After Washington v. Glucksberg and Vacco v.

Quill (Supreme Court cases that dealt

with physician-assisted suicide), U.S. health care policy regarding end-of-life
treatment decisions settled into a sort of orthodoxy.1 Patients are now allowed to
refuse treatment, including life-saving treatment, and adequate pain suppression is
emphasized even to the point of causing loss of competence and hastening death, so
long as death is not intended. On the other hand, euthanasia of conscious but
incompetent patients is prohibited. Within this framework, however, there is little
agreement on what values should guide an individual in making decisions about his
or her own endof-life care. When, if ever, should life-saving treatment be refused? Is
there a point at which analgesics should be refused? Kant’s writings on suicide are
instructive both for the guidance they provide individual decision making and for the
extent to which they challenge current health care orthodoxy. Indeed, Kantian
challenges are both far-reaching and radical. Notwithstanding Kant’s explicit
arguments against suicide, several philosophers have interpreted Kant’s writings to
permit suicide in some cases.2 Contrary to such views, I argue that Kant’s moral
philosophy prohibits suicide in medical contexts in which end-of-life treatment
decisions are made. Although Kant did not address the question of assisted suicide, it
follows from his arguments regarding suicide that assisted suicide is impermissible.
While those who decry assisted suicide will find solace in this aspect of Kant’s moral
philosophy, other aspects are disquieting. Some implications of Kant’s views are
more conservative than current policy. Not only would it be unjustifiable on Kantian
grounds to use assisted suicide to end pain and suffering, it would also be
unjustifiable to use analgesics that result in permanent loss of rationality. In addition,
refusals of life-saving treatment in order to shorten pain and suffering by ending
one’s life would also be impermissible on Kantian grounds. Other aspects of Kant’s
arguments are far more liberal than current policy. The main arguments Kant gives
against suicide do not apply to euthanasia once rational agency is permanently lost—
even if the patient is conscious. Whether this justifies adopting a policy that permits
euthanasia, however, depends on an assessment of the risks of such a policy for
those who are still rational agents and who might be mistakenly killed.
Kant’s central argument against suicide is that it violates the categorical imperative.
Kant believes that there is only one categorical imperative, although he gives various
formulations of it. Universal Law The first formulation of the categorical imperative
(the formula of universal law) admonishes us to “act only in accordance with that
maxim through which you can at the same time will that it become a universal law.”3
Kant illustrates this formulation with four examples, one of which concerns a
potential suicide who is motivated by self-love to commit suicide in order to avoid
future woes. Kant argues that the maxim that one should end one’s life out of self-
love when living longer “threatens more troubles than it promises agreeableness”
cannot be universalized because it would involve an inconsistency in willing.4 The
inconsistency results, according to Kant, because killing oneself defeats the purpose
of self-love, which is the furtherance of life.5 If taken as a general argument against
suicide, this is less than convincing. It is not clear, for instance, why we should accept
the teleological view that the purpose of self-love is furtherance of life or even the
view that selflove has some natural purpose.6 In addition, not all suicides are
motivated by self-love. As various philosophers have pointed out, suicides can be
motivated by altruistic considerations or worries about loss of rational agency.7 Hilde
Lindemann Nelson, for instance, argues that one can formulate maxims regarding
suicide that can be universalized. The particular maxim she suggests is “[O]ut of
altruism and in the face of imminent death I kill myself lest I inflict grave harm on
others as I die.”8 The sort of grave harm that Nelson has in mind is the harm that
results from the burdens placed on family members who must care for an
increasingly debilitated and terminally ill patient.
Nelson’s proposal does not, however, avoid all problems under the universal-law
formulation of the categorical imperative. A broader Kantian argument can be given
for opposing suicide by considering the nature of rational agency. Kant holds that
rational agents are committed to the unconditional value of rationality and to willing
its continued existence. As Thomas E. Hill, Jr., notes, “The practical effect of valuing
rational agency is implicit commitment to preserving it, developing it, making use of
it, and ‘honoring’ it symbolically.”10 Although Kant postulates the immortality of the
soul and the continuation of one’s rationality after death, the suicide nonetheless is
acting inconsistently in ending his or her use of rationality as a finite being.11 The
suicide is involved in an inconsistency of willing because the suicide wills to end
reasoning as a finite being, while it is also true that as a rational being he or she
necessarily wills the continuation of rationality as a finite being. It does not matter
whether the suicide’s justification is based on self-love, altruism, or some other
specific reason. 278 Martin Gunderson We can see how this general argument works
by considering more specific ways in which those who act on suicidal maxims are
involved in inconsistent willing. All rational agents, including potential suicides, have
an imperfect duty to seek their own perfection and the happiness of others.12
Failure to seek these ends, according to Kant, involves the agent in an inconsistency
of willing.13 Kant calls these “wide duties,” since they have “a latitude for doing more
or less, and no specific limits can be assigned to what should be done.”14
Nonetheless, rational agents who are committed to willing these ends are also
committed to willing the necessary means to carrying them out.15 The continued life
of the rational agent is, of course, one such necessary means. Hence, the suicide who
is committed as a rational agent to willing his or her own perfection and the
happiness of others is involved in an inconsistency of willing, because he or she fails
to will the necessary means to those ends.
