Insane Automatism: A Proposal for Reform
Alan Brudner
ont
restreint
la
Les
tribunaux
d6fense
d’automatisme sans alienation mentale de la common
/aw en d6finissant la maladie mentale, un 616ment es-
sentiel de l’automatisme avec alienation mentale, de fa-
gon assez large pour prendre au pifge toute personne
dont ‘automatisme risquerait d’entrainer ]a r6currence
d’un danger pour
le public. Cette d6finition
d’automatisme avec alidnation mentale selon le critare
de danger pour le public signifie que des personnes ju-
g~es innocentes de malfaisance sont d6tenues et possi-
blement enfermnves pour leur bien (d’aprbs les autres)
ou pour ce qu’elles risqueraient de faire k l’avenir, et
ce, en l’absence de ]a seule justification compatible
avec le respect de leur autonomie. Cette justification est
la suivante: la personne acquitt6e souffre de troubles
mentaux qui diminuent de fagon importante sa capacit6
de poser des actes autonomes, ce qui justifie le respect
moins important qui li est accord6. La d6finition
d’ali6nation mentale du Code criminel fait honneur A
cette justification, alors que la d6finition d’automatisme
avec ali6nation mentale de la common mv ne le fait
pas. En effet, une maladie mentale devrait tre redefinie
comme tout trouble mental qui rend la personne inca-
pable, de fagon gdnrale, d’apprdcier les consequences
raisonnablement prdvisibles de ses actes ou de com-
prendre linformation pertinente A la mise en ceuvre de
sa conception de bien-etre. Bien que cette ddfinition r6-
duirait le nombre d’auditions post-procis pour les per-
sonnes acquittdes, elle accorderait autant de protection
contre les personnes dangereuses qu’une soci~t6 libre
puisse permettre. Si ‘autonomie gravement diminule,
plutt que le risque de rdcurrence d’un danger pour le
public, 6tait le critare pour la defense d’automatisme
avec alienation mentale, il ne serait plus ncessaire de
rendre les defenses d’automatisme sans alienation
mentale et d’automatisme avec ali6nation mentale mu-
tuellement exclusives, ni d’imposer le fardeau de prou-
ver le caract~e involontaire de ses actes A l’accus6.
Courts have confined the common-law defence of
sane automatism by defining disease of the mind, a req-
uisite component of insane automatism, so broadly as to
ensnare anyone whose automatism might recur and lead
to violence. This definition of insane automatism in
terms of dangerousness means that persons found inno-
cent of wrongdoing are detained and possibly confined
for their own good (as others see it) or for what they
might do in the future, in the absence of the only justifi-
cation for giving force to these reasons consistent with
respect for their autonomy. That justification is that the
person acquitted is suffering from a mental disorder that
severely impairs his capacity for autonomous action,
justifying diminished respect The Criminal Code defi-
nition of legal insanity honours this justification, but the
common-law definition of insane automatism does not
Accordingly, a disease of the mind should be redefined
as any mental disorder that renders the person generally
incapable of appreciating the reasonably foreseeable
consequences of his actions or of understanding infor-
mation relevant to executing his conception of well-
being. While this definition would channel fewer ac-
quittees into the post-trial disposition hearing, it pro-
vides as much protection from dangerous persons as a
free society permits. If impaired autonomy rather than
dangerousness were the criterion of insane automatism,
there would be no need to make the pleas of sane and
insane automatism mutually exclusive or to place the
burden of proving involuntariness on the accused.
Professor, Faculty of Law, University of Toronto.
McGill Law Journal 2000
Revue de droit de McGill 2000
To be cited as: (2000) 45 McGill L. 65
Mode de rf6rence: (2000) 45 R.D. McGill 65
66
MCGILL LAW JOURNAL / REVUE DE DROJTDE MCGILL
[Vol. 45
Introduction
I. Why is Detention of the Innocent Insane Permissible?
II. Insane Automatism and the Condition for Diminished Respect
Ill. Disease of the Mind Redefined
IV. Exclusivity of Pleas and Burden of Proof
Conclusion
2000]
A. BRUDNER – INSANE AUTOMATISM
Introduction
Since the 1950s, the development of the defence of automatism in the common-
law world has been toward an ever narrower scope of applicability. Whereas in 1955,
an English judge could leave the defence to the jury in a case where the accused’s un-
conscious behaviour was precipitated by a cerebral tumour, the same case decided to-
day would almost certainly result in an instruction to the jury solely on insanity.’ Why
this is so is not difficult to determine. A successful defence of automatism means that
the accused has not acted, that his bodily movements were not expressions of a mind
or will; hence it leads to an absolute acquittal. The fear has been that someone who
falls into an automatic state because of a diseased mind and who causes harm will be
able to circumvent the restrictions (the reverse onus) and consequences (continued
detention) of the defence of insanity and that a person prone to recurrent unconscious
behaviour would thus go free, probably to inflict harm again. Accordingly, common-
law courts have increasingly confined automatism as an independent defence, taking
care that it does not encroach on the defence of insanity. They have done so by draw-
ing a distinction between sane and insane automatism. Sane automatism is uncon-
scious behaviour that is unlikely to recur or that is caused by an external event such as
a blow to the head or a psychological shock to which anyone might succumb; insane
automatism is unconscious behaviour attributable to a disease of the mind. If the evi-
dence suggests that the accused’s automatism was of the latter sort, then the judge will
instruct the jury only on insanity.
This in itself is quite unobjectionable. However, the courts have narrowly circum-
scribed the defence of sane automatism by defining very broadly the term “disease of
the mind” and so too the category of insane automatism In Bratty v. Northern Ire-
land (A.G.),’ Lord Denning defined disease of the mind as “any mental disorder
which has manifested itself in violence and is prone to recur”.! And such disorders
might well include not only psychological illnesses but also diseases of a purely
physical or neurological nature such as a brain tumour, arteriosclerosis,’ epilepsy, or
diabetes.” This definition of mental disease was expanded further in R. v. Rabey,7
‘R. v. Charlson (1955), 39 Cr. App. R. 37. For an approving commentary on this development, see
J.LJ. Edwards, “Automatism and Criminal Responsibility” (1958) 21 Mod. L. Rev. 375.
2 See K.L. Campbell, “Psychological Blow Automatism: A Narrow Defence” (1980-81) 23 Crim.
L.Q. 342; W.H. Holland, “Automatism and Criminal Responsibility” (1982-83) 25 Crin. L.Q. 95.
3 [1963] A.C. 386 (H.L.) [hereinafterBratty].
4TbidL at 412. InR. v. Parks, [1992] 2 S.C.R. 871 at 903, 95 D.L.R. (4th) 27, 15 C.R. (4th) 289, 75
C.C.C. (3d) 287 [hereinafter Parks cited to S.C.R.], Justice LaForest criticized the view that likeli-
hood of recurrence is an exclusive criterion of insane automatism, but it seems that it is nonetheless
conclusive.
