Catholic Health Care Advocates Papers

 Print Page

 Email Page

The Preeminence of Autonomy in Bioethics

Janet E Smith, PhD.

Upon reading case books on medical ethics or simply news reports in the media, one cannot escape noticing that the medical professions are currently involved in practices that some few decades ago would have been unthinkable, not only because of the level of technology involved, but because of the then prevailing moral evaluation of the practices. One reads of Dr. Kevorkian and his death machine; of a sixty-one year old woman having a baby conceived with another woman's ova in a petri dish; of embryos created solely for experimental purposes; of vital organs taken from living anencephalic infants; of millions of abortions yearly. Clearly a revolution of some kind, beyond the merely technological, has taken place.

Again, some few decades ago all of these procedures would have horrified most individuals for there was a widespread consensus that such practices violate some fundamental good, such as the laws of nature, or the dignity of persons, or God's laws or that they were indicative of bad character -- that those who performed them were venal or malicious. Perhaps there is not yet a consensus that these practices are morally good. Many factors, beyond a favorable moral evaluation, undoubtedly contribute to the growing acceptance and legalization of such practices. Certainly a major contributing factor is the importance that our culture and bioethicists in particular have come to put on respect for autonomy.

There seems to be a growing consensus that respect for the autonomy of individuals disallows the prohibition of practices previously thought morally objectionable -- even were there to be a widespread consensus that such practices are wrong. Abortion may be a good example of this phenomenon; some studies show that most Americans oppose abortions done for the reasons most women put forth to justify their choice, yet despite this opposition, most Americans support a woman's 'right to choose.' Although the growing primacy of the value of autonomy in bioethics has not gone unchallenged, virtually no bioethicist denies that the increasing respect for the autonomy of patients has been a very salutary development in medicine. The debate rages largely over what other principles, if any, compete with and might limit the principle of respect for autonomy (PRA). What I would like to do in this essay is to show how, in spite of claims to the contrary, respect for autonomy has begun to eclipse all other values in bioethics. I will show that advocates of the PRA have come to promote it no longer as the prima facie primary principle of bioethics, but as an absolute value that trumps all others.

The importance of respect of an individual's autonomy can be seen in a shift in the issues that dominated bioethical discussion: whereas bioethical texts and journals used to be occupied largely with assessing the moral pros and cons of procedures, an increasingly large proportion of bioethical literature now focuses on determining means to ensure competency, protection of confidentiality and privacy, informed consent, and voluntariness. In discussion of cases, bioethicists will often argue in a way that suggests that once it is ensured that the patient's choice is autonomous, the work of the bioethicist has been done. A bioethicist concerned that a patient make a good moral choice would be attempting to explain the principles which have lead the bioethicist to judge a choice to be moral and finding just ways to persuade the patient to accept these principles and then to make the moral choice. Less and less of bioethical literature argues the morality of procedures and practices; more and more is directed towards finding means to ensure that choices are autonomous. The autonomous choice has supplanted the good moral choice as the primary concern of bioethics. What explains this dominance of the PRA and related concerns in bioethics? It might be supposed that the ascendancy of the PRA is best explained by a growing respect for the self-determining powers of the human person. I will argue, however, that moral pluralism and skepticism have in fact been the forces pushing us in the direction of a nearly unfettered respect given to autonomy. And I will suggest that skepticism in particular threatens the very enterprise of bioethics. Whereas ethical debate may be possible -- though frustrating and perhaps ultimately indecisive -- in a climate of moral pluralism, skepticism eliminates the need for debate about the moral acceptability of different practices and policies. I shall propose that efforts to justify the PRA by appeal to moral skepticism, the view that universal standards for ethical judgment cannot be known, have seriously eroded the capacity of bioethics to assess the moral acceptability of different practices and procedures.

Bruce Miller's much cited article "Autonomy and the Refusal of Lifesaving Treatment" offers a core definition of autonomy: 'the right to autonomy is the right to make one's own choices, and ... respect for autonomy is the obligation not to interfere with the choice of another and to treat another as a being capable of choosing.' As he describes four senses of 'autonomy,' he sketches criteria that are generally considered requisite for a choice to qualify as an autonomous choice: autonomy considered as 'free action' requires that the choice be voluntary and intentional; autonomy as 'authenticity' requires that a choice be consistent with a patient's 'attitudes, values, dispositions, and life plans'; autonomy as 'effective deliberation' requires that the patient was 'aware of the alternatives and the consequences of the alternatives, evaluated both, and chose an action based on that evaluation' and that the patient 'rationally weight' the alternatives; autonomy as 'moral reflection' means that one has personally appropriated the values that govern one's choices.