Humanity as an End Whatever the merits of Kantian arguments under the universal-
law formulation of the categorical imperative, the potential suicide also needs to
consider the categorical imperative stated as the formula of humanity. This
formulation states that one must act so as to treat humanity in oneself and others as
an end and never merely as a means.16 Humans have dignity in virtue of their
humanity. As John Rawls notes, “Kant means by humanity those of our powers and
capacities that characterize us as reasonable and rational persons who belong to the
natural world.”17 This includes the sort of reasoning ability that is required for moral
agency. It is not the life of the individual human that has dignity. Rather Kant states
that “morality, and humanity insofar as it is capable of morality, is that which alone
has dignity.”18 We respect persons by respecting their humanity and hence the
rational agency that constitutes humanity. For Kant, this dignity has worth beyond
price.19 Kant also says that dignity has a worth that is “infinitely above all price, with
which it cannot be brought into comparison or competition at all without, as it were,
violating its holiness.”20 This is best interpreted as meaning that dignity is
incommensurable with other values and not to be measured against or traded for
other values. Dignity is not to be dealt with in terms of costbenefit analysis. As
Barbara Herman notes, Kantian deliberation does not involve weighing amounts of
value, because the value of rational willing is not scalar.21 According to Kant, suicides
who take their lives to avoid suffering treat themselves merely as a means to
achieving a life that has a tolerable balance of happiness over suffering.22 The
suicide destroys his or her rational agency in order to attain a tolerable state of affairs
and thereby trades what is priceless for what has a price. This is a violation of the
formula of humanity. Nelson suggests that altruistic suicide is compatible with the
formula of humanity, because the suicide is not motivated by self-serving ends and
hence is not using himself or herself merely as a means to an end.23 As previously
noted, Kant claims that there is a wide duty to adopt the goal of seeking the Kantian
View of Suicide and End-of-Life Treatment 279 happiness of others. One of the ways
we respect humanity is by adopting the ends of others as our own, and this is an
imperfect duty. In this respect, Nelson is correct that we need to consider the
burdens we impose on others. This does not, however, justify suicide. The imperfect
duty to help others achieve their ends and to adopt the happiness of others as a goal
does not override the perfect duty not to destroy the conditions of one’s continued
moral agency.
Hill provides a different Kantian argument for suicide under some conditions when
he argues that committing suicide is permissible for a person who is faced with a loss
of rational agency through dementia or pain that overwhelms the ability to engage in
rational thought.24 Although he is explicating a modified version of Kant’s moral
theory, Hill’s view is instructive for considering Kant’s own views. The problem with
applying this view to Kant’s actual position is that the suicide who reflects on his or
her growing dementia is still in possession of enough rationality to act as a moral
agent even though the end of rationality and hence moral agency is near. Thus, the
suicide is destroying what is left of his or her rational agency in order to avoid
reaching a point at which rationality disappears. The suicide is sacrificing something
that has worth beyond price in order to prevent future loss of dignity. But the
prevention of this does not have worth beyond price. It is important to distinguish
between violating one’s dignity and losing one’s dignity through no fault of one’s
own. The progressively demented person is not violating the moral law, but simply
losing over time the ability to think in terms of reasons. By contrast the suicide who
acts on the maxim of preventing future loss of dignity violates his or her present
dignity and thereby treats humanity in himself or herself merely as a means. Hill’s
suggested maxim also faces problems under the universal-law formulation of the
categorical imperative. If a suicide is motivated by the preservation of dignity at least
in the sense of avoiding life without dignity, he or she is sacrificing dignity in order to
maintain dignity and thereby involved in an inconsistency in willing. In addition,
reason necessarily wills its own continuation, but the suicide who is motivated by the
future loss of rationality defeats this, even though the defeat is only temporary, since
rationality will be lost in the near future in any case. It might seem that this is not a
powerful objection to Hill’s analysis, since those who are facing dementia may not
have much time left in which they can function as rational agents or accomplish
significant projects. What matters, for Kant, however, is whether a person’s rational
agency is destroyed, not how long that person has to live. The view that rational
agency may not result in significant projects is not decisive for Kant, because the
worth of rational agency does not dependent on its utility.25 It should also be noted
that even those who are near death can do such morally momentous acts as
forgiving alienated family members, expressing love, and reaffirming or rejecting
religious commitments. There are also practical issues to consider. It is difficult to
calculate how much time is left before dementia permanently destroys rational
agency. People with Alzheimer’s, for instance, do not simply become severely
demented at a particular time. They drift in and out of dementia for some time with
good days and bad days. It is extremely difficult for a currently competent person
who has been diagnosed with Alzheimer’s to know just how much time of mental
clarity is left.
