48 Crim LQ474

DATE DOWNLOADED: Thu Nov 16 05:33:34 2023
SOURCE: Content Downloaded from HeinOnline
Citations:
Please note: citations are provided as a general guideline. Users should consult their preferred
citation format's style manual for proper citation formatting.
Bluebook 21st ed.

Steven Penney, What's Wrong with Self-Incrimination - The Wayward Path of
Self-Incrimination Law in the Post-Charter Era - Part III: Compelled Communications,
the Admissibility of Defendants' Previous Testimony, and Inferences from Defendants'
Silence, 48 CRIM. L.Q. 474 (2004).
ALWD 7th ed.

Steven Penney, What's Wrong with Self-Incrimination - The Wayward Path of
Silence, 48 Crim. L.Q. 474 (2004).
APA 7th ed.

Penney, S. (2004). What's wrong with self-incrimination the wayward path of
self-incrimination law in the post-charter era part iii: compelled communications,
the admissibility of defendants' previous testimony, and inferences from defendants'
silence. Criminal Law Quarterly, 48(4), 474-532.
Chicago 17th ed.

Steven Penney, "What's Wrong with Self-Incrimination - The Wayward Path of
Silence," Criminal Law Quarterly 48, no. 4 (May 2004): 474-532
McGill Guide 9th ed.

Steven Penney, "What's Wrong with Self-Incrimination - The Wayward Path of
Silence" (2004) 48:4 Crim LQ 474.
AGLC 4th ed.

Steven Penney, 'What's Wrong with Self-Incrimination - The Wayward Path of
Silence' (2004) 48(4) Criminal Law Quarterly 474
MLA 9th ed.

Penney, Steven. "What's Wrong with Self-Incrimination - The Wayward Path of
Silence." Criminal Law Quarterly, vol. 48, no. 4, May 2004, pp. 474-532. HeinOnline.
OSCOLA 4th ed.

Steven Penney, 'What's Wrong with Self-Incrimination - The Wayward Path of
What's Wrong with Self-Incrimination?
The Wayward Path of Self-Incrimination
Law in the Post-Charter Era
Part III: Compelled Communications,
the Admissibility of Defendants'
Previous Testimony, and Inferences
from Defendants' Silence
Steven Penney*
6. Compelled Communication
As we have seen, police have no legal power to compel criminal
suspects to answer questions. To obtain self-incriminating
evidence, they must either convince suspects to speak voluntari-
ly or employ surreptitious methods like undercover questioning
or electronic surveillance. In some circumstances, however, the
law empowers governmental authorities other than police to
compel potentially self-incriminating communications.' This
section describes and critiques the legal rules associated with these
powers.
Associate Professor, Faculty of Law, University of New Brunswick. This is the
third and final part of the article. Part one can be found at (2003), 48 C.L.Q. 249
and part two at (2003), 48 C.L.Q. 280.
1. The recent adoption of the investigative hearing provisions in the Anti-terrorism
Act, S.C. 2001, c. 41, s. 4, created an exception to this principle. These provi-
sions, which are discussed in more detail infra, footnote 45, permit police to
apply to a judge for an order compelling persons to respond in court to questions
relating to terrorism offences. See Criminal Code, R.S.C. 1985, c. C-46, ss. 83.28
to 83.29.
2004] What's Wrong with Self-Incrimination? Part 11I 475
(1) Testimonial Compulsion

(a) The State of the Law
Criminal defendants are not compellable witnesses at their
own trials.2 But like any other person, a criminal suspect may be
compelled to give testimony in other legal proceedings. Suspects
may be ordered to testify as witnesses at other persons' criminal
trials, at civil trials or at any other civil proceeding empowered
to issue subpoenas, such as coroners' inquests and regulatory
inquiries. If they fail to comply, they may be punished for
contempt. This raises the possibility that the state could do
indirectly what it is forbidden to do directly - compel suspects'
testimony for use against them in future criminal proceedings.
Over the years, courts and legislators have adopted different
strategies to prevent this. At common law, witnesses were
entitled to refuse to answer questions that might incriminate
them.' This was the so-called "privilege" against self-
incrimination. Parliament abrogated the privilege by enacting
s. 5(1) of the Canada Evidence Act, which requires witnesses to
answer potentially self-incriminating questions.' Section 5(2) of
the Act, however, prevents the state from adducing those answers
in later criminal proceedings against those witnesses, as long as
they were under compulsion and claimed the Act's protection at
the previous proceeding.' The enactment of s. 13 of the Charter
2. Section 11 (c) of the Charter grants a person charged with an offence the right "not
to be compelled to be a witness in proceedings against that person in respect of
the offence". At common law, criminal defendants were not competent to testify
for either the defence or prosecution. In 1893, Parliament amended what is now
s. 4(1) of the CanadaEvidence Act, R.S.C. 1985, c. C-5, to allow accused persons to
testify in their own defence. See Canada Evidence Act, 1893, S.C. 1893, c. 31, s. 4.
3. See E. Ratushny, Self-incrimination in the CanadianCriminal Process (Toronto:
Carswell, 1979), p. 78.
4. Section 5(1) states: "No witness shall be excused from answering any question on
the ground that the answer to the question may tend to criminate him, or may tend
to establish his liability to a civil proceeding at the instance of the Crown or of
any person." Similar rules apply to proceedings governed by the various provin-
cial evidence statutes. See J. Sopinka, S. Lederman and A. Bryant, The Law of
Evidence in Canada,2nd ed. (Toronto: Butterworths, 1999), p. 823.
5. Section 5(2) states:
Where with respect to any question a witness objects to answer on the ground that
his answer may tend to criminate him, or may tend to establish his liability to a civil
476 Criminal Law Quarterly [Vol. 48
eliminated these prerequisites. 6 Section 13 prohibits the

prosecution from adducing defendants' previous testimony to
incriminate them, even if they testified voluntarily or failed to
invoke the statutory protection.7
Jurists have argued, however, that these provisions do not do
enough to protect witnesses from incriminating themselves.8 They
do not forbid the state from prosecuting people who have pro-
vided compelled testimony at previous proceedings. And while
testimony compelled from defendants is in itself inadmissible, the
provisions do not stop authorities from using it to discover and
adduce other inculpatory evidence. So while both the Charter and
the CanadaEvidence Act give criminal defendants "use immunity",
neither enactment expressly gives them either "transactional
immunity" (prohibiting authorities from charging a witness with
any offence relating to the content of their testimony) or "deriva-
tive use immunity" (excluding evidence that would not have been
found but for compelled testimony).
Before the enactment of the Charter, the Supreme Court of
Canada (the Court) strained to fill these gaps by reading s. 5 of
the Canada Evidence Act narrowly. In Batary v. Saskatchewan
(Attorney General), it held that a coroner could not compel a
person charged with murder to testify at an inquest into the
killing.9 Section 5 did not apply, the court reasoned, because
proceeding at the instance of the Crown or of any person, and if but for this Act,
or the Act of any provincial legislature, the witness would therefore have been
excused from answering the question, then although the witness is by reason of
this Act or the provincial Act compelled to answer, the answer so given shall not
be used or admissible in evidence against him in any criminal trial or other crim-
inal proceeding against him thereafter taking place, other than a prosecution for
perjury in the giving of that evidence or for the giving of contradictory evidence.
See also R. v. Marcoux, [1976] 1 S.C.R. 763 at pp. 768-69, 24 C.C.C. (2d) 1, 60
D.L.R. (3d) 119.
6. It states: "A witness who testifies in any proceedings has the right not to have any
incriminating evidence so given used to incriminate that witness in any other pro-
ceedings, except in a prosecution for perjury or for the giving of contradictory
evidence."
7. R. v. Dubois, [1985] 2 S.C.R. 350 at p. 360, 22 C.C.C. (3d) 513,48 C.R. (3d) 193.
8. See Canadian Committee on Corrections, Toward Unity: Criminal Justice and
Corrections (Ottawa: Queen's Printer, 1969) at p. 68; Ratushny, supra, footnote
3, at pp. 87 and 349-52.
9. [1965] S.C.R. 465, [1966] 3 C.C.C. 152, 46 C.R. 34.
2004] What's Wrong with Self-Incrimination? PartII1 477
witnesses facing criminal charges are subject to the rules

regarding accused persons, not ordinary witnesses. The provin-
cial statute purporting to compel the accused's testimony was
legislation relating to criminal law and was therefore ultra vires
the legislature.
Batary promised criminal suspects expansive immunity from
legal compulsion.' 0 But its tortured logic did not stand the test of
time. The Court eventually confined Batary to its particular facts,
holding that uncharged suspects are compellable witnesses at
coroners' inquests" and other inquiries.' 2
The enactment of the Charter presented another opportunity to
bolster the protection provided by the Canada Evidence Act. As
discussed, s. 13 effected only modest reform.' 3 But lawyers soon
began asserting that legislation compelling suspects' testimony
violated the "principle against self-incrimination" implicit in s. 7
of the Charter. After some initial confusion,'4 the Supreme Court
concluded in R. v. S. (R.J.) that in most cases witnesses can be
compelled to provide self-incriminating testimony.'5 Section 7
requires, however, that they be given a limited form of derivative
10. See Ratushny, supra, footnote 3, at p. 81; W. Henkel, "Competency,

Compellability and Coroners' Courts" (1970), 12 C.L.Q. 166.
11. R. v. Faber,[1976] 2 S.C.R. 9, 27 C.C.C. (2d) 171, 65 D.L.R. (3d) 423.
12. Di lorio v. Quebec (Warden of the Common Jail of Montreal), [1978] 1 S.C.R.
152, 33 C.C.C. (2d) 289, 73 D.L.R. (3d) 491; Quebec (Attorney General) v.
Canada(Attorney General), [1979] 1 S.C.R. 218, 43 C.C.C. (2d) 49, 6 C.R. (3d)
145. See generally R. v. S. (R.J.), [1995] 1 S.C.R. 451 at paras. 123-30, 96 C.C.C.
(3d) 1, 36 C.R. (4th) 1.
13. See also D.M. Paciocco, "Self-Incrimination and the Case to Meet: The Legacy
of Chief Justice Lamer" (2000), 5 Can. Crim. L. Rev. 63 at p. 68.
14. In Thomson Newspapers v. Canada (Director of Investigation and Research,
Restrictive Trade Practices Commission), [1990] 1 S.C.R. 425, 54 C.C.C. (3d)
417, 76 C.R. (3d) 129, each member of the five-panel bench gave reasons.
Everyone agreed that s. 7 could in some circumstances provide residual protec-
tion against self-incrimination beyond that contained in s. 13. But they could not
agree on its scope. La Forest J. and Wilson J. agreed that when triggered, s. 7 pro-
vided derivative use immunity. But Wilson J.'s conception of derivative evidence
was more expansive than La Forest J.'s. See ibid. at pp. 483-84, Wilson J. and at
pp. 548-63, La Forest J. L'Heureux-Dub6 J. held that s. 7 provided use immuni-
ty only. See ibid., at pp. 583-85. And Sopinka J. concluded that it empowered wit-
nesses to resist compulsory self-incrimination altogether. See ibid., at pp. 603-06.
Lamer J. did not speak to the issue.
15. S. (R.J.), supra, footnote 12.
Criminal Law Quarterly [Vol. 48
use immunity' 6 encompassing "evidence which could not have

been obtained, or the significance of which could not have been
appreciated, but for [their testimony]".' 7 This conclusion was
justified, the court reasoned, by the "case-to-meet" principle and
the Charter's "basic distaste for self-conscription"."
The Court in S. (R.J.) also concluded that in limited circum-
stances compelling testimony would in itself violate s. 7. It
could not reach a consensus, however, on either the nature of
those circumstances or the appropriate remedy for a violation. 9
But soon after, in British Columbia Securities Commission v.
Branch, it held that the determining factor is "whether the
predominant purpose for seeking the evidence is to obtain
incriminating evidence or rather some legitimate public
16. The Court held that courts should exclude derivative evidence pursuant to their
common law discretion (constitutionalized by the fair trial right expressed in
s. 11(d) of the Charter) to exclude evidence that is more prejudicial than proba-
tive: ibid., at paras. 196-99. See also Thomson Newspapers, supra, footnote 14,
at pp. 560-61, La Forest J.
17. Ibid., at para. 191. Though this phraseology uses the word "could", Justice
Iacobucci, who wrote for the majority on this point, later specified that "[tihe
important consideration ... is whether the evidence, practically speaking, could
have been located. That is, would the evidence, on the facts, have otherwise come
to light?": ibid., at para. 195 (emphasis added). This formulation is very similar
to La Forest J.'s proposal in Thomson Newspapers, supra, footnote 14, at p. 561.
Justice lacobucci also held that while the ultimate burden to prove a s. 7 violation
lies with the party claiming the breach, generally the accused will only have to
"demonstrate a plausible connection between the proposed evidence and the prior
testimony". Thereafter the practical burden will rest with the prosecution, "since
it is the Crown which can be expected to know how evidence was, or could have
been, obtained": S. (R.J.), supra, footnote 12, at para. 203. In British Columbia
Securities Commission v. Branch, [1995] 2 S.C.R. 3 at para. 5, 97 C.C.C. (3d)
505, 38 C.R. (4th) 133 (hereafter Branch), the Court reiterated that once the
accused establishes a plausible connection, the "Crown will have to satisfy the
court on a balance of probabilities that the authorities would have discovered the
impugned derivative evidence absent the compelled testimony". See also R. v.
Primeau, [19951 2 S.C.R. 60, 97 C.C.C. (3d) 1, 38 C.R. (4th) 189; R. v. Jobin,
[1995] 2 S.C.R. 78, 97 C.C.C. (3d) 97, 97 W.A.C. 23.
18. S. (R.J.), ibid., at para. 83.
19. L'Heureux-Dubd J.(Gonthier J. concurring) concluded that s. 7 is violated when
the authority's predominant purpose in compelling the witness is "to build or
advance its case against that witness instead of acting in furtherance of those
pressing and substantial purposes validly within the jurisdiction of the body com-
pelling the testimony": ibid., at para. 278. But in most cases, she asserted, the
appropriate remedy would be to stay the proceedings (if and when they arise)
2004] What's Wrong with Self-Incrimination? PartIII 479
purpose"." Where the state's primary objective is legitimate,

witnesses must testify.2' And even where the predominant
purpose is to incriminate, witnesses are still compellable if the
"only prejudice is the possible subsequent derivative use of the
testimony". 22 If, on the other hand, testifying would threaten the
fairness of a witness's subsequent trial, then the witness may
refuse to testify.23
The Court also clarified in Branch that the issue of compella-
bility may arise at the time that the witness is called to testify (the
subpoena stage) as well as at subsequent criminal proceedings
against that witness (the trial stage). 4 If a s. 7 violation occurs at
the subpoena stage, then the remedy is to quash the subpoena.
Though the Court did not specify the appropriate remedy for
trial-stage violations, in most cases it will presumably be a stay
of proceedings. 5
against the witness rather than quashing the subpoena: ibid., at paras. 285-88.
Iacobucci J. (La Forest, Cory and Major JJ. concurring) agreed that the inquiry
should focus on purpose, but his elaboration of the point was obscure. He
declined to comment on the remedy issue: ibid., at paras. 146-54. Sopinka J.
(McLachlin J. concurring) argued that a witness should be able to resist compul-
sion where "prejudice to his or her interests overbears the necessity of obtaining
the evidence". He set out a list of factors judges should consider in balancing
these interests: ibid., at para. 326. Lamer J. agreed generally with Sopinka J.on
this point: ibid., at para. 4.
20. Supra, footnote 17, at para. 7. See also Primeau,supra, footnote 17; Jobin, supra,
footnote 17.
21. Branch, ibid.
22. Branch, ibid., at para. 9.
23. Ibid. In his concurring minority opinion in Phillips v. Nova Scotia (Commission
of Inquiry into the Westray Mine Tragedy), [1995] 2 S.C.R. 97 at paras. 82 and
85-86, 98 C.C.C. (3d) 20, 39 C.R. (4th) 141, Justice Cory (lacobucci and Major
JJ. concurring) mistakenly reiterated these holdings. Without purporting to con-
tradict Branch, in PhillipsCory concluded that where the witness establishes that
the predominant purpose of compulsion is to elicit self-incriminatory testimony,
the witness is never compellable. It is only where self-incrimination is not the pre-
dominant purpose that prejudice need be considered. In Branch the Court (in an
opinion written by Justices Sopinka and Iacobucci) stated clearly that where the
witness proves that the predominant purpose of the compelled testimony is self-
incrimination, the "witness should not be compelled unless the party seeking to
compel the witness justifies the compulsion": Branch, ibid., at para. 11 (empha-
sis added).
24. Ibid., at para. 10.
25. See R. v. Liakas (2000), 144 C.C.C. (3d) 359 (Que. C.A.) at para. 42.
(b) Critique
How well does this doctrine identify and balance the competing
interests raised by testimonial compulsion? Consider first the
reliability rationale. Even if we assume that compelling testimony
creates a significant risk of false confessions (a question that I
deal with later), this risk is eliminated by the use immunity
granted by s. 13 of the Charter. By definition, evidence derived
from compelled testimony has an independent existence. So the
fact that testimony leading to its discovery was compelled can
have no impact on its reliability. Rules limiting the admissibility
of evidence derived from compelled testimony, then, do not help
to prevent wrongful convictions.
But such rules do help to improve adjudicative accuracy. As I
have discussed, witnesses who have committed unlawful acts
will often be reluctant to testify truthfully. In many cases, they
will rationally perceive that the expected cost of lying (that is,
the probability x magnitude of punishment for perjury or giving
contradictory evidence) is less than the expected cost of telling
the truth (that is, the probability x magnitude of punishment for
offences revealed by their testimony). Prohibiting the state from
using witnesses' testimony against them in future proceedings
reduces the cost of truth-telling, thereby increasing the likelihood
that factually guilty witnesses will testify truthfully. And since
many of these witnesses would not have testified truthfully with-
out it, giving them immunity does not deprive the state of much
evidence of guilt. The net result is more socially beneficial, reli-
26
able evidence than would have been available without immunity.
Though it is rarely recognized by Canadian jurists, this simple
calculus explains a great deal of self-incrimination doctrine.27
26. There are circumstances, however, in which use immunity may induce witnesses to
testify falsely. This may occur when a witness sympathetic to the accused, aware of
the protection granted by s. 13 of the Charter, falsely testifies that he or she (and not
the accused) was the perpetrator of the offence. In R. v. Jabarianha,[2001] 3 S.C.R.
430, 159 C.C.C. (3d) 1, 47 C.R. (5th) 97, the Court held that in most circumstances
prosecutors may not attempt to discredit such witnesses by asking them about their
knowledge of s. 13. The prejudicial effect of such questioning would generally out-
weigh its probative value, the Court reasoned, because most witnesses would not be
confident that s. 13 would fully protect them from being prosecuted and convicted.
27. See generally J. McNaughton, ed., Wigmore on Evidence, vol. 8 (Boston: Little,
Brown, 1961), § 225; R. Cross and C. Tapper, Cross on Evidence, 6th ed.
(London: Butterworths, 1985), p. 383.
2004] What's Wrong with Self-Incrimination? PartII 481
The common law privilege against self-incrimination, s. 5(2) of

