Assisted Suicide, Causality and the Supreme Court of Canada
Edward W. Keyserlingk”
The recent majority judgment of the Su-
preme Court of Canada in Rodriguez v. Can-
ada (A.G.) reflects what could be called a
“standard causality” position regarding the
causing of death in various circumstances in
medicine. The author compares that position to
what he labels the “empirical causality” view
and the “normative causality” view. He con-
cludes that the “standard causality” position
reflected in this decision is deficient in that it
denies but-for causality of death in contexts
other than assisted suicide and euthanasia by a
facile equation of but-for causality and crimi-
nal liability, and by its relative inattention to
the normative reasons other than causality
which can justify maintaining the illegality of
assisted suicide. As for the “empirical causal-
ity” view, it is deficient since it assumes that,
because but-for causality of death is present in
instances of withdrawal of treatment, assisted
suicide and euthanasia, they are all equally
morally and legally acceptable or unaccept-
able. Only a “normative causality” analysis
does not equate but-for causality and criminal
liability and focuses on the prior normative
and policy considerations which do or should
influence the determination of the existence of
normative causality, illegality and criminal lia-
bility in cases of assisted suicide.
La r~cente ddcision majoritaire de la Cour
supreme du Canada dans l’affaire Rodriguez c.
Canada (P.G.) traite de ce que l’on pourrait
qualifier de ocausalit standard> A l’6gard de
ce qui est la cause de la mort dans diff~rentes
circonstances en m~decine. L’auteur compare
cette position avec deux autres approches qu’il
intitule <(causalit6 empirique)> et (causalit6
normative . II conclut que la position de la
Cour quant A la ((causalit6 standard,
telle que
refldt6e dans cette d6cision, est d6ficiente en
ce que, premi~rement, elIe exclut la causalit6
ad6quate (but-for causality) pour d~terminer ]a
cause de la mort dans des contextes autres que
le suicide assist6 et l’euthanasie en traitant
comme 6quivalentes la causalit6 adequate et la
responsabilit6 criminelle, et deuxi~mement,
elle se prdoccupe peu des raisons normatives,
autres que la causalit6, qui pourraient justifier
le maintien du caract~re ill6gal du suicide
assist6. En ce qui conceme la
que, puisque la
pas la <(causalit6 addquateo A la responsabilit6
criminelle et s'int6resse aux considdrations
normatives et politiques qui influencent ou
devraient influencer la d6termination de la
sabilit6 criminelle dans le cas du suicide
assist6.
* LL.M., PhD, McGill Centre for Medicine, Ethics and Law, McGill University. The research for
this paper was supported in part by grant #806-92-0008, awarded by the Social Sciences and
Humanities Research Council of Canada.
McGill Law Journal 1994
Revue de droit de McGill
To be cited as: (1994) 39 McGill L.J. 708
Mode de r6fdrence: (1994) 39 R.D. McGill 708
1994]
CASE COMMENTS
Synopsis
Introduction
The “Standard Causality” Position
I.
II. The “Empirical Causality” Position
III. A “Normative Causality” Approach
IV. Normative Concerns about Assisted Suicide
V.
VI.
Conclusion
The “Standard Causality” Position in Rodriguez
Intention and Normative Causality
Introduction
Physicians, nurses and family members are often hesitant about terminat-
ing a patient’s life support measures, either when requested (now or in advance)
by the patient, or when supporting a permanently unconscious patient’s life is
judged by others to have become “futile”. Their reluctance can have a variety
of explanations, the most obvious being the enormity of the decision in view of
the patient’s resulting death. Concern may also arise with regard to the issue of
futility: Has life support really become futile in this case? What are or should
be the criteria for determining futility? Who should make such a determination?
A third concern which frequently arises is that stopping life support is
often seen as equivalent to causing the patient’s death. It looks and feels, at least
to some of those involved, especially the team members who actually stop the
respirator or remove the feeding tubes, as if one is killing the patient. After all,
it is sometimes said, if we had not stopped the life support, the patient would
have continued to live, or at least would not have died then. Whatever qualms
exist will typically be moral ones, though at times accompanied by vague or
explicit fears about criminal or civil liability.
