Automatism Confined
Patrick Healy*
This article begins by briefly tracing the development
of automatism in Canadian criminal jurisprudence. Most re-
cently, R. v. Stone is remarkable not only for the pronounce-
ments of the majority, but also for the differences between
them and the reasons of the minority. The majority restates
the substantive law of automatism and some of its evidentiary
aspects. According to the majority, evidence of involuntari-
ness that apparently arises from some form of automatism
raises a presumption of mental disorder against the accused.
To displace that presumption and benefit from a defence of
non-insane automatism, the accused must, first, satisfy an un-
usually weighty evidential burden before the trial judge and,
second, meet a legal burden by persuading the jury on the
balance of probabilities. Also unusually, the majority requires
that the trial judge instruct the jury as to factors to consider in
weighing the evidence. The initial presumption that automa-
tism results from mental disorder is far removed from medi-
cal understanding of the subject. The majority states that a de-
fence of non-insane automatism would be good only if the
average person would have reacted to the external events in
the same way. The chance of demonstrating this is, by defini-
tion, almost nil and so effectively eliminates the applicability
of the defence. The reverse legal burden on the accused cre-
ated in Stone violates the presumption of innocence as pro-
tected in the Canadian Charter of Rights and Freedoms. The
majority then justifies its own Charter violation under section
1. This shows a bold understanding of the Court’s role in ad-
dressing problems it perceives. The article suggests that it
seems that Canadian law cannot afford the full presumption
of innocence. The Supreme Court of Canada in Stone adjudi-
cated a constitutional question without notice and submis-
sions. In effect, the Court used the Charter to legislate ordi-
nary law. This article urges greater restraint by the Court in
use of the Constitution.
Cet article ddbute en esquissant le ddveloppement de
l’automatisme dans Iajurisprudence pdnale canadienne. R c.
Stone, une ddcision rdcente de ia Cour supreme du Canada,
est remarquable non settlement pour les propos de la majoritd,
mais aussi pour les diffdrences entre ceux-ci et l’opinion ml-
noritaire. La majoritd refait le droit substantif de Ia ddfense
d’automatisme, ainsi que quelques dldments de sa preuve.
Selon elle, une prdsomption de troubles mentaux face h
l’accusd est soulevde ds qu’il y a preuve d’un acte involon-
tae lid A une certaine forme d’antomatisme. Afir de renver-
ser cette prdsomption et de bdndficier de la defense
d’automatisme sans alidnation mentale, l’accusd dolt d’abord
satisfaire un fardeau de preuve particulirement lourd pour
ddmontrer an juge de premi&re instance qu’il existo une
preuve suffisante pour que la ddfense puisse &re soumise an
jury, pour ensuite persuader le jury selon la preponderance
des probabilitds qu’il a agi involontairement. De plus, Ia ma-
joritd exige que le juge de prem~re instance donne des in-
structions an jury quant aux dldments h prendre en compte
lors de t’dvaluation de la preuve. Cette prsomption initiale
slon laquelle l’automadsme rdsulte de troubles mentaux est
incompatible avec l’opinion du monde mddical et scienti-
fique. La majoritd soutient que Ia ddfense d’automatisme sans
alidnation mentale n’est applicable que si une personne nor-
male aurait ragi aux dvdnements extemes de la meme fagon.
Les chances de ddmontrer cela 6tant, par ddfinition, quasi
nulles, la possliblit6 d’invoquer cette ddfense sera illusoim
dans la majortd des cas. Le renversement du fardeau de la
preuve dans Ia ddcision Stone porte atteinte an droit de
l’accus6 A la prdsomption d’innocence garanti par la Charte
canadienne des droits et libertr. La majoritd justifie sa pro-
pre enfreinte en vertu de l’article 1 de la Charte. Cotte
ddmarche dnote une conception large par la Cour de son role
dans la rsolution de probl~mes dont elle prend conscience.
Selon l’auteur, il smble que le droit canadien ne puisse plus
foumir ]a pleine protection du droit d’etre prdsum6 innocent
La Cour supreme a tranch6 une question constitutionnelle
sans prdavis et sans qu’elle ait 6t6 soulevde par les parties in-
tdressdes et s’est servie de la Charte pour mettre en place une
rfgle de droit ordinaire. L’anteur pr~ne une plus grande re-
tenue dans l’emploi de la Constitution par la Cour.
. Faculty of Law & Institute of Comparative Law, McGill University; Counsel, Shadley Battista,
Montreal. My thanks to Guy Counoyer, Robert Leckey, Yves-Marie Morissette, and David Paciocco
for helpful comments. Above all, thanks to M.P.
McGill Law Journal 2000
Revue de droit de McGill 2000
To be cited as: (2000) 45 McGill LI. 87
Mode de rdfdrence : (2000) 45 RD. McGill 87
88
MCGILL LAW JOURNAL/REVUEDEDROITDE MCGLL
[Vol. 45
Introduction
I. Evidential Burden
II. Mental Disorder?
Internal Cause
A.
B. Recurrence
Ill. Legal Burden and the Constitution
Conclusion
2000]
R HEALY – AUTOMATISM CONFINED
Introduction
In R. v. Rabey,’ Dickson J. (later C.J.C.) noted that automatism had recently made
its way into criminal jurisprudence in Canada and elsewhere. The premise for its in-
troduction was that the common law should recognize some forms of a dissociated
mental state as distinct from insanity (later “mental disorder”) but sufficient for ac-
quittal. Complex questions of medical science and legal policy went with this, of
course, particularly with regard to “internal” and “external” causes of automatism.
But the premise itself was settled law, soundly based on the principle that criminal re-
sponsibility requires proof of voluntary conduct.
Following the opinion of Martin J.A. in the Ontario Court of Appeal,2 a majority
all but excluded the possibility that automatism that has an internal cause could be
classified as non-insane. They acknowledged that automatism with an external cause
would probably be characterized in this way, but they did not exclude the possibility
that such causes might induce insane automatism. In dissent, Dickson J. was of the
opinion that the characterization of automatism should not be determined by an un-
questioned distinction between internal and external causes, but rather by as precise
an explanation of the cause of automatistic involuntariness as the evidence will allow.
The distinction between internal and external causes might therefore be helpful if the
evidence is otherwise uncertain, but the distinction itself should not preclude the pos-
sibility that non-insane automatism might arise from evidence of a mental disturbance
that is not the result of mental disorder within the meaning of section 16 of the Crimi-
nal Code. That non-insane automatism could be caused by a mental condition that is
not a disease of the mind was and is, to many, a proposition that should be hedged
with suspicion.