Refusing Treatment Kant is no vitalist who defends life at all costs. Respecting
humanity requires not only preserving the conditions of humanity, but abiding by the
moral law, which consists of perfect duties such as not making false promises and
imperfect duties such as seeking the happiness of others. For Kant, there is a
difference between the value of life and the value of humanity (rationality). The
Lectures on Ethics record Kant as saying that although suicide is forbidden under any
condition, it is better to sacrifice one’s life to avoid a disreputable act.26 In particular,
Kant states, according to the Lectures on Ethics, that one ought not to extend one’s
life if it means loss of dignity.27 If, for example, one is threatened with death if one
does not commit an immoral act, one should still refuse to act immorally. In the
Critique of Practical Reason, when discussing pure morality and pure virtue, Kant
gives the example of “an honest man whom someone wants to induce to join the
calumniators of an innocent but otherwise powerless person (say, Anne Boleyn,
accused by Henry VIII of England).”28 Kant makes it clear that morality requires that
the honest man not participate in the evil conspiracy even if it costs him his life.29
The honest man in Kant’s example is not committing suicide, however. Rather he is
refusing to act immorally even though it means he will be unjustly killed. Kant also
accepts risks to life when they are taken in accord with dignity. The Lectures record
Kant as saying, [W]hen a soldier pits himself against the foeman’s steel, he is merely
risking his life; the seaman likewise, the fisherman and other people with dangerous
occupations risk their lives, and if they did not, one might indeed take it that they
were expending less effort than they might, to sustain themselves, or to maintain a
limited use of their freedom.30 In short, one can will the preservation of one’s
rational agency and hence one’s life while still willing to risk one’s life, as long as the
risk is reasonable. This is, of course, true in medical contexts as well. The ill person
who creates burdens for others can refuse the help of others even though it
increases the risk of death. For example, a person could justifiably insist on being
cared for in a nursing home at public expense in order not to impose severe burdens
on family members, even though the care there is less adequate and there is an
increased risk of dying sooner. In these contexts there is no formula for determining
precisely when a risk is acceptable. One respects humanity, and this requires taking
account of the degree of risk to one’s life and the sorts of ends that one will be
supporting or thwarting in risking or not risking one’s life. For similar reasons, a
person might refuse to consume an unjust share of medical resources even though it
puts his or her life at grave risk. Justifiable refusals of treatment need to be
distinguished from two other sorts of cases. In the first case, a person refuses
treatment in order to avoid suffering by shortening his or her life. Whether one
decides to end one’s life by killing oneself or by refusing life-saving treatment, one is
still bringing about the end of one’s life and hence the end of one’s rational agency in
order to avoid suffering. For Kant, as was previously noted, one who wills an end also
necessarily wills the necessary means to that end. The person who refuses life-saving
treatment in order to end suffering by hastening death wills death, even if it is not
directly intended. The person thereby wills an end to rationality (which is beyond
price) in order to avoid suffering (something that is not beyond price). In the second
sort of case a person decides to refuse life-saving treatment in order to avoid further
suffering resulting from the treatment but does not want to die, even though he or
she knows that death will follow. Death is not a means to avoid suffering but is a
known consequence of refusing live-saving treatment. As noted previously, Kant
holds that we have a duty to seek our own perfection. Hence actions taken to avoid
suffering must be compatible with the underlying duty of seeking one’s own
perfection. When death is the foreseeable result of refusing life-saving treatment,
refusing the treatment is not compatible with the duty to seek one’s own perfection

Pain Control
Kant is quite clear that pain does not constitute a reason for suicide. According to the
Lectures on Ethics, Kant states, So neither the greatest advantages, nor the highest
degree of well-being, nor the most excruciating pains and even irremediable bodily
sufferings can give a man the authority to take his own life, to escape from anguish
and enter earlier upon a hoped-for higher happiness.32 Since the reason suicide is
not permitted is because it destroys rational agency, the permanent use of analgesics
that destroy rationality would also be prohibited, on Kant’s view.33 This has
relevance for current medical practice. At the present time, pain medications such as
morphine may be used to reduce pain and suffering even to the point at which the
person can no longer function as a moral agent. On the Kantian view described here,
pain control should not be administered so as to permanently destroy rational