the Canada Evidence Act, and s. 13 of the Charter all provide
some assurance to witnesses that telling the truth will not
hurt them.28 The calculus also explains why courts have supple-
mented the use immunity granted by these rules with further
protections. Some witnesses will rationally conclude that use
immunity will not fully protect them from the damage that their
truthful testimony could cause them in future proceedings. Some
of these witnesses will only be prompted to tell the truth if they are
given a more extensive form of immunity.
The question, then, is whether the derivative use immunity
granted by the Supreme Court in S. (R.J.) is the best way to
induce truthful testimony. We want a rule that, in relation to both
present proceedings and future criminal trials, elicits more
socially beneficial, truthful testimony than any alternative rule.
The rule should therefore provide the minimum level of protec-
tion necessary to prompt witnesses to testify truthfully in current
legal proceedings, thereby maximizing the state's ability to use
the fruits of that testimony in future proceedings against them. 9
Although it was not conceived with this objective in mind,3"
S. (R.J.)'s "discoverability" rule is likely to come closer to
achieving it than any other alternative. It excludes evidence that
would not have been obtained but for the witness's testimony.
In theory, this eliminates the disincentive to truth-telling that
28. See Dubois, supra, footnote 7, at p. 384, where Justice McIntyre stated the
following in his dissenting reasons:
There is a social interest in encouraging people to come forward to give evi-
dence, not only in court but on other occasions in the tribunals and proceed-
ings referred to above. That interest is not served where witnesses in testify-
ing expose themselves to the danger of self-incrimination because of such
testimony. It is suggested that it was a recognition of this fact, together with
a recognition of the inadequacy of the law relating to self-incrimination and
the inadequacy of provincial powers in this respect that caused the framers
of the Charter to include the very greatly strengthened Charterprovisions
relating to self-incrimination.
See also R. v. Nodl, [2002] 3 S.C.R. 433 at paras. 22-24, 168 C.C.C. (3d) 193, 5
C.R. (6th) 1.
29. See A. Mewett, "The Right to Silence" (1990), 32 C.L.Q. 273 at p. 274.
30. Somewhat remarkably, not once in his 98-page majority reasons did Justice
Iacobucci advert to the notion that providing immunity against self-incriminating
testimony helps to induce truthful testimony.
compulsion would otherwise create. If authorities would have

found evidence derived from a witness's compelled testimony
wholly apart from that testimony, then testifying truthfully would
not make the witness's position any worse than testifying
falsely. Given the possibility, however remote, of prosecution
and punishment for lying, witnesses should therefore tell the
truth. The S. (R.J.) rule also (again theoretically) relieves the
prosecution of the burden of deciding whether calling a witness
will unduly prejudice any future prosecution of that witness. If
prosecutors are only prohibited from using evidence that could
not have been found apart from the testimony, then calling the
witness carries no risk - without the testimony, evidence
immunized by S. (R.J.) would not have been available.
Of course, the problem with derivative use immunity lies in
deciding whether evidence really would have been found apart
from the compelled testimony. Because this is a speculative
determination, it cannot be made with precision." Courts will
inevitably make mistakes.32 It is this difficulty that has animated
the debate over the type of immunity required by the fifth
amendment of the United States Constitution.33 Some argue that
transactional immunity is needed to ensure that compelled
testimony does not result in self-incrimination. From guilty
witnesses' perspective, the discoverability-based formulation of
derivative use immunity set out in S. (R.J.) is especially prob-
lematic. They may perceive that the risk of incorrect findings of
discoverability is greater than the risk of punishment for lying.
The opposite problem may also arise. Authorities may be
reluctant to call witnesses when they fear that courts may later
incorrectly decide that derivative evidence was not discoverable.
Despite these difficulties, S. (R.J.)'s limited derivative use
immunity is superior to any stronger form of protection, whether
full derivative use immunity (where the prosecution is prohibited
31. See S. (R.J.), supra, footnote 12, at para. 324, Sopinka J.
32. See I. Bloom, J. Nagy and G. Campbell, "The Residual Protection Against
Self-Incrimination in Canada: The Road Not Taken" (1995), 5 N.J.C.L. 363 at
pp. 392-93.
33. See e.g. R. Ghio, "The Iran-Contra Prosecutions and the Failure of Use
Immunity" (1992), 45 Stan. L. Rev. 229; J. Murphy, "The Aftermath of the Iran-
Contra Trials: The Uncertain Status of Derivative Use Immunity" (1992), 51 Md.
L. Rev. 1011.
2004] What's Wrong with Self-Incrimination?Part III 483
from adducing any evidence obtained as a result of the com-

pelled testimony) or transactional immunity. For most factually
guilty witnesses, the combination of s. 13's use immunity and
S. (R.J.)'s derivative use immunity, along with other incentives
for truth-telling (such as social norms promoting honesty, the
testimonial oath, and the threat of punishment for lying) will
likely be sufficient to induce honesty. Some witnesses, of course,
will only testify honestly if they are guaranteed greater protec-
tion against self-incrimination.34 However, in many of these
cases the authority contemplating compelling the testimony will
be in a position to provide such protection; that is, when it
believes that a witness will only testify truthfully under the pro-
tection of transactional or full derivative use immunity, and when
it believes that it is in the public interest to do so, the authority
can grant a witness greater immunity. 35 Limiting constitutional
protection against compelled self-incrimination to use immunity
plus discoverability-based derivative use immunity does not, in
other words, prevent authorities from offering greater protection
when it is needed to elicit truthful testimony. Transactional
immunity and full derivative use immunity are therefore
inappropriate. They would give more protection than is neces-
sary to induce most guilty witnesses to testify honestly. The state
would in some cases consequently be deprived of reliable
derivative evidence. Strong immunity rules, in summary, would
very likely produce a lesser quantity of trustworthy evidence
than S. (R.J.)'s limited derivative use immunity.
The question remains, however, whether the S. (R.J.) rule is
superior to a rule offering less protection against self-incrimination.
If limited derivative use immunity is better than full derivative or
transactional immunity, then it is possible that the best option would
be to offer compelled witnesses use immunity only.36 For many
witnesses, use immunity would be sufficient to remove the dis-
incentive to truthfulness created by the fear of self-incrimination.
34. Some witnesses, it should be noted, will not testify honestly no matter what
incentive is provided.
35. See Ratushny, supra, footnote 3, at pp. 353-54.
36. Note that I have not considered the possibility of eliminating the blanket provi-
sion of use immunity. This alternative would require both a radical break with
longstanding doctrine and a constitutional amendment and is therefore not
politically feasible.
When it would be insufficient, the compelling authority could

elect to offer stronger protection. A "use immunity only" rule
would therefore maximize the availability of derivative
evidence, permitting authorities to determine on a case-by-case
basis whether further immunity would be required to prompt
truthful testimony and whether it would be in the public interest
to grant it.
Although this alternative was the law for many decades, it
presents several practical challenges. The first is the difficulty of
determining whether a witness is likely to tell the truth only if
granted additional immunity. Not all factually guilty witnesses
will be suspected of criminal activity at the time that they are
questioned. In such cases authorities will not be aware of the
need to consider whether to offer them enhanced immunity.
Second, even when authorities are aware of the need for
enhanced immunity, they may not be in a position to offer it. This
is most likely to occur in the context of non-criminal, inquisitor-
ial proceedings. Administrative regulators and criminal prosecu-
tors often have different priorities. In cases where any criminal
prosecution would be carried out by an authority other than the
one compelling the testimony, prosecutors may be unwilling to
approve or honour immunity agreements proposed by the com-
pelling authority.37 Legislation could empower compelling
authorities to grant immunity on a case-by-case basis notwith-
standing the agreement of criminal prosecutors. But this would
only be effective for federal authorities. As a result of the consti-
tutional division of powers, immunity granted pursuant to
provincial legislation does not apply to the prosecution of federal
offences.3 8
Third, in some cases compelled communications are made
before authorities have a chance to offer immunity. As I discuss
in the next section, authorities sometimes compel statements and
37. The same problem may also occur even when witnesses are compelled to testify
in criminal proceedings. As Ratushny points out, immunity may be promised by
police, who are not legally entitled to fulfil that promise. Prosecutors may subse-
quently refuse to honour the promise of immunity. See Ratushny, supra, footnote
3, at p. 353.
38. See R. v. Walker, [1939] S.C.R. 214; Bell v. Klein (No. 1), [1955] S.C.R. 309;
R. v. White, [1999] 2 S.C.R. 417, 135 C.C.C. (3d) 257, 24 C.R. (5th) 201. See also
Ratushny, ibid., at p. 401.
2004] What's Wrong with Self-Incrimination? Part II 485
other self-incriminating communications outside the context of

formal legal proceedings. In many of these cases authorities will
have no opportunity to assess whether anything more than use
immunity is needed to induce honesty.
Of course, these problems may arise even under the S. (R.J.)
rule, which as we have seen gives limited derivative use immu-
nity to all compelled witnesses. To maximum truth-telling, we
would need to give all witnesses transactional immunity. But as
I have already argued, such a rule would be too broad. Choosing
between the S. (R.J.) rule and the "use immunity only" rule is
more difficult. The former produces more reliable evidence at
the time compelled testimony is given; the latter produces more
reliable evidence for use in future prosecutions. In my estimation
the S. (R.J.) rule is likely to produce a greater quantity of trust-
worthy evidence overall. Many guilty witnesses, especially those
competently instructed by counsel, know that their testimony,
while immunized from direct use, could lead authorities to other
incriminating evidence. Under a "use immunity only" rule, many
of these witnesses would be likely to take this knowledge into
account in relating their testimony. The state would consequently
gain neither the truthful testimony at the initial proceeding nor
any derivative evidence that it could have used to subsequently
prosecute those witnesses. The S. (R.J.) rule, in contrast, gives
factually guilty witnesses considerable assurance that the fruits
of their testimony will not be used against them later. The state
accordingly gains reliable testimony at the initial proceeding.
And if we assume that in the absence of derivative use immunity
guilty witnesses would not have testified truthfully, then the state
does not lose any derivative evidence that the S. (R.J.) rule would
have excluded.
Without empirical validation this conclusion can only be ten-
tative. It is possible that we would end up with a greater quantity
of reliable evidence from the "use immunity only" rule. It is also
possible (although in my view much less likely) that a rule
guaranteeing full derivative use or transactional immunity would
produce the greatest quantity of reliable evidence. In any case,
we can see that despite its rhetorical preoccupation with free
choice thinking, the Supreme Court's approach is fully consistent
with the reliability rationale. It also demarcates a reasonable
486 CriminalLaw Quarterly [Vol. 48
compromise between two incompatible, reliability-based objec-

tives: maximizing truthful compelled testimony and minimizing
the loss of derivative evidence.
Compelling testimony from non-accused witnesses also raises
concerns about governmental overreaching. As I have discussed,
requiring witnesses to answer questions in a public, supervised
proceeding is not intrinsically abusive.39 It exhibits few of the
hallmarks of coercive, incommunicado police interrogation:
relentless questioning, physical deprivation, emotional exploita-
tion, psychological manipulation, abuse of trust, physical
domination and the like. But as many jurists have recognized,
compelling testimony does raise the possibility of a different
kind of abuse. Forcing witnesses to answer questions in public
forums about possible illegal activity will often embarrass and
stigmatize them. The state should not be able to subject
people to such treatment at its whim. It should be required to
demonstrate a compelling public interest."n
Use and derivative use immunity provide some degree of pro-
tection against this kind of abuse. If the state is prohibited from
using compelled testimony or its non-discoverable fruits in
future proceedings against a witness, then it will generally only
compel testimony if it has a legitimate reason for doing so, such
as regulating economic activity, ensuring public safety or con-
victing criminals. But occasionally government authorities may
compel testimony for less savoury motives. They may wish, for
a variety of purposes, to harass witnesses without any expecta-
tion of generating admissible evidence.4" Or they may believe
that compelled testimony might produce such evidence despite
the provision of legal immunity. Many of the most egregious
examples of this type of abuse have arisen in the context of
inquisitorial fishing expeditions designed to both intimidate and
incriminate witnesses vaguely suspected of criminal activity.42
39. See S. (R.J.), supra, footnote 12, at para. 80.
40. See Thomson Newspapers, supra, footnote 14, at p. 480, Wilson J.
41. The adverse publicity generated by the compelled self-incriminating evidence of
a witness may also prejudice the fairness of any future trial of that witness. See
Ratushny, supra, footnote 3, at pp. 349-51 and 359-61.
42. Ibid., at p. 349; S. (R.J.), supra, footnote 12, at paras. 142 and 319. As Ratushny
notes, such practices were prevalent in certain provinces (especially Quebec) in
the 1960s and 1970s: ibid., at pp. 358-84.
2004] What's Wrong with Self-Incrimination?PartIII 487
To prevent these abuses, the Supreme Court of Canada has

held that in addition to providing derivative use immunity to
supplement s. 13's use immunity, s. 7 of the Charter also
empowers compelled witnesses, in limited circumstances, to
resist compulsion altogether. Unfortunately, the "predominant
purpose" test set out in Branch does not adequately protect
witnesses against abusive inquisitorial proceedings. To begin, it
is too indeterminate. In many cases the distinction between
compulsion designed to incriminate and that designed to further
a legitimate policy objective will be extremely difficult to
discern. Branch itself is a good example. The predominant
purpose of allowing securities investigators to compel testimony
is to facilitate the regulation of capital markets. But that
regulation is effected in large measure by imposing penalties
(including penal sanctions) for activities (like defrauding
investors) that threaten the integrity of the regulatory scheme. To
say that compelling self-incriminating testimony is only
"incidental" to the primary goal of ensuring the viability of
securities markets is misleading. By this logic, compelling self-
incriminating testimony from suspected drug traffickers could be
said to be incidental to the primary goal of preventing the
destructive consequences of the illegal narcotics trade. For
inquiries investigating conduct that may eventually lead to penal
sanctions, the distinction between legitimate and self-incriminating
purposes is chimerical. It is more accurate to say that the purpose
of compelling testimony in cases like Branch is to induce self-
incriminating testimony that will lead to the conviction and
punishment of individuals who have violated securities law,
which will deter others from doing so, thereby ensuring the
efficient operation of the securities industry.
In any case, the test for determining non-compellability should
not turn on the characterization of the inquiry's purpose. Even if
it were clear that the primary purpose of compelling testimony in
a particular case was not to induce self-incrimination, the
compulsion might still unjustifiably prejudice the witness's
interests. In all cases, the compellability determination should
therefore focus on the second part of the Branch test: whether the
witness is likely to suffer prejudice that outweighs the state's
interest in collecting the evidence.43 The definition of "pre-