With R. v. Parks, the significance of the distinction between internal and external
causes was modified, at least for the purpose of making a distinction between special
verdicts of mental disorder under section 16 of the Code and general verdicts based
upon the common law of non-insane automatism. The decision in Parks did not re-
solve all the complexities, but a majority of the Court recognized that, with appropri-
ate caution, the courts should be allowed to follow the evidence at trial to its natural
conclusion. If the evidence supported a reasonable doubt that the accused acted invol-
untarily in a dissociated or automatistic state caused by a disturbance other than
mental disorder, an acquittal should be returned. To do otherwise would serve no
valid penal purpose.
‘ [198012 S.C.R. 513, 15 C.R. (3d) 225 [hereinafter Rabey].
2R. v. Rabey (1977), 17 O.R. (2d) 1, 37 C.C.C. (2d) 461 (C.A.) [hereinafter Rabey (C.A.) cited to
O.R.].
3 R.S.C. 1985, c. C-46, s. 16 [hereinafter the Code].
4 [1992] 2 S.C.R. 871, 15 C.R. (4th) 289 [hereinafter Parks].
MCGILL LAW JOURNAL / REVUE DE DROITDE MCGILL
[Vol. 45
The result in Parks recognized non-insane automatism caused by a mental distur-
bance that does not amount to mental disorder within the meaning of section 16 of the
Code. Everything would turn on the evidence, notably expert psychiatric evidence,
and its import for legal purposes. This conclusion was qualified on grounds of legal
policy by the proposition that non-insane automatism should be excluded from con-
sideration where protection of the public militates against it. In Parks, then, the Court
ruled, first, that the classification of automatism as insane or non-insane should be
governed by the evidence and, second, that the classification of non-insane automa-
tism should be constrained by justifiable limitations based upon prudence. Accord-
ingly, even a case of non-insane automatism should not lead to acquittal if the under-
lying cause of the automatistic episode persists or if such an episode is otherwise
likely to recur.
The scope of the decision in Parks has been viewed in two ways, one narrower
than the other. One approach is that the position of the majority in Rabey remains,
subject to an exception for cases of somnambulism in which the evidence excludes
mental disorder.! On this view the distinction between internal and external causes re-
tains its vigour, with the result that an internal cause that is not somnambulism in this
sense is presumptively considered mental disorder. The other view is that the dissent
of Dickson J. in Rabey was endorsed by a majority of the Court in Parks. On this view
the distinction between internal and external causes is not determinative and, in prin-
ciple, there might be a range of internal causes for automatistic involuntariness that
could not properly be classified as mental disorder. Thus the form of somnambulism
considered in Parks should be considered as but one example of a form of mental
disturbance in that range of possible internal causes.
Since Rabey, much controversy has surrounded the question whether a severe
psychological blow could also be a cause of non-insane automatism. According to the
view expressed by Dickson J., the answer to this difficult question lies not in the dis-
tinction between internal and external causes, but in the nature and quality of the
emotional shock and in the nature and quality of its effect upon the mind of the ac-
cused actor. The majorities in both Rabey and Parks were obviously apprehensive
about this possibility, but Dickson J. had little fear that spurious or specious claims
would get past the stout common sense of a Canadian jury.
And now R. v. Stone,’ which is remarkable not only for some of the pronounce-
ments’ of the majority of five’ but also for the differences between them and the mi-
nority of four.’ For the dissenters, Stone was about the sufficiency of evidence ad-
‘This is the approach preferred by the Court of Appeal for British Columbia in R. v.Stone (1997), 6
C.R. (5th) 357, 113 C.C.C. (3d) 158 (B.C.C.A.).
6 [1999] 2 S.C.R. 290, 24 C.R. (5th) 1 [hereinafter Stone cited to S.C.R.].
Also considered in Stone were questions relating to disclosure and sentence. Neither of these
points will be considered in this text.
‘The majority consisted of L’Heureux-Dub6, Gonthier, Cory, McLachlin, and Bastarache JJ.
9The minority consisted of Lamer CJ.C. and Iacobucci, Major, and Binnie JJ.
2000]
P HEALY – AUTOMATISM CONFINED
duced at trial and, consequently, the adequacy of the trial judge’s direction to the jury
on issues of automatism. These four judges were of the opinion that non-insane
automatism should not have been withdrawn by the trial judge because he had ruled
that there was sufficient evidence to support an inference that the accused was “un-
conscious” throughout the stabbing. Moreover, psychiatrists called by the prosecution
and the defence agreed that in their medical judgment the accused was not suffering,
at the time of the act, from a disease of the mind. Thus, since there was- some evidence
of automatism, the minority was of opinion that the issue should have been resolved
by the jury. They noted that it might have been, in the jury’s view, a weak case for
non-insane automatism, and hence a case in which this defence might easily be dis-
missed in the deliberations, but according to the authorities it was nonetheless a suffi-
cient case for the jury’s consideration.
Bastarache J., giving reasons for the majority, thought there was more to this case
and undertook to restate the substantive law of automatism and some of its eviden-
tiary aspects. Several features of these reasons are complex; others are striking. For
convenience, the principal conclusions can be stated in the following propositions:
1. Where involuntariness is raised by the evidence, the judge must determine, as
a first question of law, whether a jury could find on a balance of probabilities that
such involuntariness was caused by automatism.”
2.
If yes, the judge must then decide a second question of law, whether such
probable automatism was caused by mental disorder. The trial judge should presume
that the accused suffered from a disease of the mind and then decide whether the evi-
dence distinguishes the instant case from mental disorder. For this purpose the judge
should consider two points. “Under the [first], the internal cause theory, the trial judge
must compare the accused’s automatistic reaction to the way one would expect a
normal person to react in order to determine whether the condition the accused claims
to have suffered from is a disease of the mind”‘ Thus a claim of automatistic invol-
untariness must be considered a claim of mental disorder in every case except one in
which evidence of an extremely shocking nature would establish that a normal person
would have reacted to it by entering into an automatistic state. Second, any claim of
‘” A note about terms is required. The majority said that “[i]nvoluntary action which does not stem
from a disease of the mind gives rise to a claim of non-insane automatism. If successful, a claim of
non-insane automatism entitles the accused to an acquittal” (Stone, supra note 6 at 368). This is too
broad, because there can be involuntary action that has nothing to do with automatism in the sense
that it does not occur in a “dissociated state” as that term is commonly used. The ultimate question is
whether the conduct of the accused was voluntary. This question might be answered by determining
whether his conduct was automatistic, that is, performed in a dissociative state. If there is sufficient
evidence that it was, it must also be determined whether the dissociative state was caused by a disease
of the mind or some other cause. Thus, involuntary conduct might include automatism, but some
forms of involuntariness, such as reflexive action, are better not described as automatistic. In many
such instances another label is more appropriate, such as accident.
