R. v. Hodgson, [1998] 2 SCR 449

R. v. Hodgson, [1998] 2 S.C.R. 449

Michael Colin Hodgson                                                                     Appellant

v.

Her Majesty The Queen                                                                   Respondent

and

The Attorney General of Canada,

the Attorney General of Quebec,

the Attorney General of British Columbia

and the Attorney General for Alberta                                              Interveners

Indexed as:  R. v. Hodgson

File No.:  25561.

1998:  March 24; 1998:  September 24.

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

on appeal from the court of appeal for ontario

Criminal law ‑‑ Evidence ‑‑ Confessions ‑‑ Accused admitting crime to complainant and her parents when confronted by them ‑‑ Accused held at knife-point after making statement ‑‑ Out-of-court statement admitted without voir dire and without objection by defence counsel ‑‑ Extent to which person in authority requirement should remain part of confessions rule -- Whether voir dire should have been held to determine whether statements made to person in authority and whether statements made voluntarily ‑‑ Whether onus always rests on defence to request voir dire ‑‑ If not, circumstances under which trial judge should hold voir dire ‑‑ Evidence which triggers trial judge’s obligation to hold voir dire.

The trial judge admitted into evidence certain out‑of‑court statements allegedly made by the accused and convicted him of sexual assault.  The complainant and her parents had confronted the accused at work and all testified that he admitted the sexual assaults.  The complainant’s mother then went to call the police and, on her return, struck the accused.  At some point after receiving the statement, the complainant’s father held the accused at knife-point, allegedly to prevent him from leaving before the police arrived.  At trial, the accused denied making a confession but testified that he was neither frightened nor threatened during the confrontation.  His counsel raised no objection to the admission of the out-of-court statement at trial.

At issue is whether the trial judge erred in failing to direct a voir dire of his own motion to determine whether the statements made to the complainant and her family were given to persons in authority and if so, whether the statements were made voluntarily.  Several subsidiary issues required consideration.  First, does the onus always rest with the defence to request a voir dire to test the voluntariness of an accused’s out‑of‑court statements?  If not, when and under what circumstances should a trial judge hold a voir dire of his or her own motion?  Further, is the trial judge’s obligation to hold a voir dire triggered only where the receiver of the confession is a “conventional” person in authority, or should the obligation be construed more broadly?  Lastly, to what extent should the “person in authority” requirement remain part of the confessions rule?

Held:  The appeal should be dismissed.

Per Lamer C.J. and Gonthier, Cory, McLachlin, Iacobucci, Major and Binnie JJ.:  A statement made by an accused to a person in authority must be made voluntarily and must be the product of an operating mind.  This rule is based upon two fundamentally important concepts:  the need to ensure the reliability of the statement and the need to ensure fairness by guarding against improper coercion by the state.  This results in the requirement that the admission must not be obtained by either threats or inducements.  The person in authority requirement is grounded in the underlying rationales for the confessions rule and should remain part of the rule. 

The definition of “person in authority” typically refers to those formally engaged in the arrest, detention, examination or prosecution of the accused and so applies to police officers and prison officials or guards.  When an accused makes a statement to a police officer or prison guard a voir dire should be held to determine its admissibility as a voluntary statement, unless the voir dire is waived by the defence.  In addition, those persons whom the accused reasonably believes are acting on behalf of the state and could therefore influence or control the proceedings against him or her may also be persons in authority.  That issue must be resolved by considering it from the viewpoint of the accused.  The receiver’s status as a person in authority arises only if the accused has knowledge of that status.  In addition, there must be a reasonable basis for the accused’s belief that the person hearing the statement was a person in authority.  This issue will not normally arise in relation to undercover police officers because, from the accused’s viewpoint, they will not usually be viewed as persons in authority.

The defence must raise the “person in authority” issue with the trial judge.  This is appropriate for it is only the accused who can know that the statement was made to someone regarded by the accused as a person in authority.  On the ensuing voir dire, the accused will have the evidential burden of demonstrating that there is a valid issue for consideration.  If the accused meets the burden, the Crown will then have the persuasive burden of demonstrating beyond a reasonable doubt that the receiver of the statement was not a person in authority, or if it is found that he or she was a person in authority, that the statement was made voluntarily.

In extremely rare cases, the evidence adduced during a trial, viewed objectively, may indicate that the issue as to whether the receiver of a statement of the accused was a person in authority should be explored by way of voir dire.  In those cases, the trial judge must of his or her own motion direct a voir dire.  Evidence which clearly demonstrates that the receiver of the statement was closely connected to the authorities should alert the trial judge to the need for a voir dire on the issue.  This evidence progresses along a spectrum.  Where the receiver of the statement is a “conventional” person in authority, such as a police officer or prison guard, the trial judge must clearly proceed to a voir dire to test the voluntariness of the statement.  Similarly, where the evidence discloses a close connection between the receiver and the authorities such that, if known to the accused, he or she could reasonably have believed that the receiver was acting as an agent of the prosecuting authorities, the trial judge should inquire whether the defence is prepared to discharge its evidential burden on the person in authority issue or whether it waives a voir dire on this issue.  However, the further away the receiver of the statement is from the conventional person in authority, the less likely it will be that the evidence will alert the trial judge to the need to hold a voir dire and the greater the obligation of the accused to raise the issue.

If the trial judge is satisfied that the receiver of the statement was not a person in authority but that the statement of the accused was obtained by reprehensible coercive tactics such as violence or credible threats of violence, then a direction should be given to the jury.  The jury should be instructed that if they conclude that the statement was obtained by coercion, they should be cautious about accepting it, and that little if any weight should be attached to it.

The evidence at trial did not disclose any evidence that was sufficient to trigger the trial judge’s obligation to hold a voir dire.  When the statements were admitted into evidence, there was nothing to suggest that the complainant or her family members had spoken to the police or anyone else in authority or were even considering making a complaint.  Similarly, there was nothing to suggest that the accused subjectively believed the complainant’s family to have control over criminal proceedings. The trial judge properly used the evidence of the confession to confirm the complainant’s testimony.

Per L’Heureux-Dubé and Bastarache JJ.: The result of Cory J. was agreed with but his test with respect to the proper interpretation of “persons in authority” was not agreed with. 

The confessions rule is a specific and delimited exception to the basic rule that it is for the trier of fact to assess the truth of statements in all the surrounding circumstances, as well as the recognition that statements made by the accused against his or her interest are inherently reliable. Interpretation of its integral concept of “person in authority” must be bound by applicable policy rationales which justify the confessions rule’s exclusion of relevant evidence.  Essentially where policy concerns  historically related to the confession’s reliability, modern rationales of the confessions rule now focus on state conduct and individual fairness.  Basic fairness in the criminal process unifies the two rationales of reliability and state deterrence, and questions of truth have given way to a focus on voluntariness.  The confessions rule is only concerned with voluntariness, however, where statements are made to persons in authority.  Indeed, the general unfairness of utilizing involuntary statements resulting from private coercion has never been the focus of the confessions rule, even when the rule was justified by policy concerns for reliability.  Its modern rationales explicitly affirm that the rule is concerned only with voluntariness within the relationship between the state and the individual.

As such, legislative change to the confessions rule altering the requirement of a person in authority is neither appropriate nor desirable.  The best rule is still that, once it is accepted that the confession of the accused was not made to a person in authority, it is properly admissible without any requirement for the Crown to establish that it was voluntary. Strict retention of this rule promotes clarity, is consistent with modern rationales justifying the exclusion of relevant evidence and serves the general public interest by facilitating the pursuit of truth. The “person in authority” requirement is, and should remain, the pivotal concept in the confessions rule.

The proper test for  “person in authority” examines first the objective status of the person to whom the confession or statement was made, and only where they are identified as someone formally engaged in the arrest, detention, interrogation or prosecution of the accused is it then necessary to examine whether the accused believed that the person could influence or control the proceedings against him or her.  Only in the rarest of cases will non-traditional persons in authority fall into this category, and it still must be objectively established on the facts that such persons had actual control in the proceedings.  This approach furthers the modern principles underlying the confessions rule, and does not erect additional unwarranted barriers to the admission of relevant evidence. It recognizes that a subjective approach to the test for “persons in authority” was primarily adopted to address the circumstance of undercover state agents.  The necessary prerequisite, however, remains an actual person in authority or “instrumentality” of the state.

As a general rule, the defence has the onus of  requesting a voir dire,  raising the issue that the accused’s statement was made to a “person in authority” and thus putting the statement’s admissibility into question.  In exceptional circumstances, the trial judge may have the duty to conduct a voir dire where the evidence reveals the realistic potential that the admission was made to a representative of the state and the accused might objectively have known of this authority.  Only evidence available on the record prior to the admission of the accused’s statement is relevant to the determination of what the trial judge’s duty was.   This duty will automatically arise in cases of obvious persons in authority, subject to informed waiver by the accused, since in such cases it is reasonable to infer the accused’s knowledge of this authority.  In the case of non-obvious persons in authority, however, the evidence must reveal both the reasonable possibility that the person was an agent of the state and that the accused could have known of this status.

Here, there was no realistic possibility that the complainant and her immediate family constituted persons in authority for purposes of the confessions rule, and therefore the trial judge had no duty to host a voir dire.

Cases Cited

By Cory J.

Distinguished:  Thongjai v. The Queen, [1998] A.C. 54; considered:  Erven v. The Queen, [1979] 1 S.C.R. 926; referred to:  R. v. Wells, [1998] 2 S.C.R. 517; R. v. A.B. (1986), 26 C.C.C. (3d) 17; Ibrahim v. The King, [1914] A.C. 599; Prosko v. The King (1922), 63 S.C.R. 226; Boudreau v. The King, [1949] S.C.R. 262; Ward v. The Queen, [1979] 2 S.C.R. 30; Horvath v. The Queen, [1979] 2 S.C.R. 376; R. v. Warickshall (1783), 1 Leach 263, 168 E.R. 234; DeClercq v. The Queen, [1968] S.C.R. 902; R. v. Buric (1996), 28 O.R. (3d) 737, aff’d [1997] 1 S.C.R. 535; R. v. Charemski, [1998] 1 S.C.R. 679; Commissioners of Customs and Excise v. Harz, [1967] 1 A.C. 760; R. v. Sang, [1979] 2 All E.R. 1222; Piché v. The Queen, [1971] S.C.R. 23; Rothman v. The Queen, [1981] 1 S.C.R. 640; R. v. Hebert, [1990] 2 S.C.R. 151; R. v. S. (R.J.), [1995] 1 S.C.R. 451; R. v. Whittle, [1994] 2 S.C.R. 914; Deokinanan v. R., [1968] 2 All E.R. 346; Watkins v. Olafson, [1989] 2 S.C.R. 750; R. v. Salituro, [1991] 3 S.C.R. 654; Bow Valley Husky (Bermuda) Ltd. v. Saint John Shipbuilding Ltd., [1997] 3 S.C.R. 1210; Winnipeg Child and Family Services (Northwest Area) v. G. (D.F.), [1997] 3 S.C.R. 925; R. v. Todd (1901), 4 C.C.C. 514; R. v. Roadhouse (1933), 61 C.C.C. 191; R. v. Berger (1975), 27 C.C.C. (2d) 357; R. v. Trenholme (1920), 35 C.C.C. 341; R. v. Wilband, [1967] S.C.R. 14; R. v. Downey (1976), 32 C.C.C. (2d) 511; R. v. Sweryda (1987), 34 C.C.C. (3d) 325; R. v. Scott (1984), 1 O.A.C. 397; Morris v. The Queen, [1979] 2 S.C.R. 1041; R. v. McKenzie, [1965] 3 C.C.C. 6; R. v. Postman (1977), 3 A.R. 524; R. v. Sweezey (1974), 20 C.C.C. (2d) 400.

By L’Heureux-Dubé J.