. v. Kemp (1956), [1957] 1 Q.B. 399, [1956] 3 All E.R. 249, [1956] 3 W.L.R. 724 [hereinaf-
6 See R. v. Hennessy, [198912 All E.R1 9, 1 W.L.R. 287 (C.A.).
7 [1980] 2 S.C.R. 513, 114 D.L.R. (3d) 193, 15 C.R. (3d) 225, 54 C.C.C. (2d) 1 [hereinafter Rabey
SSee
ter Kemp cited to All E.R.].
cited to S.C.R.].
MCGILL LAW JOURNAL / REVUE DE DROIT DE McGILL
[Vol. 45
where it was held that any subjective disposition to which an automatic episode could
be attributed constituted a diseased mind whether or not the medical profession would
consider it pathological. And in R. v. Stone,s the Supreme Court of Canada pushed this
trend to the limit by articulating two propositions that render the defence of sane
automatism practically inaccessible. One is that an automatic episode is itself a mental
disorder, so that there is a presumption of a diseased mind whenever sufficient evi-
dence to ground a plea of automatism is introduced. The other is that, in assessing the
likelihood of recurrence, the judge is to consider not only the psychiatric history and
emotional make-up of the accused but also the likelihood that the triggering event will
itself recur.’ Together, these propositions effectively submerge the category of sane
automatism, for, according to the first, any case of unconscious behaviour is by defi-
nition behaviour evincing a diseased mind unless an external cause can be identified;
and, according to the second, even if an external cause can be identified, there may
nonetheless be disease of the mind if it is likely that the external cause will recur and
irrespective of whether there is evidence of a peculiar susceptibility in the accused.
Thus, if Rabey decided that a disease of the mind need not be a disease, Stone decides
that it need not even be predicable of a mind.
Clearly, the definition of mental disease elaborated by the courts in the aforemen-
tioned cases bears little resemblance to any that a psychiatrist might proffer. It is a le-
gal rather than a medical definition-one carefully crafted with a view to a policy of
controlling persons thought to be dangerous.” This definition ensures, first of all, that
anyone accused of a crime who lacked conscious choice because of a dangerous and
potentially recurrent disorder or event will not go free but will be subject to continued
detention and confinement after acquittal. Obversely, it ensures that only those whose
lack of conscious choice was caused by an external and probably non-recurrent event
will have the benefit of a defence that leads to absolute acquittal and a return to society.
However, the policy of controlling dangerous persons has produced a law of
automatism that, I shall argue, is seriously discordant with justice. For it has led to a
concept of a diseased mind that no ordinary person would recognize as such and
hence to the detention of innocent and reasonably well-functioning persons for their
own welfare and that of society-reasons for coercion that we normally admit only in
the case of the genuinely insane. Suppose, to take an extreme example, that a per-
fectly healthy football player inflicts harm during an unconscious episode resulting
[1999] 2 S.C.R. 290, 173 D.L.R (4th) 66, 24 C.R. (5th) 1, 134 C.C.C. (3d) 353 [hereinafter Stone
cited to C.C.C.
‘Ibid. at 440.
“See Bratty, supra note 3 at 410, Lord Denning:
Suppose a crime is committed by a man in a state of automatism or clouded conscious-
ness due to a recurrent disease of the mind. Such an act is no doubt involuntary, but it
does not give rise to an unqualified acquittal, for that would mean that he would be let
at large to do it again. The only proper verdict is one which ensures that the person who
suffers from the disease is kept secure in a hospital so as not to be a danger to himself
or others.
2000]
A. BRUDNER – INSANE AUTOMATISM
from a blow to the head suffered during a game. If, as Stone decides, the unconscious
episode itself raises a presumption of mental disease, and if the external cause cannot
dislodge the presumption because of the high likelihood of its repetition, then the
player’s only defence is insane automatism unless he intends to retire. True, there is
little prospect that such a person would actually be confined to a mental hospital once
acquitted. Nevertheless, he has been stigmatized as not criminally responsible because
of mental disorder, detained even though innocent pending his review, and subjected
to a process of review for dangerousness whose reservation for the insane shows that
we think it unfitting for the mentally competent.
If we regard Stone as the end-point in the decline of automatism as an independ-
ent defence, then its extreme inflation of the concept of mental disease may be viewed
as a reason for Parliament to review the entire law on insane automatism and to em-
bark on a fresh approach. Such an approach should rest on a more principled under-
standing of the concept of a diseased mind than that hitherto shown by the courts,
should bring the law of insane automatism into line with the statutory defence of in-
sanity, and should accomplish the policy objectives of the present law without violat-
ing any constitutional rights of the accused. In this essay, I shall suggest such an ap-
proach. To see what is needed, however, we must first understand how the present law
unjustifiably violates rights. It does so, I will argue, by detaining innocent persons for
reasons of social protection and their own welfare in the absence of the only justifica-
tion for giving force to these reasons consistent with their right to liberty. A satisfac-
tory criterion of insane automatism will provide this justification and will afford as
much social protection from dangerous persons as a free society permits. Throughout
the essay, I will use the idea of a free society or a society of free and equal persons as
a norm by which to test the admissibility of various justifications for coercion. By that
idea I mean a society whose moral foundations include (but are not necessarily ex-
hausted by) the premise that individual persons are free to pursue their welfare as they
see fit provided they do not interfere with, or impose excessive risks upon, the equal
liberty of others.
I. Why is Detention of the Innocent Insane Permissible?
To see what is wrong with the current law of insane automatism, we have to ask
more generally what, if anything, justifies confining those who have been acquitted of
wrongdoing on the ground of mental disorder. After all, these people are subject to
coercion by the state for reasons that would not be considered valid for most people.
Having been found undeserving of punishment, they are nonetheless forcibly con-
fined for their own good as others see it and to protect others from what they might do
in the future. In a society of free and equal persons, such treatment of a sane individ-
ual would be regarded as an unjust deprivation of liberty, because by coercing him for
our good or to implement our conception of his good, we violate a duty of respect
owed to his autonomy. To be sure, punishment also protects us from the convict for a
time and affords us an opportunity to reform him according to our conception of his
best interests. However, provided the traditional common-law criteria of guilt have
been satisfied, punishment is paradoxically respectful of the convict’s autonomy; for
if he has himself intentionally or recklessly violated a right to liberty, then he has de-
MCGILL LAW JOURNAL/REVUE DEDROITDE MCGILL
[Vol. 45
nied that there is such a right, and so has implicitly authorized the use of force against
him. By choosing crime, he has willed his own coercion and so too his subjection to
the good and good intentions of others. If, however, he has not chosen crime, then in-
tentionally coercing him for the sake of our interests is externally imposed force diffi-
cult to distinguish from crime itself. Accordingly, no one would think that a liberal
state could confine someone of sound mind who had neither committed nor attempted
to commit a crime simply because the authorities were of the view that he had dan-
gerous proclivities or that he would be better off without his liberty. Yet such an Or-
wellian fate routinely befalls the insane. How can it be justified?”