“Ibid. at 390.
MCGILL LAW JOURNAL /REVUE DEDROIT DE MCGILL
[Vol. 45
automatism must be considered mental disorder if the underlying condition presents a
continuing danger.
3. All this having been done, the judge may leave sane or non-insane automa-
tism with the jury, but not both, and in either case the judge must instruct the jury that
a verdict of not guilty cannot be returned on the basis of automatism unless it is
proved on a balance of probabilities.
Three points arise immediately and obviously from these propositions. First, it is
highly improbable that Mr. Parks could have raised his defence if the trial judge had
been obliged to apply the test described by the majority in Stone. Second, this test is
more hostile to claims of non-insane automatism than was the position of the majority
in Rabey. Third, it is not clear that the result in Stone itself would have been the same
if the trial judge had followed the rules prescribed by the majority in the Supreme
Court.
In other words, if the varnish is removed from all this, evidence of involuntariness
that apparently arises from some form of automatism raises a presumption of mental
disorder against the accused.2 That presumption can be displaced if the accused can
first persuade the judge, as a matter of law, that there was no disease of the mind-
even though this decision makes the definition of disease of the mind more sweeping
and less coherent-and then only if the jury can be persuaded to believe the
contradicting evidence on a balance of probabilities. A claim of non-insane
automatism will succeed only where the automatistic involuntariness of the accused in
the act was the mental state that any normal person would experience, provided that
there is no danger of recurrence.
In this new approach to a difficult question of law, the troublesome gap between
concepts considered by both lawyers and doctors has now widened. A disease of the
mind has seemingly become the legal explanation for involuntary behaviour that can-
not be proved otherwise: “If he acted involuntarily, he must have been mentally disor-
dered in law, unless it is proved in law and in fact that he was not mentally disor-
dered:’ Stone widens the legal concept of mental disorder to vaporous indeterminacy,
narrows sane automatism to picayune indeterminacy, and creates a “reasonable” vio-
lation of the presumption of innocence by imposing a legal burden on claims of sane
automatism. The reasons of the majority are suffused with suspicion concerning
claims of non-insane automatism. These developments merit comment.
I. Evidential Burden
Discharge of the evidential burden on any issue, by the prosecution or the de-
fence, is a question of law that in common sense must be aligned with the standard of
” This point has been made and discussed in D.M. Paciocco, “Death by Stone-ing: The Demise of
the Defence of Simple Automatism” (1999) 26 C.R. (5th) 273 at 278-79.
2000]
R HEALY – AUTOMATISM CONFNED
proof on the general issue.’3 Thus the prosecution should fail on a motion for directed
verdict if the judge determines that the trier of fact could not find an essential element
of the prosecution case proved beyond reasonable doubt.”‘ Similarly, no matter of de-
fence should go to the trier of fact unless the evidence for it is sufficient to support an
inference of reasonable doubt.” In the normal run of cases, however, appellate courts
do not attempt to prescribe the evidentiary (as distinct from the substantive) ingredi-
ents for successful discharge of the burden.” In the exceptional case where a legal
burden is attached to a matter of defence, it follows that it should not be given to the
jury unless it could be proved on a balance of probabilities by the evidence, but the
sufficiency of such evidence has rarely been the subject of sustained judicial com-
ment.
On this point the majority states its conclusions:
To sum up, in order to satisfy the evidentiary or proper foundation burden in
cases involving claims of automatism, the defence must make an assertion of
involuntariness and call expert psychiatric or psychological evidence confirm-
ing that assertion. However, it is an error of law to conclude that this defence
burden has been satisfied simply because the defence has met these two re-
quirements. The burden will only be met where the trial judge concludes that
there is evidence upon which a properly instructed jury could find that the ac-
cused acted involuntarily on a balance of probabilities. In reaching this conclu-
sion, the trial judge will first examine the psychiatric or psychological evidence
and inquire into the foundation and nature of the expert opinion. The trial judge
will also examine all other available evidence, if any. Relevant factors are not a
closed category and may, by way of example, include: the severity of the trig-
gering stimulus, corroborating evidence of bystanders, corroborating medical
history of automatistic-like dissociative states, whether there is evidence of a
motive for the crime, and whether the alleged trigger of the automatism is also
the victim of the automatistic violence. I point out that no single factor is meant
to be determinative. Indeed, there may be cases in which the psychiatric or
psychological evidence goes beyond simply corroborating the accused’s ver-
sion of events, for example, where it establishes a documented history of auto-
matistic-like dissociative states. Furthermore, the ever advancing state of medi-
cal knowledge may lead to a finding that other types of evidence are also in-
dicative of involuntariness. I leave it to the discretion and experience of trial
judges to weigh all of the evidence available on a case by case basis and to de-
termine whether a properly instructed jury could find that the accused acted in-
voluntarily on a balance of probabilities.’1
This invites further observations about the nature of an evidential burden. To begin,
however, it will be noted that these comments go far beyond remarks commonly made
” J. Sopinka, S.N. Lederman & A.W. Bryant, The Law of Evidence in Canada, 2d ed. (Toronto:
Butterworths, 1999) at 61-63.
14IbidL
5 P v. Osolin, [1993] 4 S.C.R. 595,26 C.R. (4th) 1 [hereinafter Osolin].
,6 See, e.g., R. v. Haslam (1990), 78 C.R. (3rd) 23,56 C.C.C. (3d.) 491 (B.C.C.A.).
‘7 Stone, supra note 6 at 384-85.
MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL
[Vol. 45
about the evidential burden in cases of non-insane automatism. Canadian courts now
apply the wispy standard of the “air of reality”.’8 Nobody can really say what this
means, although the formal logic in this situation would seem to dictate that it is evi-
dence in which a properly-instructed jury, acting judicially, could find a reasonable
doubt. Trial judges will generally recognize an “air of reality” when their sense of
prudence tells them that to do so would avoid needless excursions to the courts of ap-
peal.