Considered:  Rothman v. The Queen, [1981] 1 S.C.R. 640; referred to:  Ibrahim v. The King, [1914] A.C. 599; R. v. Khan, [1990] 2 S.C.R. 531; D.P.P. v. Ping Lin, [1976] A.C. 574; R. v. Seaboyer, [1991] 2 S.C.R. 577; Hardy’s Trial (1794), 24 State Tr. 199; Clarkson v. The Queen, [1986] 1 S.C.R. 383; R. v. Whittle, [1994] 2 S.C.R. 914; Lego v. Twomey, 404 U.S. 477 (1972); R. v. Paonessa (1982), 66 C.C.C. (2d) 300, aff’d [1983] 1 S.C.R. 660; R. v. A.B. (1986), 26 C.C.C. (3d) 17, leave to appeal refused [1986] 1 S.C.R. v; R. v. Stewart (1980), 54 C.C.C. (2d) 93; R. v. Fowler (1982), 4 C.C.C. (3d) 481; R. v. Collins (1975), 29 C.C.C. (2d) 304; R. v. Todd (1901), 4 C.C.C. 514; R. v. McIntyre (1993), 135 N.B.R. (2d) 266, aff’d [1994] 2 S.C.R. 480; R. v. Hebert, [1990] 2 S.C.R. 151; R. v. Amyot (1990), 58 C.C.C. (3d) 312; R. v. Frewin (1855), 6 Cox C.C. 530; R. v. McKenzie, [1965] 3 C.C.C. 6; R. v. Sweryda (1987), 34 C.C.C. (3d) 325; R. v. Trenholme (1920), 35 C.C.C. 341; R. v. Kyle (1991), 68 C.C.C. (3d) 286; Wilband v. The Queen, [1967] S.C.R. 14; R. v. Unger (1993), 83 C.C.C. (3d) 228; R. v. Corbett, [1988] 1 S.C.R. 670; R. v. Lomage (1991), 2 O.R. (3d) 621; R. v. Sweezey (1974), 20 C.C.C. (2d) 400; Erven v. The Queen, [1979] 1 S.C.R. 926; R. v. Wells, [1998] 2 S.C.R. 517; R. v. Pettipiece (1972), 7 C.C.C. (2d) 133; Powell v. The Queen, [1977] 1 S.C.R. 362; Park v. The Queen, [1981] 2 S.C.R. 64.

Statutes and Regulations Cited

Evidence Act 1995, 1995 (Australia), No. 2, s. 84.

Evidence Act 1995, 1995 (N.S.W.), No. 25, s. 84.

Police and Criminal Evidence Act 1984, 1984 (U.K.), c. 60, s. 76.

Authors Cited

Berger, Mark.  “The Exclusionary Rule and Confession Evidence:  Some Perspectives on Evolving Practices and Policies in the United States and England and Wales” (1991), 20 Anglo‑Am. L. Rev. 63.

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

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

Gillies, Peter.  Law of Evidence in Australia, 2nd ed. Sydney:  Legal Books, 1991.

Herman, Lawrence.  “The Unexplored Relationship Between the Privilege Against Compulsory Self‑Incrimination and the Involuntary Confession Rule (Part I)” (1992), 53 Ohio St. L.J. 101.

Kaufman, Fred.  The Admissibility of Confessions, 3rd ed.  Toronto:  Carswell, 1979.

McWilliams, Peter K. Canadian Criminal Evidence, vol. 2, 3rd ed.  Aurora, Ont.: Canada Law Book, 1988 (loose-leaf updated April 1998).

Sopinka, John, Sidney N. Lederman and Alan W. Bryant.  The Law of Evidence in Canada.  Toronto:  Butterworths, 1992.

United Kingdom.  Criminal Law Revision Committee.  Eleventh Report, Evidence (General).  Cmnd. 4991.  London:  H.M.S.O., 1972.

APPEAL from a judgment of the Ontario Court of Appeal (1996), 91 O.A.C. 298 (sub nom. R. v. M.C.H.), 107 C.C.C. (3d) 327, [1996] O.J. No. 2366 (QL), dismissing an appeal from conviction by Paisley J. sitting with jury.  Appeal dismissed.

Irwin Koziebrocki, for the appellant.

Ian R. Smith, for the respondent.

S. David Frankel, Q.C., for the intervener the Attorney General of Canada.

Joanne Marceau and Jacques Gauvin, for the intervener the Attorney General of Quebec.

John M. Gordon, for the intervener the Attorney General of British Columbia.

Written submission only by Martin W. Mason, for the intervener the Attorney General for Alberta.

//Cory J.//

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

1                                   Cory J. ‑‑ The same issues must be resolved in both this appeal and that of R. v. Wells, [1998] 2 S.C.R. 517.

2                                   In both cases, defence counsel did not request a voir dire to test the voluntariness of certain out‑of‑court statements allegedly made by the accused  and, as a result, the statements were admitted into evidence.  The appellants contend that the trial judge erred in failing to direct a voir dire of his own motion to determine whether the statements were given to a person in authority and if so, whether they were made voluntarily.

3                                   In order to determine whether the trial judge erred it is necessary to consider several subsidiary issues.  First, does the onus always rest with the defence to request a voir dire to test the voluntariness of an accused’s out‑of‑court statements?  If not, when and under what circumstances should a trial judge hold a voir dire of his or her own motion?  Further, is the trial judge’s obligation to hold a voir dire triggered only where the receiver of the statement is a “conventional” person in authority, or should the obligation be construed more broadly?  Lastly, to what extent should the “person in authority” requirement remain part of the confessions rule?

I.  Factual Background

4                                   The appellant was a friend of the complainant’s family and occasionally babysat the complainant and her siblings.  The complainant, who was sixteen years old at the time of trial,  testified that commencing when she was approximately seven or eight years of age and continuing until she was approximately eleven years of age, the appellant sexually assaulted her on several occasions.  The complainant testified that she never told anyone about the incidents because she was afraid and because the appellant told her that she would get in trouble if she did.

5                                   The complainant testified that in 1993, she finally told her mother about these incidents.  When the allegations were revealed, the complainant, her mother, her father, and her stepfather went to the appellant’s place of employment and confronted him.  They all testified that the appellant confessed to having sexually assaulted the complainant on several occasions, that the appellant had said he was sorry, and that he had said he “knew it would catch up with him”.  The complainant’s mother went to call the police, and when she returned she struck the appellant.  At some point, the complainant’s father pulled out a knife and held it to the appellant’s back.  The father, stepfather and mother testified that the father pulled the knife after the appellant confessed in order to prevent the appellant from leaving before the police arrived.

6                                   At trial, the appellant testified that he was confronted at work by the complainant and her family about the sexual assaults, but he denied making a confession.  He stated that he was stunned, shocked and upset by the confrontation and did not want the situation to get blown out of proportion, but that he was neither frightened nor threatened during the confrontation.

7                                   At trial, the appellant raised no objection to the admission of the confession evidence.  The trial judge relied on this evidence and convicted the appellant.

II.  Judgments Below

A.  Ontario Court (General Division) (Paisley J.)

8                                   The trial judge noted that the appellant had previously been of good character, and he found that this enhanced the appellant’s credibility.  Nevertheless, the trial judge found the confession evidence “overwhelming”, notwithstanding the appellant’s denial.  Paisley J. also found the testimony of the complainant to be credible.  The trial judge recognized that even if he disbelieved the evidence of the appellant, he was still bound to assess the weight of the Crown’s evidence to see whether or not it had proven its case beyond a reasonable doubt.  He concluded that the Crown had discharged its burden, and that there was no rational basis for coming to any other conclusion than that the appellant was guilty of the acts of sexual assault.  The appellant was convicted and sentenced to four and a half years’ imprisonment.

B.  Court of Appeal (1996), 91 O.A.C. 298 (Finlayson J.A. for the court)

9                                   Finlayson J.A. relied on R. v. A.B. (1986), 26 C.C.C. (3d) 17 (Ont. C.A.), for the proposition that in some circumstances, persons other than those engaged in the arrest, detention, examination or prosecution of an accused may be considered persons in authority.  In particular, he noted that the court in A.B., supra, found that in certain circumstances, the parent of an infant complainant may be considered to be a person in authority.

10                               However, Finlayson J.A. noted that to determine whether some or all of the persons confronting the appellant in this case were persons in authority would require a careful factual investigation, and found that it would not be proper to decide the matter on the limited record available.  Moreover, he held that when a statement is made to a person not ordinarily engaged in the arrest, detention, examination or prosecution of an accused person, the defence has a responsibility to raise the issue at trial and ask for a voir dire to determine whether the receiver of the confession was a person in authority.  Finlayson J.A. found that the defence bears this responsibility because the subjective reaction of the accused to his or her interrogator is essential to the ultimate determination of voluntariness.  Absent the presence of a conventional authority figure at the time when an inculpatory statement is made, the trial judge must be placed on notice that the voluntariness of the statement is a live issue.  Thus the trial judge did not err in failing to hold a voir dire to test the voluntariness of the appellant’s confession.

11                               Finlayson J.A. held that in this case, the question as to whether any or all of the family members confronting the appellant were persons in authority was not a live issue.  The admissibility of the statement was not challenged, and the only issue at trial was the weight to be given to it.  The trial judge gave it considerable weight, and found it to be important, as he was entitled to do.  The trial judge found that the complainant was credible, and it was open to him to use the evidence of the appellant’s confession as confirmation of the complainant’s testimony.  The appeal from conviction was dismissed.

III.  Analysis

12                               It “can now be taken to be clearly established in Canada that no statement made out of court by an accused to a person in authority can be admitted into evidence against him unless the prosecution shows, to the satisfaction of the trial judge, that the statement was made freely and voluntarily”.  See Erven v. The Queen, [1979] 1 S.C.R. 926, at p. 931, per Dickson J., as he then was.  This, of course, is the confessions rule.

13                               The basic issue in this appeal is whether the trial judge erred in failing to hold a voir dire of his own motion to test the voluntariness of certain out‑of‑court statements made by the accused before admitting them.  In order to resolve this issue, it is appropriate to consider whether the confessions rule should continue to apply only to statements made to persons in authority, or whether it should be expanded so as to capture the out‑of‑court statements made by the accused in this case.  It will therefore be helpful to begin by examining the history of the confessions rule generally, and the person in authority requirement in particular, in order to understand the purpose and function of the rule in the criminal law.

A.  The Confessions Rule and its Relation to the Person in Authority Requirement

14                               Evidence of a confession has always been accorded great weight by triers of fact.  This is a natural manifestation of human experience.  It is because of the tremendous significance attributed to confessions and the innate realization that they could be obtained by improper means that the circumstances surrounding a confession have for centuries been carefully scrutinized to determine whether it should be admitted.  A confession is not excluded, however, simply because of the risk that a conviction may result, but because of the greater risk that the conviction will be unfairly obtained and unjust.  The unfairness of admitting a confession has historically been addressed by a consideration of two factors.  First, the voluntariness of the statement; and second, the status of the receiver of the statement, that is to say, whether the receiver was a person in authority.

15                               As to the first factor, a statement is said to be voluntary when it is made without “fear of prejudice or hope of advantage”:  see Ibrahim v. The King, [1914] A.C. 599 (P.C.), at p. 609, adopted in Canada in Prosko v. The King (1922), 63 S.C.R. 226.  In Boudreau v. The King, [1949] S.C.R. 262, at p. 269, Rand J. explained that “the rule is directed against the danger of improperly instigated or induced or coerced admissions”.  Voluntariness also requires that the statement must be the product of an operating mind:  see Ward v. The Queen, [1979] 2 S.C.R. 30, at p. 40, per Spence J. and Horvath v. The Queen, [1979] 2 S.C.R. 376, at p. 425, per Beetz J.  Voluntariness is determined by a careful investigation of the circumstances surrounding the statement of the accused, and involves a consideration of both objective and subjective factors.

16                               Second, the person in authority requirement generally refers to anyone formally engaged in “the arrest, detention, examination or prosecution of the accused”:  see, e.g., A.B., supra, at p. 26.  This definition may be enlarged to encompass persons who are deemed to be persons in authority as a result of the circumstances surrounding the making of the statement.  For the moment, however, let us consider the purpose of each of these factors as they pertain to the admissibility of statements of the accused.

17                               Historically the insistence that a confession must be voluntary related to concerns about the reliability of the evidence.  Indeed, the basis for the admission of a statement of the accused as an exception to the rule against hearsay is that what people freely say which is contrary to their interest is probably true.  However, where a statement is prompted by a threat or inducement held out by a person in authority, it can no longer be presumed to be true.  Initially when considering the admissibility of confessions some judges focused exclusively on reliability concerns as the sole rationale for the confessions rule.  This is made readily apparent in the case of R. v. Warickshall (1783), 1 Leach 263, 168 E.R. 234, at p. 263 and at pp. 234‑35, respectively:

It is a mistaken notion, that the evidence of confessions and facts which have been obtained from prisoners by promises or threats, is to be rejected from a regard to public faith:  no such rule ever prevailed.  The idea is novel in theory, and would be as dangerous in practice as it is repugnant to the general principles of criminal law.  Confessions are received in evidence, or rejected as inadmissible, under a consideration whether they are or are not intitled [sic] to credit.  A free and voluntary confession is deserving of the highest credit, because it is presumed to flow from the strongest sense of guilt. . . .  [Emphasis added.]

18                               There is also strong historical precedent for the proposition that the confessions rule is rooted in a concern for the administration of justice and fundamental principles of fairness, in particular the principle against self‑incrimination.  In a treatise on the law of evidence written by Lord Chief Baron Gilbert and published in 1754, the author makes the following comment:

. . . the voluntary Confession of the Party in Interest is reckoned the best Evidence; for if a Man’s swearing for his Interest can give no Credit, he must certainly give most Credit when he swears against it; but then this Confession must be voluntary and without Compulsion; for our Law differs from the Civil Law, that it will not force any Man to accuse himself; and in this we do certainly follow the Law of Nature, which commands every Man to endeavor his own Preservation; and therefore Pain and Force may compel Men to confess what is not the Truth of Facts, and consequently such extorted confessions are not to be depended upon.  [Emphasis added.]