Responses from ethics and law to questions about causality in the context
of life and death decisions may arguably be grouped into three major views or
strands. They are not in every respect mutually exclusive, and in some respects
that which distinguishes the views and their proponents is more a question of
emphasis and level of analysis than of diametrically opposed positions. With
that proviso, the views can legitimately be separated for purposes of analysis
and comparison. The first position will be labelled in this paper as the “standard
REVUE DE DROIT DE McGILL
[Vol. 39
causality” view or explanation. The second will be designated the “empirical
causality” view. The third will be referred to here as the “normative causality”
position.
Of interest in this paper are the causality stances adopted in the majority
judgment of the Supreme Court of Canada in Rodriguez v. Canada (A.G.).’ By
focusing almost exclusively on the matter of causality, this comment will not
directly address the core of the Rodriguez judgment, which dealt primarily with
constitutional issues. Nevertheless, in that decision the assumptions and stances
on the matter of criminal liability for causing death implicitly or explicitly con-
tributed to the conclusions about whether assisted suicide should be allowed for
Sue Rodriguez. Although the three positions to be examined are as much posi-
tions on ultimate criminal liability as they are positions on causality, it is the lat-
ter, and its role in criminal liability, which is of primary interest in this paper.
I. The “Standard Causality” Position
What will be described under the label of “standard causality” is not,
strictly speaking, a “position” on causation, and certainly not a coherent one. It
could more accurately be characterized as a series of somewhat incoherent ele-
ments and stances comprising the present state of the law on criminal liability
in this area. Among the elements in addition to but-for causality are those of
intent and distinctions between act/omission, natural/artificial and lawful/
unlawful. The “standard causality” view can best be described by applying it to
the five different circumstances which follow. A competent patient is entitled to
refuse or terminate life-supporting treatment, and to continue to provide it in the
face of such a (present or advance) refusal would constitute assault. To termi-
nate treatment is therefore not interpreted as causing the patient’s death, but
simply as allowing the patient to die by respecting his or her autonomy or right
to self-determination. To stop a respirator in this situation, for example, is
regarded as letting the patient die, not as killing him or her; as an omission, not
as an act; as simply passive euthanasia, not as active euthanasia.2
If it is established that a patient is permanently unconscious, then a family
member is entitled to request that futile life support measures be stopped. The
cause of death in such cases is said to be the patient’s disease or condition,
which medicine is powerless to alter, not the act of stopping the respirator. What
is involved here is commonly said to be more in the nature of an omission than
an act.3
A terminally ill patient may (possibly “must”) be provided with an appro-
priate form and amount of pain medication to control that patient’s pain, even
‘[1993] 3 S.C.R. 519, (sub nona. Rodriguez v. British Columbia (A.G.)) 107 D.L.R. (4th) 342
[hereinafter Rodriguez cited to S.C.R.]. Due to space limitations, this comment will not consider
the causality stances of the three minority judgments.
2Nancy B. v. Hbtel-Dieu de Quibec, [19921 R.J.Q. 361. 86 D.L.R. (4th) 385 (Sup. Ct.): Mallette
v. Shulman (1990), 72 O.R. (2d) 417. 67 D.L.R. (4th) 321 (C.A.); Cruzan v. Director; Missouri
Department of Health, 497 U.S. 261 (1990).
3Airedale N.H.S. Trust v. Bland. [1993] 1 All E.R. 821. [1993] 2 W.L.R. 316 fhereinafter Aire-
dale].
19941
CHRONIQUES DE JURISPRUDENCE
if death is thereby “indirectly” hastened. Because the physician has a duty to
alleviate pain, and because it is the alleviation of pain that he or she intends and
not to cause the patient’s death, then here too the cause of death is said to be
the patient’s disease or condition, not the pain medication.4
The “standard causality” explanation goes on to distinguish all three of the
above situations from those of assisted suicide and voluntary euthanasia.
Assisted suicide is prohibited by.paragraph 241(b) of the Criminal Code and is
considered by many to be immoral because to so assist is at least to participate
in causing another’s death. As for voluntary active euthanasia, the final act caus-
ing death is that of the doctor, nurse or family member. Death from assisted sui-
cide includes an “act” of assistance, not merely an omission. Death from active
euthanasia is by the “act” of another who is no longer simply “omitting”, for
example, respirator support. In both cases, death is not natural, that is, it is not
caused by the patient’s disease or condition.
II. The “Empirical Causality” Position
The second view, labelled in this paper as the “empirical causality” posi-
tion, takes a different tack. Though ethically, legally and logically persuasive for
many, the first or “standard causality” explanation is not uniformly reassuring
for health care practitioners at the deeper experiential level. Some continue to
feel that it is counter-intuitive and does not account for the widely shared belief
that their decisions and activities in all five circumstances are in some manner,
to at least some extent, causative of the deaths which follow.