judice", moreover, should incorporate not only Branch's concern
for the fairness (i.e. reliability") of any potential subsequent trial,
but also a concern for the fairness of the proceeding at which the
testimony is compelled. "Fairness", in the latter sense, means
any unjustified harassment, stigmatization, or other abuse that
testimonial compulsion is likely to induce. It is here that the
existence of a colourable, self-incriminating purpose is relevant.
Ultimately, however, it is the effect of the compulsion that should
be determinative. Baseless inquisitions should not be permitted,
even when they would not increase the possibility of wrongful
convictions. Of course, the question of whether a proceeding is
unduly harassing, stigmatizing or otherwise abusive is not
capable of precise formulation. Relevant factors include the
nature of the proceeding; its legislative and regulatory context;
the degree of concretized suspicion attaching to the witness; the
availability of other investigative techniques; the importance of
the testimony; and the procedural protections attaching to the
proceedings, such as whether the hearing is held in public or
in camera, the availability of publication bans, and whether the
witness has a right to be represented by counsel.45
43. This is essentially the test proposed by Justice Sopinka in S. (R.J.). He intended,
however, for non-compellability to be an alternative to derivative use immunity,
not a complement. See S. (R.J.), supra, footnote 12, at para. 326. It is also con-
sistent with the approach articulated by Cory J. in his concurring minority reasons
in Phillips, supra, footnote 23, at para. 82, although he again failed to acknowl-
edge that this approach differed from that set out in Branch ("[E]ven where the
purpose of compelling testimony is valid, it is necessary to assess the prejudicial
effect of such compulsion upon the witness.").
44. The concern over the fairness of any subsequent trial is really about reliability, not
abuse. Adverse publicity stemming from compelled testimony may unduly influ-
ence the trier of fact, and hence impugn the reliability of the eventual verdict. In
his concurring minority reasons in Phillips, supra, footnote 23, Cory J. suggested
that remedies for adverse publicity created by compelled testimony can be sought
pursuant to ss. 11(d) and 24(1) of the Charter. This approach is limited by the fact
that s. 11(d) does not apply to persons not yet "charged with an offence" who are
likely to be charged later.
45. By these measures there is nothing objectionable about the investigative hearing pro-
visions in the Anti-terrorism Act, supra, footnote 1, which give authorities limited
powers to compel testimony from persons believed to have terrorism-related
information. These hearings have been criticized for departing from the principle
that criminal suspects are entitled to refuse to speak to authorities. See K. Roach,
2004] What's Wrong with Self-Incrimination? Part I1 489
As the Supreme Court rightly held in S. (R.J.) and Branch,

courts may consider the compellability issue at either the
subpoena or trial stage. Congruent with the scheme outlined
here, concerns for abusive questioning should in most cases be
dealt with before testimony is given. Concerns for trial fairness,
in contrast, should most often be dealt with if and when a trial
"The Dangers of a Charter-Proof and Crime-Based Response to Terrorism", in
R.J. Daniels, P. Macklem and K. Roach, eds., The Security of Freedom: Essays on
Canada'sAnti-terrorism Bill (Toronto: University of Toronto Press, 2001), p. 131 at
p. 135; M. Shaffer, "Effectiveness of Anti-Terrorism Legislation: Does Bill C-36
Give Us What We NeedT', ibid., p. 195 at p. 200. Certain aspects of the investigative
hearing process are admittedly novel. Police apply to obtain the orders, for example,
and questioning is conducted before judges. As we have seen, however, there are
many different ways in which state agents can compel testimony from non-accused
suspects. See also I. Coder, '"hinking Outside the Box: Foundational Principles for
a Counter-Terrorism Law and Policy," ibid., p. 111 at p. 127; M. Friedland, "Police
Powers in Bill C-36", ibid., p. 269 at pp. 276-77; S.A. Cohen, "Safeguards in and
Justifications for Canada's New Anti-terrorism Act" (2002), 14 N.J.C.L. 99 at p. 115.
The important question is not whether investigative hearings depart from tradition,
but rather whether they adequately protect suspects from abusive inquisitorial
practices. There is reason to doubt whether investigative hearings will induce many
uncooperative individuals to reveal truthful, material information. See Roach, ibid.,
at pp. 137-38; Friedland, ibid., at pp. 278. The process is not, however, inherently
abusive. See generally Friedland, ibid., at pp. 277-78; Cohen, ibid., at pp. 114-19.
The legislation specifies that a person is compellable only if a judge determines that
there are reasonable grounds to believe that a terrorism offence was or will be com-
mitted and that the person is likely to have material information about that offence.
It also requires the Attorney General to approve the application and, in the case of
investigations into possible future offences, police to have first attempted to con-
vince the individual to provide information voluntarily. Persons subject to this pro-
cedure cannot be compelled to disclose privileged information and, as one would
expect, their testimony is protected by use and derivative use immunity. They also
have the right to be represented at the hearing by counsel. It is difficult to imagine
circumstances in which individuals properly compelled by this procedure could
fairly said to have suffered unjustified harassment or stigmatization. See S. Penney,
"What's Wrong with Self-Incrimination? The Wayward Path of Self-Incrimination
Law in the Post-Charter Era, Part I" (2003), 48 C.L.Q. 249 at pp. 256-66. If such cir-
cumstances did arise, then s. 7 of the Charter would still be available to quash the
order or stay subsequent criminal proceedings. Section 7 is also available to remedy
the effects of adverse pre-trial publicity, as is s. 83.28(5)(e) of the Criminal Code,
which specifically empowers the supervising judge to "include any other terms or
conditions that the judge considers desirable, including terms or conditions for the
protection of the interests of the person named in the order...". Shaffer suggests that
investigative hearings could be used to discriminate against racial and religious
groups identified with terrorism: ibid., at p. 200. Discriminatory profiling is most
materializes. At the subpoena stage, the most common remedy

will be to quash the subpoena. But this will not always be
necessary. For example, where the only concern is the risk that
adverse publicity will make it impossible to select an impartial
jury in a future trial, it may be more appropriate to simply order
a publication ban.46 Similarly, while a stay of proceedings is the
most obvious remedy at the trial stage, less dramatic remedies
may be sufficient in some cases.
My proposal does not call for any dramatic changes to existing
law. Its chief effects are to: (i) eliminate the "predominant
purpose" criterion as a condition precedent for resisting compul-
sion; and (ii) broaden the scope of prejudice to encompass
abusive inquisitorial practices. I suspect, moreover, that the
results generated by this approach would in most cases be the
same as those resulting from the current doctrine. Given the
malleability of the predominant purpose test and the inherent
breadth of the concept of prejudice, creative judges are likely
able to decide in favour of non-compellability even where, on its
prevalent, however, in the exercise of broad, discretionary powers that are not
subject to prior judicial authorization and involve minimal costs to law enforcement
(as is the case, for example, with street and vehicle stops and customs searches). See
D. Tanovich, "Using the Charter to Stop Racial Profiling: The Development of an
Equality Based Conception of Arbitrary Detention" (2002), 40 Osgoode Hall L.J.
145; S.R. Gross and K.Y Barnes, "Road Work: Racial Profiling and Drug
Interdiction on the Highway" (2001), 101 Mich. L. Rev. 651; U.S. Customs Service,
Better Targetingof Airline Passengersfor PersonalSearches Could Produce Better
Results: Report to the HonourableRichard J. Durbin, U.S. Senate / United States
General Accounting Office (Washington, D.C.: The Office, 2000). Investigative
hearings do not exhibit these characteristics. Moreover, the investigative hearing
provisions are subject to an annual reporting requirement that is likely to reveal any
systemic abuses: CriminalCode, s. 83.31(1). They are also subject to a sunset clause
that mandates their expiry five years after coming into force unless Parliament
decides to renew them: CriminalCode, s. 83.32. At the time of writing, the Supreme
Court of Canada has reserved judgment in a case arising out of the Air India
bombings that will give it the opportunity to resolve a number of issues surrounding
the investigative hearing provisions. The judge at first instance found that the provi-
sions were constitutional. See In the Matter of an Application under s. 83.28 of the
CriminalCode, 2003 BCSC 1172, leave to appeal to SCC granted August 11, 2003,
[2003] S.C.C.A. No. 326 (QL).
46. See Phillips, supra, footnote 23, at paras. 34-35, L'Heureux-Dubd J., and at
para. 134, Cory J. Other possible remedies include ordering an in camera hearing
and delaying the taking of testimony until the conclusion of the witness's trial.
2004] What's Wrong with Self-Incrimination? Part III 491
face, the Branch doctrine would not permit it. This approach, it
should be stressed, does not call for a broad right to resist testi-
monial compulsion. As under the Branch test, witnesses who
appear at inquiries serving legitimate public functions and who
receive adequate protection from adverse publicity would rarely
be entitled to resist compulsion. Regulatory investigations, ad
hoc commissions of inquiry, and prosecutions of co-accused
would not grind to a halt. But the changes that I have proposed
would make it easier for courts to protect witnesses against the
excesses of pretrial inquisitions. They would also discourage
inquisitorial bodies from abusing their considerable powers.
On the whole, the S. (R.J.)/Branch framework represents a
reasonable compromise between the competing interests
implicated by testimonial compulsion. It permits the state to
compel testimony for legitimate purposes, provides an adequate
incentive for truthful testimony, and allows discoverable deriva-
tive evidence to be admitted in future prosecutions against
compelled witnesses. It also enables courts to forbid compulsion
in exceptional circumstances. It is only in this last feature that
there is a need for modest reform. The proposal set out here
would help to ensure that the right to resist compulsion is
available when needed to prevent inquisitorial abuses.
(2) Non-Testimonial Compulsion
Defence lawyers have also attempted to use the principle
against self-incrimination inhering in s. 7 of the Charter to pre-
vent prosecutors from using defendants' documents and non-
testimonial statements to incriminate them in criminal and quasi-
criminal proceedings. For the most part, these attempts have
failed. The Supreme Court of Canada addressed the question of
compelled documents in Branch.7 It found that documents
produced prior to and independently of compulsion do not
implicate self-incrimination concerns. 8 It noted, however, that
the self-incrimination principle might apply to non-testimonial
communications "brought into existence by the exercise of
47. Branch, supra,footnote 17.
48. Ibid. at paras. 43-46. See also Thomson Newspapers, supra, footnote 14, at
p. 608, Sopinka J. and p. 588, L'Heureux-Dubd J.
compulsion by the state".4 9 Numerous legislative provisions

require individuals to make statements or keep records for
regulatory purposes. In some cases the statute compelling the
communications provides use immunity. But at common law
compelled, non-testimonial communications are admissible in
criminal and regulatory prosecutions against the person who
made them." Nor is there anything to prevent this in the express
language of the evidence statutes or the Charter. As we have
seen, the use immunity and non-compellability rules set out in
those enactments apply only to testimony.' Section 7 could be
interpreted to provide use and derivative use immunity, but in
Branch the Court declined to comment on precisely when such
protection might arise. 2
The Court confronted the issue directly in R. v. Fitzpatrick,53 a
quasi-criminal prosecution under the FisheriesAct. 4 Prosecutors
wished to adduce documentary fishing records and oral "hail
reports" of daily catches that the accused was required to submit
to regulatory officials. The admission of statutorily compelled
communications, the Court concluded, does not always violate
the self-incrimination principle.5 Courts must approach the
question contextually, paying particular attention to the societal
interests at play in the regulatory environment. 6 Factors to be
considered include the nature of the relationship between the
parties, the purpose of the compulsion, the existence of coercion,
49. Branch, supra, footnote 17, at para. 43.

50. See Walker, supra, footnote 38.
51. Recall that s. 13 of the Charter provides use immunity to a "witness" who testi-
fies in "any proceedings". Section 5(2) of the Canada Evidence Act also applies
only to a "witness". Section 11 (c) of the Charter protects persons "charged with
an offence" from being "compelled to be a witness in proceedings against that
person in respect of the offence". See generally D. Paciocco, CharterPrinciples
and Proofin Criminal Cases (Scarborough: Carswell, 1987), pp. 465-66 (noting
that s. 13's language does not support the extension of immunity to persons com-
pelled by statute to disclose information in informal settings) and pp. 485-89 (col-
lecting decisions rejecting arguments that s. 11(c) applies to obligations to pro-
vide physical and oral evidence outside of trial proceedings).
52. Branch, supra, footnote 17, at para. 48.
53. [1995] 4 S.C.R. 154, 102 C.C.C. (3d) 144, 43 C.R. (4th) 343.
54. R.S.C. 1985, c. F-14.
55. Fitzpatrick,supra, footnote 53, at para. 21.
56. Ibid., at paras. 27-30.
whether compulsion impacts the reliability of the communication,

and the potential for the abuse of state power. Applying these
factors, the Court concluded that admitting the evidence would
not violate the self-incrimination principle. At the time the
records are made, fishers and the state are not in an adversarial
relationship; the predominant purpose of the record-keeping
requirement is to facilitate efficient regulation, not to compel
self-incrimination; fishers participate in the industry voluntarily;
any incentive to falsify records exists despite the absence of use
immunity; and fishers are not subject to psychological or
emotional pressure or forced to divulge intimate information.
The Court came to the opposite conclusion in R. v. White.57
The defendant there was charged under the Criminal Code with
failing to stop at the scene of an accident. 8 At trial, the prosecu-
tion attempted to introduce self-incriminating statements
compelled by provincial motor vehicle legislation. That legis-
lation purports to provide use immunity for such statements in
subsequent proceedings. 9 But division of powers principles
prevent this immunity from applying to prosecutions for federal
offences.' Considering the same factors outlined in Fitzpatrick,
the Court determined that admitting the statements would violate
s. 7 of the Charter.6 ' In its view, the police's presence creates an
environment of "pronounced psychological and emotional
pressure"62 that could induce drivers to "provide a more
extensive statement to the police than legally required under the
Act",63 the reports' use in criminal prosecutions is not integral to
the regulatory scheme, and the fear of criminal prosecution could
induce false statements.
57. White, supra, footnote 38.

58. Criminal Code, R.S.C. 1985, c. C-46, s. 252(1)(a).
59. Motor Vehicle Act, R.S.B.C. 1979, c. 288, s. 61(7), as amended.
60. White, supra, footnote 38, at para. 35; Walker, supra, footnote 38. See also Bell,
supra, footnote 38.
61. The Court upheld the trial judge's decision to exclude the evidence pursuant to s.
24(1) of the Charter. It stated, however, that exclusion could also have been
effected, as it was in S. (R.J.), supra, footnote 12, by the use of courts' common
law duty to exclude evidence "whose [sic] admission would render the trial
unfair": White, supra, footnote 38, at para. 89.
62. White, supra, footnote 38, at para. 58.
17 - 48 C.L.Q.
Most recently, in R. v. Jarvis'4 and R. v. Ling,65 the Court

considered provisions requiring taxpayers to answer questions
and provide documents related to their income tax reporting.'
Authorities conducting regulatory audits, the Court concluded,
may use these powers freely. But once the "predominant
purpose" of the investigation becomes an inquiry into penal
liability, they may no longer compel answers and must obtain
warrants to seize documents.67 It is only where taxpayers face the
possibility of criminal punishments, the Court reasoned, that
they enter into an "adversarial relationship" with the state.68 To
determine whether such a relationship arose, courts should
consider whether authorities had reasonable grounds to believe
that a criminal offence was committed,69 attempted to collect
evidence of mens rea,7 ° and shifted primary responsibility for the
investigation from auditors to criminal investigators.7 '
(b) Critique
Like its jurisprudence on testimonial compulsion (and despite
its continued reliance on free choice reasoning)," the Court's
treatment of non-testimonial compulsion accords reasonably
well with the reliability and abuse-prevention rationales.
64. [2002] 3 S.C.R. 757, 169 C.C.C. (3d) 1, 6 C.R. (6th) 23.
65. [2002] 3 S.C.R. 814, 169 C.C.C. (3d) 46, 6 C.R. (6th) 64.
66. Income TaxAct, R.S.C. 1985, c. 1 (5th Supp.), ss. 231.1(1) and 231.2(1).
67. Jarvis,supra, footnote 64, at paras. 2, 46, 88 and 99. In S. (R.J.) and White, it was
the admission of self-incriminating evidence at trial that would violate s. 7, not its
collection. Because the evidence was not "obtained in a manner" that infringed
the Charter, courts could not exclude it pursuant to s. 24(2). As discussed, the
Court consequently excluded the evidence by means of either the (constitutional-
ized) common law power to exclude unfairly prejudicial evidence (S. (R.J.),
supra, footnote 12) or the general remedial power in s. 24(1) of the Charter
(White, supra, footnote 38). In Jarvis,however, the Court concluded that Charter
protection against self-incrimination is triggered once it is apparent that the
primary purpose of the investigation is to determine criminal liability. Evidence
obtained in violation of these protections may therefore be excluded under
s. 24(2) of the Charter: Jarvis, ibid., at para. 105.
68. Jarvis, ibid., at paras. 84, 88.
72. See S. Penney, "The Continuing Evolution of the s. 7 Self-Incrimination
Principle: R. v. White" (1999), 24 C.R. (5th) 247.
Concerns about the reliability of compelled, non-testimonial