See Lawrence Herman, “The Unexplored Relationship Between the Privilege Against Compulsory Self‑Incrimination and the Involuntary Confession Rule (Part I)” (1992), 53 Ohio St. L.J. 101, at p. 153, citing Sir Geoffrey Gilbert, The Law of Evidence (1769).  Thus, it is apparent that from its very inception, the confessions rule was designed not only to ensure the reliability of the confession, but also to guarantee fundamental fairness in the criminal process.

19                               Of particular significance is the relationship between these two concerns of reliability and fairness.  It must be recognized that the purpose of the confessions rule is to exclude putatively unreliable statements, not actually unreliable statements.  In other words, the confessions rule excludes statements obtained by force, threat or promises as somehow inherently unreliable, but does not inquire into the actual truth or falsity of the statement.  If the concern of the confessions rule were truly the reliability of the statement, then the court’s inquiry would focus on objective corroboration of the confession evidence; if additional evidence confirmed the confession was accurate, it should be admitted under a reliability rationale.

20                               Instead, the confessions rule asks only if the statement was voluntary, not if the statement is true.  DeClercq v. The Queen, [1968] S.C.R. 902.  This focus on voluntariness allows a court to analyse the circumstances surrounding the statement and effectively acts as a check on the abuse of state power.  In other words, if the state were left with the option of simply corroborating forced confessions, there would be little incentive to refrain from reprehensible investigative measures.  That is why the confessions rule automatically excludes involuntary statements, regardless of their veracity.  As stated by Professor Mark Berger in “The Exclusionary Rule and Confession Evidence:  Some Perspectives on Evolving Practices and Policies in the United States and England and Wales” (1991), 20 Anglo‑Am. L. Rev. 63, at p. 71:

. . . it is inescapable that the decision to reject all involuntary confessions incorporates policies that find fault with the use of coercive tactics to extract statements, independent of their impact on reliability.  In short, the exclusion of involuntary confessions, at least as practised in the United States and formerly in England, is designed as a response to improper police interrogation tactics as much as, if not more than, it is aimed at ensuring evidence reliability.

This aspect of the confessions rule ‑‑ which focuses on voluntariness over truth ‑‑ indicates that the rule is not concerned solely with accuracy or reliability.

21                               This approach to the rule determining the admissibility of a statement of the accused also accords with the view that the quality, weight or reliability of evidence is a matter for the jury, and that the admission of evidence which may be unreliable does not per se render a trial unfair:  see, e.g., R. v. Buric (1996), 28 O.R. (3d) 737 (C.A.), aff’d [1997] 1 S.C.R. 535, and R. v. Charemski, [1998] 1 S.C.R. 679.  The confessions rule does not force a trial judge to exclude “unreliable” evidence that is highly probative of guilt.  Rather it focuses on putative reliability, by analysing the circumstances surrounding the statement and their effect on the accused, regardless of the statement’s accuracy.  Thus the “reliability” rationale and the “fairness” rationale for the confessions rule blend together, so as to ensure fair treatment to the accused in the criminal process by deterring coercive state tactics.

22                               Indeed, when considering this notion of fairness, several courts have found that the confessions rule is based upon the principle against self‑incrimination.  Two decisions of the House of Lords, Commissioners of Customs and Excise v. Harz, [1967]  1 A.C. 760, and R. v. Sang, [1979] 2 All E.R. 1222, suggest that this is the modern basis for the rule.  The relationship between the principle against self‑incrimination and the confessions rule has also been noted in dicta in a number of Canadian decisions.  See, e.g., DeClercq, supra, at p. 923, per Hall J. (dissenting); Piché v. The Queen, [1971] S.C.R. 23, at p. 26, per Cartwright C.J.; Rothman v. The Queen, [1981] 1 S.C.R. 640, at pp. 653‑54, per Estey J. (dissenting).  More recently, McLachlin J. in R. v. Hebert, [1990] 2 S.C.R. 151, at p. 173 specifically linked the confessions rule to fundamental notions of fairness and the principle that accused persons should not be conscripted to provide evidence against themselves.  She put it in this way:

. . . one of the themes running through the jurisprudence on confessions is the idea that a person in the power of the state’s criminal process has the right to freely choose whether or not to make a statement to the police.  This idea is accompanied by a correlative concern with the repute and integrity of the judicial process.  This theme has not always been ascendant.  Yet, its importance cannot be denied.  It persists, both in Canadian jurisprudence and in the rules governing the rights of suspects in other countries.

23                               I recognize, as did McLachlin J. in Hebert, supra, at p. 173, Iacobucci J. in R. v. S. (R.J.), [1995] 1 S.C.R. 451, at pp. 500‑501, and Sopinka J. in R. v. Whittle, [1994] 2 S.C.R. 914, at p. 932, that the self‑incrimination basis for the confessions rule “must be historically qualified” (S. (R.J.), at p. 499) and that “[i]n Canada, a rationale for the confessions rule extending beyond trustworthiness has not always been easy to locate” (S. (R.J.), at p. 500).  Nevertheless, I must recognize, as did my colleagues, that in a modern sense, the confessions rule has clearly been associated with these ideas.  Indeed, in the Report of the Federal/Provincial Task Force on Uniform Rules of Evidence (1982), the Task Force concluded at p. 175 that “the clear common law principle that the Crown must establish its case without the assistance of the accused . . . is the primary rationale of the Confessions Rule today”.

24                               For this reason, the person in authority requirement is properly seen as an integral component of the confessions rule.  The emphasis on voluntariness has two main effects:  it both avoids the unfairness of a conviction based on a confession that might be unreliable, and has a deterrent effect on the use of coercive tactics.  This deterrent effect is properly focused upon the prosecutorial authority of the state, not the personal authority of private individuals.  It cannot be forgotten that it is the nature of the authority exerted by the state that might prompt an involuntary statement.  As Estey J. stated in Rothman, supra, at pp. 650‑51, “their very authority might, by promise or threat, express or implied, produce a statement whether or not the accused was truly willing to speak” (emphasis added).  In other words, it is the fear of reprisal or hope of leniency that persons in authority may hold out and which is associated with their official status that may render a statement involuntary.  The rule is generally not concerned with conversations between private citizens that might indicate guilt, as these conversations would not be influenced or affected by the coercive power of the state.  This limitation is appropriate since most criminal investigations are undertaken by the state, and it is then that an accused is most vulnerable to state coercion.

25                               On a practical level, the Crown would obviously face an overwhelming burden if it had to establish the voluntariness of every statement against interest made by an accused to any person.  See the Law Reform Commission of Canada, Report on Evidence (1975), at p. 62.  In particular, as the intervener the Attorney General of Canada notes, the elimination of the person in authority requirement would have serious consequences for undercover police work and for the admissibility of wiretap evidence, where the identity of the receiver of the accused’s statement is often unknown.  For example, if the Crown were to intercept a phone call between an accused and a confederate who is senior to him in a criminal hierarchy, the Crown would obviously have difficulty tendering the requisite evidence if it were forced to prove beyond a reasonable doubt that the statements were made without “fear of prejudice or hope of advantage”.  Moreover, all statements to undercover police officers would become subject to the confessions rule, even though the accused was completely unaware of their status and, at the time he made the statement, would never have considered the undercover officers to be persons in authority.

26                               Practical considerations alone lead to the conclusion that the person in authority requirement should remain a part of the confessions rule.  Yet there can be no doubt that there may well be great unfairness suffered by the accused when an involuntary confession obtained as a result of violence or credible threats of imminent violence by a private individual is admitted into evidence.  For this reason, the British  Criminal Law Revision Committee in 1972 recommended eliminating the person in authority requirement from the confessions rule.  See Eleventh Report, Evidence (General) (1972), at p. 39.  The Committee focused on reliability concerns and realistically concluded that a false confession could just as readily result from threats and inducements offered by a private individual as from a person in authority.  The Police and Criminal Evidence Act 1984, 1984 (U.K.), c. 60, was later enacted to implement some of these suggestions.  Section 76 of that Act provides that where it is represented to the court that a confession was obtained “by oppression of the person who made it” or “in consequence of anything said or done which was likely, in the circumstances existing at the time, to render unreliable any confession which might be made by him in consequence thereof”, the Crown must prove beyond a reasonable doubt that the confession was not obtained in this manner.  “[O]ppression” is defined to include “torture, inhuman or degrading treatment, and the use or threat of violence”.

27                               In Australia, the common law confessions rule is similar to the rule in Canada, but in addition, a statement of the accused may be excluded where “it is not made in the free exercise of the will, because this will was overborne by the conduct of a person other than the accused”:  see Peter Gillies, Law of Evidence in Australia (2nd ed. 1991), at p. 537.  Moreover the Commonwealth government and the government of New South Wales have enacted legislation that specifically provides for the exclusion of a confession influenced by “violent, oppressive, inhuman or degrading conduct” or the threat of such conduct.  See Evidence Act 1995, 1995 (Australia), No. 2, s. 84, and Evidence Act 1995, 1995 (N.S.W.), No. 25, s. 84.  However, under both the common law and the relevant Australian statutes, the Crown is only required to prove voluntariness on a balance of probabilities.

28                               It is significant that these changes to the common law of England and Australia were effected through legislative reform.  Indeed, the House of Lords refused to eliminate the person in authority requirement judicially.  In Deokinanan v. R., [1968] 2 All E.R. 346 (P.C.), Viscount Dilhorne, for the court, stated as follows at p. 350:

The fact that an inducement is made by a person in authority may make it more likely to operate on the accused’s mind and lead him to confess.  If the ground on which confessions induced by promises held out by persons in authority are held to be inadmissible is that they may not be true, then it may be that there is a similar risk that in some circumstances the confession may not be true if induced by a promise held out by a person not in authority, for instance if such a person offers a bribe in return for a confession.  There is, however, in their lordships’ opinion, no doubt that the law as it is at present only excludes confessions induced by promises when those promises are made by persons in authority.

29                               The last sentence quoted reflects the present law in Canada.  The confessions rule, including the burden on the Crown to prove voluntariness beyond a reasonable doubt, is carefully calibrated to ensure that the coercive power of the state is held in check and to preserve the principle against self‑incrimination.  The elimination of the person in authority requirement would represent a fundamental change to the confessions rule, and a significant change to the common law which could bring about complex and unforeseeable consequences for the administration of justice.  This change involves the recognition of a new concept.  It does not, as in other cases, simply involve the interpretation of an amendment to a statute, such as the Criminal Code .  The unfairness of admitting statements coerced by private individuals should be recognized.  However, it is the sort of change which should be studied by Parliament and remedied by enactment.  See Watkins v. Olafson, [1989] 2 S.C.R. 750; R. v. Salituro, [1991] 3 S.C.R. 654; Bow Valley Husky (Bermuda) Ltd. v. Saint John Shipbuilding Ltd., [1997] 3 S.C.R. 1210, at para. 93; Winnipeg Child and Family Services (Northwest Area) v. G. (D.F.), [1997] 3 S.C.R. 925.  Because of the very real possibility of a resulting miscarriage of justice and the fundamental unfairness of admitting statements coerced by the violence of private individuals, I would hope that the study will not be long postponed.

30                               In the meantime I would suggest that in circumstances where a statement of the accused is obtained by a person who is not a person in authority by means of degrading treatment such as violence or threats of violence, a clear direction should be given to the jury as to the dangers of relying upon it.  The direction might include words such as these:  “A statement obtained as a result of inhuman or degrading treatment or the use of violence or threats of violence may not be the manifestation of the exercise of a free will to confess.  Rather, it may result solely from the oppressive treatment or fear of such treatment.  If it does, the statement may very well be either unreliable or untrue.  Therefore, if you conclude that the statement was obtained by such oppression very little if any weight should be attached to it.”  However, if a private individual resorts to violence or threatens violence after the statement has been made, this conduct will not as a general rule be a factor affecting the voluntariness of the statement and the suggested direction will not be needed.

B.  Limits of the Person in Authority Requirement

31                               It has been seen that the person in authority requirement is grounded in the underlying rationales for the confessions rule, and as a result it should remain part of the rule.  Consideration must now be given as to who should come within the designation “person in authority”.

32                               “Person in authority” typically refers to those persons formally engaged in the arrest, detention, examination or prosecution of the accused:  see A.B., supra, at p. 26.  However, it may take on a broader meaning.  Canadian courts first considered the meaning of “person in authority” in R. v. Todd (1901), 4 C.C.C. 514 (Man. K.B.).  In that case, the accused made a statement to two men he believed to be fellow prisoners, but who were in fact acting as agents of the police.  It was held, at pp. 526‑27, that:

A person in authority means, generally speaking, anyone who has authority or control over the accused or over the proceedings or the prosecution against him. . . . [T]he authority that the accused knows such persons to possess may well be supposed in the majority of instances both to animate his hopes of favour on the one hand and on the other to inspire him with awe, and so in some degree to overcome the powers of his mind. . . .  [Emphasis added.]