Similarly, but more emphatically, a number of philosophers, jurists and
others maintain that the withdrawal of life support is as much a cause of death
as assisted suicide and active euthanasia. After all, they claim, stopping a life-
supporting respirator is the empirical cause, the cause-in-fact, of the death. Had
the life support system not been terminated, the patient would not have died, or
at least would not have died then. It is therefore suggested by those who espouse
this second view, that assisted suicide, and possibly voluntary euthanasia as
well, should be as ethically and legally allowable as are competent refusals of
life support, family decisions to stop life support on grounds of futility, and the
provision of appropriate pain control even if death is hastened. Those espousing
this view tend not to be impressed by counter-arguments against the decriminal-
ization of assisted suicide which highlight the potential dangers of abuse or the
added pressures on the vulnerable and those who believe that they are a burden
to their family or friends.5 Others come to the opposite conclusion for the same
reason. In effect, they agree that in all these circumstances death is being caused
by the act of another and not solely by the patient’s disease or condition. How-
ever, for that very reason, they maintain that patient and family autonomy con-
cerning life-sustaining technology and pain control measures should be
4Law Reform Commission of Canada, Euthanasia, Aiding Suicide and Cessation of Treatment
(Report No. 20) (Hull, Que.: Supply & Services Canada, 1983).
5J. Rachels, The End of Life: Euthanasia and Morality (New York: Oxford University Press,
1986); D. Brock, “Voluntary Active Euthanasia” (1992) 22:2 Hastings Center Rep. 10; Note,
“Physician-Assisted Suicide and the Right to Die with Assistance” (1992) 105 Harv. L. Rev. 2021.
McGILL LAW JOURNAL
[Vol. 39
restricted because otherwise they will lead inevitably to the decriminalization of
assisted suicide and voluntary euthanasia.
M. A “Normative Causality” Approach
The third position, the one advocated in this paper, is labelled here as a
“normative causality” view. A more comprehensive label would be “a norma-
tive approach to criminal liability”. It is suggested that both the first and second
positions are inadequate and much in need of further qualification. First of all,
the “standard causality” view is deficient in that it does not acknowledge that
stopping a life-supporting respirator is indeed the empirical, scientific or “but-
for” cause of death. In such cases, it is not sufficient to claim that the disease
or the patient’s condition is the cause of death. While this is particularly evident
when the patient is not terminally ill, even if terminally ill, the patient would not
otherwise have died at that time. In the final analysis, criminal liability for caus-
ing death cannot simply be based upon distinctions between acts and omissions.
After all, in medical practice it is by no means clear whether the relevant prac-
tice, for instance, of switching off a respirator or stopping medical feeding and
hydration, is an act or an omission. In this regard, the third position takes more
seriously the experience and hesitations of some health care professionals
referred to above, though it does not stop there.
Compared to the second or “empirical causality” view, however, the third
position emphasizes not simply scientific causality, the cause-in-fact of a death,
but normative or legal causality as well. This more nuanced and comprehensive
analysis holds that empirical, scientific or “but-for” causality cannot alone
account for what distinguishes assisted suicide from voluntary euthanasia, or
what distinguishes both from withdrawal of life support or appropriate pain con-
trol resulting in death. Nor can empirical causality alone serve as the basis for
deciding upon the criminality of assisted suicide or voluntary euthanasia. Other
elements of greater moral and legal significance are, especially, those of duty,
legality and estimates of the social consequences of decriminalization.
Act/omission distinctions are not solely determinative of criminal or tort
liability. When there is a duty to act, failure to do so could make the omission
actionable. However, “whether there is such a duty to act is a legal conclusion,
not a matter of policy-free factf’mding.”6 Causality in law is, in the final ana-
lysis, a normative or policy choice. To stop a life-supporting respirator at a com-
petent patient’s request, or because life support has become futile, is legal. How-
ever, if life support is neither refused nor futile, then a court could find that the
physician’s act (or omission) was the legal cause of the patient’s death. In both
cases, the physician’s act is the empirical cause-in-fact of the death. Concluding
in the first case that the patient’s condition was the cause of death is, from the
perspective of this normative causality analysis, only a shorthand affirmation, a
policy-based conclusion, that a court would not find the physician liable or
responsible for the consequences which occurred, that the act was not the legal
or proximate cause of death?