Two possibilities suggest themselves. One is to say that detaining the innocent in-
sane is also a violation of their right, but one that is justified by the social good it pro-
duces. The other is to deny that there is a violation of right. The first line of argument
is unpromising, however, because any social good that might be obtained by violating
the rights of the insane will also be obtained by violating the rights of the sane. We
have just as much reason for imprisoning someone we know intends a crime but who
has yet to act upon his intent or whom psychological testing has revealed as a sadist as
we do for imprisoning someone innocent and dangerously insane. Similarly, if we can
violate the rights of the innocent insane for their own good, why not the rights of the
innocent spendthrift, alcoholic, or cigarette smoker? Accordingly, the first line of ar-
gument threatens to submerge the liberal state in the Orwellian one, from which it will
be saved only by ad hoc restraints. Besides, although rights can be overridden by
goods, they cannot coherently be overridden by all goods. A right to autonomy may
be overridden by goods ensuring a greater autonomy for all; but it cannot consistently
be defeasible by considerations-such as another’s good or the opinion of others as to
one’s own good-incompatible with it, no matter how strongly felt the countervailing
interest. To allow a right to autonomy to be so defeated is to deprive the concept of a
right of any distinctive meaning or force.
One might object, however, that the right of self-defence is surely available to
anyone whose life or safety is threatened by a person, and that this right justifies the
use of force against a potential assailant provided the defender uses no more force
than is necessary to protect himself.'” Detaining and confining the dangerously insane
is arguably justified by self-defence, and there is nothing illogical about declining to
exercise the right in other cases where it might also apply. Other factors-such as a
concern for liberty-might lead us to invoke the right of self-defence against the dan-
gerously insane but not, say, against someone we strongly believe intends a crime but
who has neither conspired with another nor acted upon his intent.
“In R. v. Swain, [1991] 1 S.C.R. 933, 5 C.R (4th) 253, 63 C.C.C. (3d) 481, the Supreme Court of
Canada did not have to address this larger question, for it found a violation of fundamental justice in
the previous regime’s lack of controls over the detention of the innocent insane. For example, the lat-
ter could be held in custody indefinitely without a mandatory inquiry into their dangerousness to so-
ciety and without periodic reviews of their continuing dangerousness.
,2 See Criminal Code, 1&S.C. 1985, c. C-46, s. 34.
2000]
A. BRUDNER – INSANE AUTOMATISM
However, while the instinct for self-preservation may partly (for concern for the
needs of the mentally ill is also at play) motivate our policy of maintaining control
over those acquitted of crimes because of mental disorder, the right of self-defence
will not lend its normative stamp to the gratification of this instinct. The right of self-
defence is a right of self-help applicable only when the state’s powers of law en-
forcement are not realistically available and when the defender is thus effectively in a
state of nature vis-4-vis his assailant. Formally speaking, the right cannot be invoked
by the state against an individual subject to its laws, for this would imply that the col-
lectivity is in a state of nature vis-h-vis the individual, which implication contradicts
its authority as a state. Applied to the public authority, the “right of self-defence”
against those subject to its laws is nothing more or less than its right to apprehend and
punish wrongdoers, and yet the insane we are concerned with have already estab-
lished their innocence. Moreover, even if we view the right as being invoked, not by
the state, but by all potential victims severally, its rationale would not apply against
the innocent insane. Since the defenders have no greater authority to use force than
the state has, the rationale for their right of self-defence is identical to that of the
state’s right to apprehend and punish. Like the latter, the right of self-defence justifies
the use of necessary force against a wrongdoer who has, by his intentionally wrongful
act or intentionally wrongful threat of impending harm, implicitly challenged the va-
lidity of rights and so authorized the violation of his own.’3 Those acquitted because of
mental disorder were not, however, intentional wrongdoers in the past, nor do those
whom we confine after acquittal necessarily threaten others intentionally in the pres-
ent, and those that do may be dealt with as criminals. Thus, the right of self-defence
cannot justify maintaining control over the innocent insane.
Suppose, however, one is threatened with impending harm by someone whom
one knows is acting under the delusion that he is being attacked by extraterrestrials
and must kill in order to save himself, and in response to this threat one ties up the
delusional agent. Here, surely, the right of self-defence applies even though the wrong
is unintentional, and if an individual may invoke this right, so may a class of those
similarly threatened. To this argument there are two rejoinders, one a terminological
quibble, the other a more substantive reply. The quibble is that the defender here lacks
culpability, not because he has a right of self-defence properly understood (and which
has a distinctive rationale akin to that of punishment), but because he has a defence of
necessity, which justifies the infliction of harm when necessary to protect an interest
more important to freedom than the interest harmed, or which excuses unjust acts of
self-preference in extraordinary situations of peril where persons of ordinary fortitude
would have reacted similarly.” Since the delusional assailant does nothing that de-
serves a coercive response (he would be acquitted at trial), his position is no different
from that of someone whom impersonal necessity (here the disease working on his
‘3 Thus, the statutory right of self-defence is premised on the existence of an unlawful assault; see
‘4 See R. v. Perka, [198412 S.C.R. 232, 13 D.L.R. (4th), 6 W.W.R_ 289,42 C.R. (2d) 43, 14 C.C.C.
Criminal Code, ibid.
(3d) 385.
MCGILL LAW JOURNAL / REVUE DE DROITDE MCGILL
[Vol. 45
powers of perception and judgment) has made one’s competitor for self-preservation.
The substantive reply is that the defence of necessity can neither justify nor excuse
our existing practice of confining those acquitted because of mental disorder; for this
practice requires no showing of an impending harm that would justify an immediate
response or make delay impracticable for the person of ordinary fortitude.
Accordingly, if our coercion of the innocent insane is to be legitimated, it must be
done so not through a claim that the violation of their right is justified but through a
denial that their right has been violated in the first place. Such a denial might be sup-
ported by either of two lines of argument. One could argue that coercion of the inno-
cent insane does not violate their right, because if they knew what we know about the
risks they pose and possessed full powers of deliberation and judgment, they would
consent to being held in custody for their own and others’ safety. I shall call this the
argument from constructive consent.’5 Alternatively, one could deny that the insane
fully possess the capacity for autonomous action that is the basis of the right the state
must respect. If a capacity for autonomy is the quality of a person that commands re-
spect as a matter of right, and if the insane lack this capacity, then coercing them for
reasons incompatible with respect for autonomy violates no right of the insane. I shall
call this the argument from diminished respect.’6
There is reason to view the argument from constructive consent with suspicion.
The point of this argument is to justify coercion in a way that takes rights of auton-
omy seriously as a constraint on the societal pursuit of values external to those of the
person coerced. The argument is that we can pursue these values only if the person
against whom force is used consents to the force. The idea of consenting to force will
seem less paradoxical if we recall the story of Odysseus who, to protect himself from
the allure of the Sirens’ song, commands his sailors to bind him to the mast and to dis-
regard any attempts he might make to revoke his order. In this example, of course,
Odysseus actually consents to the future application of force (much as someone might
do now by creating a power of attorney), whereas the argument under consideration
relies on a notion of constructive consent. Such an argument may nonetheless derive
more or less force from the Odysseus example depending on how closely the situation
in which force is used approximates one of actual consent. To borrow an example
(somewhat modified) of John Stuart Mill’s, if someone is unwittingly about to walk in
the path of a truck and there is no time to warn him, applying force to stop him can be
justified by a notion of constructive consent, since a reasonable person would actually
consent to such a trivial interference if it were necessary to save his life.”