Lord Denning had said in Bratty v. Northern Ireland (A.G.)’9 that a claim of non-
insane automatism would require more support than a “facile mouthing” before it
would be put to the jury. This is a vivid phrase but commonplace law. It means that
assertions in law may be considered on the general verdict only if they are supported
by relevant and sufficient evidence of facts. The majority in Stone expresses deep
skepticism of non-insane automatism, and impliedly some skepticism of jurors, by re-
citing specific requirements concerning discharge of the evidential burden on non-
insane automatism. Whatever else these rules demand, they set a standard for the dis-
charge of an evidential burden that is not only complex, but manifestly more onerous
than any statement of the evidential burden found elsewhere in Canadian law-even
in the exceptional cases in which the accused bears a legal burden on a matter of de-
fence. The standard set by the majority is one that depends upon a judicial determina-
tion of sufficient weight in the evidence and, moreover, an assessment of weight that
must conform to prescribed guidelines.
It is not strictly necessary to insist upon an assertion of involuntariness by the
defence, just as it is not necessary to insist upon an assertion of no mens rea or fault.
In most cases, of course, affirmative defences will be mounted upon an assertion by
the defence, but the law remains that the judge must direct the jury to consider any
live issue raised upon sufficient evidence, whether it is raised by the prosecution or
the defence. Complex scientific issues typically require expert evidence, but in Stone
the majority insists that, as a rule of law, automatism cannot be entertained without it.
And then more:
Although cases involving claims of automatism do not deal with complex
chemical reactions or the like, they do require judges to assess confusing and
often contradictory psychiatric evidence. In particular, when determining
whether the evidentiary burden for automatism has been satisfied, trial judges
must be careful to recognize that the weight to be given to expert evidence may
vary from case to case. If the expert testimony establishes a documented his-
tory of automatistic-like dissociative states, it must be given more weight than
if the expert is simply confirming that the claim of automatism is plausible. In
the former case, the expert is actually providing a medical opinion about the
accused. In the latter case, however, the expert is simply providing an opinion
about the circumstances surrounding the allegation of automatism as they have
been told to him or her by the accused. Trial judges must keep in mind that an
28 Pappajohn v. R, [1980] 2 S.C.R. 120, 52 C.C.C. (2d) 481; Osolin, supra note 15; R. v. Park,
[1995] 2 S.C.R. 836,39 C.R. (4th) 287.
,” [1963] A.C. 386, [1961] 3 W.L.R. 965 (H.L.).
2000]
0 HEALY – AUTOMATISM CONFINED
expert opinion of this latter type is entirely dependent on the accuracy and
truthfulness of the account of events given to the expert by the accused. Indeed,
in the present case, Dr. Janke, the defence psychiatrist, qualified his opinion by
noting that it was based almost exclusively on the accuracy and truthfulness of
the appellant’s account of events …’
And then even more, as noted above, including what appears to be something in the
nature of a requirement for corroboration and a catalogue of other considerations that
share a running theme of suspicion that the majority candidly concedes.
All this amounts to a direction from the majority that trial judges should thor-
oughly weigh the evidence of automatism before even considering whether it provides
a sufficient evidentiary foundation for further consideration. This stands in stark con-
trast to the traditional view that judges should not embark upon extensive weighing of
the evidence when considering the evidential burden, for fear of trespassing upon the
province of the jury or even usurping its authority. Indeed, it is a view of the eviden-
tial burden that stands in naked contradiction to the opinion expressed by Bastarache
J. in R. v. Charemski.2′ This is perplexing because in Charemski the issue concerned
the evidential burden of the Crown; the argument against a weighing of the evidence
when considering the evidential burden on a matter of defence is surely even more
compelling.
In Stone, however, the majority enumerates multiple grounds upon which a trial
judge will commit “an error of law” if automatism is put to the trier of fact on the
mere basis that the evidence appears to support an inference of involuntariness due to
sane or insane automatism. First among them would be the absence of strong expert
evidence, followed by questions of corroboration and any number of other issues that
have entirely to do with weight and the competence of the jury. It has always been
common wisdom, especially among trial judges, not to insist too strenuously upon the
sufficiency of evidence for discharge of an evidential burden-especially on matters
of defence-and then to leave the matter, with suitable direction, to the trier of fact.
The approach of the majority in Stone reverses this and warns trial judges that the risk
of error in leaving non-insane automatism is great indeed.’
According to the majority, the trial judge must decide whether the evidence that
supports automatistic involuntariness is evidence of sane or insane automatism. It is
not apparent why the judge must decide to classify it as one or the other but cannot in-
stead leave both in an appropriate case. It is quite possible that sufficient evidence
could be adduced on both and that both would merit consideration, even according to
the stringent standard exacted by the majority. Forcing the trial judge to decide upon
one or the other, or neither, only underscores the degree to which the majority’s ap-
proach requires the judge to determine what has traditionally been reserved for the
jury. The history of the common law in these matters demonstrates a progressive (in
Stone, supra note 6 at 380-81.
21 [1998] 1 S.C.R. 679, 15 C.R. (5th) 1 [hereinafter Charemskl].
2′ It is ironic that the majority should insist on expert evidence, but then imply that even with it there
will rarely be a good defence. Consider the situation in Stone.
MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL
[Vol. 45
both senses) trend away from special verdicts and directed findings by trial judges.
The restrictive rigour of the approach declared in Stone is not radically different in
effect from that older and discredited view of the judge’s role. Moreover, as the dis-
charge of an evidential burden is a preliminary question of law, it would seem exces-
sive to demand that the trial judge decide whether the evidence proves on a balance of
probabilities that non-insane automatism is the form of automatistic involuntariness
before the court.
II. Mental Disorder?
A.
Internal Cause
The crux of this approach is the double fiction that automatistic involuntariness is
presumptively internal in its origin and that anything in the nature of an internal men-
tal cause of automatism is presumptively mental disorder. This is an undifferentiated
view of mental phenomena, one that is invoked, not justified, on the ground that it is
necessary for prudent and protective reasons of policy. It requires the trial judge to
determine the weight of the evidence and to do so in relation to a sweepingly wide
definition of mental disorder that might well not conform to the expert evidence. It
seems excessive that to erect an adequate screen against spurious claims of sane
automatism it should be necessary to adopt the fiction that all mental disturbances are
presumptively incidents of mental disorder.