communications will rarely justify protection against self-
incrimination. Consider first the problem of false confessions
leading to wrongful convictions. This will never be a problem for
documentary evidence. Requiring individuals to prepare and
maintain records for regulatory purposes does not generate the
coercive pressures apt to induce false confessions. Obligations to
report fish catches or taxable income, for example, do not cause
people to falsely confess to overfishing or tax evasion.
Legal requirements to provide oral statements or answer
questions, however, could contribute to the making of false con-
fessions. Recall that one of the advantages of the right to silence
is that it reduces the likelihood that authorities will employ 73
coercive methods to pressure suspects into admitting guilt. If
suspects know that they are not legally obliged to speak, then
they are better equipped to resist interrogation techniques apt to
induce false confessions. The magnitude of the risk of false
confessions turns on two factors: the nature of the legal obliga-
tion to speak and the environment in which questioning takes
place. Open ended obligations are more likely to produce false
confessions than narrowly tailored ones. Investigators clothed
with broad inquisitorial powers are less likely to be satisfied by
non-confessional responses (and are therefore more likely to
use coercive methods) than those entitled only to limited
information. Similarly, questioning by police-like authorities in
environments akin to custody is more likely to induce false
confessions than questioning by bureaucrats in non-custodial
settings. Suspects are much more likely to feel intimidated and
vulnerable in the former context than in the latter.
By these measures we can see that non-testimonial compul-
sion will rarely create a significant risk of false confessions.
Fitzpatrick presents the easiest case. Oral reports of daily fish
catches, conveyed by radio to regulatory officials, are equivalent
to documentary records maintained in the ordinary course of
business and present no danger of false confessions. The
required information is extremely specific and is provided to
regulators in a setting entirely devoid of coercive pressure.
73. See S. Penney, "What's Wrong with Self-Incrimination? The Wayward Path of
Self-Incrimination Law in the Post-Charter Era, Part II" (2003), 48 C.L.Q. 280
at pp. 317-20.
The situation in White is more complicated. There, the

legislation required accident reports to be made to police, who
were instructed to obtain information relating to the "the partic-
ulars of the accident, the persons involved, the extent of the
personal injury or property damage and other information
necessary to complete a written report".74 Police sometimes
obtain this information in circumstances conducive to false
confessions, such as custodial questioning ensuing from arrests
for driving-related criminal offences.
The existence of a duty to relate the details of a vehicle acci-
dent, however, does not necessarily heighten the danger of false
confessions. In theory, once individuals relate the factual details
of the accident, their obligation to report is fulfilled and they may
thereafter remain silent. Indeed, contrary to the Supreme Court's
finding in White,75 the reporting obligation may actually reduce
the likelihood of coercion in vehicle accident investigations. In
the absence of a comprehensive right to remain silent, police
may be less likely to badger and manipulate suspects into
making false confessions.76 The reporting obligation, in other
words, may diminish the incentive to use coercive methods to
induce recalcitrant suspects to speak.
On the other hand, if in practice there is no clear separation
between mandatory reporting and voluntary cooperation,
suspects may feel an obligation to respond throughout the ques-
tioning process, thereby increasing the risk of false confessions.
This lack of separation led the Court in White to provide use
immunity to all statements made pursuant to the statutory reporting
requirement.77 As the Court intended,78 this has since prompted
police to separate the taking of accident reports from their
general investigative inquiries.79
74. Motor Vehicle Act, R.S.B.C. 1979, c.288, ss. 61(1) and (4).
75. See White, supra, footnote 38, at paras. 58-59.
76. See Penney, supra, footnote 72, at p. 262.
77. See White, supra, footnote 38, at para. 64.
79. The Royal Canadian Mounted Police's current "caution card" includes the
following statement to be read to persons investigated for vehicle accident
related offences: "I am not at this time seeking an accident report from you
pursuant to the provisions of the Motor Vehicle Act... therefore you are under
no compulsion to answer my questions. IfI do decide to take an accident report
The Court did not have to provide immunity to compelled

accident reports, however, to protect against false confessions.
Instead, it could simply have required police to explain to acci-
dent victims that their duty to report is limited and that once they
provide the required information they have no further duty to
answer questions. It could also have instructed police, in
taking accident reports, to passively elicit statements and avoid
persistent, manipulative or confrontational questioning. Police
could comply with these requirements, for example, by providing
drivers with a standard form and instructing them to fill it out and
send it directly to regulatory authorities. This procedure would
clearly demarcate the boundary between accident reporting and
investigative questioning and eliminate the risk that mandatory
reporting might contribute to the making of false confessions.
Prosecutors, however, would still be permitted to use accident
reports in criminal trials.
The danger that non-testimonial compulsion could engender
false confessions is more acute for the Income Tax Act investiga-
tions addressed in Jarvis and Ling. Section 231.1 (1)(d) of the Act
requires taxpayers to "answer all proper questions" relating to
tax investigations.8 0 This "requirement power" is much broader
than the duty to report vehicle accidents. And unlike the accident
report duty, it can not be circumscribed in the manner described
above. Investigators could therefore use it to badger and manip-
ulate innocent suspects into confessing.
Not all uses of the requirement power, however, raise the
danger of false confessions. In many cases the power is exercised
by auditors or other officials in non-coercive environments.
Auditors asking routine questions about deduction claims or
income statements, for example, are highly unlikely to induce
false confessions. People fear auditors, but not in the same
way that they fear police. But tax-related information may
occasionally be demanded by authorities - whether police
or police-like tax investigators - in environments akin to
incommunicado police interrogation. As discussed, it is in these
circumstances that the right to silence helps to prevent false
from you later, I will clearly advise you." R.C.M.P. Caution Card (on file with the
author).
80. Supra, footnote 66.
confessions. The distinction drawn in Jarvis and Ling between

ordinary compliance audits and tax evasion investigations
responds to this concern. Where the predominant purpose of an
inquiry is to obtain evidence of criminal conduct, there is a
danger that investigators will employ coercive methods that
could create false confessions. Inquiries designed only to deter-
mine tax liability and impose regulatory penalties are much less
likely to involve coercive, interrogation-style questioning.
Giving evidentiary immunity only to statements made in the
course of criminal investigations minimizes the risk of wrongful
conviction without denying the state access to reliable, self-
incriminating evidence obtained from non-coercive regulatory
inquiries.
It could be argued, however, that Jarvis and Ling are over-
broad in providing immunity to all uses of the requirement
power in criminal investigations. After all, if the inquiry is
non-coercive, then the investigation's purpose should be irrele-
vant. If questioners do not have the intimidating presence of
police interrogators, the questioning is conducted in a familiar,
non-threatening environment, and the questions are not persistent,
manipulative or deceptive, then there is little chance that com-
pulsion will induce a false confession. Conversely, it is possible
(though far less likely) for non-criminal compliance audits to
bear the hallmarks of coercive police interrogation. Why use the
predominant purpose test as an imprecise proxy for coercion
(questioning apt to induce a false confession) when it is possible
to identify it directly?
While it is possible to identify coercion in theory, it is often
very difficult to do so in practice. Determining coercion requires
not only an objective record of the questioning but also a specu-
lative evaluation of its effect on the suspect's state of mind. As
we saw in the discussion of the voluntary confession rule,'
courts have not been very successful in making this type
of determination consistently and predictably.82 There are
accordingly few cognizable, ex ante limitations on the use of
81. See Penney, supra, footnote 73, at pp. 290-305.

82. This difficulty led courts and legislatures to supplement the confession rule by
giving suspects the right to consult with counsel and requiring police to inform
them of that right: see ibid. at p. 312.
coercive methods. In the absence of such limitations, tax inves-

tigators might be tempted to "push the envelope" of the Income
Tax Act requirement power, hoping that courts would find that
confessions were made in non-coercive circumstances.
The Jarvis/Ling "predominant purpose" test, in contrast,
demarcates a relatively clear boundary. It instructs criminal
investigators to inform suspects of their right to silence before
questioning them.83 Like other prophylactic rules, the test
sacrifices ex post precision for ex ante clarity. In this case, the
sacrifice is justified. Giving taxpayers a right to silence during all
criminal tax inquiries provides significant protection against
false confessions without precluding access to alternative
sources of self-incriminating evidence. Prosecutors may still use
compelled statements from non-criminal compliance audits
(which often precede criminal investigations) as well as voluntary
statements elicited during criminal investigations.
Non-testimonial compulsion may also increase the likelihood
that factually guilty suspects will make false exculpatory state-
ments or fail to reveal truthful inculpatory information.
Wrongdoers who are obliged to make statements that may later
be used to incriminate them have a strong incentive to lie. If a
regulatory scheme depends on the mandatory submission of
reliable statements, then it may be necessary to provide eviden-
tiary immunity to maximize honesty. Specifically, immunity will
be necessary when the state's interest in promoting truthfulness
outweighs its interest in using compelled statements in criminal
or regulatory prosecutions.
In Fitzpatrick this determination was simple. Although effec-
tive fisheries regulation demands that fishers submit accurate
reports, it also requires that those reports be admissible in over-
fishing prosecutions. If they were inadmissible, the Court noted,
then regulators would be forced to use more intrusive and
expensive techniques to acquire evidence of overfishing. As the
Court also observed, fishers may be tempted to falsify reports
notwithstanding the existence of penalties for false reporting and
the absence of evidentiary immunity. The absence of immunity,
83. In practice, tax investigators will likely also inform suspects of their right to
counsel under s. 10(b) of the Charter. Technically, however, this is only required
when suspects are arrested or detained.
CriminalLaw Quarterly [Vol. 48
moreover, reveals a legislative assumption that the regulatory

scheme is viable without it. As the Court asked rhetorically,
"would the system be better if fishers could (honestly) submit
hail reports and fishing logs that indicated they had massively
overfished under the Act, secure in the knowledge that these
returns could not later be used against them?"' Indeed, in cases
where the only concern is the incentive for dishonesty created by
a lack of evidentiary immunity, courts should almost always
defer to the legislature's decision not to grant it. There being no
individual interest to protect, courts have no reason to interfere
with the legislature's choice. -
One of the few conceivable exceptions to this principle arose
in White. There, the provincial legislature attempted to confer
use immunity for criminal and other proceedings. But as noted
above, this immunity is ultra vires in relation to federal offences.
By providing immunity under s. 7 of the Charter, the Supreme
Court effectively repaired the damage to the regulatory scheme
wrought by the division of powers.85 In doing so, the Court had
to decide not only between competing legislative objectives
(encouraging accurate reporting to improve highway safety
versus convicting the guilty) but also between competing legis-
latures. There being no single state interest to defer to, the Court
was forced to choose. It decided that the effective operation of
the regulatory regime required the granting of immunity."6
Evidence necessary for criminal prosecutions could be obtained
in other ways.
Like testimonial compulsion, non-testimonial compulsion
may also raise concerns about inquisitorial abuse. In this context,
the concern for cruel interrogation practices is co-extensive with
the concern for false confessions. If suspects have no right to
remain silent, then authorities may be emboldened to use
coercive methods to extract admissions. These methods may be
intrinsically cruel, apt to induce false confessions, or both. In
84. Fitzpatrick,supra, footnote 53, at para. 44.
85. For a more extensive elaboration of this point, see Penney, supra, footnote 72, at
p. 265.
86. As I have argued elsewhere, this conclusion, while not unreasonable, is specula-
tive. It is equally plausible that an absence of use immunity would not have
unduly thwarted the government's ability to compile accurate accident statistics:
see Penney, ibid., at p. 263.
deciding whether to immunize evidence produced by non-

testimonial compulsion to prevent abusive questioning, courts
should therefore consider the same factors that I have suggested
they consider in assessing the danger of false confessions. 7 If
the power to compel answers can be used in an open-ended
inquisition akin to custodial police interrogation, then courts
should invoke s. 7 of the Charter to confer use and derivative
use immunity. This effectively gives suspects a right to silence
and diminishes the risk of both false confessions and abusive
questioning.
Powers to compel non-testimonial communications may also
lead to unjustified invasions of privacy. Authorities may use
their authority to compel statements and documents to harass,
embarrass and stigmatize individuals without first establishing a
reasonable basis for suspicion. Because non-testimonial compul-
sion typically arises in the regulatory context, however, the Court
has generally been willing to tolerate inquisitorial practices that
would not pass muster in the criminal realm."8 In this respect, the
Court's approach to compelled communications parallels its
jurisprudence on administrative searches,89 where it has generally
upheld seizures of business records in the absence of
warrants or probable cause. 0 The Court's aim in these cases has
been to balance individual privacy interests against state
regulatory objectives, considering the role of the search power in
the regulatory scheme, the magnitude of any reasonable expecta-
tion of privacy in the materials and places subject to the power,
and whether the search must be based on objective grounds or
87. See discussion in the text, supra, footnote 45.
88. See, e.g., Branch, supra, footnote 17, at paras. 52-53; Thomson Newspapers,
supra, footnote 14, at pp. 517-18, La Forest J.; Comit~paritairede l'industriede
la chemise v. Potash; Comiti paritairede l'industrie de la chemise v. Silection
Milton, [1994] 2 S.C.R. 406 at pp. 420-21, La Forest J. and pp. 441-44,
L'Heureux-Dub6 J., 91 C.C.C. (3d) 315, 61 Q.A.C. 241; R. v. McKinlay
TransportLtd., [1990] 1 S.C.R. 627 at pp. 645-47, Wilson J., 55 C.C.C. (3d) 530,
76 C.R. (3d) 283.
89. Section 8 states that "Everyone has the right to be secure from unreasonable
search or seizure."
90. See Thomson Newspapers, supra, footnote 14; 143471 Canada Inc. v. Quebec
(Attorney General), [1994] 2 S.C.R. 339,90 C.C.C. (3d) 1, 61 Q.A.C. 81, 31 C.R.
(4th) 120; Comiti paritaire,supra, footnote 88, at pp. 420-21; Fitzpatrick,supra,
footnote 53; McKinlay Transport, supra, footnote 88.
authorized by neutral arbiter.' The Court should engage in a

similar balancing exercise in determining whether non-testimonial
compulsion unduly compromises privacy under s. 7 of the Charter.
For the most part, it has. In Fitzpatrick,Jarvis, and Ling, the
Court noted that mandatory self-reporting and record keeping
were essential to the efficient operation of the regulatory
schemes at issue. Fisheries management and income tax collec-
tion would probably not be possible if authorities either had to
obtain warrants based on probable cause before obtaining such
communications or were prohibited from using them to prosecute
wrongdoers.92 In White, in contrast, the Court correctly concluded
that admitting vehicle accident reports in criminal prosecutions
would not further the legislature's objective (improving highway
safety through the compilation of accurate accident statistics). 3
The Court has also repeatedly noted that in the regulatory
context, compelled information usually relates to relatively
impersonal matters, such as business and financial records."
Such records are not likely to divulge the "state of mind,
thoughts, or opinions of the individual who has submitted the
records".95 Compelling their production will therefore rarely
create undue embarrassment or stigma.96
91. See generally Hunter v. Southam, [1984]2 S.C.R. 145, 14C.C.C. (3d) 97, 41 C.R.
(3d) 97 sub nom. Dir of Inv. & Research, Combines Inv. Branch v. Southam Inc.
and cases cited ibid.
92. See Jarvis, supra,footnote 64, at para. 51.
93. See White, supra,footnote 38, at para. 60.
94. See Jarvis, supra, footnote 64, at para. 72; Fitzpatrick, supra, footnote 53, at
para. 49. See also Thomson Newspapers, supra, footnote 14, at p. 507, La Forest
J.; 143471 Canada Inc. v. Quebec (Attorney General), supra, footnote 90, at
p. 378, Cory J.; Comiti paritaire,supra, footnote 88, at pp. 420-2 1; McKinlay
Transport,supra, footnote 88, at pp. 649-50.
95. Fitzpatrick,supra, footnote 53, at para. 51. See also Thomson Newspapers,supra,
footnote 14, at pp. 517-18, La Forest J.
96. The situation in White is more complicated. The Court noted there that a manda-
tory accident report is a "personal narrative of events, emotions, and decisions
that are extremely revealing of the declarant's personality, opinions, thoughts, and
state of mind": supra, footnote 38, at para. 66. But as the duty to provide a report
only attaches to persons involved in serious vehicle accidents, the state would
appear to have good cause to obtain this information. The reporting obligation, in
other words, is not likely to be used in fishing expeditions. The Court's holding
in White, moreover, does not prohibit authorities from invading suspects' privacy.
It merely excludes their reports from subsequent proceedings.
The Court was therefore wrong to hold in Jarvis that when an