Thus, from its earliest inception in Canadian law, the question as to who should be considered as a person in authority depended on the extent to which the accused believed the person could influence or control the proceedings against him or her.  The question is therefore approached from the viewpoint of the accused.  See also R. v. Roadhouse (1933), 61 C.C.C. 191 (B.C.C.A.), at p. 192.

33                               The subjective approach to the person in authority requirement has been adopted in this Court.  See Rothman, supra, at p. 663.  The approach adopted by McIntyre J.A. (as he then was) in R. v. Berger (1975), 27 C.C.C. (2d) 357 (B.C.C.A.), at pp. 385‑86 is, in my view, a clear statement of the law:

The law is settled that a person in authority is a person concerned with the prosecution who, in the opinion of the accused, can influence the course of the prosecution.  The test to be applied in deciding whether statements made to persons connected in such a way with the prosecution are voluntary is subjective.  In other words what did the accused think?  Whom did he think he was talking to? . . .  Was he under the impression that the failure to speak to this person, because of his power to influence the prosecution, would result in prejudice or did he think that a statement would draw some benefit or reward?  If his mind was free of such impressions the person receiving this statement would not be considered a person in authority and the statement would be admissible.

34                               However, to this statement I would add that the accused’s belief that he is speaking to a person in authority must also be reasonable, in the context of the circumstances surrounding the making of the statement.  If the accused were delusional or had no reasonable basis for the belief that the receiver of the statement could affect the course of the prosecution against him, the receiver should not be considered a person in authority.  Since the person in authority requirement is aimed at controlling coercive state conduct, the test for a person in authority should not include those whom the accused unreasonably believes to be acting on behalf of the state.  Thus, where the accused speaks out of fear of reprisal or hope of advantage because he reasonably believes the person receiving the statement is acting as an agent of the police or prosecuting authorities and could therefore influence or control the proceedings against him or her, then the receiver of the statement is properly considered a person in authority.  In other words, the evidence must disclose not only that the accused subjectively believed the receiver of the statement to be in a position to control the proceedings against the accused, but must also establish an objectively reasonable basis for that belief.  For example, if the evidence discloses a relationship of agency or close collaboration between the receiver of the statement and the police or prosecution, and that relationship was known to the accused, the receiver of the statement may be considered a person in authority.  In those circumstances the Crown must prove beyond a reasonable doubt that the statement was made voluntarily.

35                               Over the years, the courts have determined when and in what circumstances a person will be deemed a person in authority for the purposes of the confessions rule. See, e.g., R. v. Trenholme (1920), 35 C.C.C. 341 (Que. K.B.) (complainant’s father was held to be a person in authority where he has control over the prosecution of the accused); R. v. Wilband, [1967] S.C.R. 14 (psychiatrist is not a person in authority where he cannot control or influence the course of the proceedings); R. v. Downey (1976), 32 C.C.C. (2d) 511 (N.S.S.C.A.D.) (victim is a person in authority if the accused believed that the victim had control over the proceedings); A.B., supra (a parent is not, in law, a person in authority if there is no close connection between the decision to call the authorities and the inducement to a child to make a statement); R. v. Sweryda (1987), 34 C.C.C. (3d) 325 (Alta. C.A.)(a social worker is a person in authority if the accused knew the social worker was investigating allegations of child abuse and believed it could lead to his arrest).  These cases have not departed from the governing rule that defines a person in authority in relation to the accused’s perception of the receiver’s involvement with the investigation or prosecution of the crime nor have these decisions defined a person in authority solely in terms of the personal authority that a person might wield in relation to the accused.  Moreover, in concluding that the receiver of the statement was a person in authority, the courts have consistently found the accused believed the receiver was allied with the state authorities and could influence the investigation or prosecution against the accused.

36                               The important factor to note in all of these cases is that there is no catalogue of persons, beyond a peace officer or prison guard, who are automatically considered a person in authority solely by virtue of their status.  A parent, doctor, teacher or employer all may be found to be a person in authority if the circumstances warrant, but their status, or the mere fact that they may wield some personal authority over the accused, is not sufficient to establish them as persons in authority for the purposes of the confessions rule.  As the intervener the Attorney General of Canada observed, the person in authority requirement has evolved in a manner that avoids a formalistic or legalistic approach to the interactions between ordinary citizens.  Instead, it requires a case‑by‑case consideration of the accused’s belief as to the ability of the receiver of the statement to influence the prosecution or investigation of the crime.  That is to say, the trial judge must determine whether the accused reasonably believed the receiver of the statement was acting on behalf of the police or prosecuting authorities.  This view of the person in authority requirement remains unchanged.

37                               Finally, something must be said about the respective burdens which must be borne by the accused and the Crown on a voir dire to determine whether a statement of the accused to a person in authority should be admitted.  The Crown, of course, bears the burden of proving beyond a reasonable doubt that the statement was made voluntarily.  However, in relation to the person in authority requirement, the evidence required to establish whether or not a person should be deemed a person in authority will often lie primarily with the accused.  The accused therefore must bear some burden in relation to this aspect of the confessions rule.  The burden should be an evidential and not a persuasive one.  See, e.g., R. v. Scott (1984), 1 O.A.C. 397, at p. 399.  John Sopinka, Sidney N. Lederman and Alan W. Bryant, in The Law of Evidence in Canada (1992), at pp. 56‑57, explain the difference between the two burdens:

The term evidential burden means that a party has the responsibility to insure that there is sufficient evidence of the existence or non‑existence of a fact or of an issue on the record to pass the threshold test for that particular fact or issue. . . . In contrast, the term legal burden of proof means that a party has an obligation to prove or disprove a fact or issue to the criminal or civil standard.  The failure to convince the trier of fact to the appropriate standard means that party will lose on that issue.

The evidential burden on an accused in a criminal case is described as follows (at p. 138):

Where an evidential burden for an issue rests on the defendant in a criminal case, for example self‑defence, the accused has the obligation to ensure that there is some evidence on the record to make it a live issue.  The evidence necessary to satisfy an evidential burden may arise in the case for the Crown or the defence.

38                               In the vast majority of cases, the accused will meet this evidential burden by showing the accused’s knowledge of the relationship between the receiver of the statement and the police or prosecuting authorities.  For example, the fact that the statement was made to a police officer who was in uniform or identified himself or herself as a peace officer will satisfy the accused's evidential burden in relation to the person in authority requirement.  See, e.g., Morris v. The Queen, [1979] 2 S.C.R. 1041, at p. 1066.  Once the accused satisfies this evidential burden, the ultimate burden of proof rests with the Crown.  See R. v. McKenzie, [1965] 3 C.C.C. 6 (Alta. S.C.A.D.), at p. 28.  In R. v. Postman (1977), 3 A.R. 524, at p. 542, the Alberta Supreme Court, Appellate Division held, correctly in my view, that where a witness is not prima facie a person in authority (in that case, a doctor), “it is open to defence counsel to challenge the prima facie case and require evidence to be given to determine the facts of the matter”.  Thus, once the defence discharges its burden and establishes that there is an evidential basis to the claim that the receiver of a statement made by the accused is a person in authority, the burden shifts to the Crown to establish beyond a reasonable doubt either that the receiver is not a person in authority, or, if this burden cannot be discharged, that the statement was made voluntarily.

39                               The receiver’s status as a person in authority arises only if the accused had knowledge of that status.  If the accused cannot show that he or she had knowledge of the receiver’s status (as, for example, in the case of an undercover police officer) or close relationship to the authorities (as in the case of persons acting on behalf of the state), the inquiry pertaining to the receiver as a person in authority must end.  It is therefore appropriate to consider at the outset the reasonable belief of the accused.  It may not be useful to have the trial judge undertake a full analysis of the objective relationship between the receiver of the statement and the authorities, as Justice L’Heureux-Dubé suggests (para. 83), only to have those findings vitiated if the accused is later found to have no knowledge of this relationship.  In addition, it is important to recognize that focusing the trial judge’s inquiry on the reasonable belief of the accused accords with the allocation of the burden of proof on the voir dire.

C.  The Trial Judge’s Obligation to Hold a Voir Dire

40                               The issue as to whether the trial judge in this case erred in failing to hold a voir dire before admitting the confession into evidence can now be considered.

41                               The trial judge has a duty “to conduct the trial judicially quite apart from  lapses of counsel”:  see R. v. Sweezey (1974), 20 C.C.C. (2d) 400 (Ont. C.A.), at p. 417.  This includes the duty to hold a voir dire whenever the prosecution seeks to adduce a statement of the accused made to a person in authority:  see pp. 417‑18.  However, where the defence has not requested a voir dire and a statement of the accused is admitted into evidence, the trial judge will only have committed reversible error if clear evidence existed in the record which objectively should have alerted him or her to the need for a voir dire notwithstanding counsel’s silence.  Thus, the test for holding a voir dire is assessed by an appellate court’s objective review of the evidence in the record to determine whether something should have triggered the trial judge’s obligation to conduct an inquiry.  This test is different from the test applicable on the voir dire, which requires the trial judge to undertake an examination of the reasonable belief of the accused and the circumstances surrounding the making of the statement to determine both whether the receiver is a person in authority and whether the statement was made voluntarily.

42                               These tests are easily confused, as it is possible to assert that the evidence which alerts the trial judge to the need for a voir dire must relate to the accused’s state of mind, since only this type of evidence could demonstrate that the accused believed the receiver of the statement to be a person in authority.  This approach is too narrow.  The subjective view of the accused might never be ascertained unless there is a voir dire.  To hold that the trial judge has no obligation to hold a voir dire unless evidence of the accused’s mental state is already in the record would render the obligation meaningless.  Certainly evidence relating to the accused’s state of mind may well trigger the need for a voir dire, but the standard should be set lower.

43                               This is demonstrated by the Court’s decision in Erven, supra.  In that case, the Court had to decide whether a voir dire is always needed in respect of a statement made by the accused to a person in authority, or whether it is unnecessary in circumstances where the statement was “obviously volunteered” (p. 929).  Dickson J.,  writing for a plurality of the Court, held at p. 940, “a rule requiring a voir dire only after there is some evidence suggesting involuntariness is neither clear nor easily applied”.  He stated, at p. 937:

It is a most important right of the accused to call witnesses and present argument separately on the issue of the admissibility of a statement directly after the Crown’s evidence on this matter has been adduced.  This right would be totally lost if a statement of the accused could be admitted without a voir dire as part of the Crown’s evidence because nothing in that evidence cast doubt on its voluntariness. . . .  The testimony of the accused may be the only evidence which indicates involuntariness.

Although this statement of principle pertains to a voir dire held to determine voluntariness, it is also applicable to one held to determine whether the receiver of a statement made by the accused is a person in authority.

44                               Specifically, the reasoning in Erven, with which I agree, has two implications for the present case.  First, the requirement that a voir dire be held cannot be founded upon the presence of evidence which is conclusive of the very issue to be examined on the voir dire.  Second, a voir dire is required in respect of any statement made by an accused person to a “person in authority”.  See Erven, supra, at p. 931.  Thus, given the highly prejudicial nature of confession evidence, the trial judge has an obligation to hold a voir dire of his or her own motion, notwithstanding the absence of any request by counsel, wherever the Crown seeks to adduce a statement made by an accused to a person in authority.  Clearly, counsel for the accused may waive the voir dire.  Once the waiver is given, it is unnecessary to hold the voir dire.

45                               However, the trial judge’s obligation is triggered only where the evidence makes the need for a voir dire clear.  Evidence which clearly demonstrates that the receiver of the statement made by the accused was closely connected to the authorities should alert the trial judge to hold a voir dire.  This evidence progresses along a spectrum.  That is, where the receiver of the statement is a “conventional” person in authority, such as a police officer or prison guard, the trial judge clearly has an obligation to proceed to a voir dire.  In such a case, the connection to the authorities is readily apparent. Similarly, where the evidence clearly discloses a close connection between the receiver of the statement and the authorities which indicates that the receiver was, in the circumstances, acting as a person in authority, this may be sufficient to trigger the trial judge’s obligation to hold a voir dire.

46                               Specifically, the trial judge must be satisfied that if the accused had been aware of the connection between the receiver of the statement and the authorities, the accused could reasonably believe the receiver was acting as an agent of the police or prosecuting authorities or as part of the prosecution team and was therefore capable of influencing the prosecution against him or her.  If the evidence establishes this sort of connection, the trial judge should inquire whether the defence is prepared to discharge its evidential burden on the person in authority issue or whether it waives a voir dire on this issue.  Thus, evidence of close association with the authorities might provide an entry to the voir dire procedure but the inquiry on the voir dire will still focus on an analysis to determine whether the receiver of the statement, assessed from the point of view of the accused, is a person in authority and ultimately whether the statement was made voluntarily.  However, the further away the receiver of the statement is from the “conventional” case, the less likely it will be that the evidence will alert the trial judge to the need to hold a voir dire, and the greater the obligation of the accused to raise the issue.

