K.G.B. "usually carries a knife". He said that during the fight "he [K.G.B.] had the knife open a little bit". He stated that the respondent told him that "he had the knife open". At the end of the interview, both the police and his mother asked him whether he had told the whole truth and he replied that he had.

At trial, P.M. also changed his story. He testified that when he and M.T. and P.L. went to the respondent's house, the respondent told him that he had heard nothing about what had happened at Sheppard and Morningside, and the three then left. P.M. said that the statement that he had given to police earlier was "lies" to "get myself out of it". He too denied that he was lying to the court and said he was telling the truth because he was now under oath.

C. Prior Statement and the Evidence of P.L. at Trial

At the suggestion of his mother, P.L. went to the police station on May 6, 1988. He was accompanied by his brother, his mother and a lawyer who he had retained. His brother and the lawyer were with him throughout the interview. He was told that he would be given the opportunity to consult with the lawyer in private if he so wished. He was also told he was not obligated to say anything at all to the police. P.L.'s lawyer was of course permitted to intervene and did so on at least two occasions during the interview. In the videotaped interview he stated that K.G.B. had explained that the stabbing was a "fast reaction" on his part because the deceased had punched him in the eye.

At trial, P.L., like the other witnesses changed his testimony. He said that he had lied at the police station because he was "just scared" and wanted to get out of there. He thought he was telling the police what they wanted to hear. He said that he could not in fact hear what was being said in the conversation between P.M. and the respondent regarding events on the night of the killing.

II. The Courts Below

Youth Court

The findings of Judge MacDonnell at trial regarding the testimony of the three witnesses are significant. He found that the prior statements of the Crown witnesses' M.T., P.L. and P.M. independently identified the respondent as the person who possessed and wielded the knife at the time of the incident. He found that the respondent's comments to the witnesses after the incident amounted to an admission by K.G.B. that he had in fact stabbed the deceased. He concluded that the evidence of these witnesses given in court was "lies". He went on to find that the unsworn out of court statements constituted the true version of events.

I have no doubt their recantations are false. That is, I have no doubt that on this point they were telling the police the truth as they knew it about what the accused said. [Emphasis added.]

He found that "it [was] simply unbelievable" that the version given to the police by these witnesses had been manufactured or that the three had collaborated before speaking to the officers. He concluded that the interviews conducted by the police were "completely above-board, proper, non-oppressive, non-coercive". He observed that despite what "common sense might suggest" he was not at liberty to believe the version of the events given to the police in the prior statements. He noted that if the version of events given in the statements:

... did occur, there would be substantial evidence to confirm the identification evidence given earlier by Steven Wright, and the issue of the accused's use of the knife might be resolved against him.

The trial judge viewed only short extracts of the police videotape in reaching these conclusions. It would of course be preferable for a trial judge to view the entire tape. However, there is no doubt that in this case the parties consented to the trial judge making his assessment based on those extracts of the videotape which were viewed and the complete transcript which he read. He found that the uncorroborated in court "dock" identification of K.G.B. by Steven Wright could not be the basis of a conviction. Since he could not consider the prior statements as evidence, he acquitted the respondent.

Court of Appeal (1991), 1991 13906 (ON CA), 49 O.A.C. 30

The Court of Appeal upheld the acquittal and dismissed the Crown's appeal. It decided that it was bound by the decisions of this Court which held that prior inconsistent statements could only be used to test the credibility of the witness.

III.The Arguments Against the Use of Prior Contradictory Statements for Substantive purposes

Let us consider first the arguments that have been raised against the admission of prior contradictory statements as evidence.

A. The Absence of the Oath

The Chief Justice has set out in his reasons the concerns that have been expressed regarding the admission of prior contradictory statements for substantive purposes. Perhaps the foremost concern is that the prior statement is not given under oath. Indeed the Chief Justice has made the oath a prerequisite to admissibility except in exceptional circumstances where an adequate substitute may be found. Courts have always stressed the importance of the oath. In medieval times the taking of an oath to tell the truth by placing a bare hand upon a sacred object was of fundamental importance. It was firmly believed that to defy the oath by lying would lead to divine retribution that would include punishment in this earthly life and eternal damnation in the hereafter. However, the medieval fear of damnation has diminished. Similarly the influence of religion in the affairs of men and women has decreased. There can be little doubt that the taking of an oath is frequently no more than a meaningless ritualistic incantation for many witnesses. In earlier times there may have been good reason for attaching greater weight to testimony given under oath than to unsworn statements given by the same witness. Today, an increasingly secular society simply attaches less significance to taking an oath. To many witnesses, the oath adds nothing to the reliability of their evidence.