authority's predominant purpose is to determine penal liability, it
may not examine a taxpayer's documents without a warrant.97 As
discussed, the predominant purpose test usefully distinguishes
between coercive and non-coercive uses of the Income Tax Act's
requirement power, which obliges taxpayers to answer authori-
ties' queries. Obligations to maintain and produce documents,
however, do not raise concerns for abusive questioning or false
confessions. The only issue is whether production would consti-
tute an unwarranted invasion of privacy. But as the Jarvis Court
recognized, income tax records attract no reasonable expectation
of privacy, and consequently s. 8 does not oblige authorities to
obtain warrants or establish probable grounds before seizing
them. 98 The Court should therefore also have concluded that
obliging taxpayers to keep and produce such records does not
violate s. 7 of the Charter, even when authorities order produc-
tion to further criminal investigations. If it does not unduly
violate taxpayers' privacy to examine their records when there
are no grounds for suspicion (i.e. in the course of a random
audit), then how can it do so once there are (i.e. in the course of
a criminal investigation)? If anything, taxpayers deserve more
protection in the former situation than in the latter. Concluding
otherwise favours the interests of the "likely guilty" over those
of the "likely innocent". This smacks of the perverse logic of free
choice thinkers who conceive of self-incrimination law as a
shield for the guilty. Requiring warrants and probable cause for
"criminal" inquiries may also threaten the efficacy of regulation,
especially where regulators combine inspection and investiga-
tion functions in one department or class of official. 99
To summarize, it is rarely appropriate to protect persons
compelled to make non-testimonial communications from
self-incrimination. Except where officials are empowered to
compel statements in circumstances akin to custodial interroga-
tion, providing immunity is not needed to prevent abusive
questioning or false confessions. And courts should generally
97. See Jarvis, supra, footnote 64, at para. 96.
98. See Jarvis, ibid., at para. 95.
99. See David Stratas, "'Crossing the Rubicon': The Supreme Court and Regulatory
Investigations" (2003), 6 C.R. (6th) 74 at p. 80.
defer to the legislature's choice to favour the availability of

evidence for prosecution over the incentive for truth-telling that
immunity might provide, at least where they can discern a single,
coherent legislative choice. It will also rarely be necessary to
provide protection against self-incrimination to prevent fishing
expeditions. In the regulatory context in which non-testimonial
compulsion usually occurs, individuals typically do not have a
sufficient expectation of privacy to warrant restrictions on
inquisitorial intrusions.
Where there has been a violation of s. 7 in relation to non-
testimonial communications, the appropriate remedy will
almost always be to exclude that evidence (as well as any non-
discoverable derivative evidence) from the accused's trial.
Because we are dealing with communications compelled outside
of formal legal proceedings, there will rarely be an opportunity
to challenge compulsion before the communications are made.
Where such an opportunity does exist, the question of non-
compellability should be determined in the same manner that I
have proposed for testimonial compulsion.
(3) Non-Linguistic Compulsion
So far I have been discussing the statutory compulsion of
evidence (testimony, documents and out-of-court statements)
that is self-evidently linguistic and communicative. It is easy to
characterize such evidence as self-incriminating. But attempts
have also been made to affix the "self-incriminating" label to
certain types of non-linguistic evidence, such as bodily samples,
bodily impressions, and the results of identification "line-ups".
These attempts have largely failed."° Prior to the Charter, the
Supreme Court repeatedly denied that compulsory takings of
non-linguistic evidence infringed common law or statutory' °'
100. See generally D. Paciocco and L. Stuesser, The Law of Evidence, 3rd ed.
(Toronto: Irwin Law, 2002), p. 231; Paciocco, supra, footnote 13, at p. 89.
101. Section 2(d) of the CanadianBill of Rights, S.C. 1960, c. 44, states in relevant
part: "[N]o law of Canada shall be construed or applied so as to authorize a court,
tribunal, commission, board or other authority to compel a person to give evi-
dence if he is denied ... protection against self-crimination ....
2004] What's Wrong with Self-Incrimination? Part 11I 505
protections against self-incrimination.1 2 Post-Charter outcomes

have been similar.' 3 The Court has rejected Charter challenges to
compulsory breath and blood alcohol sampling" 4 and mandatory
fingerprint and photograph identification'05 without any mention
of self-incrimination concerns.' 6 In Stillman, the Court stated
102. See Quebec (Attorney General) v. Begin, [1955] S.C.R. 593 (blood sample taken
without consent admissible as rules relating to admissibility of statements do not
apply to physical evidence); Marcoux, supra, footnote 5 (exhibition of suspect to
witnesses for identification purposes does not violate privilege against self-
incrimination); R. v. Curr, [1972] S.C.R. 889, 7 C.C.C. (2d) 181, 26 D.L.R. (3d)
603 (obligation to provide breath sample for alcohol analysis does not infringe
s. 2(d) of the Bill of Rights).
103. As mentioned, the Court has considered compelled physical evidence to be self-
incriminating for the purposes of determining admissibility under s. 24(2) of the
Charter. See Penney, supra, footnote 73, at pp. 329-32; S. Penney, "Taking
Deterrence Seriously: Excluding Unconstitutionally Obtained Evidence under
Section 24(2) of the Charter" (2004), 49 McGill L.J. 105 at p. 129.
104. Section 254 of the Criminal Code gives police the power to demand breath
samples for roadside screening and breathalyzer analysis. It also authorizes them
to demand blood samples if the suspect is physically unable to provide a breath
sample. Section 254 makes it an offence to fail to comply with any of these
demands. In some circumstances, police may also obtain a warrant to take blood
samples from a suspect who is unable to consent: ibid., s. 256. The Supreme
Court has never been confronted directly with the question of whether these
forms of compulsion violate s. 7. But the Code's impaired driving provisions have
survived attacks based on Charter ss. 8, 9, and 10(b). See R. v. Thomsen, [1988]
1 S.C.R. 640 at pp. 648-49, 40 C.C.C. (3d) 411, 63 C.R. (3d) 1;
R. v. Hufsky, [1988] 1 S.C.R. 621, 40 C.C.C. (3d) 398, 63 C.R. (3d) 14; R. v.
Ladouceur, [1990] 1 S.C.R. 1257,56 C.C.C. (3d) 22,77 C.R. (3d) 110. The Ontario
Court of Appeal has recently rejected the argument that the screening device
demand infringes s. 7's protection against self-incrimination: R. v. Thompson
(2001), 151 C.C.C. (3d) 339, 41 C.R. (5th) 344, 52 O.R. (3d) 779 (C.A.).
105. Section 2 of the Identification of Criminals Act, R.S.C. 1985, c. I-I, requires per-
sons charged with indictable offences to submit to fingerprinting and photo-
graphing and empowers police to use force to effect compliance. Failure to sub-
mit constitutes an offence: Criminal Code, R.S.C. 1985, c. C-46, s. 145. The
Court upheld these provisions in R. v. Beare, [1988] 2 S.C.R. 387, 45 C.C.C. (3d)
57, 66 C.R. (3d) 97.
106. See S. (R.J), supra, footnote 12, at para. 201, where lacobucci J. noted that his
findings on the s. 7 self-incrimination principle are "directed at testimonial com-
pulsion and should not be taken as putting in jeopardy schemes relating to other
forms of compulsion such as breathalyzers, fingerprinting, blood sampling and
other similar schemes". "Those arrangements", he observed, "raise other issues
that are different from those involved in testimonial compulsion, which I need not
address". In her dissenting reasons in the same case, L'Heureux-Dub6 J. noted
cursorily that police violated s. 7 of the Charter in taking bodily

samples and impressions without statutory authorization or
consent. 7 It grounded this conclusion, however, on the fact that the
takings constituted unjustified invasions of privacy and bodily
integrity.0 8 It mentioned nothing in this context about self-
incrimination"° and went out of its way to suggest that subsequently
enacted warrant provisions authorizing police to forcibly extract
bodily samples for DNA analysis would survive constitutional
scrutiny." The Court has since confirmed the constitutionality of
these provisions. In R. v. B. (S.A.) it held that while DNA sampling
forces suspects to "provide evidence contrary to the principle
against self-incrimination","' this compulsion is justified by (among
other things) the importance and reliability of DNA evidence, the
modest degree of intrusion into bodily integrity and informational
privacy entailed by the procedure, and the prerequisite of prior
authorization based on reasonable and probable grounds." 2
(b) Critique
Should criminal suspects compelled to provide non-linguistic
evidence be protected against self-incrimination? Some jurists
claim that this type of evidence falls outside the ambit of
that the "constitutionality of demanding a breathalyzer sample on reasonable and

probable grounds [has never] been successfully challenged on the basis that it
violates a principle of fundamental justice against self-incrimination under s. 7 of
the Charter":ibid., at para. 254. See also R. v. Fischer (1998), 124 Man. R. (2d)
280 (Q.B.), leave to appeal to Man. C.A. refused 129 Man. R. (2d) 145 (C.A.).
107. R. v. Stillman, [1997] 1 S.C.R. 607 at para. 51, 113 C.C.C. (3d) 321, 5 C.R. (5th) 1.
108. The Court also found, after a much more thorough analysis, that the takings vio-
lated s. 8 of the Charter: ibid., at paras. 25-50.
109. In her dissenting reasons, McLachlin J. rejected the argument that the s. 7 self-
incrimination principle extended beyond testimonial evidence: ibid., at paras.
198-216.
110. Ibid., at para. 92. Sections 487.04 to 487.09 of the Criminal Code give judges the
power to issue warrants authorizing the taking of bodily samples for DNA analysis
in certain circumstances. Police may use reasonable force to ensure compliance.
Section 487.091 creates a similar process for obtaining handprints, fingerprints,
footprints, foot impressions, teeth impressions and other bodily impressions.
111. (2003), 178 C.C.C. (3d) 193 at para. 57, 14 C.R. (6th) 205, [2003] 2 S.C.R.
678.
112. B. (S.A.), ibid., at paras. 40-61. See also R. v. E (S.) (2000), 141 C.C.C. (3d) 225, 128
O.A.C. 329, 32 C.R. (5th) 79, sub nom. F (S.) v. Canada(Attorney General) (C.A.).
self-incrimination law because it is inherently non-communica-

tive" 3I or does not transgress the inviolate zone of privacy
reserved for products of the mind." 4 These arguments are mis-
guided. Like linguistic evidence, many types of non-linguistic
evidence communicate private, incriminating information about
criminal suspects, such as their identity or blood alcohol
content."5 Some of this information may be intensely personal.
DNA samples, for instance, can reveal intimate details or
suspects' physical and mental health and parentage." 6 This kind
of information is potentially more damaging to privacy interests
than the content of much compelled testimony.
Compelled, non-linguistic evidence may be communicative
and intimate, but it should not trigger protection from self-
incrimination. The reliability of this type of evidence is not
affected by the fact that it is compelled." 7 And while the
potential for abuse is manifest, s. 8 of the Charter adequately
shields suspects against unjustified invasions of privacy and
cruel methods without regard to the concept of self-incrimination."
In B. (S.A.) the Court wisely concluded that s. 8 - not s. 7 -
was the appropriate vehicle for determining the constitutionality
of the DNA warrant procedure." 9 But it refused to discard
self-incrimination discourse altogether, preferring to graft it
onto its s. 8 analysis.' At best, this approach is redundant. As
113. See McNaughton, supra, footnote 27, at pp. 378-79; Schmerber v. California,384
U.S. 757 at pp. 761-64 (1966).
114. See D. Paciocco, "Self-Incrimination: Removing the Coffin Nails" (1989), 35
McGill L.J. 73 at p. 88; Stillman, supra, footnote 107, at para. 204, McLachlin J.
dissenting.
115. See S.Easton, The Case for the Right to Silence, 2nd ed. (Aldershot: Ashgate,
1998), pp. 207-35; P. Arenella, "Schmerber and the Privilege against Self-
Incrimination: A Reappraisal" (1982), 20 Am. Crim. L. Rev. 31 at pp. 36-42.
116. In B. (S.A.), supra, footnote 111, at para. 49, the Court noted that it is conven-
tional for forensic scientists to examine only genetic material that does not reveal
medical, physical or mental characteristics. This does negate the fact, however,
that bodily samples extracted for identification purposes can be examined for
information relating to these other characteristics.
117. See B. (S.A.), ibid., at para. 58.
118. See F (S.), supra, footnote 112, at paras. 21-22; Paciocco and Stuesser, supra,
footnote 100, at p. 232.
119. Supra, footnote 111, at para. 35.
120. Ibid., at paras. 57-60.
CriminalLaw Quarterly (Vol. 48
the B. (S.A.) Court recognized, the DNA warrant process

utterly denies suspects freedom of choice. "A person has little
choice", it noted, "but to comply with the request for blood, hair
or saliva made under a valid DNA search warrant.""'2 Yet as I
have discussed, the Court upheld the process because it suffi-
ciently protects suspects from abusive methods. The fact that
the state seizes DNA evidence through the "compelled partici-
pation of the accused", 22' in other words, adds nothing to the
Court's s. 8 calculus, which would have been substantially the
same had the seized item been non-conscriptive, such as a
personal diary locked in a residential safe or narcotics secreted
in a bodily cavity. In each case, the Court would have gauged
the impact of the seizure on the suspect's privacy and dignity
interests and weighed that impact against the strength of the
state's justification for the intrusion, considering the nature
of the offence under investigation, the degree of suspicion
attaching to the suspect, and the presence of procedural protections
like prior authorization.
The use of self-incrimination discourse in this context, more-
over, may cause more damage than mere redundancy. It may also
tempt courts to employ free choice reasoning, which could have
disastrous results. Recall that the free choice rationale gives
suspects the freedom to refuse to provide incriminating informa-
tion, even where there are strong grounds for suspicion and
authorities do not obtain the information using methods that are
either inherently cruel or apt to produce wrongful convictions.
Nothing in the logic of the rationale justifies limiting this pro-
tection to linguistic information. But if suspects were granted a
general right to refuse to provide all forms of self-incriminating
evidence, the efficacy of many powerful investigative tools, such
as bodily impression analysis and breath and blood alcohol
testing, would be severely compromised. Fortunately, the result
reached in B. (S.A.) greatly reduces the likelihood that self-
incrimination-based challenges to these procedures will be
successful. The Court's refusal to entirely disavow free choice
thinking leaves open the possibility, however, that in cases where
the state's interest is not as obviously compelling as in B. (S.A.),
courts will exclude conscripted, non-linguistic evidence in the

absence of abusive state conduct.
7. Section 13 of the Charter and the Admissibility
of Defendants' Previous Testimony
(1) The State of the Law
As discussed, s. 13 of the Charter forbids prosecutors from
adducing defendants' previous, self-incriminating testimony.
Section 13 limits this immunity, however, to testimony that is23
"used to incriminate" a defendant in "any other proceedings".
In the years immediately after the passage of the Charter, prose-
cutors seized upon these limitations to assert that defendants'
previous self-incriminating testimony should be admissible in
certain circumstances. First, they argued that defendants' prior
voluntary trial testimony should be admissible at retrials for the
same offence on the basis that trials and retrials are part of the
same "proceedings". Courts had long accepted this argument in
interpreting s. 5(2) of the Canada Evidence Act,'24 and several
courts of appeal applied the same reasoning in deciding the first
cases under s. 13 of the Charter. 125 But in R. v. Dubois, the
Supreme Court disagreed, holding that retrials are separate
proceedings and that s. 13 accordingly forbids prosecutors from
adducing the previous testimony of defendants who decline to
testify again at their retrials.126 If retrials were merely continua-
tions of previous trials, it reasoned, then prosecutors could
conscript the accused "in discharging its burden of a case to meet"
123. Emphases added. To reiterate, the full text of s. 13 states the following: "A wit-
ness who testifies in any proceedings has the right not to have any incriminating
evidence so given used to incriminate that witness in any other proceedings,
except in a prosecution for perjury or for the giving of contradictory evidence."
124. R.S.C. 1985, c. C-5. See R. v. Brown (1963), 40 C.R. 105, 42 W.W.R. 448
(S.C.C.), revg 40 C.R. 90, [1963] 3 C.C.C. 326, 41 W.W.R. 129 (N.W.T.C.A.).
See also F. Kaufman, The Admissibility of Confessions, 3rd ed. (Toronto:
Carswell, 1979), p. 361.
125. See R. v. Sophonow (1983), 12 C.C.C. (3d) 272, 11 D.L.R. (4th) 24, 12 C.C.C.
(3d) 272, 29 Man. R. (2d) 1 (C.A.), affd [1984] 2 S.C.R. 524, 15 D.L.R. (4th)
480n, 17 C.C.C. (3d) 128n; R. v. Mannion (1984), 11 C.C.C. (3d) 503, 9 D.L.R.
(4th) 621, 53 A.R. 81 (C.A.), affd (on different grounds) [1986] 2 S.C.R. 272, 28
C.C.C. (3d) 544, 53 C.R. (3d) 193.
126. Dubois, supra, footnote 7.
and thus deny "his or her right to stand mute until a case has been
made out"." 7 The Court similarly decided in R. v. Mannion that
when defendants choose to testify again at their retrials, prose-
cutors may not attempt to incriminate them by adducing their
prior inconsistent testimony.'28
Dubois and Mannion prohibit prosecutors from introducing
defendants' prior testimony at their retrials for the purpose of
proving guilt.'29 But in neither case did the Court decide whether
this evidence is admissible to impeach their credibility.
Prosecutors argued in subsequent cases, therefore, that cross-
examining defendants on their prior inconsistent statements does
not "incriminate" them within the meaning of s. 13. This argu-
ment was initially successful. In R. v. Kuldip, the Supreme Court
concluded that trial judges may admit defendants' previous testi-
mony from any source (whether from earlier trials for the same
offence or altogether different proceedings) for the purpose of
impugning credibility.'3 ° As long as the trier of fact is made aware
that such evidence is not admissible for its truth, the Court
stated, its admission does not violate s. 13.13' The Court has since
backtracked from this position, holding in R. v. Noel that if the
real effect of admitting a defendant's prior testimony would be
incriminating, it may not be admitted even if accompanied by a
limiting instruction.'32
(2) Critique
The Court's s. 13 jurisprudence does not stand up to scrutiny.
Consider first the question of whether the prosecution should be
able to incriminate defendants at retrials by adducing their
previous testimony. As Justice Mclntrye noted in his dissent in
Dubois, the plain meaning of "proceedings" in s. 13 is "uncertain
127. Ibid., at p. 365.