However, the farther we get from a situation of imminent and certain harm to life
and limb, the more strained and artificial does the argument from constructive consent
3d ed. (Toronto: McGraw-Hill Ryerson, 1992) 398.
s See J. Woodward, “Paternalism and Justification” in W. Cragg, ed., Contemporary Moral Issues,
This was Hegel’s argument; see Hegel’s Philosophy of Right, trans. T. M. Knox (Oxford: Claren-
don Press, 1964) at 82.
‘” See J.S. Mill, On Liberty (New York: Crofts, 1947) at 97; Woodward, supra note 15 at 400.
2000]
A. BRUDNER – INSANE AUTOMATISM
become.’8 This is so because, short of that extreme situation, the point at which some-
one may prefer protection against a risk of harm to his freedom of movement involves
a subjective weighing of values that varies from one individual to another. If, within
the area of reasonable diversity of preference, we simply impute to the insane person
the values we hold or consider rational to hold, then obviously our respect for the
autonomy constraint on the pursuit of our values is a sham- We could, however, say
that coercion of the insane is justified only if the insane person would have consented
in advance given the relatively enduring content and configuration of his own values
and commitments. And we might thus require substituted consent from someone in-
timately familiar with those values. Yet, the cases of imminent harm to oneself aside,
the impossibility of projecting oneself into another’s schedule of values so as to de-
cide as he would have decided, as well as the high risk of bias in doing so, means that
here too the argument from constructive consent-and the device of substituted con-
sent in particular-is a mere figleaf for a consequentialism freed of effective con-
straints.’9 Indeed, our practice understands this, for it places no reliance on substituted
consent as a legitimating precondition for confining the innocent insane. It simply
confines them.
Accordingly, the argument from constructive consent will justify coercion of the
innocent insane only in a very narrow range of cases involving imminent and practi-
cally certain harm to interests crucial to anyone’s conception of welfare. Needless to
say, our present practice of confining the insane is not limited to such cases. There-
fore, either this practice cannot be justified in a free society or there exists another
justification.
This brings us to the argument from diminished respect. The reason we may co-
erce the innocent insane, according to this view, is that they lack the capacity for full
autonomy that grounds a right to respect for their choices. Here we make no pretense
of receiving consent for subordinating their liberty to the welfare of others or to their
own welfare as others see it. The detention is non-consensual but nonetheless permis-
sible because in the case of the insane, consent is not required.
The argument from diminished respect may have an unattractive ring, but there is
at least one element of the law of insanity for which it provides an attractive explana-
tion. I refer to the presumption of sanity and the corresponding burden on the accused
to prove insanity should he wish to plead it. Since the result of a successful plea of
mental disorder is that the accused is morally innocent of criminal wrongdoing, the
reverse onus violates his right to be presumed innocent, as a majority of the Supreme
Court held in R. v. Chaulk.: However, this violation is justified in a free society, one
8 This was recognized by Mill, see ibiL at 98.
‘9 In Ontario’s Consent to Treatment Act, S.O. 1992, c. 31, s. 13, if the substitute decision-maker
does not know of an actual wish of the incapable person with respect to the proposed treatment, he is
instructed to act in the incapable person’s best interests, and the values and beliefs of that person are
only one of several factors to be considered in deciding where his best interests lie.
‘0 [1990] 3 S.C.R. 1303,2 C.R. (4th) 1, 62 C.C.C. (3d) 193.
MCGILL LAW JOURNAL / REVUE DE DROITDE MCGILL
[Vol. 45
could argue, because an acquittal on grounds of mental disorder exposes the individ-
ual to consequences that would be degrading to a free person, and we need more than
a reasonable doubt about the accused’s sanity before we subject him to those conse-
quences. Even if the accused prefers therapeutic custody as a diminished person to
conviction as a full one, the inalienability of the person’s right to autonomy requires
that he prove the condition for diminished respect, at least on a balance of probabili-
ties. Here, then, the reason for limiting the right to be presumed innocent is consistent
with the norm of respect for persons underlying the right itself.
Let us assume that an argument from diminished respect is needed to justify de-
tention of the innocent insane for prophylactic and paternalistic reasons. A question
now arises as to whether all those who have a disease of the mind lack the capacity for
autonomous action that grounds a right to respect. Here we must distinguish between
two senses of autonomy. The latter can mean a purely formal capacity for choice in
the presence of which we can always say that an agent could have done otherwise
than he did, no matter how powerful the emotions or distorted the perceptions that
drove him to act in a certain way. It seems clear that, with the exception of a transient
episode of automatic behaviour, no mental disorder can vitiate autonomy in this weak
and formal sense, which is why insanity justifies at most a diminished respect for
agency rather than a total withdrawal. Even those insane persons deemed dangerous
to themselves or to others are entitled to-and under our positive law generally re-
ceive-as much respect for their choices as is consistent with their own and the pub-
lic’s safety.’
However, the idea of autonomy can also denote something richer and more capa-
cious than mere freedom of the will. It can refer to a capacity to act from ends them-
selves chosen after reflection on the kind of life one wishes to lead and an ability to
see in the outcomes of one’s deeds the embodiment of one’s considered values and
choices. It seems possible that mental disorder could seriously impair autonomy in
this larger sense, if only because it might pervasively distort information about the en-
vironment that one needs in order to accomplish what one chooses to do. If I want to
behead a snake that is attacking me but instead decapitate a person who appears to me
as a snake, I do not act autonomously in the capacious sense, for I do something and
face consequences I did not choose. And if as a result of an underlying pathology, I
suffer from these delusions frequently, then my overall capacity for autonomous ac-
tion is fundamentally damaged.
There is, however, no reason to think that mental disorder necessarily effects such
a radical diminution of one’s capacity for autonomy, and so we need to distinguish
between those disorders that justify diminished respect for autonomy and those that
do not. Section 16 of the Criminal Code does this quite ingeniously, for it employs a
formula that does double duty as a criterion of both moral innocence and diminished
respect. It states that “no person is criminally responsible for an act committed…while
21 See, for example, Criminal Code, supra note 12, ss. 672.54 and 672.55.
2000]
A. BRUDNER – INSANE AUTOMATISM
suffering from a mental disorder that rendered the person incapable of appreciating
the nature and quality of the act…or of knowing that it was wrong'” That the person
did not appreciate what he was doing renders him innocent of criminal wrongdoing
because he lacked the mens rea for the offence. That he did not appreciate because he
could not means that he is generally incapable of fully autonomous action, hence less
worthy of respect than those with full capacity, hence subject to continued detention
even though innocent. Similarly, that the accused did not know his action was wrong
negates criminal intent in the same way that colour of right negates guilt for theft.
That he was incapable of knowing means that he is chronically subject to delusions
that lead him to mistake criminal actions for ones that reasonable people would regard
as justified or excused. Because he did not know, he is innocent of crime and unde-
serving of punishment; because he could not know, he has a diminished capacity for
autonomy and so is detainable for reasons of safety alone.