Nor can the fear of secular punishment of imprisonment arising from perjury be considered to be a sure means of ensuring truthful testimony. There can be no better demonstration of the proposition that the swearing of a witness to tell the truth has little effect than this very case. Here, despite the taking of the oath by the three witnesses, they lied in court and were convicted of perjury for those lies. The truth was told in their videotaped statements to the police. The facts of this case make it clear that in today's society the taking of the oath by witnesses does not guarantee that they will tell the truth.

Various jurisdictions in the United States have struggled with the same issue. On the one hand, there is the desire to ensure that prior inconsistent statements are reliable so that they can be considered and assessed for all purposes by the triers of fact, on the other hand, is the need to assure a fair trial for the accused. Either as a result of legislation or decisions of the courts four American positions can be identified. First are those states where prior inconsistent statements will as a general rule be admitted for their truth value (15 states). Second, there are states where the prior statements will be admitted so long as there is some indicia presented to the court of their reliability (7 states). Third are those states and the Federal Rules of Evidence which requires the prior statement to have been made under oath in order to be admissible (23 states and the Federal Rules). Fourth, those states and the District of Columbia where the prior statements are not admissible (5 states and the District of Columbia).

It may be helpful to consider the Federal Rule and its history. That rule provides:

Rule 801. ...

(d) Statements which are not hearsay. -- A statement is not hearsay if --

(1) Prior statement by witness. -- The declarant testifies at the trial or hearing and is subject to cross-examination concerning the statement, and the statement is (A) inconsistent with the declarant's testimony, and was given under oath subject to the penalty of perjury at a trial, hearing, or other proceeding, or in a deposition.... [Emphasis added.]

It is significant that the inclusion of the oath in the present American rule, was the result of a political compromise late in the legislative process. The Advisory Committee on the new federal rules, appointed by the then Chief Justice Warren, of the United States Supreme Court, had recommended that prior inconsistent statements be generally admissible provided only that the witness had contradicted himself and could be cross-examined on the earlier statement. See M. H. Graham "Employing Inconsistent Statements for Impeachment and as Substantive Evidence: A Critical Review and Proposed Amendments of Federal Rules of Evidence 801(d)(1)(A), 613, and 607" (1977), 75 Mich. L. Rev. 1565, at pp. 1565, 1575-76). The American Senate as well expressed the view that the oath was not necessary as a precondition. The following appears from the Senate record:

The requirement that the statement be under oath also appears unnecessary. Notwithstanding the absence of an oath contemporaneous with the statement, the witness, when on the stand, qualifying or denying the prior statement, is under oath.

(Senate Report No. 93-1277, 93rd Cong., 2d Sess. (1974), reprinted in [1974] U.S. Cong. & Adm. News 7051, at p. 7062.)

Indeed Professor E. W. Clearly pointed out before the Senate committee that a rule requiring an oath could go far to destroy the utility of the proposed provision with regard to inconsistent statements. He stated:

... [it] would virtually destroy the utility of provision (A), which deals with prior inconsistent statements. If the witness has made a prior statement under oath, the threat of a perjury charge makes it highly unlikely that he will subsequently relate a different story again under oath. Hence the instances in which the rule would operate under the suggested redraft would be greatly curtailed. The problem area consists of cases in which the prior statement was not under oath, whether in the course of a judicial proceeding or not, and a rule which does not deal with these cases is of no practical significance.

(Graham, supra, at p. 1577, note 40.)

A House of Representatives bill would have required both contemporaneous cross-examination and an oath before an inconsistent statement could be admitted. As a result, a compromise was struck which reconciled the House and Senate bills by requiring that the oath be retained as a part of the rule in exchange for dropping the need for contemporaneous cross-examination.

Thus, in the United States, an oath is required at the time of the earlier statement despite the fact that the declarant, in the case of a prior inconsistent statement, must also be a witness before the court testifying under oath.

I think that the retention of the oath as a condition of admissibility should not be imported into Canada. This is especially true since neither the provisions of the Criminal Code nor the rationale of this Court in R. v. Boisjoly, 1971 153 (SCC), [1972] S.C.R. 42, could result in a conviction for perjury arising from a sworn declaration made in the course of a criminal investigation. In Boisjoly, police charged an individual with making a false statement under oath contrary to what was then Criminal Code, S.C. 1953-54, c. 51, s. 114. The police believed that Boisjoly had lied to them in a sworn statement which dealt with his involvement in the coercion of a witness at a preliminary inquiry. Fauteux C.J., for an unanimous Court, held that even if the affidavit were deceitful, the accused was not liable for making a false statement. In his opinion, s. 114 only applied to a person "permitted" by law to make sworn declarations. He found that "permitted" could not mean simply "not prohibited" since the legislative history demonstrated that the purpose of the section was to limit the circumstances in which oaths would be administered. Indeed, the equivalent section found in an earlier Canadian statute, S.C. 1874, c. 37, was headed Unnecessary Oaths Suppression. Fauteux C.J. was unable to find in any relevant legislation a provision empowering witnesses to make sworn statements to police. He concluded (at p. 51):

The legislator does not recognize as an ingredient of the indictable offence mentioned in s. 114, affidavits which he has not permitted, authorized or required, in short affidavits which have no legal meaning or scope.