6Note, ibid. at 2029. This in substance is the position of Lamer C.
in Rodriguez, supra note
1 at 561.
7 Note, ibid. at 2030. See also, D.W. Brock, “Forgoing Life-Sustaining Food and Water: Is It Kill-
19941
CASE COMMENTS
As for the provision of appropriate pain control, from the perspective of this
third normative position a similar analysis applies. In effect, our society, crim-
inal law and courts have made an implicit policy decision that it is acceptable
in view of medicine’s mandate to relieve suffering, to provide pain control mea-
sures sufficient for its alleviation, even if they also hasten death. The normative
and legal criterion is not whether death is caused, but whether the dosage is
appropriate for that person’s pain.
IV. Normative Concerns about Assisted Suicide
What then can distinguish withdrawing life support treatment from assisted
suicide? What could justify imposing criminal liability for the latter but not for
the former?- From the perspective of this “normative causality” analysis, such
justification is not found by stating that the withdrawal of life support is an
and assisted suicide is a “provision” (of death assistance). There are
arguably three main policy justifications for maintaining criminal liability for
assisted suicide.
C”omission”
One possible policy justification is the traditional ground of protecting a
vulnerable minority, the psychologically unsound! Some who accept this as a
valid concern claim that it does not justify the prohibition of assisting the sui-
cides of rational persons.’ Others respond that rational suicide may be rare and
-may often be the result of failure to recognize treatable depression or to provide
effective pain management. Furthermore, it is suggested that the social sanc-
tioning of rational suicide and assisted suicide may result in an increase in irra-
tional suicide and irrational assisted suicide, and put increased pressure on the
elderly and those who feel that they are burdensome to kill themselves or
request suicide assistance. 0
A second policy reason advanced for maintaining the criminal liability of
assisted suicide is that assisted suicide could be used as a way of disguising
murder. That was the reasoning behind the formulation in the American Law
Institute’s Model Penal Code, which made causation of suicide criminal hom-
icide if it involved force, duration or deception.” This same concern influenced
the Law Reform Commission of Canada to propose the maintenance of criminal
liability for aiding suicide.’2 The number of reported cases in which persons
ing?” in J. Lynn, ed., By No Extraordinary Means: The Choice to Forgo Life Sustaining Food and
Water (Bloomington, Ind.: University Press, 1986) 117.
SAs has been noted, historically in Anglo-American common law there were two other grounds
for the assertion of state interest in preventing suicide and prohibiting suicide assistance. One was
religious belief, which served as the basis of the denial of funeral rites to suicides. Another was
sovereign cupidity, as evidenced by the forfeiture of a suicide’s goods, in the feudal period, to his
liege lord, and by the fourteenth century, to the coffers of the Crown. In present-day secular soci-
eties, only the third historical ground, namely the protection of the vulnerable, merits considera-
tion. See M.T. CeloCruz, “Aid-in-Dying: Should We Decriminalize Physician-Assisted Suicide
and Physician-Committed Euthanasia?” (1992) 18 Am. J. Law & Med. 369 at 373-76.
9 bid. at 397.
luSee Y. Kamisar, “Are Laws against Assisted Suicide Unconstitutional?” (1993) 23:3 Hastings
Center. Rep. 32 at 37.
“Model Penal Code 210.5(1) (Proposed Official Draft 1962).
12Supra note 4.
REVUE DE DROIT DE McGILL
[Vol. 39
who caused suicide were charged with murder, or could have been, makes this
concern a legitimate one.’3
A third policy argument made against decriminalizing assisted suicide is
that distinguishing assisted suicide from voluntary euthanasia can be so difficult
that legalizing the former could lead to condoning the latter. Whether or not one
subscribes to the view of Joseph Fletcher that voluntary euthanasia is simply a
form of suicide, 4 there clearly are circumstances in which the activities
involved make it hard to distinguish the two. Consider, for example, the physi-
cian who places the glass with a lethal dosage beside the patient, or hands it to
the patient unable to lift it, or pours it down the throat of a patient unable to hold
the glass. Who performs the final act, the one that results in death?