128. Mannion, supra, footnote 125.
129. See Dubois, supra, footnote 7, at pp. 364-65; Mannion, ibid., at p. 280.
130. R. v. Kuldip, [1990] 3 S.C.R. 618, 61 C.C.C. (3d) 385, 1 C.R. (4th) 285.
131. Ibid., at pp. 634-35.
132. See Noel, supra, footnote 28, at paras. 30 and 46 to 59. Writing for the majority
at para. 54, Arbour J.stated that Kuldip is limited to cases where the reference to
the prior evidence is exclusively for impeachment purposes and carries no other
risk of incrimination. In other words, the cross-examination would be permitted
when there is no possibility that the jury could use the content of the prior
2004] What's Wrong with Self-Incrimination?Part II 511
and variable".'33 The problem must therefore be resolved by

reference to policy. Could admitting an accused's previous
voluntary testimony at a retrial for the same offence increase the
likelihood of a wrongful conviction? It is difficult to imagine
how. It is possible for innocent defendants to testify in a manner
that hurts their case. It could be argued, therefore, that the rule
excluding previous testimony diminishes the risk of wrongful
convictions by giving such defendants an opportunity at their
retrials to either recast their testimony or remain silent.
The admission of previous damaging testimony, however,
does not heighten the risk of wrongful conviction. It merely fails
to fully ameliorate a risk that always present when defendants
choose to testify in their own defence. Any impact that the
Dubois/Mannion rule might have in preventing wrongful
convictions, moreover, would be largely arbitrary. The only
defendants who benefit from the rule are those who are retried.'"
Most defendants do not get a second chance to revisit their testi-
monial decisions. To minimize the danger of poor defendant
testimony, we would also need a rule mandating at least one
retrial for all defendants convicted after testifying at their first
trial. In any case, the Dubois/Mannion rule gives very little pro-
tection to innocent defendants even in the few cases in which it
does apply. As I discuss more fully below, the two major threats
to innocent defendants' credibility are their poor presentation
skills and criminal records.'35 Even if their previous testimony is
excluded, defendants who testify again at their retrials cannot
avoid these problems. And defendants who remain silent would
have little to fear from the admission of their previous testimony.
Their criminal records would not be admissible and the effects of
poor presentation would be attenuated by the admission of their
prior testimony in transcript form. Lastly, in those rare cases in
which the admission of previous testimony would still present a
testimony to draw an inference of guilt, except to the limited extent that a finding
that the accused has been untruthful under oath could be damaging to his defence.
133. Dubois, supra, footnote 7, at p. 378.
134. Innocent defendants are admittedly more likely to be tried again, whether because
of errors leading to their conviction (which are corrected on appeal) or hung
juries. But there is no guarantee that an innocent defendant whose poor presenta-
tion skills contributed to a wrongful conviction would receive a new trial.
135. See discussion infra, footnotes 173-77 and accompanying text.
risk of wrongful conviction, the trial judge could exclude it on

the basis that its prejudicial effect would outweigh its probative
value.136
The chief effect of the Dubois/Mannion rule on adjudicative
accuracy is to make it more likely that factually guilty defen-
dants will be acquitted. Guilty defendants who perceive that the
case against them is strong often choose to take the stand,
hoping that their testimony will raise a reasonable doubt as to
their guilt. This strategy sometimes backfires. If their perception
is incorrect, their testimony may unwittingly extinguish the
reasonable doubt that would have existed had they remained
silent. There is nothing wrong with this.'37 Society is better off
when guilty defendants testify in a manner that contributes to
their conviction. But when this testimony is excluded from
retrials, these defendants are able to correct their mistakes. They
can either remain mute or change their testimony without fearing
that their previous responses will be used against them.t3 s In
summary, the Dubois/Mannionrule appreciably reduces the like-
lihood of convicting the guilty without significantly diminishing
the risk of wrongful convictions.'39
136. See Penney, supra, footnote 73 at pp. 294-95.

137. See Paciocco, supra, footnote 114, at p. 96. For a contrary view, grounded on a
faithful application of the case to meet principle, see Ratushny, supra, footnote 3,
at pp. 328-31.
138. Recall that s. 13 permits prosecutors to use an accused's testimony in a prosecu-
tion for perjury or giving contradictory evidence. These offences may be difficult
to prove, however, and may carry significantly lesser punishments than the
original offence.
139. In holding that s. 13 permits the admission at trial of defendants' preliminary
inquiry testimony, Martin J.A. stated the following in R. v. Yakeleya (1985), 20
C.C.C. (3d) 193 at p. 195, 9 O.A.C. 284, 46 C.R. (3d) 282:
It would . . . be anomalous, in our view, if an accused could testify under
oath as a witness at his preliminary hearing in respect of the charge against
him, in order to exculpate himself from the charge and upon ascertaining that
his first story did not fit the facts, could then tell a different story under oath
at his trial, and rely on s. 13 of the Charter to insulate his first story from
exposure at his trial. This would mean that the jury would be required to try
an artificial case rather than the real case. We cannot think that s. 13 of the
Charter in those circumstances precludes the exposure to the jury of the
accused's first story when he takes the stand in his own defence at his trial.
The same reasoning, of course, applies to retrials. Note that in Dubois, supra,
footnote 7, the Court expressly declined to opine on whether a preliminary inquiry
This might be acceptable if the rule were needed to prevent

abusive inquisitorial practices. But it is not. As mentioned,
defendants are adequately protected against oppressive questioning
at trial. Nor is the Dubois/Mannion rule needed to thwart
unwarranted invasions of privacy. In Dubois, the Court asserted
that the rule was mandated by the case to meet principle.' 4° There
is no compelling reason, however, why the state should have to
establish a primafacie case before requiring suspects to respond
to criminal allegations. As discussed, reasonable and probable
grounds should generally be sufficient.'"' In any case, in the
Dubois/Mannion scenario the accused is never actually com-
pelled. "42 The only pressure to testify arises from the strength of the
prosecution's primafacie case. Under the case to meet principle,
this kind of compulsion is perfectly acceptable.' 3 While the
prosecution may sometimes require the defendant's previous testi-
mony to establish another primafacie case at the retrial, this is
immaterial. What is abusive about admitting defendants' own
voluntary testimony? How much more privacy would they lose
beyond that already lost as a result of their arrest, prosecution and
initial decision to testify? Dubois and Mannion neatly illustrate
how doctrinaire adherence to the case to meet principle can generate
a rule that defies common sense and good criminal justice policy.
The Court should accordingly overrule them.
If Mannion is overruled and defendants' prior inconsistent
testimony becomes admissible at their retrials for the purpose of
incriminating them, then the question of whether such evidence
should be admissible for the purpose of impeaching their credi-
bility becomes largely irrelevant. If triers of fact can use such
evidence to determine defendants' guilt, then they should also
be able to use it to assess their credibility. The distinction
between incriminating and impeaching purposes is in any case
is "another proceeding" under s. 13. If it was, the Court added, the admission of
preliminary inquiry testimony at trial might be justified by s. 1 of the Charter.
140. Supra, footnote 7, at p. 365.
141. See Penney, supra, footnote 45, at pp. 264-66.
142. See Dubois,supra, footnote 7, at p. 385, McIntyre J. dissenting; Paciocco, supra,
footnote 114, at pp. 93-94.
143. See R. v. Noble, [1997] 1 S.C.R. 874, 114 C.C.C. (3d) 385, 6 C.R. (5th) 1,Lamer J.
dissenting; Sopinka, Lederman and Bryant, supra, footnote 4, at p. 826; Paciocco,
supra, footnote 51, at pp. 461 and 471.
an exceedingly fine one.1"' Despite limiting instructions, it will

often be difficult for triers of fact to resist the inference that
contradictory prior testimony is true. 145 It will also often be
tempting to infer guilt from the fact that at least one of the
accused's stories is untrue. But as long as defendants' prior
contradictory testimony remains inadmissible at retrials for
incriminating purposes, it is better to maintain the distinction and
allow its admission to impugn credibility. Even if triers of
fact sometimes "misuse" prior testimony by considering it for
its truth, there is no danger that this will lead to wrongful
convictions or abuse.
The situation is different when prosecutors seek to introduce
defendants' prior inconsistent testimony from proceedings other
than previous trials on the same indictment. As discussed, s. 13's
use immunity (and s. 7's derivative use immunity) encourage
non-accused witnesses to testify truthfully.'4 6 Without immunity,
non-accused witnesses at criminal trials, regulatory hearings,
commissions of inquiry and other legal proceedings would be
less likely to reveal self-incriminating information.'4 7 The
question here is whether limiting immunity to formally "incrim-
inating" uses (and allowing impeaching uses) significantly
144. See R. v. Kuldip (1988), 40 C.C.C. (3d) 11 (Ont. C.A.) at p. 23, revd supra,
footnote 130; Noel, supra, footnote 28, at para. 18.
145. See Noel, ibid., at para. 27.
146. See discussion supra, footnotes 26-28 and accompanying text.
147. See Noel, supra, footnote 28, at para. 24. The Noel Court failed to comprehend,
however, that immunity's truth-promoting effect does not apply to defendants tes-
tifying at their own criminal trials. Writing for the majority, Justice Arbour stated
at para. 25 that there is no reason "to draw a distinction between evidence given
under compulsion and evidence given voluntarily, even when the evidence is vol-
untarily given by an accused who waives his non-compellability and testifies in
his own trial". In each case, she argued, immunity is granted as a quid pro quo for
defendants' truthful testimony. If they testify falsely, the state's only remedy is to
prosecute them for perjury or the giving of contradictory evidence. Immunity
must still be provided, she concluded, to preserve the "contract". But as Justice
Arbour recognized, there is no actual contract; immunity arises from public law
rule grounded in a specific policy objective (the promotion of truthful testimony).
Immunizing defendants' voluntary trial testimony, however, does not further this
objective. Factually guilty defendants who testify truthfully will almost always be
convicted and punished. They cannot therefore be persuaded to admit their guilt
by an assurance that their testimony will not be used against them if they are
retried. The Court's failure to grasp this obvious truth is difficult to fathom.
diminishes this truth-promoting effect. Unfortunately, there is no

easy way to resolve this question. How can we estimate how
many more witnesses would have told the truth with protection
against impeachment (full immunity) than without such protection
(partial immunity)? I suspect that providing full immunity would
not produce much more trustworthy evidence. Partial immunity
assures witnesses that, as long as they remain silent in future
proceedings, their testimony cannot be used against them. On
the other side of the equation, giving defendants full
immunity would allow them to brazenly contradict their previous
testimony, safe in the knowledge that triers of fact would never
be informed of the inconsistency.'48 Partial immunity prevents
this from happening and thus increases the likelihood of con-
victing the guilty. The Court should accordingly maintain the
distinction between incriminating and impeaching purposes and
allow prosecutors to cross-examine defendants on their prior
testimony.'49 Further, it is probably not useful to attempt to
distinguish (as the Court did in Noel) between prior testimony
that "carries no risk of incrimination" and that which may have
both impeaching and incriminating effects.'5 ° It should be suffi-
cient if judges instruct the jury (and themselves) that prior
testimony may be used only to assess the accused's credibility.
148. See R. Delisle and D. Stuart, Evidence Principles and Problems, 6th ed.
(Scarborough: Carswell, 2001), p. 337; Noe'l, supra, footnote 28, at para. 132,
L'Heureux-Dub6 J. dissenting.
149. In Kuldip, supra, footnote 130, at pp. 639-42, the Court found that s. 5 of the
Canada Evidence Act does not forbid the admission of the defendants' previous
testimony for impeachment purposes. Recall that s. 5(2) states that the testimony
of witnesses who invoke its protection "shall not be used or admissible in evi-
dence against him in any ... criminal proceeding against him thereafter taking
place...". The Court changed its mind in Nol, supra, footnote 28, holding that
s. 5 prohibits the use of previous testimony for any purpose, including impeach-
ment. In my view, the text of s. 5 does not provide a definitive answer. Policy dic-
tates, however, that the Court revert to the position that it took in Kuldip. First,
Noel gives witnesses who claim the benefit of s. 5 at the original hearing greater
protection than those who do not. There is no justification for this different treat-
ment. Kuldip, in contrast, treats all witnesses equally and harmonizes the effect of
the Canada Evidence Act and the Charter. Second, as discussed in the text, pro-
viding protection against impeaching uses is not likely to induce many witnesses
who would otherwise have withheld the truth to divulge it. But it does allow
guilty defendants to contradict their previous testimony without consequence,
making it more likely that they will avoid conviction.
150. Supra, footnote 28, at para. 54.
To summarize, prosecutors should be allowed to adduce

defendants' testimony from previous trials for the same offence
for both incriminating and impeaching purposes. This position,
which is perfectly compatible with s. 13's wording, would
increase the likelihood of convicting the guilty without height-
ening the risk of abuse or wrongful conviction. Determining
whether prosecutors should be able to cross-examine defendants
on prior inconsistent testimony from other sources is more diffi-
cult. Allowing them to do so, however, would certainly promote
the conviction of the guilty. And it would probably not dissuade
very many non-accused witnesses from providing truthful testi-
mony. Defendants' prior testimony should therefore always be
admissible to impeach their credibility.
8. Inferences from Defendants' Silence
(1) At Trial
As discussed, the non-compellability rule prohibits prosecutors
from compelling criminal defendants' testimony at their own
trials. But the rule is not as great a boon to criminal defendants as
one might think. While defendants cannot be compelled to
testify, triers of fact will often expect them to do so. Innocent defen-
dants generally proclaim their innocence and relate any knowledge
that they have about the incident in question. The guilty, in contrast,
often fear that testifying (and especially cross-examination) will
expose their guilt. Triers of fact therefore often infer guilt from
silence.151 Judges understand, and juries are adamantly instructed,
that the prosecution bears the onus of proof and defendants are not
required to establish their innocence. But in many cases these
mantras are overborne by the pull of logic and common sense,
especially when the prosecution's evidence calls for the accused to
offer an explanation or alternative version of events.
Parliament chose to deal with this conundrum in an interesting
way. Section 4(6) of the Canada Evidence Act 52 forbids
151. See D. Dolinko, "Is There A Rationale for the Privilege Against Self-
Incrimination?" (1986), 33 U.C.L.A. L. Rev. 1063 at p. 1075; D. Seidmann and
A. Stein, "The Right to Silence Helps the Innocent: A Game-Theoretic Analysis
of the Fifth Amendment Privilege" (2000), 114 Harv. L. Rev. 430 at p. 446.
prosecutors and judges from commenting on defendants' failure