In Canadian law the affirmative defences of mental disturbance have been osten-
sibly twofold: mental disorder and non-insane automatism. Throughout its reasons in
Stone, the majority repeats that it is imperative as a matter of policy to restrict the de-
fence of non-insane automatism. The decision that all forms of mental disturbance in-
yoked in relation to the voluntariness of the actus reus should be deemed mental dis-
order, unless proved otherwise, is only one element of this overarching rationale for
the opinion. “Disease of the mind” is, of course, a legal term of art that has little
meaning for doctors, but that can be adapted by lawyers to meet concerns that are
considered important by them. A presumption that involuntariness arises from men-
tal disorder unless proved otherwise would seem to serve concerns of policy, but only
by the creation of an amorphous residuum called “disease of the mind”. This is certain
to cause some further chaos in the examination and cross-examination of expert
medical witnesses at trial.
Martin l.A. wrote that disease of the mind, “although a legal concept, … contains a substantial
medical component as well as a legal or policy component” (Rabey (C.A.), supra note 2 at 12). The
deemed mental disorder in Stone reduces substantially the connections between the legal term of art
and medical science.
24 Lawyer: Doctor, I must tell you that in law a dissociative state is deemed to proceed from a dis-
ease of the mind unless proved otherwise, which means that a dissociative state is deemed to be a
form of mental disorder until some other explanation is established.
Doctor: Well, I must say, that is not how we would understand these matters….
2000]
R HEALY – AUTOMATISM CONFINED
The majority describes as the “objective componenf’ in a legal test for automa-
tism that aspect concerned with whether the apparent involuntariness in the conduct
of the accused comports with “the reaction that a normal person would have in the
circumstances.’ ‘ It is not immediately apparent what the significance of a positive or
a negative answer should be, but in either case this criterion can scarcely be described
as objective in any meaningful sense of the word, although it might be construed in
some way as a standard of reasonableness. The question it appears to raise is whether
it would be reasonable for a reasonable person to have an unreasonable reaction in the
circumstances. At all events, it is clear that the viability of non-insane automatism will
be nil unless the judge decides, as a matter of law, that the average sane person would
react to the events in issue by a dissociation of mind and body as expressed in invol-
untary physical behaviour. The effect of this will be to eliminate the defence of non-
insane automatism because it is a standard that cannot be met. Defences of mental
disorder or automatism are, by definition, highly specific to the mental make-up of
individual persons. To demand that the average sane person would react to the events
in issue in a specified way is to preclude, by law, the possibility that this adcused per-
son actually did react to shocking events by a dissociation of mind and body, even if
the average sane person might not have done so. Besides, a psychiatrist would have
great difficulty in answering the question of whether the average person, placed in the
same circumstances as the accused, would have had such a reaction.
The majority says that the law presumes the voluntariness of the actus reus.n This
is not strictly accurate because what the law does is only to recognize common sense:
the prosecution must always prove the voluntariness of the act, but in almost every
case that proof.is an easy inference from evidence of the alleged conduct. It is only
when the evidence pointedly raises a question about voluntariness that the issue be-
comes more subtle. The position taken by the majority on this point is that evidence of
involuntariness creates in law a presumption of mental disorder as its cause. Thus the
accused who raises an issue of involuntariness also puts mental disorder in issue.
Moreover, it is mental disorder that will be put to the jury unless the accused can
prove that its cause was some form of mental disturbance that is not mental disorder.
The chances that he can do this are negligible because the category of mental disorder
now comprises any mental state that a normal person would not have. It seems en-
tirely odd that the test for mental disorder should turn in such large measure upon the
magnitude of the precipitating circumstances and the ordinariness or reasonableness
of involuntariness as the reaction. One can only hope that this expression of the law in
Stone will be reconsidered at the earliest possible date.
Stone, supra note 6 at 392-93.
Ibid at 373. See the comment on this point by RJ. Delisle in “Stone: Judicial Activism Gone
Awry to Presume Guile’ (1999) 24 C.R. (5th) 91 at 93.
MCGILL LAW JOURNAL/REVUE DE DROITDE MCGILL
[Vol. 45
B. Recurrence
The continuing-danger theory can be expressed in two complementary ways. One
is that if the underlying cause of automatistic involuntariness or a dissociative state is
likely to subsist, or to recur, it is more likely to have some form of mental disorder
among its effects and thus, as an empirical proposition, should be legally classified as
mental disorder within the meaning of section 16 of the Code. Another is that where
there is a risk of recurrence, the precise cause is immaterial and reasons of prudence
justify either a narrower definition of automatism or treating the danger of recurrence
with the same protective measures as are applied in cases of mental disorder. In Stone
the majority appears to accept both.
The most difficult aspect of the continuing-danger theory is the accuracy and reli-
ability of prediction. It is perhaps for this reason that the majority in Stone adopts an
approach to this issue that considerably broadens the range of cases in which non-
insane automatism would be excluded by law from the jury’s consideration. The con-
tinuing danger, says the majority, cannot lie solely in the risk that the accused would
commit another act of similar violence in a dissociative state. Nor can it lie solely in
the danger that the precipitating cause of the dissociative state would recur and thus
probably precipitate another dissociative state. It must include more generally the
cluster of attendant circumstances within which the dissociative state arose previ-
ously; and if those attendant circumstances, or similar conditions, were capable of re-
curring, non-insane automatism should be excluded for reasons of policy. On this
ground, however, it must be allowed immediately that the exclusion of non-insane
automatism is a decision that really has nothing to do with mental disorder. The wider
approach adopted by the majority would exclude a defence of non-insane automatism
if the cluster of personal and social circumstances of the accused are likely to recur,
thus making the recurrence of dissociation somewhat more likely. At the very least,
this approach begs for some form of empirical justification, but the opinion of the
majority provides none.
Ill. Legal Burden and the Constitution
Viscount Sankey’s speech in Woolmington v. D.RP included the following pero-
ration, which seemed to make plain that at common law the accused could not be re-
quired to prove a sufficient condition of acquittal: “[W]hile the prosecution must
prove the guilt of the prisoner, there is no such burden laid on the prisoner to prove his
innocence and it is sufficient for him to raise a doubt as to his guilt; he is not bound to
satisfy the jury of his innocence” 2′ Woolmington was accepted as the law in Canada,
but it is no longer quite so. At common law, the Supreme Court of Canada decided in
. v. Sault Ste. MariH ‘ to recognize a class of strict-liability offences in which the ac-
cused could avoid conviction upon proof of due diligence or reasonable mistake. This
27 [1935] A.C. 462 at 481, [1935] All E.R. Rep. 1 (H.L.) [hereinafter Woolmington cited to A.C.].