The statutory provisions relating to fitness to stand trial can be understood in a
similar way. Someone “unable on account of mental disorder to conduct a de-
fence…and, in particular, unable…to understand the nature or object of the proceed-
ings…the probable consequences of the proceedings, or commmunicate with coun-
sel”‘ is excused from trial, because the fairness of the trial would be seriously com-
promised by the accused’s inability to mount a defence or to provide counsel with the
information needed to mount one for him. But the same criterion that selects those un-
fit for trial also selects those whose mental disorder is so extreme as to render them
incapable of functioning as autonomous agents. And it is because they are incapable
of so functioning that they can be indefinitely confined without trial by virtue of their
dangerousness alone.
Accordingly, the Criminal Code provisions on mental disorder can be reasonably
interpreted as carefully selecting those who are subject to state coercion for public
safety and paternalistic reasons even though innocent of crime (or not yet proved
guilty) on the basis of a diminished capacity for autonomy. Although incapacity to
appreciate consequences and to know right from wrong might, along with past violent
conduct, strongly indicate a dangerous person, it is clear from the overall structure of
the Code provisions on mental disorder that dangerousness is not the criterion for the
Court’s remitting the innocent insane to state custody. Where dangerousness becomes
the key factor is at the post-trial disposition hearing, at which the Court (now recon-
stituted as an administrative body) or Review Board is instructed by law to impose the
least restrictive order consistent with public safety and the welfare of the mentally
disordered individual.’ At this stage a full inquiry is conducted into the present
dangerousness of the acquittee-an inquiry that would be redundant if section 16
were interpreted as providing predictors of violence. Moreover, the criterion of unfit-
ness to stand trial-inability to conduct a defence-is not even loosely connected
with a propensity for violence, nor is the suspected commission of an unlawful act,
” Criminal Code, supra note 12, s. 2.
! lbid, s. 672.54.
MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL
[Vol. 45
which may not have involved violence at all. Accordingly, section 16 and the fitness
provisions are best viewed as providing eligibility criteria for determining which in-
nocents will be vulnerable to state coercion for reasons of dangerousness rather than
as themselves providing indicia of dangerousness. So interpreted, these provisions are
fair because, by selecting for coercion for social protection and paternalistic reasons
only those innocents who, lacking a capacity for full autonomy, command diminished
respect for their agency, they violate no right of the person.
II. Insane Automatism and the Condition for Diminished Respect
Now let us turn to the law of automatism. The latter is a common-law defence,
and the limitations that have formed around it are common-law limitations. In par-
ticular, the distinction between sane and insane automatism is a judge-made distinc-
tion; and the definition of insane automatism has been developed by judges with’ a
view to cases of involuntary behaviour and without regard to how well it agrees with a
statutory definition of legal insanity developed with a view to non-intentional conduct.
Despite this difference in origin, the courts have treated the defence of insane
automatism as subsumed under the section 16 defence, in order that the legislative re-
gime for dealing with the acquitted insane may take over. This merger, however, pro-
duces a rather glaring disharmony. Although someone acquitted of a crime because of
insane automatism will have been acquitted because he was incapable of appreciating
the nature and quality of his act and will be subject to the process governing those ac-
quitted because of mental disorder, he will not have been channelled into this process
under the same criterion as others acquitted under section 16. Whereas those who
acted consciously are detained after acquittal simply because of their incapacity due
to mental disorder to appreciate what they were doing, those who behaved uncon-
sciously are detained because their involuntary behaviour was attributable to an inter-
nal disposition likely to lead to recurrent episodes. Under the common law of
automatism, therefore, dangerousness rather than a diminished capacity for autonomy
selects the innocents destined for continued detention’ Such detention is unjust, how-
ever, because it involves the coercion of someone who is innocent of crime and who
has full capacity for autonomy for the sake of interests external to his own. It thus
violates the autonomy of someone who has neither authorized such a violation
through criminal conduct nor justified a diminished respect for his autonomy through
severe mental disorder.
It might be objected, however, that the criterion of insane automatism, no less
than that of section 16 mental disorder, does double duty as a test of moral innocence
and diminished autonomy, so that no one worthy of full respect will be detained after
an acquittal because of insane automatism. Anyone who falls into an automatic state
24 This is also true of the Law Reform Commission of Canada’s proposal, which would have codi-
fied the common law essentially as it existed after Rabey, supra note 7. See Criminal Law: The Gen-
eral Part: Liability and Defences (Working Paper 29) (Ottawa: Law Reform Commission of Canada,
1982) at 67.
2000]
A. BRUDNER – INSANE AUTOMATISM
and commits a violent act from a mental disorder that is likely to recur is, one might
argue, as lacking inyfull autonomy as the person who was incapable of appreciating
what he was doing or of knowing that it was wrong. Yet a consideration of some cases
will reveal this assurance as overly sanguine.
In Kemp, the accused struck his wife with a hammer causing a serious wound.
There was strong medical evidence that at the time of the attack, the accused was not
conscious of picking up the hammer nor of striking his wife with it. Two doctors also
testified that the accused’s loss of consciousness was probably caused by an inade-
quate supply of blood to the brain, in turn attributable to arteriosclerosis. Much of
Lord Devlin’s judgment is taken up with a discussion of whether arteriosclerosis is a
disease of the mind so as to warrant an instruction on insanity rather than involuntari-
ness; and of course, once the issue is framed in these terms, the conclusion is inevita-
ble that a mental disease might have a physical as well as a purely psychic cause.
Concluding that the accused was indeed suffering from a disease of the mind, Lord
Devlin left only insanity with the jury, which accorded Mr. Kemp the dubious benefit
of an acquittal on that ground.
Surely, however, it is not enough to ask whether an episode of automatism is at-
tributable to a disease of the mind. Since an acquittal because of insanity leads to de-
tention despite innocence, one must also ask whether the mental disease was of a na-
ture and extent that would justify the diminished respect for autonomy implied by that
disposition. The facts of the case relevant to such an inquiry are mentioned by Lord
Devlin only in passing. They are that the accused was, as Lord Devlin described him,
a man “of excellent character” and that he and his wife had “always been thought to
be a devoted couple.'” Moreover, two doctors testified that, although arteriosclerosis
might lead to a massive degeneration of brain cells amounting to a disease of the mind
as they understood the term, Kemp’s condition had not yet nearly progressed to that
point. What had occurred, according to these witnesses, was a temporary interference
with the normal supply of blood to the brain, akin to that which occurs during a con-
cussion. Since the cause of that interference was an underlying physical disease, it is
reasonable to assume that the violent behaviour could happen again. Yet Kemp was
far from exhibiting the kind of mental deterioration that alone justifies the diminished
respect for autonomy signified by detention following acquittal and by the possibility
of indefinite confinement because of dangerousness alone.
The case of R. v. Quice likewise illustrates the divergence of the criteria for in-
sane automatism and diminished respect, although in this case the Court reached the
correct result by refusing to follow the logic of Lord Devlin’s position. The accused
was a nurse who assaulted a paraplegic patient causing serious bodily harm. The evi-
dence revealed that the accused was a diabetic, that he had taken insulin on the
morning of the assault, and that he had consumed little food but a considerable quan-
tity of alcohol during the day. There was also evidence that on twelve previous occa-
Kemp, supra note 5 at 251.
[1973] 3 All E.R. 347,3 W.L.R. 26 (C.A.) [hereinafter Quick].