At the time of that case, perjury was restricted to witnesses who lied in the course of a judicial proceeding (s. 112). Since then, the Criminal Law Amendment Acts of 1975 and 1985 (S.C. 1974-75-76, c. 93, s. 6, and S.C. 1985, c. 19, s. 17) have amalgamated ss. 112 and 114 and provide for one broader offence of perjury. The language considered in Boisjoly thereby finds its way into perjury under what is now s. 131(3). Section 114 stated that:

114. Every one who, not being a witness in a judicial proceeding but being permitted, authorised or required by law to make a statement by affidavit, by solemn declaration or orally under oath, makes in such a statement, before a person who is authorised by law to permit it to be made before him, an assertion with respect to a matter of fact, opinion, belief or knowledge, knowing that the assertion is false, is guilty of an indictable offence and is liable to imprisonment for fourteen years. [Emphasis added.]

This wording is now found at s. 131. Section 131(1) creates the offence of perjury and s. 131(3) provides:

(3) Subsection (1) does not apply to a statement referred to in that subsection that is made by a person who is not specially permitted, authorized or required by law to make that statement. [Emphasis added.]

The rationale of Boisjoly is therefore applicable to the current perjury provisions.

However, the 1975 amendments also created a new offence for persons, who while not specially permitted to do so, make sworn statements that are discovered to be false. This new offence is found at s. 134(1) of the Code. However it contains the following exception:

(2) Subsection (1) does not apply to a statement referred to in that subsection that is made in the course of a criminal investigation.

Thus it is not applicable to the type of statements under consideration in this case.

I note in passing that the decision in Boisjoly has been applied in a case involving a prior inconsistent statement. In Coulombe v. La Reine, [1976] C.A. 327, a witness (later charged with murder) gave to police, in the course of their investigation, a sworn statement and contradicted it at trial. The Quebec Court of Appeal held that the witness could not be found guilty of perjury for making the earlier statement.

With the greatest of respect for the contrary view, to make the oath one of the initial requirements for admissibility of a prior statement when no criminal consequences may be attached to the oath would seem to be an exercise in hollow formalism. In the case at bar, the requirement of administering the oath would render the prior statement inadmissible. Yet it was the prior statements that were true not the sworn testimony of the witnesses. The requirement would prevent the trier of fact from ever hearing the truth and would make impossible the rendering of a true verdict. I would think that the trier of fact, whether judge or jury, should be permitted to assess the truth of a prior statement provided it has met the requisite indicia of reliability in order to be admitted.

The facts of this very case make it abundantly clear that neither the taking of the oath nor the possibility of perjury charges resulting from their testimony can in themselves ensure that a witness will tell the truth. It follows that the administration of the oath should not be a prerequisite to the admissibility of a prior inconsistent statement. Rather it is the reliability of the statement that should determine its admissibility. That reliability will depend on a number of factors. One of these will be whether the witness as a result of making the statement may be subject to criminal prosecution for some offence (not just but including perjury) if the statement is false.

Alternative Sources of Criminal Liability

A prior inconsistent statement could become admissible as evidence if the witness may be subject to criminal prosecution if that prior statement were false.

A witness giving a deliberately false statement to the police may be subject to prosecution for a number of offences other than perjury. They include fabricating evidence, contrary to s. 137 of the Criminal Code, R.S.C., 1985, c. C-46, which provides a maximum of 14 years; obstructing justice, contrary to s. 139(2) providing a maximum of 10 years of imprisonment or public mischief, contrary to s. 140, with a maximum penalty of 5 years. As an example, of these offences ss. 139(2) and 140 provide:

139. ...

(2) Every one who wilfully attempts in any manner other than a manner described in subsection (1) to obstruct, pervert or defeat the cause of justice is guilty of an indictable offence and liable to imprisonment for a term not exceeding ten years.

140. (1) Every one commits public mischief who, with intent to mislead, causes a peace officer to enter on or continue an investigation by

(a) making a false statement that accuses some other person of having committed an offence;

(b) doing anything intended to cause some other person to be suspected of having committed an offence that the other person has not committed, or to divert suspicion from himself;

(c) reporting that an offence has been committed when it has not been committed; or

(d) reporting or in any other way making it known or causing it to be made known that he or some other person has died when he or that other person has not died.