47                               It should be emphasized that only rarely will a trial judge have heard sufficient evidence to trigger the need for a voir dire on the person in authority issue where the receiver of the statement is not a conventional person in authority.  This follows because the evidence must establish more than the mere status of the receiver of the statement.  Status or personal authority alone will not as a general rule provide evidence from which it can be inferred that the receiver of a confession is, in the eyes of the accused, a person in authority.  Rather, in order to demonstrate the need for a voir dire, the evidence must show that the receiver of the statement was closely associated with the authorities prior to obtaining the statement, and that there was as well a close connection in time between the contact with the authorities and its receipt.  The evidence must suggest that the receiver was acting in concert with the police or prosecutorial authorities, or as their agent, or as part of their team.  Only in these circumstances will the trial judge be obliged to hold a voir dire of his or her own motion on the person in authority issue, subject to waiver of the voir dire by counsel for the accused.

IV.  Summary

48                               Perhaps it may be of some assistance to set out in summary form the applicable principles pertaining to the admission of statements made by the accused to persons in authority and some of the factors to be taken into consideration with regard to them.

1.  The rule which is still applicable in determining the admissibility of a statement made by an accused to a person in authority is that it must have been made voluntarily and must be the product of an operating mind.

2.  The rule is based upon two fundamentally important concepts:  the need to ensure the reliability of the statement and the need to ensure fairness by guarding against improper coercion by the state.  This results in the requirement that the admission must not be obtained by either threats or inducements.

3.  The rule is applicable when the accused makes a statement to a person in authority.  Though no absolute definition of “person in authority” is necessary or desirable, it typically refers to those formally engaged in the arrest, detention, examination or prosecution of the accused.  Thus, it would apply to person such as police officers and prison officials or guards.  When the statement of the accused is made to a police officer or prison guard a voir dire should be held to determine its admissibility as a voluntary statement, unless the voir dire is waived by counsel for the accused.

4.  Those persons whom the accused reasonably believes are acting on behalf of the police or prosecuting authorities and could therefore influence or control the proceedings against him or her may also be persons in authority.  That question will have to be determined on a case-by-case basis.

5.  The issue as to who is a person in authority must be resolved by considering it subjectively from the viewpoint of the accused.  There must, however, be a reasonable basis for the accused’s belief that the person hearing the statement was a person in authority.

6.  The issue will not normally arise in relation to undercover police officers.  This is because the issue must be approached from the viewpoint of the accused.  On that basis, undercover police officers will not usually be viewed by the accused as persons in authority.

7.  If it is contended that the recipient of the statement was a person in authority in the eyes of the accused then the defence must raise the issue with the trial judge.  This is appropriate for it is only the accused who can know that the statement was made to someone regarded by the accused as a person in authority.

8.  On the ensuing voir dire the accused will have the evidential burden of demonstrating that there is a valid issue for consideration.  If the accused meets the burden, the Crown will then have the persuasive burden of demonstrating beyond a reasonable doubt that the receiver of the statement was not a person in authority or if it is found that he or she was a person in authority, that the statement of the accused was made voluntarily.

9.  In extremely rare cases the evidence adduced during a trial may be such that it should alert the trial judge that the issue as to whether the receiver of a statement made by an accused was a person in authority should be explored by way of voir dire.  In those cases, which must be extremely rare in light of the obligation of the accused to raise the issue, the trial judge must of his or her own motion direct a voir dire, subject, of course, to waiver of the voir dire by counsel for the accused.

10.  The duty of the trial judge to hold a voir dire of his or her own motion will only arise in those rare cases where the evidence, viewed objectively, is sufficient to alert the trial judge of the need to hold a voir dire to determine if the receiver of the statement of the accused was, in the circumstances, a person in authority.

11.  If the trial judge is satisfied that the receiver of the statement was not a person in authority but that the statement of the accused was obtained by reprehensible coercive tactics, such as violence or credible threats of violence, then a direction should be given to the jury.  The jury should be instructed that if they conclude that the statement was obtained by coercion, they should be cautious about accepting it, and that little if any weight should be attached to it.

V.  Application to this Appeal

49                               The appellant contends that the fact that the confession was made to the complainant and her immediate family should have alerted the trial judge to the need for a voir dire since they are capable of being persons in authority for the purpose of the confessions rule.  It is true the complainant and her family members are capable of being persons in authority.  Indeed, anyone is capable of being a person in authority where a person becomes sufficiently involved with the arrest, detention, examination or prosecution of an accused, and the accused believes that the person may influence the process against him or her.  It does not follow that simply because it has been held, in the circumstances presented in other cases, that a family member was a person in authority, that the trial judge should have been alerted to the need for a voir dire.  Virtually any category of person ‑‑ parents of the accused, parents of the complainant, teachers, psychiatrists, physicians ‑‑ may, in light of the particular evidence adduced, be considered to be a person in authority.  As the respondent observed, to hold that the trial judge committed an error on the basis that the receiver of the confession is merely capable of being a person in authority is to require a voir dire (or waiver) for every statement against interest made by every accused person to anyone.  It cannot be forgotten that it is the accused who is in the best position to demonstrate that the receiver of the statement was in his or her eyes a person in authority.

50                               In this case, the evidence at trial did not disclose any evidence that was sufficient to trigger the trial judge’s obligation to hold a voir dire.  The confrontation at the appellant’s workplace was first described by the complainant.  She testified as to the events leading up to the confrontation.  She stated (1) that her mother questioned her about whether she was pregnant and whether she had had intercourse; (2) that in the course of that conversation, she told her mother that the appellant had sexually assaulted her; (3) that her mother telephoned her father; (4) that she and her mother visited a walk‑in clinic in Mississauga where it was confirmed that the complainant was pregnant (her boyfriend at the time was the father); (5) that the complainant, together with her mother, father, stepfather and cousin went to confront the appellant.  The complainant then related, without objection by the defence, the statements made by the appellant.  Thus, when the statements were admitted into evidence, there was nothing to suggest that the complainant or her family members had spoken to the police or anyone else in authority or were even considering making a complaint.  Similarly, there was nothing to suggest that the appellant subjectively believed the complainant’s family to have control over criminal proceedings.  In those circumstances, the trial judge cannot be said to have committed an error by failing to hold a voir dire on his own motion.

51                               The appellant, in an alternative argument, relied upon the Privy Council decision in Thongjai v. The Queen, [1998] A.C. 54.  There it was held that where there is no allegation of mistreatment by the police and the defence is an absolute denial that the statement was made, it is not necessary to hold a voir dire to consider whether the statement was voluntary.  That argument could not be raised in the case at bar.  The person in authority issue was not raised in Thongjai.  In the circumstances presented in this case, it would be inappropriate to deal with this alternative argument.

52                               Finally, the appellant also contended that even if the evidence of the confession was properly before the trial judge, the judge erred by relying upon this evidence almost exclusively in convicting the appellant.  The appellant submits that the trial judge relied on the confession evidence without fully weighing and considering the reliability of the evidence in light of the circumstances surrounding the making of the confession.  These submissions cannot be accepted.  The record clearly reveals that the trial judge carefully reviewed the appellant’s testimony relating to the confrontation with the complainant’s family before concluding that the confession evidence was “overwhelming” and that “[t]here [was] no plausible basis for [his] disbelieving the evidence that [the accused] admitted his involvement”.  The trial judge also found the complainant to be credible.  In these circumstances he could properly use the evidence of the confession to confirm the complainant’s testimony.  The reasons of the learned trial judge disclose no error.

VI.  Disposition

53                               The appeal is therefore dismissed.

//L’Heureux-Dubé J.//

The reasons of L’Heureux-Dubé and Bastarache JJ. were delivered by

54                      L’Heureux-Dubé J. -- This case is about the scope of a trial judge’s duty to direct independently a voir dire in order to determine whether out-of-court statements made by an accused were to persons in authority and, if so, are voluntary.  Its resolution on the facts of this case is inextricably linked to the more general determination of who constitutes a “person in authority” under the confessions rule. 

55                      I have had the benefit of reading the reasons of my colleague Justice Cory.  While I agree with his ultimate disposition of this appeal, I differ with respect to the proper interpretation of “persons in authority”, as a pertinent component of the confessions rule, and thus the circumstances which give rise to a trial judge’s duty to direct a voir dire proprio motu for the following reasons.

I.   The Confessions Rule in Context

56                      My colleague Cory J. has set out the relevant facts, and as stated, I ultimately agree with his disposition of this appeal.  My differences with his reasons arise in relation to the proper formulation of the “person in authority” element of the confessions rule.

57                      The basic formulation of the confessions rule is uncontroversial: a statement obtained by hope of advantage or fear of prejudice held out by a person in authority must be proven voluntary beyond a reasonable doubt before admission into evidence by the Crown (see Ibrahim v. The King, [1914] A.C. 599 (P.C.)).  As the expansive body of divergent case law demonstrates, however, the rule’s precise scope and meaning have perplexed courts.  In this regard, providing some clarity and certainty to the “person in authority” component of the rule, which is the foundation of a trial judge’s duty to independently direct a voir dire, is essential. We cannot forget that common law rules are moulded by judges, and therefore, it is our province, and indeed our duty, to ensure that they serve the interests which they bind (R. v. Khan, [1990] 2 S.C.R. 531).  The challenge must be to properly construe the confessions rule -- this “rule of policy” -- and its element of “person in authority” in light of its underlying rationales as they have evolved from the 18th century into the modern criminal justice system.

58                      It is worthwhile to observe at the outset how dramatically different the original criminal context was which gave rise to the confessions rule, for with changing conditions, underlying concerns have inevitably shifted.  Lord Hailsham in D.P.P. v. Ping Lin, [1976] A.C. 574 (H.L.), at p. 600, directed his attention to its origins:

By the judiciary, though [the confessions rule] ought not to be extended, it must by no means be whittled down.  It bears, it is true, all the marks of its origin at a time when the savage code of the eighteenth century was in full force. At that time almost every serious crime was punishable by death or transportation. The law enforcement officers formed no disciplined police force and were not subject to effective control by the central government, watch committees or an inspectorate. There was no legal aid. There was no system of appeal. To crown it all the accused was unable to give evidence on his own behalf and was therefore largely at the mercy of any evidence, either perjured or oppressively obtained, that might be brought against him. The judiciary were therefore compelled to devise artificial rules designed to protect him against dangers now avoided by other and more rational means. Nevertheless, the rule has survived into the twentieth century, not only unmodified but developed. . . .

As an artificial rule designed to protect accused against dangers which are now avoided by other, more rational means like individual rights, the basic counsel that, in spite of this evolution, the confessions rule should not be extended must be taken to heart. It is a policy rule which must be carefully construed in light of modern conditions.

59                      In addition to ensuring that our understanding of the confessions rule reflects modern concerns, it is also important to take account of the rule’s artificial and exceptional character.  Pointedly, the confessions rule is an exception to the basic principle of evidence that all relevant evidence should be admitted to the trier of fact, whose role it is to assess its quality and ultimately, get to the truth of the matter.  As explained by McLachlin J. in R. v. Seaboyer, [1991] 2 S.C.R. 577, at p. 609:

It is fundamental to our system of justice that the rules of evidence should permit the judge and jury to get at the truth and properly determine the issues. . . . [E]verything which is probative should be received, unless its exclusion can be justified on some other ground. A law which prevents the trier of fact from getting at the truth by excluding relevant evidence in the absence of a clear ground of policy or law justifying the exclusion runs afoul of our fundamental conceptions of justice. . . .

Our criminal justice system has assigned triers of fact the crucial role of assessing the reliability of evidence. Unlike the early days of the confessions rule, today the defence has a full opportunity to impugn and explore factors relevant to the reliability of evidence and credibility of witnesses before the triers of fact. As judges, we must guard against interfering with this pivotal and fair process.  The confessions rule must be justified by its policy rationales, taking account of the inclusionary gist of basic evidence rules which typically do not justify exclusion on reliability concerns.

60                      In summary, the confessions rule is a specific and delimited exception to the basic rule that it is for the trier of fact to assess the truth of statements in all the surrounding circumstances, as well as the recognition that statements made by the accused against his or her interest are inherently reliable (Hardy’s Trial (1794), 24 State Tr. 199). Therefore, interpretation of its integral concept of “person in authority” must be bound by applicable policy rationales which justify the confessions rule’s exclusion of relevant evidence, to which I will now turn.