[1978] 2 S.C.R. 1299,3 C.R. (3d) 30 [hereinafter Sault Ste. Marie].
2000]
R HEALY – AUTOMATISM CONFINED
was before the Canadian Charter of Rights and Freedoms9 came into force and, in
any event, it gave the accused a benefit that did not previously exist. Perhaps the en-
cumbrance of a legal burden at common law could be justified in those circumstances
on those grounds.
Since the Charter has come into force, almost all violations of the presumption of
innocence have been considered reasonable by the Supreme Court.’ In R. v. Da-
viault,” surprisingly, the Court even took the opportunity to encumber the new com-
mon law or constitutional defence of extreme intoxication with a legal burden. This
was neither sought nor argued by any of the parties, and the attorneys general could
present no submissions on the question because they were not notified that the Court
was considering a point of constitutional law. The majority of the Court concluded
that the burden thus imposed was a reasonable violation of the presumption of inno-
cence, although no evidence was presented and no argument heard. This aspect of the
decision attracted critical comment, in part because it took the case away from the
parties and, second, because it was intrinsically ill-advised for the final court of appeal
to make legislative decisions without advice. There was also something anomalous in
the notion that the Supreme Court could correct a violation of the presumption of in-
nocence by substituting for it another violation of the presumption of innocence.
The majority in Stone reverses the onus on the common law defence of non-
insane automatism. This is surprising, for all of the reasons just given, but more ar-
resting for the forceful manner in which it is done. The minority states flatly that there
was nothing in the record that could open the issue of the burden of proof on non-
insane automatism. In the courts below, there was not even a whisper on the issue, and
that would seem to end the matter. The majority says that the respondent Crown in-
vited the Court in its written submissions to consider the question, but at its highest,
all this means is that a majority of the Court might have thought that this was an issue
raised by the Crown’s written submissions. It is surprising that the majority might
have reached this conclusion because nowhere in those written submissions can one
find a suggestion that the Court might consider the imposition of a legal burden on the
issue of non-insane automatism. If the submissions of the respondent in this case were
sufficient to raise an important constitutional question, there is no case in which the
same assertion could not equally be made. This would make constitutional adjudica-
29 part I of the Constitution Act, 1982, being Schedule B to the Canada Act 1982 (U.K.), 1982, c. 11
[hereinafter Charter].
30 See e.g. R. v. Potvin, [1989] 1 S.C.R. 525,21 Q.A.C. 258; R. v. Downey, [1992] 2 S.C.R. 10, 13
C.R. (4th) 129 [hereinafer Downey cited to S.C.R.]; R. v. Whyte, [1988] 2 S.C.R. 3, 64 C.R. (3d) 123;
R. v. Keegstra, [1990] 3 S.C.R. 697, 61 C.C.C. (3d) 1; R. v. Wholesale Travel Group, [1991] 3 S.C.R.
154, 8 C.R. (4th) 145; R. v. Chaulk, [1990] 3 S.C.R. 1303,2 C.R. (4th) 1 [hereinafter Chaulk]; Osolin,
supra note 15. The exceptions to this pattern areR. v. Oakes, [1986] 1 S.C.R. 103, 50 C.R. (3d) 1, and
R. v. Laba, [1994] 3 S.C.R. 965, 34 C.R. (4th) 360 (legal burden found unconstitutional and replaced
by a read-in evidential burden).
3, [1994] 3 S.C.R. 63, 33 C.R. (4th) 165 [hereinafter Daviault].
MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL
[Vol. 45
tion unpredictable and it would eliminate the requirement of notice of a constitutional
issue.
The import of the reverse burden on non-insane automatism is a matter of consti-
tutional law. Frankly, in the administration of criminal justice it is unlikely to make
any appreciable difference, but the same cannot be said about the constitutional sig-
nificance of this aspect of Stone. The essential points are as follows.
Section 11 (d) of the Charter proclaims the right to be presumed innocent and it is
agreed that the imposition of a legal burden on a matter of defence is inconsistent with
the presumption of innocence.” It is also agreed that presumptions that relieve the
prosecution of the burden to prove its case beyond reasonable doubt are inconsistent
with the presumption of innocence. The rationale for these two agreed points is the
same: reverse burdens and mandatory presumptions would both permit a finding of
guilt in cases where there remains a possibility of reasonable doubt at the end of the
case and on the whole of the evidence.’ Nevertheless, section 1 of the Charter pro-
claims that the rights guaranteed by it are subject to such reasonable limitations as are
demonstrably justified in a free and democratic society. As noted previously ‘ all but
two challenges to statutory violations of the presumption of innocence have been re-
jected by the Supreme Court on the ground that they are reasonable limitations of the
right. There was a startling novelty in Daviault when the Court itself created a legal
burden on a matter of defence and, in the absence of evidence or argument, pro-
nounced that this deliberate violation of the presumption of innocence was reason-
able. A decision of this kind had only once been made previously at common law by
the Supreme Court, and not without anxious compunction,35 but to see it done under
the cover of the Constitution was indeed striking. It signified many things in the fast
evolution of Canadian constitutional adjudication, and among them was the evident
ease with which the Supreme Court has assumed legislative authority. If anyone
thought that the pronouncement in Daviault concerning the burden of proof was a
momentary misjudgment of the Court’s role, Stone goes so far as to suggest that it is
the Court’s duty to do nothing less than make such changes in the law.
It was once a common dictum in the Empire and later the Commonwealth that in
deciding constitutional cases appellate courts should take care to say and to determine
no more than is necessary for the disposition of the precise matters before them?’ This
prudential doctrine of restraint no longer constrains Canadian courts; it has been sur-
passed by more energetic approaches to doing the right thing, not infrequently in-
volving recourse to the Charter as an engine of right. The Supreme Court has said
3 Woolmington, supra note 27 at 481; Sopinka, Lederman & Bryant, supra note 13 at 121-22.
“3Downey, supra note 30 at 21.
4 Supra note 30.
“Sault Ste, Marie, supra note 28.
36 See e.g. Citizens Insurance Company v. Parsons (1881), 7 A.C. 96 at 109 (EC.). See also B.L.
Strayer, The Canadian Constitution and the Courts: The Function and Scope of Judicial Review, 3d
ed. (Toronto: Butterworths, 1988) at 199-207.