In 1985, s. 140 was expanded, R.S.C., 1985, c. 27 (1st Supp.), s. 19 (formerly S.C. 1985, c. 19, s. 19), to include falsehoods that cause police to continue investigations as well as those falsehoods that cause investigations to be initiated. Unlike the offence of perjury, there is no requirement for corroboration in order to establish these offences. These offences would, I think, be as well known to members of the community, including prospective witnesses as that of perjury. Indeed, it is apparent that the Crown is more likely to prosecute under one of these provisions than it is to proceed with a charge for perjury. It is the existence of these offences and the possibility of prosecution for their breach that goes far in assuring the truthfulness and reliability of the statement.

It is true that a witness has no legal obligation to assist the police in their investigation. See the reasons of Martin J.A. in R. v. Moran (1987), 1987 124 (ON CA), 36 C.C.C. (3d) 225 (Ont. C.A.). Yet once a witness does speak to the police in the course of their investigations, they must not mislead the investigating authorities by making statements that are false. The right to say nothing cannot protect a witness from the consequences of deliberately making a false statement. See, for example, R. v. L.S.L. (1991), 1991 7820 (SK QB), 89 Sask. R. 267 (Q.B.).

There are a great many cases where persons have been prosecuted for obstruction of justice as a result of giving one version of the facts to the police, while offering a different version when testifying as a witness. In some of the cases, the factual situation was very similar to that presented in this case. See, for example, R. v. Gravelle (1952), 1952 345 (ON CJ), 103 C.C.C. 250 (Ont. Mag. Ct.), and R. v. MacGillivray (1971), 1971 1886 (PE SCTD), 3 Nfld. & P.E.I.R. 227 (P.E.I. Co. Ct.). There are a number of cases where there has been a prosecution for obstructing justice, fabricating evidence or for public mischief as result of a witness falsely accusing another of a crime. This list is put forward simply as an example of the number of cases where this type of prosecution has been brought: R. v. Feger (1989), 36 Q.A.C. 26; R v. J.(J.) (1988), 1988 7099 (ON CA), 65 C.R. (3d) 371 (Ont. C.A.), leave to appeal to the S.C.C. refused, [1989] 1 S.C.R. ix; R. v. Stapleton (1982), 1982 3331 (ON CA), 66 C.C.C. (2d) 231 (Ont. C.A.); R. v. Howard (1972), 1971 370 (ON CA), 7 C.C.C. (2d) 211 (Ont. C.A.); Lessard v. La Reine, [1965] Que. Q.B. 631; R. v. Sevick (1930), 1930 478 (NS SC), 54 C.C.C. 92 (N.S.S.C. en banc), and at the trial level: R. v. Edwards (1986), 1986 2885 (SK QB), 47 Sask. R. 303 (Q.B.); R. v. Verma (1980), 28 A.R. 233 (Q.B.); R. v. Lindstrom (1977), 1977 3211 (NS SC), 33 N.S.R. (2d) 369 (Co. Ct.); R. v. Martin (1969), 12 Crim. L.Q. 201 (B.C. Prov. Ct.); and R. v. Snider (1953), 1953 797 (ON CJ), 17 C.R. 136 (Ont. Mag. Ct.).

The number of these cases seems to make it questionable that it is essential that a warning should be given to a prospective witness of the possibility of prosecution. A person who gives a false statement to the investigating authorities causes the police to spend time in futile investigation of the false allegations. This obviously entails the expenditure of both a great deal of police time and the expenditure of public funds. Indeed by directing the attention of the police to others the real perpetrator of the crime may escape. Thus the rationale behind the offence of obstructing justice makes good common sense and can be readily understood by members of the community. Canadians would accept and indeed expect that witnesses who make deliberately false statements to the police would be subject to criminal prosecution.

This is such an obvious proposition and would appear to be such a part of the common knowledge of the community that there is no more need to warn persons that they may be liable to prosecution if the statements they give to police are false than to warn them of charges of perjury if their sworn statements are false. It follows that, although it would clearly be preferable to give a warning to a witness of the possibility of criminal proceedings if that person gives a false statement, it may not be essential to give such a warning in order to render the statement admissible.

It seems clear that criminal prosecutions for offences other than perjury are frequently brought against witnesses who make false statements. The consequences of a conviction for these offences are serious. I cannot find that there is a significant difference between the penalty for perjury for a witness lying at trial and the penalty for offences such as obstructing justice. The less stringent requirements for prosecution for obstructing justice or for public mischief (which unlike perjury do not require corroboration) and the more frequent prosecutions for these offences would, I think, have a deterrent effect upon those who might be prone to make false statements. Further since liability for these offences does not rest in any way on the oath, the requirement that it be administered could have no practical effect.