Clearly, the strength of this argument depends upon the ability to establish
the undesirability of a policy permitting voluntary euthanasia independent of
cause-in-fact similarities with assisted suicide, and even if the latter could be
justified in some respects. One plausible policy concern is that, in the case of
euthanasia, the final act leading to death is caused by the doctor (or whomever),
not the patient, whereas in assisted suicide causality is shared. There are, as
well, legitimate concerns about whether voluntary euthanasia conditions, guide-
lines, safeguards and procedures could ever be devised and enforced in a man-
ner that adequately protects vulnerable patients. Such concerns are similar to
those raised in regard to assisted suicide, but are arguably more serious in the
context of euthanasia.
V. The “Standard Causality” Position in Rodriguez
How does the majority judgment in Rodriguez deal with the matter of caus-
ing death in all five circumstances indicated above, particularly that of assisted
suicide, which is its primary focus? On which of the three views outlined above
does that judgment rely?
The majority judgment written by Mr. Justice Sopinka generally reflects
what this paper has labelled the “standard causality” position. He stated that the
assisted suicide that Sue Rodriguez was seeking was contrary to the historical
understanding of the sanctity of life, one of the values protected by section 7 of
the Canadian Charter of Rights and Freedoms.5 It is “choosing death instead
of allowing natural forces to run their course.’ 6 The implication is that, by con-
trast, when life-supporting treatment is refused or stopped, death is the result of
natural forces and is not caused by the physician who stops the respirator.
Mr. Justice Sopinka rejected the argument that paragraph 241(b) of the
Criminal Code is over-inclusive and contrary to the principles of fundamental
justice since the prohibition of assisted suicide extends to those unable to com-
13For numerous examples, see C.D. Shaffer, “Criminal Liability for Assisting Suicide” (1986)
86 Colum. L. Rev. 348 at 364-66.
“Morals and Medicine (Boston: Beacon Press, 1954).
15Part I of the Constitution Act, 1982, being Schedule B to the Canada Act 1982 (U.K.), 1982,
c. 11 [hereinafter the Charter]. See Rodriguez, supra note 1 at 585.
16Rodriguez, ibid. at 586.
1994]
CHRONIQUES DE JURISPRUDENCE
mit suicide on their own. 7 He stated that the prohibition is one aspect of an
appropriate balance between, on the one hand, the valid interest of the State in
preserving life and protecting the vulnerable, and on the other hand, the auton-
omy and dignity of the individual. That conclusion was based upon several
arguments. First of all, the blanket prohibition of assisted suicide has not been
revised to date by Parliament, and the decriminalization of attempted suicide did
not signify a societal condonation of suicide. Secondly, the blanket prohibition
of assisted suicide is the norm in Western democracies, none of which has yet
found it to be unconstitutional, and there is no consensus in favour of decrim-
inalizing assisted suicide. Thirdly, the serious concerns about appropriate safe-
guards, should exceptions be pernitted, justify the blanket prohibition. Fourthly,
on the basis of personal autonomy and dignity, there are some narrow excep-
tions permifted to the sanctity of life principle.
For our purposes we need only consider here the fourth point. Mr. Justice
Sopinka acknowledged” that courts and commentators have established the
right of patients to refuse treatment even if its withdrawal results in death, the
right of family members to have such treatment stopped if it has become “futile”
or “therapeutically useless”, and the legitimacy of appropriate palliative care
even if death is thereby hastened. Here again, the majority opinion appears to
favour the “standard causality” position in the arguments made and sources
selected to support the view that while the former are legitimate exceptions,
assisted suicide and euthanasia are not. At the same time, however, there is
some acknowledgement of the distinction between empirical and normative
causality.
Mr. Justice Sopinka cited with approval the recent House of Lords decision
in Airedale N.H.S. v. Bland, which allowed the withdrawal of medical feeding
from a patient in a persistent vegetative state with the consent of the patient’s
parents, but which rejected “the taking of active measures to cut short the life
of a terminally ill patient.”‘ 9 On the one hand, the rejection of active euthanasia
in that decision appears to be based on a presumed cause-in-fact distinction
between active and passive euthanasia, rather than on policy considerations sup-
porting the illegality of active euthanasia, regardless of whether the physician’s
act is an empirical cause of death in both active and passive euthanasia.
On the other hand, there is a suggestion that despite the professed Rubicon
between active and passive euthanasia on cause-in-fact grounds, it could never-
theless be crossed in some instances were it not for the inability to prevent
abuses. Mr. Justice Sopinka noted approvingly that the basis for the rejection of
active euthanasia in Airedale (and of assisted suicide by the Law Reform Com-
mission of Canada) “is twofold it seems –
first, the active participation by one
individual in the death of another is intrinsically morally and legally wrong, and
second, there is no certainty that abuses can be prevented by anything less than
a complete prohibition.”20
‘7lbid. at 584.