to testify. This guarantees that jurors will not be told that they
may draw an adverse inference from the accused's silence. But it
also means that the judge cannot tell jurors that they must not
infer guilt from silence.'53 Jurors are thus free to make whatever
inferences seem appropriate. In many cases jurors will count
defendants' silence as evidence of guilt.'54 There is no legislative
provision, moreover, prohibiting judges in non-jury trials from
doing the same. As Ratushny puts it, the non-compellability rule
"does not protect the accused from having adverse inferences
drawn against him by the trier of fact when he chooses not to
testify".'55 The purpose behind s. 4(6), it seems, is simply to
prevent judges and prosecutors from drawing 156
"unfair attention"
to the accused's decision not to testify.
In R. v. Noble, however, the Supreme Court drew heavily on
the free choice rationale in concluding that triers of fact may not
give independent weight to the failure of the accused to testify.'57
To do so, it held, would violate the presumption of innocence and
the accused's right to silence. In the Court's view, considering
silence effectively conscripts "communicative evidence of guilt"
from defendants, denying them the right to choose whether to
speak to the authorities. 5 s
153. Section 4(6) does not apply to defence counsel, who may therefore inform the jury
of the accused's right to remain silent. This is a poor substitute, however, for a
specific judicial direction to refrain from considering silence in determining guilt.
154. The Supreme Court has recognized this fact on numerous occasions. See R. v.
Corbett, [1975] 2 S.C.R. 275 at 280-81, 14 C.C.C. (2d) 385, [1976] 2 W.W.R.
524; Marcoux, supra, footnote 5, at p. 775; R. v. Vizeau, [1977] 2 S.C.R. 277 at
p. 288, 28 C.C.C. (2d) 81, 34 C.R.N.S. 309; R. v. McConnell, [1968] S.C.R. 802
at p. 809, [1968] 4 C.C.C. 257, 4 C.R.N.S. 269; R. v. Frangois, [1994] 2 S.C.R.
827 at p. 835, 91 C.C.C. (3d) 289, 31 C.R. (4th) 201.
155. Ratushny, supra, footnote 3, at p. 73.
156. Noble, supra, footnote 143, at para. 36, Lamer J. dissenting.
157. Ibid. The Court recognized that injury trials this rule may be impossible to enforce.
As discussed, s. 4(6) of the CanadaEvidence Act (the constitutionality of which the
Noble court did not decide) prevents judges from informing jurors of the rule, and
the absence of reasons in jury trials prevents appellate courts from determining
whether the rule was violated. Despite this, the Court maintained that in principle it
is "an error of law for the jury to become convinced of guilt beyond a reasonable
doubt as the result of the silence of the accused at trial": ibid., at para. 96.
158. Ibid., at paras. 74-75.
(b) Critique
If we believed that criminal defendants have an absolute right
to silence, and that any state-imposed pressure to speak violates
this right, then we might be convinced of the Noble Court's
reasoning. The possibility of an adverse inference does generate
a practical "compulsion" to testify. And if this kind of compul-
sion were morally wrong, then we might agree that it is a "'snare
and a delusion' to grant the accused a right to remain silent at
trial yet then proceed to use the silence to find him or her
guilty".'59 On the other hand, we might point out that in many
cases a practical compulsion to testify will exist despite the
Noble rule. As discussed, Noble does not prevent jurors from
drawing adverse inferences from silence. And even judges may
choose to favour common sense over legal prescription, although
after Noble they are unlikely to be so candid as to admit this in
their reasons. But as I have argued throughout this article, the
free choice rationale is theoretically untenable. So it is not
necessary to decide whether the Noble rule makes sense in free
choice terms.'" Instead, we should determine whether it can be
justified by either the abuse-prevention or reliability rationales.
While the Court's rationale for the "no adverse inference" rule
is couched primarily in free choice terms, it also evinces some
concern for state abuse. 6 It is difficult to imagine, however,
what this abuse consists of. As discussed, abusive inquisitions
come in two flavours. The state may attempt to extract incrimi-
nating information from suspects using cruel, inhumane
methods. This is clearly not an issue at trial. Examination takes
place in a public, recorded proceeding where defendants are
protected by their counsel, the presiding judge and evidentiary
rules limiting cross-examination to relevant, non-repetitive and
non-harassing questions.'62 Responding to a difficult cross-
examination may not be pleasant, but it is not abusive.
160. At a minimum, the free choice rationale would seem to demand the replacement
of s. 4(6) of the Canada Evidence Act with a requirement that judges warn jurors
that they may not draw adverse inferences from defendants' silence. See Noble,
supra, footnote 143, at paras. 36-38, Lamer J. dissenting.
162. See R.J. Delisle, "Silence at Trial: Inferences and Comments" (1997), 1 C.R.
(5th) 313 at p. 318.
Inquisitions may also be abusive when they are instigated

without sufficient cause for suspicion. But again, this is not an
issue at trial. Authorities must first establish reasonable and
probable grounds before bringing an accused to trial. 63
Moreover, courts have always required prosecutors to establish
at least a primafacie case before permitting triers of fact to draw
adverse inferences from silence. 'I In other words, the pressure to
testify generated by the possibility of an adverse inference arises
not from a fishing expedition, but rather "a specific accusation of
a pre-defined offence, supported by a judicial ruling that the
Crown has met a certain pre-defined standard of proof in accor-
dance with pre-established rules of evidence and procedure, in a
public trial at which the accused was represented by counsel,
who has had the opportunity to attack any possible weaknesses
in the Crown's case".16 In these circumstances, defendants'
privacy interests are vastly outweighed by the society's interest
in eliciting reliable evidence of guilt.'"
If abuse is not relevant, what about reliability? Would the
absence of the Noble rule increase the risk that innocent defen-
dants would be convicted, either on the basis of adverse
inferences or testimony that they would not have given under the
protection of the rule? To answer this question, we must first ask
another: why might an innocent person fear the consequences of
testifying? As mentioned, we generally expect innocent defen-
dants to testify. Most do so. But some of them may have good
reasons to remain silent.'67 They may wish to conceal embarrassing
or otherwise private information.'" Though this may lead to
wrongful convictions in a tiny number of cases, it does not justify
the Noble rule. As Ratushny points out, non-accused witnesses
163. At least reasonable and probable grounds are required to arrest, charge and com-
pel the appearance of suspects before a court. See Criminal Code, ss. 494, 495,
504 and 507.
164. See Ratushny, supra, footnote 3, at p. 75; Paciocco, supra, footnote 51, at p. 495;
Paciocco, supra, footnote 13, at pp. 82-84; Paciocco, supra, footnote 114, at p. 92.
165. Ratushny, ibid., at p. 332.
166. See Delisle, supra, footnote 162, at p. 320.
167. Se R.J. Delisle, "Annotation" (1983), 34 C.R. (3d) 386.
168. Some defendants who are innocent of the offences for which they are facing trial
may be guilty of unrelated offences. But as their testimony is protected by use and
derivative use immunity, they need not fear the consequences of testifying
truthfully.
cannot be excused from testifying in order to protect their

privacy.6 9 Uncooperative witnesses may be remanded into custody
or punished for contempt.'7° If the state is justified in breaching
their privacy, then as discussed immediately above, it is certainly
justified in breaching defendants' privacy. So if properly
instructed and mentally competent defendants would rather risk
conviction than testify about private matters, then they should be
allowed to make that choice.
Some innocent defendants may elect not to testify because
they would make bad witnesses. 7 ' Many criminal defendants are
inarticulate. Some have mental health problems or personality
traits that could alienate the trier of fact. Others may have
difficulty recalling the details of key events. And defendants
whose race, culture or communicative styles differ from those of
jurors may be misunderstood or be subjected to negative stereo-
typing.'72 Prosecutors can exploit any of these characteristics in
cross-examination or argument and lead jurors to make improper
inferences. To some extent, the effects of poor witness
presentation can be mitigated by effective representation.
Defence lawyers can prepare their clients for the rigours of the wit-
ness box, protect them from unfair cross-examination, and point
out the proper basis for assessing their client's testimony. But it is
likely that triers of fact will occasionally misuse this testimony and
in a small number of cases this could lead to wrongful convictions.
Innocent defendants may also wish to remain silent to prevent
the admission of their criminal records. Section 12(1) of the
Canada Evidence Act permits defendants (as well as other wit-
nesses) to be examined on their criminal records. 3 Although
169. See Ratushny, supra, footnote 3, at pp. 332-33.
170. See Criminal Code, ss. 9, 545 and 708(1).
171. See S. Schulhofer, "Some Kind Words for the Privilege Against Self-Incrimination"
(1991), 26 Val. U. L. Rev. 311 at p. 330; D. Ellis, "Vox Populi v. Suprema Lex: A
Comment on the Testimonial Privilege of the Fifth Amendment" (1970), 55 Iowa
L. Rev. 829 at p. 846; C. Bradley, "Griffin v. California:Still Viable After All These
Years" (1981), 79 Mich. L. Rev. 1290 at p. 1294; S. Easton, "Legal Advice,
Common Sense and the Right to Silence" (1998), 2 Int'l J. Evidence & Proof 109
at p. 114; Delisle, supra, footnote 162, at p. 319. See generally 0. Wellborn IIl,
"Demeanor" (1991), 76 Cornell L. Rev. 1075 at pp. 1078-91.
172. See generally S. Johnson, "The Language and Culture (Not to Say Race) of
Peremptory Challenges" (1993), 35 Wim. & Mary L. Rev. 21.
such evidence is relevant only to credibility, and is not admissible

for the purpose of showing that the defendant is more likely to
have committed the offence, defendants may fear that triers of
fact will be unduly influenced by their criminal past in
determining guilt. 7 4 Empirical evidence indicates that this fear is
well founded and commentators have therefore argued that
defendants' criminal records should not generally be admissible.'75
Lastly, defendants may decline to testify to preserve their right
to address the jury last. When the defence examines no witnesses,
the Criminal Code gives prosecutors this right. 76' Last address is
an undoubted advantage. Given the structure of adversarial
criminal trials, which require defendants to respond to the
prosecution's case and which purport to give defendants every
reasonable advantage, the defence should have the right of last
address in all cases. 77 Parliament should therefore reverse this
rule.7' But it is difficult to believe that innocent defendants,
whose testimony would otherwise assist their case, would
decline to testify simply in order to address the jury last.
174. In Corbett, supra, footnote 154, at p. 692, the Supreme Court concluded that
s. 12(1) did not violate the Charter, but held that judges may exclude prior
convictions "in those unusual circumstances where a mechanical application of
s. 12 would undermine the right to a fair trial".
175. See M. Friedland, "Criminal Law - Evidence - Cross Examination on Previous
Convictions in Canada- Section 12 of the CanadaEvidence Act" (1969), 47 Can.
Bar Rev. 656; Law Reform Commission of Canada, Report on Evidence (Ottawa:
Information Canada, 1975) at pp. 95-96; A. Doob and H. Kirshenbaum, "Some
Empirical Evidence on the Effect of s. 12 of the Canada Evidence Act upon an
Accused" (1972-73), 15 C.L.Q. 88; V. Hans and A. Doob, "Section 12 of the
Canada Evidence Act and the Deliberations of Simulated Juries" (1975-76), 18
C.L.Q. 235; Ratushny, supra, footnote 3, at pp. 335-43; R. Wissler and M. Saks,
"On the Inefficacy of Limiting Instructions: When Jurors Use Prior Conviction
Evidence to Decide on Guilt" (1985), 9 Law & Hum. Behav. 37. See also Corbett,
supra, footnote 154, at pp. 724-29, La Forest J. dissenting.
176. Section 651(3).
177. In R. v. Rose, [1998] 3 S.C.R. 362, [1996] 3 S.C.R. 362, 1 C.R. (5th) 199, a
narrow majority of the Supreme Court of Canada found that s. 651(3) did not
violate ss. 7 or 11(d) of the Charter. A differently constituted majority agreed,
however, that there were at the very least compelling reasons for Parliament to
allow defence counsel to respond to the prosecutor's closing address: ibid., at
paras. 16-57, Binnie J. dissenting and para. 59, L'Heureux-Dub6 J. concurring.
178. See Law Reform Commission of Canada, Report on the Jury (Ottawa: LRCC,
1982) at p. 68.
This brings us to the question of whether the no adverse

inference rule is needed to protect innocent defendants from the
harmful effects of testifying. In my view it is not.179 The question
would be even clearer if defendants could testify without being
cross-examined on their criminal records. But while this danger,
coupled with the bad witness problem, might justify the non-
compellability rule,' 81 it does not justify Noble. Defendants are
confronted with the dilemma of either testifying or facing the
possibility of an adverse inference only when the prosecution
establishes a primafaciecase. The only innocent defendants who
could be harmed by the possibility of an adverse inference are
consequently those who: (i) face a strong case against them; and
(ii) reasonably fear that the benefits of testifying will be out-
weighed by their criminal records or poor presentation. The
number of innocent defendants who meet these criteria is likely
minute.' 81 And even in those very few cases where innocent
defendants would be harmed by the possibility of an adverse infer-
ence, the Noble rule is not likely to protect them. Poor presentation
and criminal records are most likely to be misused by juries, not
judges. As we have seen, Noble does not prevent jurors from
drawing adverse inferences from silence.' 82 As Ratushny notes, the
only way to prevent such inferences altogether would be to83
resurrect the rule precluding criminal defendants from testifying.
In summary, Noble's prohibition on inferring guilt from
silence is justified by neither the reliability nor abuse-prevention
179. I admit to changing my view on this point. See S. Penney, "A Concern for
Innocence: Justice La Forest and the Criminal Law", in R. Johnson et al., eds.,
Gdrard V La Forestat the Supreme Court of Canada:1985-1997 (Winnipeg: The
Supreme Court of Canada Historical Society, Canadian Legal History Project,
2000), p. 217 at p. 221, note 19.
180. The question of whether the non-compellability rule is justified by the reliability
rationale is hotly contested. See e.g. Schulhofer, supra, footnote 171; M. Green,
"The Privilege's Last Stand: The Privilege Against Self-Incrimination and the
Right to Rebel Against the State" (1999), 65 Brooklyn L. Rev. 627. As non-com-
pellability is here to stay, there is little point in resolving this debate here.
181. See Green, ibid., at pp. 641-42.
183. Ratushny, supra, footnote 3, at p. 333. The Canada Evidence Act could be
amended, of course, to permit judges to caution jurors not to draw adverse
inferences, but as discussed there is no way to guarantee that they will heed this
caution.
rationales. Failing to testify is sometimes probative, reliable

evidence of guilt. The law should recognize this and expressly
authorize what triers of fact are likely to do in any case - draw
an adverse inference against defendants who fail to respond to
a prima facie case against them. 8 4 There should be no
constitutional impediment, moreover, to amending the Canada
Evidence Act to permit judges to tell jurors that they may draw
adverse inferences from the accused's silence in appropriate
circumstances.' 85
(2) Before Trial
If triers of fact should sometimes be permitted to infer guilt
from an accused's failure to testify at trial, then should they also
be allowed to infer guilt from the accused's silence? It will often
be logical, for instance, to infer guilt from a suspect's failure to
give a timely explanation to police. The inference may be
especially strong when a defendant offers an innocent explana-
tion for the first time at trial. But in R. v. Chambers,'86 the
Supreme Court concluded that such inferences violated the right
to silence inhering in s. 7 of the Charter. As in Noble, it reasoned
that it "would be a snare and a delusion to caution the accused
184. Ibid.
185. The Law Reform Commission of Canada recommended that s. 4(6) be abolished
and replaced with a provision authorizing judges, prosecutors and defence
lawyers to comment on the failure of the accused to testify and allowing triers of
fact to draw reasonable inferences therefrom. Law Reform Commission of
Canada, supra, footnote 175, at p. 89. See also Paciocco, supra, footnote 51, at
pp. 495-96. The CriminalJustice and Public OrderAct (U.K.), 1994, c. 33, s. 35,
permits adverse inferences to be drawn from defendants' silence. In R. v. Cowan,
[1995] 4 All E.R. 939 at p. 945 (C.A.), the Court held that jurors must be
instructed that the defendant is entitled to remain silent and they should not draw
an adverse inference from a defendant's silence unless the prosecution has
established a "case to answer" and "despite any evidence relied upon to explain
his silence or in the absence of any such evidence ... silence can only sensibly
be attributed to the defendant's having no answer or none that would stand up
to cross-examination". In Murray v. United Kingdom (1994), 18 Eur. Ct. H.R.
CDI, the European Court of Human Rights held that the right to silence did not
prohibit the drawing of adverse inferences in situations that clearly called for
an explanation.
186. [19901 2 S.C.R. 1293, 59 C.C.C. (3d) 321, 80 C.R. (3d) 235.
that he need not say anything in response to a police officer's

question but nonetheless put in evidence that the accused clearly
exercised his right and remained silent in the face of a question
which suggested his guilt".'8 7
(b) Critique
Did the Court's reliance on free choice thinking lead it to make
the same mistake that it made in Noble? Consider first the
reliability rationale. The Chambers Court implied that innocent
suspects might be "trapped" into remaining silent by the police
caution. While most innocent suspects respond to criminal
allegations as soon as possible, some may not. They may refuse
to cooperate, for example, because they distrust police or wish to
protect their or others' privacy. Without the Chambersrule, triers
of fact could draw adverse inferences against such defendants,
possibly leading to wrongful convictions.
Steps could be taken to prevent this. But they are unlikely to
be wholly successful. Police could be required to caution sus-
pects that remaining silent could result in an adverse inference
being drawn against them at trial.' 88 The effectiveness of this
warning, however, would hinge on the suspect's capacity to
fully understand the ramifications of remaining silent. Faced
with the possibility of an adverse inference, the rational
innocent suspect would always choose speech. But the stressful,
inherently coercive atmosphere of interrogation is not conducive
187. Ibid., at p. 1296. Note that there is an exception to this rule for alibi evidence. The
trier of fact may draw an adverse inference against defendants who fail to provide
reasonable notice to the prosecution of alibi evidence that they intend to present
at trial. See R. v. Cleghorn, [1995] 3 S.C.R. 175, 100 C.C.C. (3d) 393, 41 C.R.
(4th) 282. An accused may also cross-examine a co-defendant who has attempt-
ed to incriminate the accused on the co-defendant's pretrial silence. Such cross-
examination, however, is relevant only to credibility. See R. v. Crawford, [1995]
1 S.C.R. 858, 80 C.C.C. (3d) 421, 20 C.R. (4th) 331.
188. In England and Wales judges and juries may draw adverse inferences from
suspects' failure to mention facts that they later rely on at trial when they could
have reasonably been expected to mention such facts. See CriminalJustice and
Public OrderAct (U.K.), 1994, c. 33, s. 34. Police must warn suspects prior to ques-
tioning that while they do not have to say anything: "it may harm your defence if
you do not mention when questioned something which you later rely on in court".
See Police and Criminal Evidence Act 1984 (U.K.), c. 60 (Codes of Practice),
Code C: Code of Practice for the Detention, Treatment and Questioning of
Persons by Police Officers, s. 10.4.
to effective communication and rational decision-making. To