In summary, the administering of an oath or solemn affirmation should no longer be regarded as an essential safeguard for ensuring the veracity of a statement. Neither the threat of divine retribution resulting in eternal damnation nor a prosecution for perjury can guarantee truthfulness. If the witness giving the statement is subject to criminal prosecution for making a deliberately false statement, that will serve as a deterrent to mendacity and encouragement to veracity. A requirement that the police administer an oath to the witness cannot further deter the witness from lying and would seem to be superfluous. The absence of the oath should not presumptively stand in the way of the admissibility for substantive purposes of a prior inconsistent statement.

B. Absence of Contemporaneous Cross-examination

It is true also that at the time of the statement the witness was not subjected to the rigours of cross-examination. However, if the prior statement is ruled admissible then the witness will be subject to cross-examination at trial and the trier of fact will be able to study the demeanour of the witness throughout the testimony and can assess what weight if any should be attached to all the evidence of the witness including the prior statement. Two diametrically opposed views may be taken of the prior inconsistent statement. First it might be thought that the earlier statement resulted from coercion or pressure brought to bear on a witness by the investigating authorities. Yet it may just as well be concluded that the courtroom recantation arose from pressure brought upon the witness prior to the trial by the accused, his friends or associates. Either of these scenarios may be fully explored in examination and cross-examination and assessed by the trier of fact.

Alternatively the court may be confronted with a witness who forgets the earlier statement and the events related in that statement. Once again assuming that there is a complete and reliable transcript of the earlier statement, the nature of the memory loss can be explored at trial with the sworn witness. The voir dire procedure which will be outlined later will provide protection for the accused and the witness. If the prior statement has met the other conditions of reliability that I will refer to, then it should be admitted, for it may better enable a verdict to be reached on the basis of the truth derived from all the relevant evidence. A true verdict reached upon properly admissible evidence must be the goal of all court proceedings.

C. The Lack of Evidence as to Demeanour

Those who have argued against the admissibility of a prior inconsistent statement have put forward the position that the trier of fact is unable to see the witness at the time the statement was made. It is the lack of opportunity to observe the witness that is the basis for the traditional reticence of appellate Courts to interfere with the findings of fact and especially the findings of credibility made by triers of fact. Thus deprived of the ability to observe the demeanour of the witness, the trier will be unable to reach any conclusion as to the truth of the prior statement.

There is some inconsistency in this position for the witness at the time of trial is indeed before the trier of fact. At that time a conclusion can be reached as to which of the two statements is true. That conclusion will no doubt be based in part upon the in court demeanour of the witness.

The refusal to admit a prior inconsistent statement based on the inability of the jury to observe the witness at the time the prior statement was made, has been considered and rejected by the United States Supreme Court. In California v. Green, 399 U.S. 149 (1970), the United States Supreme Court considered a California law authorizing the substantive use of all prior inconsistent statements. It was argued that this legislation violated the right of the accused to confront all the witnesses called against him guaranteed by the Sixth Amendment. The majority of the court concluded that the legislation did not infringe the constitutional right. On this issue the majority stated (at pp. 160 and 188):

... some demeanor evidence that would have been relevant in resolving this credibility issue is forever lost.

[But] the witness may be examined at trial as to the circumstances of memory, opportunity to observe, meaning, and veracity.... I think it fair to say that the fact that the jury has no opportunity to reconstruct a witness' demeanor at the time of his declaration, and the absence of oath are minor considerations.

The same conclusion had been reached earlier by Learned Hand J. in Di Carlo v. United States, 6 F.2d 364 (2d Cir. 1925). He observed at p. 368 that the recanting witness:

... is present before the jury, and they may gather the truth from his whole conduct and bearing, even if it be in respect of contradictory answers he may have made at other times.... If, from all that the jury see of the witness, they conclude that what he says now is not the truth, but what he said before, they are none the less deciding from what they see and hear of that person and in court. [Emphasis added.]

Thus even without a videotape reproduction of the taking of the statement there is strong authority for concluding that the mere fact that the witness could not be observed when the original statement was made should not be a ground for making that statement inadmissible. The rights of the accused are protected by the cross-examination of the witness at the time of trial. It will often arise that, as a result of cross-examination, the witness will, as in this case, deny the truth of the earlier statement. No better result could be expected of the most brilliant cross-examination. It will then be for the trier of fact to determine which version is correct.

In any event, in this case we are dealing with videotaped evidence. The videotape shows the witness sitting at a table. In a corner of the screen there is a close-up of the witness's face. At the bottom of the picture there is a time clock sequence which ensures the integrity of the tape. The videotaping of the statement provides such a complete record not only of the questions posed and the answers given, but of the demeanour of the witness throughout the interview that little can be said to warrant its exclusion based on the inability to assess the demeanour of the witness when the statement was made.