II.   The Cardinal Role of the “Person in Authority” Element

61                      My colleague Cory J. carefully reviews the history of the confessions rule and appropriately sets its underlying policies in a contemporary context, and there is no need to repeat this exercise.  Essentially where, historically, policy concerns related to the confession’s reliability, modern rationales for the confessions rule now focus on state conduct and individual fairness.  Basic fairness in the criminal process unifies the two rationales of reliability and state deterrence, and questions of truth have given way to a focus on voluntariness.

62                      The confessions rule is only concerned with voluntariness, however, where statements are made to persons in authority. This latter concept becomes all the more critical when, as my colleague has properly held, we accept that regardless of reliability, the predominant modern rationales for the confessions rule are individual fairness in the criminal process and state accountability.  In my view, these rationales explicitly endorse a focus on the relationship between individuals and the state or its representatives, and should define our notion of “person in authority”. 

63                      While judicial recognition of these rationales has been cautious thus far, Cory J.’s reasons reflect their cogency and ultimate acceptance. Modern developments of the confessions rule bolster this conclusion.  In Clarkson v. The Queen, [1986] 1 S.C.R. 383, at p. 391, Wilson J. assessed for the majority how the “operating mind” test as a component of the confessions rule logically revealed:

. . . a concern not so much for the probative value of the statement as for adjudicative fairness in the criminal process and for control of police conduct in interrogating accused persons.

Similarly, in its study, the Report of the Federal/Provincial Task Force on Uniform Rules of Evidence (1982) the Task Force concluded at p. 175 that the modern-day rationale for the confessions rule is:

. . . based on the principle that the Crown cannot require the accused to provide evidence against himself.  The rationale puts the Rule in a party and party context, and accordingly does not apply to outsiders.  The Crown therefore is not precluded from taking advantage of relevant evidence obtained as the result of actions by someone other than a Crown agent or representative. The Task Force unanimously concludes that the Confessions Rule should only apply to statements made to persons in authority, and that inducements emanating from other sources, where there is no implied authority from the Crown, should only affect the weight of any answer given by the accused. [Emphasis added.]

My colleague accepts this view which implies that the confessions rule has no justification without the policy of deterrence.  In other words, without state presence as required by the “person in authority” component, the confessions rule would stand in direct conflict with the basic rule of evidence that the reliability of evidence is a matter for the trier of fact.  Given the rule’s infidelity to the principle of reliability, it is important to both affirm its inextricable relationship with concerns about the integrity of the justice system, the fairness of criminal procedure and the conduct of state agents, and reflect this affirmation in our interpretation of the “person in authority” element.

64                      While accepting these modern rationales for the confessions rule, my colleague, surprisingly, does not similarly endorse the Task Force’s strong conclusion regarding “persons in authority” such that, absent Crown conduct, a lack of voluntariness should only affect the weight of statements and not their admissibility.  Indeed, Cory J. allows that “the person in authority requirement should remain a part of the confessions rule” (para. 26); however, he also suggests that Parliament should consider and ultimately enact its elimination in order to redress the “unfairness of admitting statements coerced by private individuals” (para. 29).  I respectfully disagree that such a recommendation is appropriate or desirable, and maintain that general involuntariness should not affect the admissibility of statements.

65                      First, this suggestion contradicts the affirmation that the confessions rule is now primarily directed toward deterring improper use of prosecutorial state authority, which ultimately implies that some presence of the state or its agents is crucial.  Therefore, any instances of private coercion fall beyond the scope of the confessions rule.  The general unfairness of utilizing involuntary statements resulting from private coercion has never been the focus of the confessions rule, even when the rule was justified by policy concerns for reliability.  Its modern rationales explicitly affirm that the rule is concerned only with voluntariness within the relationship between the state and individual.

66                      Indeed, I find the recommendation to eliminate the person in authority requirement to be inconsistent with Cory J.’s own acknowledgment (at para. 24) that “it is the nature of the authority exerted by the state that might prompt an involuntary statement”(emphasis added); and further (at para. 29), how the confessions rule “is carefully calibrated to ensure that the coercive power of the state is held in check and to preserve the principle against self-incrimination”.  These rationales, and thus the confessions rule itself, are concerned with state action, as affirmed by Sopinka J. in R. v. Whittle, [1994] 2 S.C.R. 914,  at p. 932:

The preoccupation of the common law and Charter cases in preserving for the suspect the right to choose has been in relation to state action. Did the action of police authorities deprive the suspect of making an effective choice by reason of coercion, trickery or misinformation or the lack of information? [Emphasis added.]

Essentially, an expansion of the confessions rule to private instances of coercion would not serve the modern principles which justify its exclusionary character and, therefore, I do not agree with my colleague’s suggestion in this regard.  The confessions rule is not concerned with private conversations, but rather a state’s criminal investigations in which private persons are most vulnerable to the coercive power of the state.

67                      Finally, we must consider how an expansion of the confessions rule to private instances of coercion should account for the public interest in placing probative evidence before juries.  In assessing an expansion of the confessions rule, White J. for the U.S. Supreme Court in Lego v. Twomey, 404 U.S. 477 (1972), fittingly observed at pp. 488-89:

Without good cause, we are unwilling to expand currently applicable exclusionary rules by erecting additional barriers to placing truthful and probative evidence before state juries and by revising the standards applicable in collateral proceedings. . . . [T]he exclusionary rules are very much aimed at deterring lawless conduct by police and prosecution and ... the public [has an] interest in placing probative evidence before juries for the purpose of arriving at truthful decisions about guilt or innocence.

Indeed, not only does an expansion of the confessions rule to private instances of coercion not serve its modern underlying rationales, but it would also erect additional barriers keeping relevant evidence from the triers of fact and thus unjustifiably impede their task of ascertaining the truth of the matter before them.  Failing to serve any legitimate purpose within the confines of this policy rule, I conclude that the expansion suggested by my colleague is inappropriate and undesirable. Any private threats and inducements which affect the reliability of confessions are for the trier of fact to assess.

68                      I similarly find that the special jury directions provided by my colleague for cases of private coercion (para. 30) unnecessary in any categorical sense. These statements will be subject to cross-examination and refutation by the defence before the trier of fact, who scrutinizes them carefully in all the surrounding circumstances before assigning them with any weight. Special warnings are no more necessary in these circumstances to supplement the basic fairness of the existing process than other types of evidence like, for example, alibi or identification evidence. The prospect of special warnings in every instance of evidence prejudicial to the accused does not foster an effective administration of justice.

69                      In summary, legislative change to the confessions rule altering the requirement of a person in authority is neither appropriate nor desirable.  The best rule remains, as stated by Martland J. in Rothman v. The Queen, [1981] 1 S.C.R. 640, at p. 664:

Once it is accepted that the confession of the appellant was not made to a person in authority, it was properly admissible without any requirement for the Crown to establish that it was voluntary.  The confession was in the same position as if it had been made to someone other than a police officer.

Strict retention of this rule promotes clarity, is consistent with modern rationales justifying the exclusion of relevant evidence and serves the general public interest by facilitating the pursuit of truth. At the end of the day, the “person in authority” requirement is, and should remain, the pivotal concept in the confessions rule.  The next relevant question is how to interpret this concept properly.

III.   Defining “Person in Authority”

70                      The general definition of “person in authority” is “someone [formally] engaged in the arrest, detention, examination or prosecution of the accused” (emphasis added) (see R. v. Paonessa (1982), 66 C.C.C. (2d) 300 (Ont. C.A.), at p. 306, aff’d [1983] 1 S.C.R. 660).  This includes both formal representatives and agents of the state.  In my view, this remains the basic test and I have difficulties with my colleague’s largely subjective approach to this concept, which, I fear, leaves undue scope for encompassing private persons.  Given my conclusion as to the cardinal role of the concept of “persons in authority” in the confessions rule, and how it directs the latter toward state conduct, I find that a broad and subjective interpretation of “persons in authority” has the danger of excluding evidence derived from purely private conversations and thus indirectly achieving what we have rejected above as inappropriate. Indeed, as examined below, logic and the applicable authorities persuade me that the analysis of “person in authority” should begin with an objective test that mirrors the general definition, and only subsequently turn to examine the subjective belief of the accused.

71                      According to my colleague, the broad principled test for “persons in authority” is: “persons whom the accused reasonably believes are acting on behalf of the police or prosecuting authorities and could therefore influence or control the proceedings against him or her may also be persons in authority” (para. 48, principle 4, emphasis added).  This issue should be resolved by “considering it subjectively from the viewpoint of the accused”, although there must be “a reasonable basis for the accused’s belief that the person hearing the statement was a person in authority” (para. 48, principle 5).  This reasonableness element involves, in his view, a consideration of the “circumstances surrounding the making of the statement” (para. 34), and appears to necessitate an objective assessment of the accused’s particular perceptions (at para. 34):

. . . where the accused speaks out of fear of reprisal or hope of advantage because he reasonably believes the person receiving the statement is acting as an agent of the police or prosecuting authorities and could therefore influence or control the proceedings against him or her, then the receiver of the statement is properly considered a person in authority.

On his test, anyone is capable of being a person in authority -- parents of the accused or complainant, teachers, psychiatrists, physicians -- depending on the particular evidence adduced and what it reveals of the accused’s perceptions (para. 49).  One can immediately see the vagueness of the test and the uncertainty that this approach will generate.

72                      There is considerable disagreement in the authorities as to the nature of the “person in authority” test.  My colleague, like a number of lower courts, draws on the highly subjective test advocated by Fred Kaufman in The Admissibility of Confessions (3rd ed. 1979), at p. 81, which asks:

Did the accused truly believe, at the time he made the declaration, that the person he dealt with had some degree of power over him? [Emphasis in original.]

See, for example, R. v. A.B. (1986), 26 C.C.C. (3d) 17 (Ont. C.A.), leave to appeal refused [1986] 1 S.C.R. v; R. v. Stewart (1980), 54 C.C.C. (2d) 93 (Alta. C.A.); R. v. Fowler (1982), 4 C.C.C. (3d) 481 (Nfld. C.A.); and R. v. Collins (1975), 29 C.C.C. (2d) 304 (Alta. S.C.). 

73                      At the other hand of the spectrum, in Paonessa, supra, a highly objective approach was adopted.  In the Court of Appeal, Zuber J.A. for the majority, in a judgment affirmed by our Court, held at p. 306 that a “person in authority” was only someone actually engaged in criminal proceedings against the accused:

The confession rules apply only where the accused makes a statement to a “person in authority”, i.e., someone engaged in the arrest, detention, examination or prosecution of the accused.  Paquette’s evidence at his bail hearing was a public statement in open Court presided over by a Magistrate (Provincial Court Judge). . . . The Magistrate, as investigator or interrogator is no longer a part of the Canadian system of criminal justice and as a result the modern Magistrate is not a “person in authority” so as to attract the operation of the rules of evidence relating to confessions. [Emphasis added.]

Such emphasis on the objective status of “persons in authority” is warranted. Indeed, it should serve as a threshold requirement for the application of the exclusionary rule, before the subjective belief of the accused comes into play.  This approach has support in the authorities, and more importantly, is required by the rule’s modern rationales directed toward state conduct.  In addition, a purely subjective test is inappropriate because it effectively collapses the “person in authority” requirement into an assessment of subjective voluntariness.  I will review these arguments in turn.

74                      Cory J. relies on the majority judgment of this Court in Rothman, supra, for his conclusion that the relevant test is based on “the extent to which the accused believed the person could influence or control the proceedings against him or her” (para. 32 (emphasis added)).  However, the subjective test adopted by Martland J. must be understood in its particular context, for Rothman addressed the admissibility of a statement made to an undercover police officer during incarceration. Therefore, the objective requirement of state presence, a traditional person in authority, was never in doubt which explains why both parties conceded to the application of a subjective test in the case.  Indeed, Martland J. himself, at p. 664, specified that:

.  . . a subjective test should be applied in the circumstances of this case. I also agree with [the] conclusion that McKnight [the undercover police officer] was not a person in authority because he was not regarded as such by the appellant.

75                      Martland J., at p. 663, quoted Jessup J.A., who in turn had quoted Kaufman, The Admissibility of Confessions, to the effect that notwithstanding the presence of a person who, “from a purely objective point of view, was in a position of undoubted authority”, the rules which attach to this status under the confessions rule will only apply if the accused believed that the person could make good on the promises or threats.  In such cases, a consideration of the effect of the inducement on the mind of the accused in the known circumstances is called for.  This subjective effect only became relevant once the objective threshold test was first met -- the statement was made to a person formally engaged in the detention of the accused, without doubt otherwise, a person in authority.

76                      This view of the Court’s test in Rothman is bolstered by the evident and laudable concern of the majority not to impede police investigations unduly, nor keep clearly relevant information from the trier of fact.  Lamer J., as he then was, outlined this germane concern in concurring reasons, at p. 697:

It must also be borne in mind that the investigation of crime and the detection of criminals is not a game to be governed by the Marquess of Queensbury rules. The authorities, in dealing with shrewd and often sophisticated criminals, must sometimes of necessity resort to tricks or other forms of deceit and should not through the [confessions] rule be hampered in their work.