MCGILL LAW JOURNAL / REVUE DE DROITDE MCGILL
[Vol. 45
sions, the accused was admitted to hospital either unconscious or semi-conscious
from hypoglycaemia, a condition produced by too low a level of blood sugar. On four
of these occasions, the accused was violent. A doctor also testified that Mr. Quick’s
ingestion of insulin combined with his having consumed little food and much alcohol
could have produced hypoglycaemia, in the advanced stages of which the sufferer
may become unconscious and violent. On this evidence the trial judge ruled that the
accused’s only defence was insanity, whereupon the accused pleaded guilty and ap-
pealed.
Applying Lord Denning’s definition of mental disease (any mental disorder that
has manifested itself in violence and is prone to recur) to this case, the conclusion is
inescapable that the trial judge’s ruling was correct. Unconsciousness produced by
hypoglycaemia is no different than unconsciousness produced by an epileptic seizure
or by arteriosclerosis; moreover, the disorder had manifested itself in violence, had
occurred several times before, and was likely to occur again. Nevertheless, the Eng-
lish Court of Appeal resisted the pull of this logic in the name of common sense.
What, after all, would be the point of sending Quick to a mental hospital? Since, how-
ever, one could ask the same question about Kemp, Lord Lawton needed a principle
to stop the logic of Lord Devlin. The one he devised drew a distinction between un-
consciousness produced by an external factor (such as a blow to the head or the in-
gestion of a drug) and that caused by an internal one. The former, which is a transitory
and likely isolated incident, gives rise to a defence of sane automatism; the latter,
which is more likely to recur, supports only a defence of insanity. In this case, said
Lord Lawton, the accused’s automatic state was produced by an external factor-the
injection of insulin.
From the point of view of Lord Denning’s dangerousness criterion, however, the
external/internal dichotomy is unstable, because unconsciousness may be produced
by an external factor working upon an internal susceptibility. A thin-skulled person
may be just as prone to unconsciousness from minor blows to the head as an epileptic
is from internal causes. Quick himself took insulin only because he was a diabetic. To
say that his unconsciousness was produced by an external factor ignores the internal
condition that ensures the regular repetition of the external cause. Accordingly, the
dangerousness test of insane automatism would, were its logic pursued relentlessly,
detain and possibly confine indefinitely an innocent person whose brittle skull or dia-
betes made him a risk to others. Neither a brittle skull nor diabetes, however, funda-
mentally impairs autonomy. And if it is objected that this logic needn’t concern us,
because no purpose would be served by placing a diabetic in a mental hospital, the re-
ply is simple: if the point is public safety, why must it be a mental hospital or, indeed,
even a hospital?
The breakdown of the external/internal dichotomy is the theme of Rabey, the
leading Canadian case on psychological blow automatism. There the accused was a
somewhat introverted university student who had conceived a romantic interest in a
2000]
A. BRUDNER – INSANEAUTOMATISM
woman with whom he shared various student activities. One day he found a letter she
had written to a girlfriend in which she declared a sexual interest in someone else and
described the accused as a “nothing.’ The next time the accused saw her, he inquired
as to her feelings about him, and when told that she regarded him simply as a friend,
he fell (or so he said) into a semi-conscious state and began striking her with a rock
and choking her. When he realized what he was doing, he stopped. At trial, the ac-
cused raised the defence of sane automatism, claiming that the psychological shock
received on reading the letter induced a dissociative state that was not a disease of the
mind nor symptomatic of any underlying pathology. The Crown argued that a disso-
ciative state is itself a mental disease and that the accused’s only defence on the evi-
dence offered was insanity. Both arguments had the support of psychiatric opinion.
The trial judge, however, accepted the view of the defence psychiatrist that the ac-
cused was not insane and acquitted the accused on the ground of sane automatism.
The Crown successfully appealed the ruling on insanity and the accused then ap-
pealed to the Supreme Court of Canada.
A majority of the Court dismissed the appeal on reasoning borrowed from Martin
J.A. of the Ontario Court of Appeal. Justice Martin distinguished between a malfunc-
tioning of the mind caused primarily by factors internal to the accused and that attrib-
utable to an external cause. Where the event triggering a dissociative episode is a psy-
chological blow, one must inquire into the nature of the trauma in order to locate the
primary cause of the malfunctioning. If the shock were sufficiently extreme to pre-
cipitate the same reaction in an ordinary individual, then the cause is the external
shock, and the appropriate plea is sane automatism. However, if the triggering event
is, as in this case, part of the ordinary stresses and disappointments of life, then the
primary cause is an unusual emotional sensitivity of the accused, and the only defence
is insane automatism. Ritchie J., who wrote the majority judgment in the Supreme
Court, found the key to the accused’s extreme reaction in the latter’s introversion and
inexperience in amorous relationships, in which psychological setting the infatuation
with the woman created an “abnormal condition” amounting to disease of the mind.’
In Rabey, accordingly, the logic of the likelihood of recurrence test that was re-
sisted by Lord Lawton in the case of an insulin injection for diabetes was pursued in
the case of a psychological shock working on an unusual susceptibility. The result
was that introversion, inexperience with the opposite sex, and an infatuation were held
to indicate a diseased mind for which the accused could be detained despite having
been found innocent of criminal wrongdoing. Although the Crown’s psychiatrist testi-
fied that a dissociative episode was itself a mental disorder, there was no evidence in
Rabey of any underlying pathology, either mental or organic, that would warrant the
diminished respect for autonomy implied by detention of the innocent for reasons of
safety. In response to Justice Dickson’s dissenting argument that a verdict of not
guilty by reason of insanity would be unfair to an accused in whom no pathology had
Rabey, supra note 7 at 517.
2b lid at 521.
MCGILL LAW JOURNAL / REVUE DE DROITDE MCGILL
[Vol. 45
been found, Ritchie J. pointed to the legal provisions permitting an absolute or condi-
tional discharge of someone found by a Board of Review to be no longer dangerous.
However, this response misses the point. The mere detention of an innocent person for
the purpose of assessing his dangerousness as well as his subjection to a procedure
that makes his liberty conditional on predictions of harm all involve affronts to human
dignity for which some threshold justificatory condition must be met. Public safety
cannot itself be that justification, because detention for prophylactic reasons alone is
precisely the thing to be justified. Nor can it be an answer to the offence to dignity
implicit in the very process of administrative review for dangerousness that the person
acquitted may at the end of the process be set free.
III. Disease of the Mind Redefined
The cases of Kemp, Quick, and Rabey amply illustrate the possibility of diver-
gence between the common-law criterion of insane automatism and that for dimin-
ished respect for autonomy. Both the continuing danger test and the internal cause test
(which is itself ultimately concerned with dangerousness) will routinely lead to the
detention and confinement of innocent persons who exhibit no marked impairment of
cognitive and deliberative powers and who, though they may suffer periodic episodes
of semi-conscious behaviour, are as generally capable of autonomous choice and ac-
tion as anyone without their infirmity. For these people, the disrespect for their auton-
omy implied in continued detention for paternalistic and prophylactic reasons in-
volves an unjustified violation of their rights, for nothing in their condition warrants it.
Accordingly, we need a new definition of disease of the mind to distinguish in-
sane from sane automatism. Like the present one, this definition should be a legal
rather than a purely medical one, because its work is normative rather than diagnostic.