In passing, I would observe that the videotape serves as well to monitor the conduct of the police during the interview. It goes far to ensuring that nothing untoward happened in the course of the interview. The usefulness of the videotape of an interview was aptly described by A. Heaton-Armstrong and D. Wolchover in "Recording Witness Statements", [1992] Crim. L.R. 160, at p. 169:

The more accurate and comprehensive the record of a statement the stronger the case becomes for introducing it as evidence of the facts contained in it. There is a world of difference between a conventional witness statement signed by the maker but written by a police officer and a video-recorded interview with a witness which will include the questions and other potentially important features such as witness's demeanour, in modern parlance non-verbal communication or, as it is popularly known "body language".

It may on occasion be cogently argued that police pressure was exerted upon the witness prior to the taping. That need not be considered in this case as parents, lawyers or family members were present with the witnesses from the time they came to the police station until the interview was completed.

The videotaped statement with its complete and comprehensive record of the questions posed, the answers given and the demeanour of the witness, will often serve as a complete answer to the issues of reliability and voluntariness of the statement. However, it is not essential that a statement be videotaped in order to be admissible. It may well be that other situations will be presented where a complete and comprehensive record will be preserved of the statement together with satisfactory evidence of the circumstances of the interview and the demeanour of the witness so that all the requirements on this count will be met. For example a recorded statement made by a witness in front of the police and independent, reliable persons may well meet all the criteria of reliability and be just as acceptable for purposes of admissibility as a videotaped statement. If the prior statement, while not videotaped, meets all the criteria of reliability it should be admissible.

The argument based on the inability to assess the demeanour of the witness at the time of the original statement should not be accepted as grounds for excluding the videotaped statement with which we are concerned.

D. Possibility of Coercion

The common law has a long history of resistance to the use of testimony that was compiled from the depositions and letters of persons who were never called to testify. There is a healthy fear that the authorities may be so anxious to secure a conviction that they will exert unwarranted pressure on a witness to obtain the statement which they desire. The rule against the substantive admissibility of a prior inconsistent statement had the salutary effect of allowing a witness to reconsider the earlier statements in court where the presence of a judge would overcome any perceived coercion by the police.

A new rule that would admit the statements if certain conditions are met should ensure that the prior statement of the witness was made voluntarily. It should still permit the witness to recant and to explain that the earlier statement resulted from police pressure or coercion. Yet it must be remembered that the police are not the only ones who can exert pressure on a witness. The accused, or friends and supporters of the accused can as well influence a witness to change a statement. This may be particularly true in drug cases and cases of spousal assault. Obviously, the incentive of an accused to have a witness change his version of events is substantial if the present rule is maintained since the retraction by itself can effectively end the prosecution's case. Frequently this will frustrate any possibility of the resolution of the charge on its merits.

The Supreme Court of Georgia observed in Gibbons v. State, 286 S.E.2d 717 (1982), that a rule allowing for the substantial admissibility of prior declarations means that (at p. 722):

... both sides are assured a measure of protection against efforts to influence the testimony of a witness, as the prior declaration is no longer effectively revocable at the will of the witness.

For the same considerations, witnesses are protected from improper attempts to influence testimony -- the potential gain from impropriety being diminished substantially by the adoption of this rule. [Emphasis in original.]

It will always be vitally important to ensure that there has been no impropriety or abuse of the power of the state in obtaining the prior statement. Thus the court must make certain that a statement of a witness that may be later used against the accused despite its repudiation of it by the witness at trial, has been freely and voluntarily made. This can be accomplished in much the same way that courts presently ensure that confessions were not improperly obtained. I would therefore recommend that before the prior statement can be found to be admissible under s. 9 of the Canada Evidence Act that a voir dire be held to ensure that the statement was freely and voluntarily made. In the case at bar, the trial judge found that the police conduct was "completely above-board, proper, non-oppressive, non-coercive". That I think is the proper standard to be applied in reviewing the statement.

At the conclusion of the voir dire the trial judge should be satisfied beyond a reasonable doubt that the statement was given voluntarily in an atmosphere that was as free of oppression as possible and did not result from the hope of reward or the fear of penalty.

IV. The Conditions for Admissibility Which Must be Satisfied on the Voir Dire

There are a number of matters that must be explored on the voir dire. It may be useful to set out at this point the manner in which it should be conducted and the issues it should resolve.

The voir dire should proceed in the usual way. As the Chief Justice noted, the prior statement must be otherwise admissible and not excluded by the operation of some other evidentiary rule such as those reviewed by him. In the case at bar, the prior inconsistent statements of the witnesses set out incriminating statements made by the respondent, which would ordinarily be admissible.