This policy concern, the potential for the confession rule to hamper police investigation, and the objective reality of state involvement, precipitated the Court’s examination of the accused’s subjective belief.  In other words, in order to avoid giving the confessions rule an overly broad scope, the objective threshold test for person in authority was nuanced with a subjective element.  As assessed by Sopinka, Lederman and Bryant in The Law of Evidence in Canada (1992), at p. 351,  A.B., supra, determined that “even if it can be established on objective criteria that the person who obtains the statement could be characterized as an authority figure, that is not determinative of the issue”.

77                      The dissenting comments of Estey J. in Rothman are also instructive.  Embracing the more modern rationales of the confessions rule, Estey J. affirmed, at p. 650, the irrelevance of subjective belief where there is not, objectively, a state representative:

The principle [of voluntariness] itself, of course, requires (and this is an absolute requisite) that the statement in fact be made to a person in authority; and if this qualification is not met, then it matters not whether the person is known to the accused to be one in authority. This is because the principle adopted for the protection of the integrity of the administration of justice is founded upon the realization that persons in authority, instrumentalities of the State, must observe certain basic rules. [Emphasis added.]

In my view, this passage properly demonstrates how the confessions rule is, first and foremost, concerned with state conduct. However, given the concern of unduly hampering legitimate investigative work, the majority properly determined that subjective belief also becomes relevant where there is, in fact, an “instrumentality of the State” present. In Rothman, although the disguised police officer was in fact a person in authority, he was not to be regarded as such under the confessions rule because the accused did not believe him to be so.

78                      Rothman, therefore, stands for the proposition that, in some instances, a person with undoubted objective authority will not retain this status under the confessions rule because the accused was unaware of this authority and it cannot be said that the latter bore upon the voluntariness of the admission.  In such circumstances, the confessions rule promulgates the fiction that a police officer is not a “person in authority”.  This fiction is only necessary, as Lamer J. noted, at p. 680, where the rule is couched in terms of voluntariness as opposed to reliability, and the subjective impact of the actual authority held out is assessed.  In summary, I conclude that the adoption in Rothman of a subjective test must be taken in its evident state context, where there was never any question that the objective threshold of a traditional person in authority was met.  The same reasoning applies to the case of R. v. Todd (1901), 4 C.C.C. 514 (Man. K.B.), which also involved undercover agents of the state.

79                       I, therefore, cannot agree with my colleague’s assertion at para. 32 that “from its earliest inception in Canadian law, the question as to who should be considered as a person in authority depended on the extent to which the accused believed the person could influence or control the proceedings against him or her”.  The authorities suggest on the contrary, in my view, that the first requirement is that such a person has actual authority or control over the proceedings against the accused, and that  the subjective belief is an additional requirement.  Pointedly, Rothman depicts the utility of a subjective test in undercover cases, whereby the scope of “persons in authority” is narrowed, the fiction propagated that objective persons in authority were not to be recognized as such, and the relevant evidence admitted before the trier of fact.

80                      The pivotal relationship of the confessions rule with objective state conduct was affirmed in R. v. McIntyre (1993), 135 N.B.R. (2d) 266 (C.A.), aff’d [1994] 2 S.C.R. 480.  There the accused had been released by the police, and five months later made statements to undercover officers who passed themselves off as criminals.  Considering the concern expressed in Rothman that police may sometimes of necessity resort to tricks, the Court of Appeal held at p. 277 that there was no reason to exclude the statement in the circumstances because:

. . . there was no reason to protect the appellant from the power of the State. He was free in his comings and goings and he was in no way being restricted by the police. There was no coercion in this case. The appellant could have left the police officers and had nothing further to do with them at any time. The statement should therefore be received in evidence. It would then be the responsibility of the jury to determine the weight to be given to the statement.

81                      McLachlin J.’s examination of the confessions rule in R. v. Hebert, [1990] 2 S.C.R. 151, also suggests this approach.  In examining the nature of the choice protected by the confessions rule, she elaborated at p. 165 that the detained person only has a negative right:

... the right not to be tortured or coerced into making a statement by threats or promises held out by a person who is and whom he subjectively believes to be a person in authority. [Emphasis added.]

Clearly both objective and subjective elements are present in this definition.  However, a consideration of subjective factors is relevant only in so far as they are linked to the police methods employed (R. v. Amyot (1990), 58 C.C.C. (3d) 312 (Que. C.A.), at p. 322) because it is the “person in authority” factor which triggers concerns about voluntariness. See also R. v. Frewin (1855), 6 Cox C.C. 530; R. v. McKenzie, [1965] 3 C.C.C. 6 (Alta. S.C.A.D.); Collins, supra.  Therefore, the test is not only whether the person is one whose promise or threat would influence the accused, but also, and above all, whether that person had control over the course of the prosecution.

82                      This test for “person in authority” preserves the original sense of the confessions rule that voluntariness need only be established where statements are made to representatives or agents of the state because it does not resort to definitions of authority that are structured by factors of voluntariness. In this regard, I am persuaded by the cogent reasoning of Laycraft C.J. in R. v. Sweryda (1987), 34 C.C.C. (3d) 325 (Alta. C.A.), as to the problems with premising the “person in authority” test on the subjective belief of the accused. He identifies, at pp. 329-30, the difficulty with a highly subjective test to be that it reverses the proper method of analysis.

. . . the [subjective] rule . . . is that if the accused believes the maker of the threats or promises can make good on them, that fact alone transforms the maker into a person in authority regardless of his previous status. The usual method of analysis is thus reversed. One first uses voluntariness as the test to determine whether the maker of the threat or promise is a person in authority. If a threat or promise induced the statement, the analysis concludes at once because the maker thereby becomes a person in authority. Indeed, the rule as formulated might be stated in simpler terms without going through the formality (or the subterfuge) of finding the maker of the threat or promise to be a person in authority: If a statement is induced by a promise or a threat which the accused thinks the maker could make good, the statement is inadmissible. [Emphasis in original.]

In other words, if we begin the analysis by focusing on the effects of threats or inducements in relation to subjective beliefs, and allow this to define “persons in authority”, we effectively collapse the cart into the horse, as it is the provision and effect of inducements which come to define inappropriately who are “persons in authority”. Such an approach not only unduly expands the exceptional rule of evidence, but is also inconsistent with our recognition of the essential role played by the concept of “person in authority” in the confessions rule.

83                      In summary, the proper test for “person in authority” examines, first, the objective status of the person to whom the statement was made, and only where they are identified as someone formally engaged in the arrest, detention, interrogation or prosecution of the accused, is it then necessary to examine whether the accused believed that the person could influence or control the proceedings against him or her.  While Cory J. adds a “reasonableness” element to the latter subjective element, I respectfully find this to be an inadequate objective requirement that continues to focus on the belief of the accused.

84                      Furthermore, I think that my colleague’s remarks at para. 39 concerning my test play both ways and also apply to his proposed procedure.  Indeed, according to his view, a trial judge may similarly see his findings relating to the reasonable belief of the accused “vitiated” if the receiver is found afterwards not to be a person in authority.  The “vitiating” risk exists in both inquiries.  Therefore, in my opinion Cory J.’s comments are not convincing and do not alter my conclusion that the assessment of the objective status of the receiver should be made prior to the examination of the accused’s belief.

IV.   Categories of “Persons in Authority”

85                      Who then might fall into the category of “persons in authority” under the confessions rule?  As mentioned earlier, Cory J. on the basis of his test concludes at para. 49  that “[v]irtually any category of person -- parents of the accused, parents of the complainant, teachers, psychiatrists, physicians -- may, in light of the particular evidence adduced, be considered to be a person in authority”.  In contrast, upon the threshold objective test outlined above, I find that the concept may not encompass persons beyond those formally engaged, including agents, in the arrest, detention, examination or prosecution of the accused.  Only in the rarest of cases will non-traditional persons in authority fall into this category, and it still must be objectively established on the facts that such persons had actual control in the proceedings.

86                      In this regard, where early cases like R. v. Trenholme (1920), 35 C.C.C. 341 (Que. K.B.), extended the status of “person in authority” to a complainant’s parents, this perspective must be confined to its historical realities.  As Peter K. McWilliams observes, this case reflects the social condition of the time where there were no modern police forces and the complainants had to prosecute crimes for themselves (Canadian Criminal Evidence, vol. 2 (3rd ed. 1988 (loose-leaf)), at para. 15:10730). I agree that cases which, today, find the complainant to be a person in authority are inappropriate carry-overs from such days. 

87                      Indeed, such authority must be objectively proven on the facts. For example, in Sweryda, supra, the Alberta Court of Appeal reasonably found that a social worker was a “person in authority” on the basis that she was investigating a criminal act and had the statutory power to institute proceedings against the accused.  The court correctly began with this objective fact concerning the authority vested in the social worker under the Child Welfare Act, before then examining the subjective perception of the accused.

88                      Today, however, complainants, their parents or witnesses do not have the discretion to decide whether or not to proceed against an accused, and thus they typically cannot constitute persons in authority.  And where they are, in fact, formally acting as agents of the state, the accused must have subjective knowledge of this relationship before they can be considered “persons in authority” under the confessions rule.

89                      Thus, the case of A.B., supra, properly held that an accused child’s subjective belief that his mother was a person in authority did not render her to be one. In this instance, Cory J.A., as he then was, found at p. 28 that there was no “realistic and close connection” between the mother’s decision to call the authorities and the offered inducement to make a statement. While I agree with this conclusion, I find the “close connection” test not only too weak an objective requirement of state involvement, but also a vague and confusing criterion, fraught with uncertainty in the law and difficult to apply.

90                      In R. v. Kyle (1991), 68 C.C.C. (3d) 286, on the basis of A.B.’s close connection test, the Ontario Court of Appeal held that a 15-year-old complainant who had tape-recorded a confession of the considerably older accused was a person in authority because he had the intention to go to the police armed with this statement and made threats on this basis.  In my view, this case demonstrates the shortcoming not only of the “close connection” test, but also of an approach to the definition of “person in authority” which requires no more than the power to influence proceedings against the accused. The reality is that every complainant typically has the power to “influence” proceedings in so far as they may only ever commence upon his or her complaint.  Such powers of influence similarly reside with all witnesses and various judicial actors, including undercover agents, jailhouse informants and state doctors.  All can be said to have a “realistic and close connection” to the prosecuting state.

91                      Indeed in McKenzie, supra, at p. 28, it was acknowledged that witnesses, complainants and doctors always have the power to influence the course of the prosecution by withholding or colouring evidence, even to the extent of affecting the verdict.  However, this did not make them “persons in authority”.  See also Wilband v. The Queen, [1967] S.C.R. 14, where the Court determined that the confessions rule did not apply to statements made to psychiatrists examining the convicted offender under Criminal   Code  provisions.  The crucial point is that not only do such persons not have control over proceedings, but also the police themselves have no control over the conduct of either the accused or complainant.  Therefore, the exclusion of privately induced statements at that stage would not serve any principle of deterrence or fair treatment at the hands of the state.  Essentially, state involvement must be explicit.  It is not enough to allege that a person might “influence” proceedings.

92                      The basic power to influence proceedings cannot be enough to satisfy the objective threshold of state involvement, and there is no need to examine the accused’s subjective beliefs in such a scenario.  Regardless of a person’s prospective intention to contact the authorities, as in Kyle, the accused’s reasonable apprehension of the complainant’s power over him, and the evident involuntariness of his statements, it would not serve the interests of the confessions rule to exclude such statements. The defence has ample opportunity to impugn their reliability in the course of the trial, and the statements should be admitted, like any other evidence, subject to the scrutiny and weight attached to them by the triers of fact.

93                      As this discussion shows, I cannot agree with the “close connection” test set out in A.B., supra.  Such a test has the potential, I fear, to reappear in the test for “person in authority” adopted by my colleague in this case, which invokes the same “power of influence” standard and rests on an ambiguous assessment of what might make a subjective belief “reasonable”.  In practice, the two tests have the likely potential to amount to the same uncertain standard, which I find to be an unacceptable proposition.  And indeed, as examined below, this precise test surfaces in my colleague’s approach to when the trial judge has a duty to conduct a voir dire.  If it is coercive state action which the confessions rule aims to deter, how is this principle served by excluding evidence acquired by the questionable conduct of private parties?  In my view, it is not.