Its purpose is to screen those innocents whose liberty the state may, without violation
of right, severely restrict for the good of others or for their own good as others see it.
Though novel, the definition should nonetheless be indigenous to the legal tradition,
akin to the criterion of incapacity recognized in other statutes governing the mentally
incompetent. The definition I propose, which adapts language from section 16 of the
Criminal Code and from Ontario’s Substitute Decisions Ace and Consent to Treat-
inent Act,-‘ is this: a disease of the mind is any impairment of the powers of cognition,
foresight, deliberation, and judgment that renders the individual generally (and not
merely episodically) incapable of appreciating the reasonably foreseeable conse-
quences of his actions or of understanding information relevant to executing his con-
ception of well-being. Anyone whose condition satisfies this definition of mental dis-
ease commands a diminished respect for autonomy, because such a person will ha-
bitually produce outcomes and face consequences he did not choose.
S. 0. 1992, c. 30, ss. 6,45.
“o Supra note 17, s. 6.
2000]
A. BRUDNER – INSANE AUTOMATISM
Several objections to this definition may be anticipated. First, if the criterion of
insane automatism were narrowed to mental impairment destructive of autonomy,
many more people would win absolute acquittals whose dissociative episodes are
likely to recur and who thus pose a continuing danger to others. This would not only
increase the risk of harmful behaviour but also undermine public confidence in the
administration of justice, which will be seen to favour the rights of the accused over
the security of the public. Second, it might be argued that the proposed definition,
with its requirement of a general and not merely episodic incapacity for autonomous
action, would counter-intuitively exclude those whose automatic episodes occur fre-
quently as well as those whose severe psychiatric disorders are normally controllable
by medication. These people would, it seems, go free if morally innocent of wrong-
doing. Third, it could be argued that the definition is incongruent with the language of
section 16, which nowhere speaks of an incapacity to execute a conception of well-
being. And finally, the definition is arguably too broad, for everyone has a nephew
whose plans always seem to misfire through lack of judgment and foresight.
To the first concern there are at least two responses, one that staunchly invokes
principle and one that meets the concern on its own ground of expediency. The prin-
cipled response is that any decrease in public safety that might result is in itself no ar-
gument against the narrower criterion, for a society of free and equal persons is enti-
tled only to as much public safety as is consistent with its moral premises. From a
public safety standpoint, there is as much reason to subject every convict whose
prison term has expired to a review for likelihood of recidivism as there is to hold in-
nocent people in custody under the present criterion of insane automatism. Yet we
would all recognize such a practice as inconsistent with a liberal legal order. Moreo-
ver, if public opinion ridicules the administration of justice for being true to its princi-
ples of freedom and equality, then, as Justice LaForest observed in Parks,3′ the opinion
may be discounted.
However, there is also a response that more directly meets the concerns for safety
and public confidence and that, realistically, a legislative body responsible to the pub-
lic is entitled to make. The risk of freeing the dangerous can be lowered without any
compromise of principle if those knowingly prone to unconscious episodes were
charged with criminal negligence if they failed to take reasonable precautions for oth-
ers’ safety, and if the evidentiary burden on the accused to raise a reasonable doubt
about the voluntariness of his actions were toughened’ The presumption that human
agents act voluntarily is, after all, a strong one and should not be capable of being
easily rebutted. Accordingly, trial judges could be instructed by legislation on the
kinds of considerations that weigh against putting automatism of any kind to the jury.
3, Supra note 4 at 908, (1992), 140 N.R. 19.
32In R v. Shaw, [1938] O.R 269, 3 D.L.R 140 (C.A.), the Court ordered a new trial because the
judge dismissed the charge of criminal negligence at the end of the case for tfie Crown, where the ac-
cused drove a car knowing he was subject to fainting spells and killed two passengers. The accused in
Quick, supra note 26 would seem to have been an obvious candidate for such a charge as well.
MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL
[Vol. 45
Bastarache J. listed a number of such considerations in Stone.” An expert witness
adds little to the accused’s bare assertion of involuntariness if he simply testifies to the
plausibility of the accused’s account of events and to the general nature and aetiology
of dissociative states rather than to the actual psychiatric condition and history of the
accused. If there is no witness to the event to corroborate the accused’s account, then
the absence of personalized expert testimony should be fatal to any plea of involun-
tariness. Also, the fact that the accused has an obvious motive for the assault or that
the victim is the very person whose insult triggered the alleged dissociative episode
casts serious doubt on a claim of involuntariness, since we would expect violent acts
committed while reason is asleep to be inexplicable in terms of clear and evident mo-
tives.
The second objection voices a concern that those who have frequent periods of
automatic behaviour or who generally control their severe symptoms with medication
will, under the proposed definition of mental disease, go free once acquitted. The re-
sponse is that they will go free only if their illnesses do not seriously interfere with
their capacity to carry out their life goals. Someone who experiences automatic epi-
sodes that occur (or are expected to occur) so frequently as to significantly impair his
overall capacity for autonomy will not go free under the proposed definition, nor will
someone who repeatedly neglects to take his medication. Of course, it is impossible to
determine with precision the frequency of unconscious behaviour at which someone
may be said to suffer from a general incapacity for self-directed action, but indetermi-
nacy of this kind is no stranger to the law, and juries are well enough equipped with
common sense to deal with it, or so we assume.
The third objection (that the proposed definition is foreign to the language of sec-
tion 16) we have already implicitly answered. Section 16 says nothing about an inca-
pacity to execute a conception of well-being, because its criterion of diminished re-
spect functions also as a criterion for innocence of the criminal charge, whereas the
proposed definition is not hampered by this requirement; whether the act is involun-
tary and whether the involuntariness manifests a disorder of the required severity are
separate questions. However, incapacity to appreciate the nature and quality of one’s
act or to know that it is wrong is the best available proxy for an incapacity for
autonomous action, given that the criterion of diminished respect is here tied to the
conditions of innocence. Finally, we need not be concerned that the proposed defini-
tion will sweep in every ne’er-do-well, since it requires an incapacity to execute a
conception of well-being rather than a mere failure to do so-one attributable to a se-
vere impairment of mental faculties.
IV. Exclusivity of Pleas and Burden of Proof
Adopting the proposed definition of mental disease (or one like it) would have
implications for two problems that have arisen ancillary to the one concerning the
” Supra note 8 at 428-30.
2000]
A. BRUDNER – INSANE AUTOMATISM
criterion of insane automatism: whether the defences of sane and insane automatism
should be conjunctive or mutually exclusive and where the burden of proof should lie.
As long as a policy of controlling the dangerous underlies the definition of insane
automatism, there will be an impetus to ensure that the category of insane automatism
is inescapable for those who qualify for it. Thus, in Bratty, the House of Lords ruled
that once the jury had rejected a plea of insane automatism, the accused could not
plead sane autoynatism on the same evidence. This ruling, however, carried implica-
tions embarrassing for the courts. Since the burden of proving insanity is on the party
raising the issue, while the burden of proving voluntariness (an element of the actus
reus) is on the Crown, the rule in Bratty meant that the jury, having rejected the ac-
cused’s insanity defence, would be forced to convict even if they had a reasonable
doubt about the voluntariness of the accused’s actions. This, however, would violate
the accused’s right to be presumed innocent and (since the judge has withheld from
the jury a defence logically open to him) his right to a trial by jury. To render the cate-
gory of insane automatism logically airtight, therefore, one further adjustment was
necessary. If the presumption of innocence were abolished in cases where involuntari-
ness is pleaded (that is, if the accused had to prove involuntariness on a balance of
probabilities), then disallowing recourse to sane automatism on the same evidence
upon which the jury had already concluded that the accused’s actions were probably
voluntary would no longer seem anomalous.