2000]
R HEALY – AUTOMATISM CONFINED
many times that it must be guided in its interpretation of the Charter by the Charter’s
general objects and by the specific purposes sought by its various provisions.” It has
also said that in reaching legislative decisions about the meaning of constitutional
rights it must give due regard to the context in which the issue for decision arises.’
The mandate for legislative reform derived from the Constitution is onerous and it
is for this reason that restraint should be cautioned afresh. Perhaps the first article of
restraint should be that the courts ought not to decide a constitutional question that
they are not asked, and if they perceive a need to answer such an unasked question
they should remit the matter for adequate preparation and submissions. A constitu-
tional answer to an unasked question might be incomplete. It is more likely, however,
that the net effect will be to trivialise the Constitution itself. Without doubt, the inter-
pretation of the Constitution forces courts to make difficult decisions, but the deci-
sions concerning the burden of proof in Daviault and Stone were unnecessary.
Before addressing these decisions in particular, it should be said that the scorecard
on reasonable violations of the presumption of innocence in Canada is long enough to
suggest that adherence to this principle might be a public value that Canadians cannot
afford at full price. If almost all statutory violations of the presumption of innocence
have been sustained as reasonable limitations, it would appear that the rigour of the
right is too rich for the smooth running of Canadian law. Indeed, it is not inapt to
paraphrase the general economy of the Charter by saying that it guarantees rights that
the courts consider reasonable in a free and democratic society. That is exactly the
conclusion that is intended by inclusion of a saving provision such as section 1, which
legitimates the notion that guaranteed rights can be reasonably violated.
On this point there has been some intriguing word-play in the Supreme Court. It
has been suggested that where the Court finds a reasonable limitation upon a consti-
tutional right there is no violation of the Constitution.” In the result this might be ac-
curate, because it recognizes that such limitations might be constitutionally lawful, but
it collapses the chain of reasoning that leads to the result. That is, there is no need to
consider whether a limitation is reasonable unless and until a violation of a constitu-
tional right has been established. It may be entirely sound that the violation is a rea-
sonable limitation, but that does not mean that the violation is retroactively made to
disappear.
In Daviault and Stone, the Court decided to create a violation of the presumption
of innocence and to assert its confidence in the reasonableness of the decisions. Taken
together these decisions demonstrate, paradoxically, that the presumption of inno-
cence under the Charter has less force than it had at common law and, further, that the
3′ Hunter v. Southam Inc., [1984] 2 S.C.R. 145, 41 C.R. (3d) 97; R. v. Big M Drug Mart, [1985] 1
S.C.R. 295, 18 C.C.C. (3d) 385.
3 See e.g. R. v. Mills, [1999] S.CJ. No. 68 at para. 64, online: QL (SCJ).
39 R. v. Swain, [1991] 1 S.C.R. 933 at 997, 63 C.C.C. (3d) 481, Lamer CJ.C. [hereinafter Swain
cited to S.C.R.].
MCGILL LAW JOURNAL / REVUE DE DROITDE MCGILL
[Vol. 45
Constitution itself is the source of this erosion. The Court’s decisions to reverse the
burdens in Daviault and Stone show the Constitution being used for legislative pur-
poses as if it were no different as a source of law than the common law. It is different,
of course, because the Constitution is and should be comparatively inflexible. It pro-
vides “a rule for making rules” and should not be the instrument for quotidian con-
cerns. That is, it should rarely be invoked by the courts ex proprio motu to amend
what the legislator has done or to do what the legislator has chosen not to do. It should
be invoked only where the ordinary law of the land fails to conform to the basic law.
To use the Charter as an engine for making ordinary law is to exploit the Constitution
and transfer ordinary law-making powers of the legislator away from their proper
source.
What was done in Daviault and Stone raises a point of considerable interest. In
both cases the Court knowingly and deliberately created a rule that is inconsistent
with the presumption of innocence, and with the Court’s interpretation of the pre-
sumption of innocence in section 11(d) of the Charter. The Court then pronounced
that this violation of the presumption of innocence was a reasonable limitation under
section 1. The point of interest is whether it is permissible for a court to create a rea-
sonable limitation upon constitutional guarantees by first creating a rule that violates
one of those guarantees. ‘ One answer is that there is no objection to this because the
Charter must be read as a whole, including, of course, section 1, and thus the courts
have the authority under the Constitution to develop the law in a manner that is, on the
whole, consistent with the Charter. This is problematic, however, because the ortho-
dox approach to questions under section 1 of the Charter is that they do not arise un-
less and until it is determined that a rule of law (statutory or common law) breaches a
right guaranteed by the Constitution. What was done in Daviault and Stone cannot be
accommodated within this orthodox approach. Indeed, what was done there amounts
to an assertion that the courts can and indeed must find a reasonable solution to a
problem, such as extreme intoxication or automatism, even if it requires the creation
of a rule that violates the Charter, as long as it does so in a reasonable manner.
‘ The Court in Swain, ibid, chose not to discuss its ability to apply s. I to a new rule that it is cre-
ating that it already realizes breaches a Charter right from the outset. Lamer CJ.C., for the majority,
contemplated the reformulation of the existing common law to conform with s. 7 of the Charter.
Failing a reformulation that did not infringe a constitutionally protected right or freedom, he would
turn to see if “the common law rule” (presumably the original rule, and not any reformulation) could
be upheld under s. 1 (at 978-79). Later, Lamer C-.C. said, “It is not enough to say that the newly
formulated common law rule is less intrusive than the previous rule … If this Court is to enunciate a
new common law rule to take the place of the old rule, it is obliged to consider the status of that new
rule in relation to all relevant aspects of the Charter. In my view, the only other provision of the
Charter which is directly applicable to the new common law rule is s. 15″ (at 989). He makes no ref-
erence to the possibility that a Charter infringement in the new rule could be upheld under s. 1, or that
s. 1 was a relevant aspect of the Charter. In contrast, Wilson J. did subject the proposed new rule to s.
1 scrutiny, against which she argued it failed (at 1035).