‘$lbid. at 598-601, 606-607.
‘9lbid at 598, citing Airedale, supra note 3.
20Ibid. at 601.
McGILL LAW JOURNAL
[Vol. 39
What this somewhat inconsistent view appears to mean is that, on the one
hand, a third party’s empirical causing of death from active euthanasia makes
it intrinsically wrong, but if a way could be found to prevent abuses, then
despite the empirical causing of death and the inherent immorality, that cause-
in-fact would no longer be considered the legal or proximate cause of death. A
more straightforward analysis, from the perspective of “normative causality”,
would not assign immorality or illegality on the basis of third party empirical
causing of death, but on the basis of other normative policy reasons such as the
danger of abuses.
There is yet another example of ambiguity in the majority opinion regard-
ing causality of death and the distinction between active and passive forms of
treatment. Justice Sopinka stated that there are two competing viewpoints.2’ One
view finds that in both withdrawal of life support and assisted suicide, death is
foreseen and does follow as a result of the action taken. This is essentially what
this paper has labelled the “empirical causality” position. The other view claims
that in withdrawal of life support, “the death is ‘natural’ –
the artificial forces
of medical technology are removed and nature takes its course,”‘ whereas in the
case of assisted suicide, “the course of nature is interrupted, and death results
directly from the human action taken.” This is in effect what this paper has
designated the “standard causality” position.24
Regardless of which view one adopts, Sopinka J. concluded that the phy-
sician must nonetheless respect the patient’s instructions and “is therefore not
required to make a choice which will result in the patient’s death as he would
be if he chose to assist a suicide or to perform active euthanasia.”‘
In effect,
however, this conclusion is consistent only with the “standard causality” posi-
tion. After all, even though requested by the patient, the physician’s withdrawal
of life support is the empirical cause-in-fact of death and remains a choice that
will “result in the patient’s death.’ ‘” If, on the other hand, Mr. Justice Sopinka
meant that the physician’s withdrawal of life support in that case is not the
“legal” cause of the patient’s death, since the normative policy in this case gives
precedence to patient autonomy, then his position would be consistent with what
has been designated in this paper as the “normative causality” position.
It may not be far-fetched to claim that the very fact that the majority judg-
ment placed as much emphasis as it did on the matter of societal consensus is
an implicit acceptance of the “normative causality” position. Although Mr. Jus-
tice Sopinka concluded that there is at present no societal consensus in favour
of decriminalizing assisted suicide, he implied that should there be evidence of
2t bid. at 606.
2Ibid.
23Ibid. [emphasis in original].
24Mr. Justice Sopinka cited (ibid. at 606) a study paper written by this writer for the Law Reform
Commission of Canada in support of the “standard causality” approach, namely, Sanctity of Life
or Quality of Life in the Context of Ethics, Medicine and Law (Ottawa: Supply & Services Canada,
1979). That was indeed the analysis this writer proposed at that time. However, as readers of this
comment will be aware, that is no longer the case.
2Rodriguez, ibid.
261bid.
1994]
CASE COMMENTS
such a consensus, the blanket prohibition against it could be modified.27 The
judgment appears to accept that in assisted suicide the physician’s act is a cause-
in-fact of the patient’s death (a choice resulting in the patient’s death). There-
fore, to envisage the possibility of its decriminalization, in at least some circum-
stances, means that in the event of decriminalization the physician’s assistance
would not constitute a culpable, normative or legal cause of death. By Sopinka
J.’s own reasoning, it could become an instance of a justifiable cause of death.
VI. Intention and Normative Causality
We come, finally, to the subject of intent. Mr. Justice Sopinka acknow-
ledged that, “[t]he administration of drugs for pain control in dosages which the
physician khows will hasten death constitutes active contribution to death by
any standard.”‘2 He went on to distinguish such palliative care from assisted sui-
cide on the basis of intent: “However, the distinction drawn here is one based
upon intention –
in the case of palliative care the intention is to ease pain,
which has the effect of hastening death, while in the case of assisted suicide, the
intention is undeniably to cause death.”29
This analysis is in one respect similar to the “normative causality” view.