In Stone, the Supreme Court of Canada took this step. The accused had stabbed
his wife forty-seven times after he had been subjected to a stream of verbal abuse
from her. His defence was that the psychological shock he received induced a disso-
ciative state during which his physical movements were entirely independent of his
conscious will, and two psychiatrists testified to the plausibility of this explanation.
The trial judge left insane automatism and provocation with the jury, which rejected
the former but accepted the latter and convicted the accused of manslaughter. The ac-
cused appealed the conviction, arguing that his rights to be presumed innocent and to
a trial by jury had been violated by the judge’s refusal to leave sane automatism with
the jury.
Writing for a 5-4 majority, Justice Bastarache finessed the accused’s constitu-
tional challenge to the mutual exclusiveness of the defences of sane and insane
automatism by shifting the burden of proving involuntariness on a balance of prob-
abilities to the accused.’ This infringement of the accused’s right to be presumed in-
nocent is justified, he argued, because a burden of proof on the Crown would be too
onerous given that automatism is easily feigned and that “all knowledge of its occur-
rence rests with the accused.” According to Bastarache, J. there exists a presumption
that human agents act voluntarily, one that reflects a policy of relieving the Crown of
too heavy a burden of proof. This policy would be defeated, he argued, if the accused
‘ Neither the respondent Attorney-General of British Columbia nor any of the intervening Attor-
neys-General (Canada, Ontario, and Alberta) proposed a reversed onus.
” Stone, supra note 8 at 425.
MCGILL LAW JOURNAL / REVUE DE DROITDE MCGILL
[Vol. 45
need only introduce sufficient evidence to raise a reasonable doubt about the volun-
tariness of his actions.”
Bastarache J.’s reasons for violating the right to be presumed innocent take up
one short paragraph of his judgment. In this passage he overrides a constitutional right
to be presumed innocent for the sake of a consideration–ease of prosecution-that
the right is specifically meant to trump (proving guilt beyond a reasonable doubt is
always onerous), reflecting as it does a societal commitment to minimize the risk of
punishing the innocent even at the cost of incurring a high risk of freeing the guilty.
As Woolmington v. D.RP.3″ long ago decided, the presumption of voluntariness is not a
legal presumption reflecting a policy of solicitude for the Crown nor any other policy
for that matter; it is a common-sense presumption rooted in an everyday understand-
ing of the nature of human agents. Thus, it may be dislodged by evidence raising a
reasonable doubt as to whether the accused conforms to that general understanding.
Moreover, automatism is not more easily feigned nor is knowledge of its occurrence
more exclusively the property of the accused than other mental conditions such as
mistake of fact or drunkenness. Hence Bastarache J.’s reasoning carries an implicit
threat to the presumption of innocence whenever the accused claims that a mental
element of the offence is lacking-an implication so at odds with the values of a lib-
eral legal order as to render extravagantly purchased any resultant advantage to prose-
cutors.
The violation of the rights to be presumed innocent and to trial by jury is the end-
result of a law of automatism based on a policy of controlling the dangerous. If an
impaired capacity for autonomy rather than dangerousness were the criterion of in-
sane automatism, there would be no need for mutually exclusive pleas nor for a uni-
form burden of proof as between sane and insane automatism. No policy would be
subverted and no principle violated if a jury, having decided that the accused probably
did not evince a mental disorder warranting diminished respect, nevertheless acquitted
because they held a reasonable doubt as to whether the accused’s actions were volun-
tary. And because there would be no need for mutually exclusive pleas, there would
be no need to place the burden of proving involuntariness on the accused in order to
remove the anomaly of disparate burdens created by exclusivity.
Conclusion
In their concern for controlling dangerous persons, courts applying the common
law have produced a law of automatism that massively violates constitutional rights.
Not only does it deprive acquitted persons of their liberty without proper justification;
it also runs a high risk of convicting the innocent and violates the accused’s right to a
trial by jury. In Stone, the policy of controlling the dangerous has reached its zenith;
for that case all but abolishes the defence of sane automatism, while producing the
16Ibid.
” [1935] All E.R 1, [1935] A.C. 462 (H.L.).
2000]
A. BRUDNER – INSANE AUTOMATISM
spectacle of a Court legislating on its own initiative a reverse onus in automatism
cases and then, judging in its own cause, declaring its violation of rights to be justified
in a free and democratic society (even though the United States, The United King-
dom, Australia, and New Zealand apparently do well with the ordinary evidential
burden). The common-law jurisprudence on automatism is perhaps too solidly en-
trenched to allow hope for reform by the courts. However, in a constitutional democ-
racy characterized by co-operation between courts and legislatures, it may not be too
naive to hope that Parliament will in this case initiate the changes required by the
Charter.
Accordingly, legislation is needed to put the law of automatism on a principled
footing independent of a policy of controlling the dangerous. This is all the more pos-
sible now that public safety concerns are given full play where they belong-in the
legislative scheme for the post-trial disposition of mentally disordered individuals.
Since the acquittee’s continuing dangerousness is decided on all the evidence at the
post-trial stage, the law of insane automatism can focus exclusively on the question of
principle, namely, who among innocent persons is properly subject to detention and
confinement for public safety and paternalistic reasons alone? The answer, I suggest,
is those whose capacity for autonomy has been severely impaired by disease.
The law of automatism should therefore be reformed along the following lines.
Where the accused raises sufficient evidence to render a plea of automatism plausible,
the judge should put the defence of insane automatism to the jury only if he concludes
that the evidence for the unconscious episode would, if believed by a jury, warrant a
finding that the accused was probably suffering from a mental disorder that renders
him generally incapable of appreciating the reasonably foreseeable consequences of
his actions or of understanding information relevant to executing his conception of
well-being. It would be irrelevant whether such a disorder had a psychic or physical
cause. It would then be for the jury to determine whether they had a reasonable doubt
that the action was voluntary and, if so, whether the episode probably manifested a
mental disorder of the required severity. If they find the action voluntary, the accused
is guilty. If they have a reasonable doubt about voluntariness but there is insufficient
evidence to establish mental disorder on a balance of probabilities, the verdict is an
absolute acquittal. If they have a reasonable doubt about voluntariness and conclude
that the accused probably has a disease of the mind in the sense proposed, the verdict
is that the accused is not criminally responsible because of mental disorder. Under this
scheme, there would be no exclusivity in the pleas and no burden on the accused to
prove involuntariness. Whenever insane automatism is put to the jury, a verdict of
sane automatism would also be available, for the jury may decide that the accused be-
haved involuntarily but did not have a disorder of the severity needed to justify di-
minished respect.