2000]
P HEALY – AUTOMATISM CONFINED
The Supreme Court has made clear that it will reformulate the common law to
make it consistent with the Charter, even where there is no government actor present
in the sense of section 32 of the Charter.’ Where government is a party to the litiga-
tion and the Charter is thus generally applicable to the litigation in question, the
Charter applies more strictly to any common law rules, in the same way it does to
statutes and regulations.”‘ It is difficult indeed to understand where a court could de-
rive the lawful authority to approve a reasonable violation of the Charter by first cre-
ating a rule that plainly breaches the Charter. If this is done by the Supreme Court of
Canada, there is certainly no forum in which the lawfulness of this practice can be
challenged, but nonetheless, it is submitted that the Charter affords the courts no basis
upon which to create a rule of law that deliberately breaches the Constitution, even if
it is justifiable as a reasonable limitation under section 1. In the absence of any other
forum in which to test this submission, perhaps when the rules in Daviault and Stone
next come before the Supreme Court it should be submitted to the Court that those
rules were created per incuriam and are of no authority.
The reasons given for the majority in Stone concerning the reverse burden on
non-insane automatism lack persuasive force as justifications for violating the pre-
sumption of innocence, not least because there was no evidence before the Court to
demonstrate a limitation of this right. The most curious of these reasons is the sug-
gestion that there should be consistency in respect of the legal burden on claims of
mental disorder (section 16 of the Code), extreme intoxication (Daviault) and non-
insane automatism. 3 The rationale for this, apparently, is that all three are claims of
involuntariness and, because such claims should be viewed with suspicion, they will
be appropriately handicapped by a reverse burden. That is, the suspicion is strong
enough to offset the presumption of innocence.
As already noted, this does not say much for the presumption of innocence, or
perhaps it is more accurate to say that it says less and less for the presumption of in-
nocence. What we have here is the extension of the reverse burden from insanity to
extreme intoxication and now to non-insane automatism. The first of these claims has
been recognized in Anglo-Canadian law for two hundred years, the second for five
years in Canada alone, and the third for almost fifty years in Canada and elsewhere. It
is appropriate to recall, however, that even in Woolmington, Viscount Sankey observed
that the reverse burden on insanity at common law was an anomaly and, further, it was
only because it had become so deeply entrenched in the common law of England that it
would not be disturbed.” This anomaly was incorporated and preserved in the Code and
a challenge to it was dismissed when the Supreme Court held in Chaulk that the reverse
burden on insanity is a reasonable limitation of the presumption of innocence.
4’ Dagenais v. Canadian Broadcasting Corp., [1994] 3 S.C.R_ 835, 34 C.R. (4th) 269; Hill v.
Church of Scientology, [1995] 2 S.C.R. 1130,184 N.R. 1.
42 Swain, supra note 39 at 968.
41 Stone, supra note 6 at 370-71.
“Supra note 27 at 475; Delisle, supra note 26 at 95 makes the same point
MCGILL LAW JOURNAL / REVUE DE DROITDE MCGILL
[Vol. 45
By itself, of course, this is not a justification for the extension of a reverse burden
to other forms of involuntariness, especially when those forms of involuntariness are
not strictly comparable. The defence of extreme intoxication after Daviault allows a
person to assert a constitutional right to acquittal for self-induced involuntariness and
irresponsibility. There is an argument that there should be no such defence at common
law, but there is no point to rehearsing this argument again because the law has been
settled in favour of such a defence. It will suffice to note that the defence of extreme
intoxication is categorically distinct from the other two forms of automatistic invol-
untariness in that the other two are concerned with involuntariness arising from causes
beyond the control of the accused.
Automatism is thus concerned with a dissociation of the mind arising from a
cause beyond the control of the accused. If it is caused by mental disorder within the
meaning of section 16 of the Code, the special verdict and the consequences of psy-
chiatric detention are self-evidently justified. If it is not, and the automatistic involun-
tariness of the accused’s conduct is blameless, there is no legitimate purpose served
by attributing criminal responsibility to such a person. The rules prescribed by the
majority in Stone not only obfuscate these important principles, but also raise the
strong possibility that the jury would be obliged to return a special verdict against a
person whose automatistic involuntariness was not caused by mental disorder.
A final point on the reverse burden: When discussing the burdens that should at-
tach to claims of automatism, the majority refers to a White Paper published but not
tabled in Parliament by the Minister of Justice: “In her 1993 Proposals to amend the
Criminal Code (general principles) the Minister of Justice recommended that the le-
gal burden of proof in all cases be on the party that raises the issue on a balance of
probabilities: “‘ It is with this reference that the majority then proceeds through a dis-
cussion of the burdens on extreme intoxication and insane automatism, leading to the
conclusion that there should indeed be a legal burden on non-insane automatism.
A word about this White Paper is in order.’ No Minister of Justice has committed
the Government of Canada to a policy that the accused should bear a legal burden on
the issue of non-insane automatism. The White Paper went nowhere and it was fol-
lowed by two subsequent initiatives concerning reform of the General Part of the
criminal law. Neither of these led to legislation, but it is more important to note that in
one of them there was no mention of a reverse burden on automatism 7 and in the
other the option for a reverse burden was one option among several.’
“Stone, supra note 6 at 375.
“Canada, Minister of Justice, Proposals to Amend the Criminal Code (General Principles) (28
June 1993). See the reference to this point by Delisle, supra note 26 at 93.
“Canada, Department of Justice, Reforming the General Part of the Criminal Code (1994) 15-16.
“Canada, Department of Justice, Toward a New General Part of the Criminal Code of Canada
(1994) 34-37.
2000]
Conclusion
R HEALY – AUTOMATISM CONFINED
The ratio decidendi of Stone is that the trial judge was right to leave insane
automatism with the jury and made no error in not leaving non-insane automatism.
More specifically, no substantial wrong or miscarriage of justice occurred in the case.
The majority dismissed the appeal, and thus the conviction for manslaughter stood.
That conviction would appear to be based upon the jury’s acceptance of a defence of
provocation.
It is clear that the trial judge and the Court of Appeal did not observe the rules on
automatism laid down by the majority in Stone, including the presumption of mental
disorder now imposed when evidence of automatistic involuntariness is adduced. If
those rules had been observed, it is far from clear that the result at trial or in the Court
of Appeal would have been the same. It is entirely possible, perhaps even likely, that
the jury would have returned a special verdict of acquittal by reason of mental disor-
der. Thus, if the majority’s new rules on automatism need not have been applied to the
case itself, and indeed they were not applied, this leaves a perplexing conclusion with
respect to those rules. They are all obiter dicta and they have only prospective persua-
sive force. This means that they are not authoritative at the moment and can be re-
argued before the Supreme Court at a future date.