It acknowledges in effect that a physician’s pain control treatment, if it hastens
death, is for that reason a cause-in-fact of the patient’s death, just as it would
be in assisted suicide. It is, of course, the case that, in the psychological sense,
a physician would typically intend something different in each case: pain con-
trol in the first, and death in the second.
However, from the perspective of causality and criminal liability, what
makes intent culpable in the legal sense is the intent to commit an unlawful act,
not merely the intent to cause death per se. The normative question is whether
or not causing death in a particular instance is legally justified or not, and
whether or not the physician is acting in a legally protected manner. The prior
normative policy choice made by our society, criminal law and courts is that
causing (hastening) death in the pain control circumstance is acceptable in view
of the medical mandate to alleviate suffering, and is therefore not an unlawful
act. Similarly, causing death by terminating life support at a competent patient’s
request is not unlawful in view of that patient’s right to refuse it.
27The reliance upon and determinations about societal consensus on the subject of issues as con-
tentious and evolving as assisted suicide is laudable in principle, but in practice has its dangers and
limits. A first problem concerns the weight that should be assigned to such consensus in formu-
lating and revising law and public policy, assuming consensus can be found. A second problem
is where one should look to find it. It is by no means evident that legislation, court decisions or
law reform commissions in this country or other western democracies reflect present-day societal
views. Current polls may seem to be more reliable sources. They appear to reflect considerable and
growing support for the decriminalization of assisted suicide, and suggest that legislation on this
subject in Western democracies is increasingly unreflective of public consensus. On the other hand,
polls may not themselves be reliable in view of ambiguities in both questions and responses. A ref-
erendum may appear to be an accurate indication of public stances, but it too can be prone to the
same ambiguities and reflects only the views of those who voted.
28Rodriguez, supra note 1 at 607.
291bid.
REVUE DE DROIT DE McGILL
[Vol. 39
What relieves the physician of culpable intent is not simply that there is no
intent to cause death, but more to the point, there is no intent to commit an ille-
gal act. Death is in fact being caused, and the physician knows it and may even
desire it, but assuming the dosage is appropriate for that patient’s pain control,
there is no intent to commit an unlawful act. Equally, what at present makes the
intent culpable in assisted suicide is not the intent to assist the taking of life per
se, but the intent to commit a presently unlawful act, to act in a legally prohi-
bited manner. What brands that act as lawful or unlawful at a point in time are
the same normative or policy-based considerations which lead to deciding
which of several causes of death will be considered “the” cause of death for
legal purposes. A society’s values and norms are not static and will always be
subject to challenges and evolution. On the one hand, defining what ought to be
criminal acts merely by reference to what is presently unlawful is inadequate
and circular. On the other hand, new normative and social considerations may
well justify the existing prohibition.
Conclusion
What then does the “normative causality” analysis add in comparison to
the essentially “standard causality” approach reflected in the majority judgment
of Mr. Justice Sopinka? In some respects, not a great deal. No claim can reason-
ably be made that the Supreme Court would have concluded otherwise had a
more overt “normative causality” approach been adopted. Nor would the rea-
sons advanced in support of that conclusion necessarily have been significantly
different. After all, normative policy factors, such as concerns about abuse,
played at least as significant a role in the decision to maintain the blanket pro-
hibition of assisted suicide as did assigning (shared) causality for death to the
assisting physician. As for the “normative causality” analysis, merely because
it finds empirical causation of death in withdrawal of life support, assisted sui-
cide and active euthanasia, does not necessarily mean that this approach is
inherently more inclined to promote the decriminalization of assisted suicide
and active euthanasia. Normative reasons such as fears of abuse, or an increased
risk to vulnerable patients, can be just as, or even more, persuasive from this
“normative causality” outlook.
Nevertheless, there are arguably at least two important correctives which
a “normative causality” analysis could have added to this decision. First of all,
by acknowledging empirical causality in all five circumstances considered, the
judgment would have been more reflective of the perceptions of many physi-
cians and family members who have to make the decisions about stopping life
support and administering appropriate but lethal pain control. Secondly, by
acknowledging that the common denominator in all five circumstances is the
physician’s act (or omission) causing death in an empirical, but-for sense, the
focus could have been more directly on how they are arguably distinct for other
normative reasons. They are not likely to be seriously weighed by law reform-
ers, legislators and couris until a more sophisticated and cumpiehienive
approach to the relationship between causality and criminal liability is adopted.