agency. When the point is reached that controlling pain requires so much morphine
that increased doses will make the patient unable to function rationally, increases
should cease even though the pain becomes greater. Of course, it may happen that
the pain itself becomes so great that it makes rational agency impossible, and at that
point the pain can be suppressed without worry about rational agency. This provides
a rough principle for determining when pain is sufficiently extreme to justify
administration of pain medication that suppresses rational thought. The point is
reached when the pain is so extreme that it will itself suppress the rationality
required for moral agency. The use of pain medication or anesthetics that render one
temporarily incapacitated as a moral agent is, however, permissible when there is
adequate reason for it. Kant states, for instance, that the use of opium as a medicine
is permissible.34 Use of anesthetics that cause temporary loss of consciousness
would also be permissible on Kantian grounds. Such medications are used to
effectuate cures or surgery so as to restore one to better functioning as a moral
agent. Similarly, sleep is necessary to maintain one’s moral agency, even though one
is unable to function as a moral agent while asleep. The problem arises in the case of
terminally ill patients for whom pain control may mean a permanent loss of rational
agency. Terminal sedation is one way this can happen. The most extreme form of this
would be sedating a person into a coma and then allowing the patient to die of
dehydration, although a person can be sedated into semiconsciousness or even
unconsciousness while the body is kept functioning through nutrition and hydration.
Euthanasia Although Kant’s arguments against suicide cannot be sidestepped within
the framework of Kant’s central tenets, Kant’s arguments against suicide do not rule
out euthanasia for people who have permanently lost their rational agency.35 Once
someone has irretrievably lost rational agency, that person no longer has the sort of
Kantian dignity that would justify a duty not to perform euthanasia. When a person is
capable of functioning as a rational agent, however, he or she should not be
euthanized, even if he or she requests euthanasia in order to relieve suffering, since
killing him or her would treat humanity in that person merely as a means to an end.
It would be permissible, on this argument, to write an advance directive refusing life-
saving treatment, and even requesting euthanasia, in the event that one were to
suffer a permanent loss of rational agency. This would not violate Kant’s statements
applying the categorical imperative to suicide. The maxim of one’s actions could be
universalized. It is not inconsistent to will that it become a universal law that people
sign advance directives enabling their death after they have permanently lost their
rational agency. The motive here is not any sort of self-love that would be
inconsistent with the purpose of self-love generally, as in Kant’s example. Rather, the
motive is respect for the integrity of one’s dignity as a rational agent. Nor would one
be treating oneself merely as a means to an end, for at the point at which one died,
one would lack rationality and hence the dignity that makes persons ends in
themselves. Finally, an advance directive permitting euthanasia would not run afoul
of the Kantian duty to seek one’s perfection, since the person who has permanently
lost the ability to function rationally is, of course, unable to seek perfection.
Interestingly, the Kantian argument outlined here permits euthanasia even for
persons who have not made an advance directive requesting it so long as there is no
reason to believe that euthanasia would thwart what the person willed as a rational
agent. If a person expresses a desire not to be killed in the event that he or she
suffers a permanent loss of rationality, then there is reason to respect that desire,
just as there is reason to respect the directives of the deceased regarding property
distribution. Respect for rational agency requires both sustaining the conditions of
rational agency and respecting decisions made by rational agents. A supposed
kingdom of ends in which Kantian View of Suicide and End-of-Life Treatment 283
rational decisions were not respected after the loss of competence or death would
be less respectful of humanity and hence not a genuine kingdom of ends. In short,
Kant’s arguments against suicide do not present us with an a priori case against
euthanasia for those who have permanently lost their rational agency. We do,
however, need to take account of empirical considerations. Even if a maxim that we
will euthanize persons who have permanently lost rational agency and have not
previously expressed a wish to the contrary passes muster under the various
formulations of the categorical imperative, it might still fail of application in a
particular case. It might, for example, be simply too difficult to tell whether an
individual has in fact permanently lost rational agency. Obviously a maxim to perform
euthanasia in cases in which we are not certain whether rational agency has been
permanently lost would be incompatible with respect for rational agency. Empirical
considerations also provide room for state regulation.