Upon the voir dire the trial judge must be satisfied beyond a reasonable doubt that the conditions for admitting the prior inconsistent statement have been fulfilled. I would suggest that the conditions are these:

(1) That the evidence contained in the prior statement is such that it would be admissible if given in court.

(2) That the statement has been made voluntarily by the witness and is not the result of any undue pressure, threats or inducements.

(3) That the statement was made in circumstances, which viewed objectively would bring home to the witness the importance of telling the truth.

(4) That the statement is reliable in that it has been fully and accurately transcribed or recorded.

(5) That the statement was made in circumstances that the witness would be liable to criminal prosecution for giving a deliberately false statement.

If those conditions are met then the prior inconsistent statement should be admitted for all purposes.

The statements to be considered on such a voir dire will fall somewhere along a continuum of reliability. For example, at one end it may be established that the statement had been videotaped; the witness had access to counsel; the witness had a relative or responsible person present throughout the interrogation; the witness had been made aware of the gravity of the investigation and the importance of telling the truth; and the witness gave the statement voluntarily. Statements made in those circumstances may be readily admitted. Other statements may still be found to be reliable and admissible although they do not present every one of the foregoing indicia of trustworthiness. For example, there may not be a videotape, but a recording of the interview made in the presence of dependable and reliable independent witnesses who attended the interview and who attest to the demeanour and deportment of the witness and the police. At the other end of the continuum may be a prior inconsistent statement contained in the hurried notes of a single police officer who was paraphrasing the words of the witness speaking to him in hectic and difficult circumstances. It might well be difficult to conclude that such a statement should be admitted.

A. Standard of Proof on the Voir Dire

The Chief Justice would decide the question of admissibility on a balance of probabilities. The jurisprudence on the burden to be satisfied during a voir dire is meagre.

This Court in deciding what standard of proof is to be adopted in particular circumstances -- beyond a reasonable doubt or on a balance of probabilities -- has taken a flexible approach. The standard selected is the result of a policy decision which is based upon a balancing of the interests of the accused and of the public which are at stake. See R. v. Oakes, 1986 46 (SCC), [1986] 1 S.C.R. 103, at pp. 137-38, and R. v. Mack, 1988 24 (SCC), [1988] 2 S.C.R. 903, at pp. 973-76. For example, the impact of a confession on a trial requires that its voluntariness be proved beyond a reasonable doubt before it is admitted: R. v. Pickett (1975), 1975 1428 (ON CA), 28 C.C.C. (2d) 297 (Ont. C.A.).

A confession is an inculpatory statement made by an accused to a person in authority. In the case at bar the admission of the respondent was made to his friends who were clearly not persons in authority. While the admissions were inculpatory, they were not confessions. The concern that statements were forced out of the accused by police is simply not present in this case.

However, the witnesses when they attended at the police station were in a sense under police control. Therefore, in the interests of fairness and uniformity, the same high standard should be applied to all prior statements. It would not only be in the best interest of the accused but also ensure the fairness of the trial process if the judge presiding over the voir dire were required to be satisfied beyond a reasonable doubt as to all the aspects of the statement. If the reliability of the prior statement is so tenuous or marginal that only the civil standard would secure its admissibility, it is not the type of evidence that should be placed before the trier of fact in a criminal trial.

The Chief Justice leaves open the question whether a confession by an accused that forms part of a prior inconsistent statement should be admissible on a balance of probabilities. In my view, it matters little whether the confession is directly admitted or goes before the jury in the form of a prior inconsistent statement -- in both cases the test to be met on the voir dire should be proof beyond a reasonable doubt.

The judge conducting the voir dire should not weigh the credibility of the statement any more than would the judge conducting a voir dire concerning a confession determine whether it is truthful. The voir dire simply ensures that the circumstances surrounding the making of the statement are such, that it should be admitted. It is for the jury to determine whether the prior statement should be accepted as true.

B. Instructions to the Jury

If at the conclusion of the voir dire the prior inconsistent statement is ruled admissible for all purposes then at some time the trial judge should give instructions to the jury regarding the statement.

The jury should be advised that although the statement has been ruled admissible it is up to them to decide what weight if any they should attach to it. The jury should be instructed that they may consider that the statement should be given less weight because it was not subject to cross-examination at the time it was made and because there was not the same opportunity to assess the demeanour of the witness as there would have been had the statement been made in court. Nonetheless, the statement is to be treated like any other admissible evidence. It can be accepted as the truth of what it relates, it can be accepted in part and rejected in part; or it can be rejected completely. In assessing the statement the jury should consider all the circumstances in which it was made.