some suspects, the caution may appear contradictory and
obscure. 89 Despite being warned of the dangers of silence, some
innocent suspects may therefore elect to remain mute.
To some extent, this problem could be ameliorated by requiring
suspects to consult with counsel before deciding whether to
speak. Lawyers would presumably advise clients to offer any
innocent explanations to police. This would not guarantee, how-
ever, that all suspects would do so. Nor could suspects be forced
to consult with counsel in the first place. Adverse inferences
could be made contingent upon consultation, but then guilty
suspects could avoid inferences by refusing either to speak to
police or to consult with counsel.
An innocent suspect could make a poor decision, moreover,
even after talking to a lawyer. From the information conveyed
during a brief consultation, the lawyer might not be convinced
that the suspect is innocent. Fearing that speaking to police
would be more harmful than an adverse inference, the lawyer
could therefore recommend silence. Alternatively, lawyers could
advise suspects to remain silent on the basis that police had
insufficient evidence to justify a response or failed to sufficiently
disclose their case. Lawyers could also recommend silence when
they believed that their clients were suffering from a mental or
physical impairment that could prevent them from presenting a
convincing explanation.'" In any of these situations, it could be
difficult to prevent an adverse inference from being drawn
against an innocent suspect. As the English experience
illustrates, courts would not decline to infer guilt solely
because the defendant's lawyer recommended silence.'91 To do so
would defeat the purpose of permitting adverse inferences, as
lawyers would then always advise suspects to remain silent. The
question would then become whether the lawyer's advice
was "tactical", which would not preclude an adverse inference
(because it would be based solely on a desire to shield the
189. See M. Zander, "Silence in Northern Ireland" New Law Journal 151:6969
(February 2, 2001), p. 138, quoting J. Jackson, M. Wolfe, and K. Quinn,
Legislating againstSilence: The Northern Ireland Experience (Belfast: Statistics
and Research Branch, 2000), pp. 199-20.
190. See, e.g., R. v. Condron (1997), 1 Cr. App. 185 (C.A.).
191. Ibid.
18 - 48 C.L.Q.
guilty) or "non-tactical", which would preclude one (because

there were other, legitimate reasons for the suspect to remain
silent).'92 Such an inquiry would threaten the integrity of solicitor-
client privilege'93 and compromise advising lawyers' ability to
represent their clients at trial. 94 More fundamentally, it would
fail to eliminate the possibility of wrongful convictions, as courts
would not always be able to accurately distinguish between
legitimate and illegitimate advice. 95
The risk of wrongful conviction could be further reduced by pro-
hibiting adverse inferences in the absence of both aprimafaciecase
and a credible, innocent explanation for the defendant's silence."
But if an defendant's explanation for silence appeared unreasonable
and the prosecution's case was not otherwise sufficient to establish
guilt beyond a reasonable doubt, then there would still be a danger
that silence could be used to convict the innocent.
To summarize, permitting adverse inferences from pretrial
silence increases the risk of wrongful convictions. But given the
availability of protective mechanisms, it is difficult to determine
whether this risk is significant. On the other side of the equation,
it is also difficult to say whether permitting adverse inferences
would significantly increase the probability of convicting the
guilty. The available empirical evidence indicates that it may
not. 97 This suggests that permitting adverse inferences from
pretrial silence is not worth the risk.
Permitting courts to draw adverse inferences from pretrial
silence would also increase the possibility of abuse. 19 People are
192. Ibid., at p. 192.
193. English courts have suggested that a suspect's decision to oppose the drawing of
an adverse inference on the basis of legal advice to remain silent constitutes a
waiver of privilege. See ibid., at p. 196.
194. A lawyer cannot be a witness in proceedings against his or her client. See G.
McKenzie, Lawyers and Ethics: Professional Responsibility and Discipline, 2nd
ed. (Scarborough: Carswell, 1999), § 4.19. This is less problematic in jurisdic-
tions like England where there is a divided profession.
195. See Condron v. United Kingdom (2001), 31 E.H.R.R. 1 (Eur. Ct. H.R.). The
English regime also poses acute difficulties for lawyers, who must determine
whether to advise their clients to remain silent. See G. Van Kessell, "European
Perspectives on the Accused as a Source of Testimonial Evidence" (1998), 100 W.
Va. L. Rev. 799 at pp. 828-29.
196. See Condron, supra, footnote 190.
197. See Zander, supra, footnote 189.
198. See Royal Commission on Criminal Procedure (U.K.), Report (London:
H.M.S.O., 1981) at para. 4.52.
2004] What's Wrong with Self-Incrimination? Part 111 527
generally amenable to police questioning. Most feel an obliga-

tion to assist in the investigation of crime, and as mentioned
innocent suspects are usually anxious to provide any information
that may clear them of wrongdoing. But in cases where there is
little basis for suspicion, and the questioning relates to intimate
or embarrassing information, individuals may legitimately feel
that cooperation would unjustifiably invade their privacy. The
fight to silence protects against such invasions. But if suspects
were aware that their silence could result in an inference of guilt,
they could feel that they had no choice but to speak.
Allowing adverse inferences from pretrial silence could also
increase the prevalence of abusive interrogation practices. As dis-
cussed, the rights to silence and counsel help to insulate suspects
from oppressive questioning. They assure recalcitrant suspects that
they are not legally obliged to endure the interminable badgering
and psychological manipulation that often accompany interroga-
tion. They also remind police that they are not entitled to con-
fessions. Suspects facing the possibility of adverse inferences,
however, could rationally perceive that they would be better off
forgoing their rights. And if police are not satisfied by their
answers, they could be tempted to use abusive methods to induce
them to admit guilt. We could eliminate the possibility of this kind
of abuse by requiring questioning to take place in the presence of
a lawyer. But as mentioned, this would likely dramatically reduce
the possibility of obtaining confessions from guilty suspects.
So despite its superficial similarity to Noble, Chambersis jus-
tified by both the reliability and abuse-prevention rationales. At
trial, the possibility of an adverse inference arises only after the
prosecution has established a case to meet.'99 Defence lawyers
are fully informed of defendants' circumstances and can ensure
that they choose wisely between silence and speech. And where
defendants choose speech, the inherently non-coercive environ-
ment of courts minimizes the possibility of abusive questioning.
None of these protections exist before trial."° Intrusive questioning
may occur without reasonable grounds for suspicion, innocent

200. See generally R. v. Crawford, [1995] 1 S.C.R. 858 at pp. 876-77, 80 C.C.C. (3d)
421, 20 C.R. 331, where the Court stated that:
[A] clear distinction exists between the right of silence at trial and pre-trial
silence. Prior to or on arrest, the accused is in a much more vulnerable position
suspects may not receive good advice on whether to remain

silent, and police may use abusive interrogation tactics. If the
ability to infer guilt from pretrial silence were critical to the
truth-seeking function of the criminal process, then we would be
much better off with mandatory pretrial judicial examination,
where suspects could be protected against fishing expeditions,
abusive questioning and wrongful convictions. But in the
absence of such a process, we need the Chambers rule.
9. Conclusion
The law has undoubtedly been correct to view self-incriminating
evidence with suspicion. Bad consequences often flow from the
state's attempts to pressure individuals to provide information
that may be used to convict and punish them. Authorities may
use pressure tactics that are apt to induce suspects to confess
falsely or cause them more pain than we are willing to tolerate.
We may also object when authorities invade individuals' privacy
by requiring them to respond to criminal accusations in the
absence of objective grounds for suspicion, even when the means
used to compel answers are neither cruel nor likely to cause
wrongful convictions. Lastly, compelling individuals to provide
potentially self-incriminating answers may prompt them to
withhold trustworthy information, which may thwart the truth-
seeking function of legal proceedings.
Some jurists, however, have argued that compelled self-
incrimination is always problematic, even when it avoids these
consequential harms. These arguments centre on the notion that
criminal suspects should be able to freely choose whether to
cooperate with authorities. Theoretical justifications for this right
against the coercive power of the state. The environment in the police station
is different from that of the courtroom where procedural rules protect the
accused. In the police station, the accused may not be represented and he may
be overwhelmed by the whole experience. The police possess considerably
greater power than the accused and there are no disclosure obligations. The
police can disclose some or misleading information or no information at all.
Evidential use of silence forces the suspect to cooperate with his interrogators
without a reciprocal exchange of information and without placing proper
limits on the power of the police to demand cooperation. In contrast, in the
courtroom, the accused is represented, he knows the case that he has to meet
(due to disclosure) and there are rules regarding admissibility of evidence.
2004] What's Wrong with Self-Incrimination? Part I1 529
of free choice vary, but none is convincing. Those grounded on

the self-preservation instinct overestimate the psychological
harm associated with self-incrimination. Those based on privacy
concerns are too broad. Compelled self-incrimination does
threaten privacy, but like other invasive investigative methods, it
is justified when the state establishes a reasonable basis for
inquiry.
The Supreme Court of Canada has unfortunately permitted
free choice reasoning to contaminate much of its post-Charter
self-incrimination jurisprudence. While the Court has not entirely
ignored the reliability and abuse-prevention rationales, it has
concentrated its efforts on preserving suspects' freedom to resist
compulsion, at least until the prosecution establishes a case that
impels them to (voluntarily) take the stand. The Court has never
really articulated a coherent normative justification for this
position, but the roots of its stance are easy to discern. The free
choice and case-to-meet principles have been important features
of the criminal process in common law jurisdictions since the
mid-19th century."° In this "accusatorial" model, suspects are
generally not compellable during either the investigative or trial
stages of the process. The Court's mistake, however, has been to
infer from this descriptive truth that ensuring free choice is
normatively imperative. The Court has incorrectly assumed, in
other words, that preserving free choice is an end of the criminal
process, and not simply a (historically contingent) means of
achieving the process's ultimate objectives, which in the context
of self-incrimination law are determining truth and deterring
governmental abuse. As many commentators have illustrated, it
is not difficult to imagine alternative models that would pursue
these objectives as robustly as the accusatorial model without
guaranteeing freedom from compelled self-incrimination.2
This is not to say that we should disavow the accusatorial
model and construct something different in its place. The
201. See generally J. Langbein, "The Historical Origins of the Privilege Against Self-
Incrimination at Common Law" (1994), 92 Mich. L. Rev. 1047; E. Moglen,
"Taking the Fifth: Reconsidering the Origins of the Constitutional Privilege
Against Self-incrimination" (1994), 92 Mich. L. Rev. 1086; S. Penney, "Theories
of Confession Admissibility: A Historical View" (1998), 25 Am. J. Crim. L. 309
at pp. 314-23.
202. See Penney, supra, footnote 45, at p. 256, note 7.
Charter, which quite naturally instantiates the basic features of

that model, would not permit this. And it is not evident that
inquisitorial alternatives would achieve a better balance between
truth-promotion and abuse-prevention than the current system. 03
However, by treating free choice as an end (as opposed to a
means), the Court has failed to maximize the accusatorial
model's capacity to achieve these objectives. This fixation on
free choice has produced three types of errors. First, it has
deflected attention away from the need to deter cruel and false
confession-inducing interrogation methods. In its recent
jurisprudence re-examining the voluntary confession rule, the
Court has failed to require the audiovisual recording of interro-
gations, recognize courts' discretion to exclude unreliable
confessions coerced by non-authorities, and set out clear ex ante
limits on coercive interrogation tactics. It has similarly failed to
articulate a determinate, ex ante test for determining the onset of
detention for the purposes of s. 10 of the Charter, which dimin-
ishes the ability of that provision to protect suspects from
coercion. The Court has also declined to prohibit police from
questioning suspects who have invoked their s. 10(b) rights,
which also invites the possibility of coercive questioning.
Second, it has led the Court to ignore significant risks to sus-
pects' privacy. Thanks to its valorization of the case-to-meet
principle, the Court has generally been cognizant of the need to
prevent unjustified invasions of privacy. But this has not always
been the case. The Court has not recognized that suspects in the
field deserve more protection from attempts by undercover
agents to elicit confessions than suspects in custody. And by
making non-accused witnesses compellable unless the predomi-
nant purpose of the inquiry is to incriminate them, it has failed to
adequately protect them from abusive inquisitions.
Third, by insulating suspects from compulsion even when
there is no danger of wrongful convictions or abuse, the Court's
reliance on the free choice theory has unjustifiably enhanced the
ability of factually guilty suspects to evade conviction. The
Court has prohibited police from asking non-coercive questions
of intoxicated suspects who decline to consult with counsel. It
has prevented undercover agents from gently eliciting reliable
203. See discussion supra, footnote 180 and accompanying text.
2004] What's Wrong with Self-Incrimination? PartIll 531
confessions from detained suspects. It has mandated the near-

automatic exclusion of self-incriminating evidence that police
obtain by violating the Charter, even when the infringement is
minor and exclusion would be unlikely to deter future violations.
It has forced criminal investigators to obtain warrants to seize
statutorily compelled business records, even when those records
carry no reasonable expectation of privacy and can be seized
without authorization or cause by regulatory inspectors. It has
barred prosecutors from incriminating or impeaching defendants
at retrials with their previous voluntary testimony. And it has
forbidden triers of fact from inferring guilt from defendants'
unreasonable failures to respond to a prima facie case against
them.
It is time, therefore, for the Court to revisit its theoretical
approach to self-incrimination law. It must rid itself of the notion
that there is something inherently wrong with denying suspects
the right to choose to provide self-incriminating information to
the state. That argument is no more compelling now than it was
when Bentham first criticized it.2" Pressuring criminal suspects
to provide self-incriminating evidence can, however, often
produce undesirable consequences. Self-incrimination law
should focus on those consequences, on not on the mere fact of
compulsion.
204. See J. Bentham, Rationaleof JudicialEvidence (London: Hunt and Clarke, 1827),
pp. 207-83.

48 Crim LQ474

Uploaded by

Copyright:

Available Formats

48 Crim LQ474

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

48 Crim LQ474

Uploaded by

Copyright:

Available Formats

DATE DOWNLOADED: Thu Nov 16 05:33:34 2023

SOURCE: Content Downloaded from HeinOnline

Bluebook 21st ed.

ALWD 7th ed.

APA 7th ed.

Chicago 17th ed.

McGill Guide 9th ed.

AGLC 4th ed.

MLA 9th ed.

OSCOLA 4th ed.

(1) Testimonial Compulsion

eliminated these prerequisites. 6 Section 13 prohibits the

witnesses facing criminal charges are subject to the rules

10. See Ratushny, supra, footnote 3, at p. 81; W. Henkel, "Competency,

use immunity' 6 encompassing "evidence which could not have

purpose"." Where the state's primary objective is legitimate,

The common law privilege against self-incrimination, s. 5(2) of

compulsion would otherwise create. If authorities would have

from adducing any evidence obtained as a result of the com-

When it would be insufficient, the compelling authority could

other self-incriminating communications outside the context of

compromise between two incompatible, reliability-based objec-

To prevent these abuses, the Supreme Court of Canada has

interest in collecting the evidence.43 The definition of "pre-

As the Supreme Court rightly held in S. (R.J.) and Branch,

materializes. At the subpoena stage, the most common remedy

compulsion by the state".4 9 Numerous legislative provisions

49. Branch, supra, footnote 17, at para. 43.

whether compulsion impacts the reliability of the communication,

57. White, supra, footnote 38.

Most recently, in R. v. Jarvis'4 and R. v. Ling,65 the Court

Concerns about the reliability of compelled, non-testimonial

The situation in White is more complicated. There, the

The Court did not have to provide immunity to compelled

confessions. The distinction drawn in Jarvis and Ling between

81. See Penney, supra, footnote 73, at pp. 290-305.

coercive methods. In the absence of such limitations, tax inves-

moreover, reveals a legislative assumption that the regulatory

deciding whether to immunize evidence produced by non-

authorized by neutral arbiter.' The Court should engage in a

The Court was therefore wrong to hold in Jarvis that when an

defer to the legislature's choice to favour the availability of

protections against self-incrimination.1 2 Post-Charter outcomes

cursorily that police violated s. 7 of the Charter in taking bodily

that the "constitutionality of demanding a breathalyzer sample on reasonable and

self-incrimination law because it is inherently non-communica-

the B. (S.A.) Court recognized, the DNA warrant process

courts will exclude conscripted, non-linguistic evidence in the

127. Ibid., at p. 365.

and variable".'33 The problem must therefore be resolved by

risk of wrongful conviction, the trial judge could exclude it on

136. See Penney, supra, footnote 73 at pp. 294-95.

This might be acceptable if the rule were needed to prevent

an exceedingly fine one.1"' Despite limiting instructions, it will

diminishes this truth-promoting effect. Unfortunately, there is no

To summarize, prosecutors should be allowed to adduce

prosecutors and judges from commenting on defendants' failure

Inquisitions may also be abusive when they are instigated

cannot be excused from testifying in order to protect their