State Regulation
Kant is a traditional political liberal. He holds that people have a right to be free to
pursue their own vision of happiness as long as they can do so in accord with
universal law.36 The bedrock of Kant’s political philosophy is the universal principle
of right, which states, “Any action is right if it can coexist with everyone’s freedom in
accordance with a universal law, or if on its maxim the freedom of choice of each can
coexist with everyone’s freedom in accordance with a universal law.”37 Kant rejects
state paternalism in which the state attempts to coerce citizens into adopting a
particular vision of happiness and goes so far as to claim that “a paternalistic
government... is the greatest despotism thinkable.”38 Kant also holds that the mere
fact that an action will harm someone is not a justification for coercion if the action is
consented to by the person harmed.39 In addition, the state cannot force people to
be moral, because it cannot force people to act on the basis of certain maxims—the
state can merely coerce external behavior. Restraint on freedom is justifiable,
however, in order to protect freedom. Coercion is justifiable only if it is able to
“coexist with the freedom of everyone in accordance with universal laws.”40 This
ensures that those whose freedom is restricted by state regulation are coerced on
the basis of principles that they themselves have reason to accept. This provides the
framework for regulating euthanasia. In those cases in which we can tell that rational
agency has indeed been permanently lost, prohibition of euthanasia is not justified
on Kantian grounds. People who write advance directives permitting such euthanasia
can be viewed as acting on their own vision of happiness without violating the
universal principle of right. In fact, it may be important to a competent person’s
present view of happiness to know that he or she will not pose problems for family
and loved ones after he or she ceases to be a moral agent. It may also be important
to the person’s vision of his or her own dignity not to be kept alive after the
possibility of rational agency has ceased. Moreover, such actions can coexist with
everyone’s freedom in accordance with universal law. Such people are pro284 Martin
Gunderson tected by Kant’s liberal political theory. For the state to enforce its own
vision of happiness on such people would be just the sort of paternalism Kant
decries. Even if carrying out the advance directive would harm the no-longerrational
human (e.g., by depriving the person of enjoyable mental states), Kant’s political
theory would not justify prohibition, because the action had been consented to
when the advance directive was written. In the case of persons who have lost
rational agency but who have not written an advance directive permitting
euthanasia, however, whether they would be harmed by euthanasia is a
consideration. This should be dealt with on a case-by-case basis and is not an
appropriate matter for state prohibition. There is, however, room for some restrictive
regulation. The state must take account of the risks that a policy permitting
euthanasia would pose for those who are, at least potentially, moral agents and who
might be mistakenly killed. In the case of persistent vegetative state, for example,
there is risk of misdiagnosis. In addition, people who are accurately diagnosed as
being in a persistent vegetative state sometimes regain consciousness. As result, a
person must remain in a persistent vegetative state for a significant amount of time
before it can be determined with reasonable certainty that the vegetative state is
permanent.41 Kantian argument would, I believe, require a high degree of certainty
that a person had in fact permanently lost rational agency, though reasonable risks,
as we have seen, are justifiable on Kantian grounds. Where the lines should be drawn
depends on subtle empirical issues. Some have claimed that euthanasia should be
prohibited because of the potential for abuse.42 On this argument, some will kill
persons who are still rational agents and then claim that it was an instance of
euthanizing a person who had permanently lost rational agency. By prohibiting
euthanasia in all cases, the argument goes, the state is using coercion to protect the
freedom of persons and is thus justifiable on Kantian grounds. This argument
succeeds only if it is impossible to detect such abuses. If such abuses are detectable,
the state can protect freedom by prohibiting the abuses and adopting adequate
enforcement mechanisms. Once again, where the regulatory lines should be drawn
depends on the resolution of empirical issues.
Conclusion
Kant’s basic moral philosophy is compatible with, and even requires, his claim that
suicide is immoral, at least in medical contexts where end-of-life decisions are made.
Moreover, Kant’s ethics prohibits assisted suicide. For similar reasons, Kant’s
arguments would prohibit the use of pain control that results in permanent loss of
moral agency. Kant’s arguments do not, however, provide an a priori reason for
thinking that euthanasia is immoral in the case of persons who have permanently
lost their rational agency, such as those in a persistent vegetative state or the later
stages of Alzheimer’s disease. What regulations are justified and whether
prohibitions are justified depends on an assessment of empirical issues. Kant’s
arguments against paternalism, however, make it clear that the burden is on those
who would propose restrictive regulations. Kantian View of Suicide and End-of-Life
Treatment 285 Previous versions of this paper were presented to the Minnesota
Monthly Moral Philosophy Meetings and the College of St. Catherine. I am also
indebted to Geoffrey Gorham and David Mayo for their comments on earlier versions
of the paper