V. The Test in Smith and Khan

Before giving a demonstration of how a voir dire in this case might proceed, I should touch upon the second facet of the Chief Justice's reasons which deals with statements where one, several, or all of the three initial preconditions (oath, warning and videotape) are missing. For the Chief Justice when such a situation arises, "alternatively" in "exceptional circumstances" the principles enunciated in Smith and Khan should apply. That is to say the judge, on the voir dire, is to search for substitute indicia of reliability.

There is no doubt that these cases provide an alternative justification for changing this Court's oft-criticized ruling in Deacon, supra. I have no quarrel with utilizing the approach to hearsay evidence set out in Khan and Smith to prior inconsistent statements provided the threshold tests for reliability I have suggested are made a part of it.

VI. Application of the Conditions of Admissibility to the Case at Bar

Let us see how these suggested conditions of admissibility could be applied to this case. I do so not to reach any conclusion on the question of admissibility but to demonstrate how the analysis I have proposed might be applied in this case.

(1) Otherwise Admissible

The evidence contained in the prior statements would be admissible if given in court.

(2) Voluntariness

In this case each witness was advised that he had a right to counsel. Each witness had a lawyer or a parent or sibling or a combination of those persons present during the interrogation. The videotape of the interviews certainly did not disclose evidence of any impropriety on the part of the police. The trial judge congratulated the police for their sensitive handling of the matter. Indeed the police did act in an exemplary manner. They scrupulously and carefully advised the witnesses of their rights, the nature of the investigation and the importance of telling the truth. There is no evidence that undue influence was exerted upon the witnesses, nor were they threatened or offered a reward of any kind. It follows that it would be open to the judge presiding at the voir dire to find that the prior statements were made voluntarily.

(3) The Importance of Telling the Truth

All witnesses attended at the police station with family members and one with a lawyer as well. They were told that the police were investigating the death by stabbing of Joseph Wright. They were all advised of the importance of telling the truth. The police went so far as to tell one witness of the possibility of criminal charges being brought against him if he was not telling the truth. When all the circumstances are considered, it would seem that it should have been apparent to the witnesses that it was important to tell the truth. Certainly it would be open to a trial judge sitting on a voir dire to reach such a conclusion.

(4) Reliability of the Statement

The videotape demonstrates the reliability of the statement. It is a complete and accurate recording of the questions posed and the answers given. It reveals the tone of voice and the facial expression of the witnesses. The tape provides such a complete and reliable record of the statement and the circumstances in which it was given that it would be open to the judge conducting the voir dire to find the statements were reliable.

(5)Witness Subject to Criminal Prosecution if a Deliberately False Statement Was Given

All the witnesses were subject to criminal prosecution for obstructing justice or public mischief if their statements were deliberately false. One witness was advised of this in general terms. All were advised of the importance of telling the truth. Here the statements were made in circumstances that could give rise to criminal prosecution of the witnesses if their statements were deliberately false.

In this case, it would be open to a judge conducting a voir dire at a new trial to find that the prior inconsistent statements met all the conditions for admissibility.

VII. New Trial

The respondent argued that if it were found that the prior inconsistent statements were admissible for all purposes that there would still not be sufficient evidence to convict the accused. Reliance was placed upon Vézeau v. The Queen, 1976 7 (SCC), [1977] 2 S.C.R. 277. I cannot agree with that submission. The words of the trial judge on this issue are significant. He stated that if the witnesses' statements to the police could be considered

there would be substantial evidence to confirm the identification evidence given earlier by Steven Wright, and the issue of the accused's use of the knife might be resolved against him.

This clearly indicates that on a new trial, if the prior inconsistent statements were found to be admissible, the result could well be different, that is to say the verdict would not necessarily be the same.

It must be remembered that a young man died as a result of a senseless stabbing. The community has a real and pressing interest in having the guilt or innocence of the respondent established on the basis of the truth. The earlier acquittal was based upon perjured evidence and upon an interpretation of s. 9 of the Canada Evidence Act that withheld the truth -- in the form of the prior statements -- from the consideration of the trial judge. Those statements might well have been considered. A trial must always be a quest to discover the truth. Irrational and unreasonable obstacles to the admission of evidence should not impede that quest. In order to reach a true verdict a court must be able to consider all the relevant admissible evidence. It is only on a new trial, when a court has had the opportunity to consider anew the admissibility of the prior statements, that an informed decision may be reached based upon all the admissible evidence.

VIII. Disposition

In the result, I would allow the appeal, set aside the acquittal and the order of the Court of Appeal upholding the acquittal and direct a new trial.

Appeal allowed and new trial ordered.

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

Solicitor for the respondent: Keith E. Wright, Toronto.