94                      This objective threshold test for “persons in authority” reflects the practical reality of the law as it has long been applied in Canada and conforms to the modern rationales of the confessions rule. Significantly, it is clear and straightforward to apply. Without some state presence, there is no legitimate purpose served in excluding evidence.  In fact, the trier of fact’s important task of ascertaining the truth is impeded and this imposes a substantial cost on the societal interest in law enforcement. The public interest of placing probative evidence before juries for the purpose of arriving at truthful decisions concerning guilt or innocence should not be wholly misplaced by the confessions rule.  The discussion in R. v. Unger (1993), 83 C.C.C. (3d) 228 (Man. C.A.), at p. 247, which quotes the majority in R. v. Corbett, [1988] 1 S.C.R. 670, is apt on this basic point made in Corbett, at p. 697:

. . . basic principles of the law of evidence embody an inclusionary policy which would permit into evidence everything logically probative of some fact in issue, subject to the recognized rules of exclusion and exceptions thereto. Thereafter the question is one of weight. The evidence may carry much weight, little weight, or no weight at all. If error is to be made it should be on the side of inclusion rather than exclusion and our efforts in my opinion, consistent with the ever-increasing openness of our society, should be toward admissibility unless a very clear ground of policy or law dictates exclusion. [Emphasis added.]

95                      In conclusion, the proper “person in authority” test explicitly includes both an objective and subjective element. First, the objective status of the person to whom the statement was made must be established, and only where they are identified as someone formally engaged in the arrest, detention, interrogation or prosecution of the accused, is it necessary to then examine the subjective belief of the accused in this regard.  This approach furthers the modern principles underlying the confessions rule, and does not erect additional unwarranted barriers to the admission of relevant evidence. It recognizes that a subjective approach to the test for “persons in authority” was primarily adopted to address the circumstance of undercover state agents.  However, it remains a necessary prerequisite to find that the statement was made to an actual person in authority or an “instrumentality” of the state.

V.   The Trial Judge’s Obligation to Direct a Voir Dire

96                      Having set out this clear position of what constitutes the proper test for “persons in authority” under the confessions rule, I turn now to the question of when a trial judge has an obligation to hold a voir dire and determine whether a statement was made to a “person in authority”, and if so, was voluntary.  This obligation should be precisely defined in order to promote certainty and help facilitate the judge’s difficult task of judicially conducting a trial.

97                      As a general rule, the defence has the onus of requesting a voir dire raising the issue that the accused’s statement was made to a “person in authority” and thus putting the statement’s admissibility into question.  In the voir dire itself, as elaborated by my colleague in para. 37, the accused has an evidential burden of ensuring there is enough evidence on the record to make the issue of whether the statement was made to a “person in authority” a live one such that voluntariness must be established.

98                      The imposition of this basic burden on the accused is reasonable because it encourages parties to raise the issue of admissibility of evidence at trial, and saves counsel from seeking a voir dire or waiver concerning every statement made by the accused to anyone, in order to avoid the possibility of a successful appeal on that basis.  More importantly, the determination of who is a “person in authority” derives from a close analysis of the facts, to which the subjective belief of the accused is relevant. As such, the imposition of this onus on the defence follows from the nature of the inquiry.

99                      In addition, this onus is consistent with the basic role of defence counsel in our adversarial system, whose competence in making tactical decisions and responsibly representing their client’s best interests we have recognized (R. v. Lomage (1991),  2 O.R. (3d) 621 (C.A.)).  In this regard, a failure to object to the admission of evidence may be based on a conscious decision of what best serves the interests of the defence. The basic duty to request a voir dire is consistent with this reality, and reflects the importance of some judicial deference to the competence of defence counsel and the avoidance of any undue interference with their trial strategy.

100                    Notwithstanding this responsibility borne by the accused, a trial judge still has the basic duty of ensuring the fair conduct of a criminal trial apart from lapses of counsel.  Thus, in exceptional cases, she or he may have the duty to conduct a voir dire into the voluntariness of a confession even if there is no request from counsel.  As Martin J.A.  held in R. v. Sweezey (1974), 20 C.C.C. (2d) 400 (Ont. C.A.), a voir dire should be held either where it is requested, or absent such a request, where there are circumstances that cast doubt on the statement’s voluntariness.  In Erven v. The Queen, [1979] 1 S.C.R. 926, a plurality of this Court clarified that the very making of a statement to a person in authority is sufficient to cast doubt on the statement’s voluntariness.  In other words, there is no need for evidence of actual involuntariness.  In this regard, the trial judge’s duty to conduct a voir dire, absent the accused’s waiver of this right, can arise on the facts before him or her.

101                    The critical issue upon which my colleague and I disagree is what circumstances trigger the trial judge’s obligation to hold a voir dire -- where does the evidence reveal the realistic possibility that a statement has been made to a person in authority.  This issue centres on the “person in authority” test, as it is the factor which triggers the application of the confessions rule. Logically, therefore, the circumstances under which the duty to hold a voir dire will arise should consistently mirror its definition.

102                    As a basic proposition, the duty to hold a voir dire is triggered where the evidence shows the reasonable possibility that the accused’s statement was made to a person in authority.  Mirroring the test for “person in authority” which I have adopted above, if the evidence reveals the reasonable possibility that first, the admission was made to someone formally engaged in the arrest, detention, examination or prosecution of the accused, and second, the accused might objectively have known of this authority, then a voir dire must be conducted.  At this stage, there is an air of reality to the claim that the confession was made to a person in authority, voluntariness becomes a live issue and the trial judge has a duty to hold a voir dire.

103                    Despite adopting a highly subjective approach to the person in authority requirement, my colleague articulates a purely objective test for when the trial judge’s obligation to conduct an inquiry is triggered.  In his view, the evidentiary threshold is met where the evidence demonstrates “that the receiver of the statement made by the accused was closely connected to the authorities” (para. 45, emphasis added).  There is no need for any evidence as to the subjective belief of the accused.

104                    My concerns with this approach are two-fold. First, the vagueness of the “close connection” test generates unwarranted uncertainty. As examined above, it does not adequately serve the principle of state deterrence which justifies the confessions rule, and is difficult to apply.  As such, it fails to provide a bright line defining the trial judge’s responsibility to raise the issue of the confession’s admissibility.  As virtually anyone has the potential to be closely connected with the authorities, judges will be inclined to canvass the issue in a voir dire whenever an out-of-court statement by the accused is tendered.  To enforce such a practice is unacceptable because it contradicts basic rules of evidence, might unduly interfere with defence counsel tactics and is contrary to the proper administration of justice.

105                    These concerns are in fact realized in the companion case of R. v. Wells,  [1998] 2 S.C.R. 517, where my colleague finds that two visits to the RCMP constituted a “close connection” sufficient to give rise to the trial judge’s duty to hold a voir dire.  On this approach, any connection with the authorities becomes enough to constitute a “close connection”.  In my respectful view, this conclusion is incompatible with my colleague’s assertion, at para. 47, that the evidence must clearly show the person to be “acting in concert with the police or prosecutorial authorities” and demonstrates the unacceptable ambiguity of the “close connection” test.  On this approach, the exceptional duty to conduct a voir dire without requests by counsel has the potential to arise in a number of unwarranted situations.

106                    My second concern with the objective “close connection” test of my colleague is that it fails to include any recognition of the subjective element in the “person in authority” definition.  The subjective knowledge of the accused is assumed in every case.  This assumption strikes me as unreasonable and unprincipled.  In the situation of undercover agents, where the status of the person is hidden from the accused, there is no reasonable basis for such an inference. In other words, there is no air of reality to the claim that the statement was made to a “person in authority” or was involuntary.  Along these lines, in R. v. Pettipiece (1972), 7 C.C.C. (2d) 133 (B.C.C.A.), Branca J.A. properly held that a voir dire was not necessary for statements made to an undercover agent in jail, as he was not a person in authority.  Indeed, to oblige a trial judge to hold a voir dire automatically absent any evidence that the accused might have known of the person’s authority serves neither the rationale justifying the exclusionary rule nor the pursuit of truth and is ultimately not in the interest of the administration of justice.

107                    The need for some evidence that the accused might have believed he was speaking with a person in authority is not onerous, and would not render the trial judge’s obligation meaningless as alleged by my colleague (para. 42).  Instead it consistently reflects the basic definition of “person in authority”, and reinforces the general onus on the defence to raise the issue of voluntariness.  This approach is preferable to the “close connection” test which categorically calls for a voir dire even if there is no possibility that the accused could have been influenced by any authority.

108                    I conclude, therefore, that the trial judge’s duty to hold a voir dire arises where the evidence reveals the realistic potential that the admission was made to a representative of the state and the accused might objectively have known of this authority.

109                    Before canvassing the different circumstances in which this duty will arise, it is worth emphasizing that only evidence available on the record prior to the admission of the accused’s statement is relevant to the determination of what the trial judge’s duty was. Evidence introduced after its admission does not bear upon this duty.  My colleague implies this requirement in his statement, at para. 41, that “the trial judge will only have committed reversible error if clear evidence existed in the record which objectively should have alerted him or her to the need for a voir dire notwithstanding counsel’s silence”.  Obviously, a trial judge cannot anticipate forthcoming evidence and what duty it might impose on him or her, and a basic evidentiary burden on the accused in this regard is justifiable.

110                    In practical terms then, when might the trial judge’s duty arise?  In the most straightforward situation, where a statement is made to an obvious representative of the state, and nothing indicates the accused would be unaware of this fact, the evidential burden is met. The voluntariness of the statement is a live issue, and the trial judge must hold a voir dire to establish the statement’s admissibility, regardless of whether or not it is requested by counsel.  Essentially, where the accused has made a statement to a police officer, or to other obvious “persons in authority”, it is reasonable to infer the accused’s knowledge of this authority, and the trial judge is obliged to hold a voir dire unless this right is waived by the accused (see Powell v. The Queen, [1977] 1 S.C.R. 362; Erven, supra, and Park v. The Queen, [1981] 2 S.C.R. 64).

111                    The more complicated situation involves non-obvious persons in authority, for the evidential burden on the accused to make the admissibility of the statement a live issue is not automatically met in such cases.  Not only can we not automatically assume the accused’s knowledge of any authority, but also there must be a realistic potential that this person is acting as an agent of the state.  In the court below (R. v. M.C.H. (1996), 91 O.A.C. 298), Finlayson J.A. addressed this situation as follows (at p. 302):

When a statement is made to a person or persons not ordinarily engaged in the arrest, detention, examination, or prosecution of accused persons, it appears to me that the defence has a responsibility to raise the issue at trial and ask for a voir dire to determine if the person to whom the confession was made was in fact a person in authority.  The defence bears this responsibility because the subjective reaction of the accused to his or her interrogator is essential to the ultimate determination of voluntariness. Absent the presence of a conventional authority figure at the time when an inculpatory statement is made, the trial judge must be placed on notice that the voluntariness of the statement is a live issue.

I agree with Finlayson J.A. that as a general rule the accused has the duty to request a voir dire in the case of non-obvious persons in authority.  Where requested, the trial judge has an obligation to grant the accused the opportunity to establish that the statement was made to a person in authority.

112                    In the rarest of circumstances, however, even without any request by counsel, there may be enough evidence on the record prior to the statement’s admission which should have alerted the trial judge to the need for a voir dire.  As outlined above, there must be the realistic potential that the person was formally acting for the state and the accused could have known of this relationship. Otherwise, it would serve no legitimate purpose to interrupt the trial to conduct a voir dire.

113                    In summary, the trial judge has a duty to hold a voir dire where there is a reasonable possibility that the statement was made to a person in authority.  This will automatically arise in cases of obvious persons in authority, subject to informed waiver by the accused.  In the case of non-obvious persons in authority, however, the evidence must reveal both the reasonable possibility that the person was an agent of the state and the accused could have known of this status. As a general rule, the accused bears the onus of requesting this trial within a trial. However, in exceptional circumstances, the evidence itself will require the trial judge to conduct a voir dire proprio motu.

VI.   Application to the Facts

114                    In the facts of this appeal, like my colleague, I have no trouble concluding that there was no realistic possibility that the complainant and her immediate family constituted persons in authority for purposes of the confessions rule, and therefore the trial judge had no duty to hold a voir dire.  There was absolutely no evidence available to the trial judge prior to the admission of the statements that these persons were acting as agents of the state and had control over any prospective proceedings against the accused.  Indeed, it is clear that they were not persons in authority with any control over the arrest, detention, examination or prosecution of the accused. 

115                    Therefore, I agree with my colleague that the evidence failed to trigger the trial judge’s obligation to hold a voir dire and the appeal should be dismissed.  The evidence was properly admitted to the trier of fact for assessment of its reliability in all the surrounding circumstances.

Appeal dismissed.

Solicitor for the appellant:  Irwin Koziebrocki, Toronto.

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

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

Solicitor for the intervener the Attorney General of Quebec:  The Attorney General of Quebec, Sainte‑Foy.

Solicitor for the intervener the Attorney General of British Columbia:  The Attorney General of British Columbia, Victoria.

Solicitor for the intervener the Attorney General for Alberta:  The Attorney General for Alberta, Edmonton.