The emphasis on the PRA is relatively new in bioethics. Paternalism is generally recognized as the historical rival value to autonomy and the one that the PRA has dethroned. Beauchamp and Childress define paternalism as 'the intentional overriding of one person's known preferences or actions by another person, where the person who overrides justifies the action by the goal of benefiting or avoiding harm to the person whose will is overridden.' Many have observed that while it can be and has been abused, the principle of paternalism seems quite well-suited to medicine, given the concern that medicine has with health and given that doctors likely know better than patients what is conducive to their health. Many speak of illness as a condition of reduced autonomy, in which it is quite proper that a patient submit to the authority of others. The necessity of paternalism for at least the incompetent, or those deemed incapable of exercising autonomy either because of age or condition (known as 'weak paternalism'), is recognized even by the strongest advocates of the PRA. The importance of the PRA to medicine is perhaps not as obvious as the suitability of paternalism. Few would argue that the PRA is invoked to promote the health of patients. Given the complexity of medicine, and the dependency of patients, it should be no surprise that in the clinical setting paternalism still reigns; patients want to trust their doctors.

The PRA clearly challenges paternalism by making patients responsible for choosing what care they receive. Whereas paternalism has been advocated as a means to ensure the best medical care for a patient, the widespread acceptance of the PRA stems from quite different concerns. It is advocated not so much as a means of restoring health, but rather as a means of ensuring good medical ethics. But it is no longer simply a means to good medical ethics; the PRA has assumed such centrality to the practice of bioethics, that some bioethicists now believe that the very goal of medicine should be changed 'from restoring health or reducing suffering' to 'restoring the autonomy of the patient.' The literature on the topic of autonomy suggests many reasons for the elevation of the PRA as the reigning principle in bioethics. Several have to do with developments within the health professions:

1. Exposes of the treatment of the subjects of medical research and experimentation brought about an interest in autonomy. Subjects of experiments were exposed to risks about which they had no knowledge. It was thought it would be minimally decent to inform them of what risks they were undertaking when they volunteered to undergo experimental treatment.
2. Many medical procedures have been introduced which, while offering some chance of cure or prolongation of life, are burdensomely painful or expensive. Patients facing horrendously painful procedures should have the option of declining such treatment.
3. Medicine is no longer considered an art, the physician no longer an artist. Medicine has become a contractual enterprise with patients as consumers and the doctor as a technician for hire.
4. Some proponents of the PRA are reacting against utilitarianism, which tended to override the wishes and good of the individual in favor of the good of the community.
5. Given the complexity of medical options today, many doctors invoke the PRA in order to 'avoid traditional responsibilities.'
6. At one time the curative power of medicine was greatly dependent upon the authoritativeness of the physician who actually had little genuine scientific knowledge. Advancing medical knowledge and the corresponding advancement in education of the patient as well, permits involving the patient in decision-making.

   The reasons cited above for the elevation of the PRA are largely historical and situational. That is, developments in medicine have made it possible and necessary to give greater respect to the choices of the patient.

   Until recent years, however, the PRA has not been treated as a principle that "trumps" all other principles. Certainly, two limitations to the PRA are perhaps obvious, uncontroversial, and ineluctable. There are certain biological limitations to the range of choices that a patient has, even if those limitations might eventually be overcome, e.g., Arnold Schwarzennegger films notwithstanding, as of yet males cannot gestate fetuses. Medical resources and cost containment also necessarily limit a patient's choices; e.g., not all those wanting to conceive babies through in vitro fertilization will be able to afford the procedure. These limitations seem more to do with practical considerations than moral ones.

   The primary moral reasons generally given for limiting a competent's individual's exercise of autonomy are 1) that a patient's choice does clear and serious harm to the patient himself (the first principle of medicine being 'do no harm'; 2) that the community may have values the violation of which it does not wish to tolerate; e.g., it may forbid abortion should the large majority of the citizenry consider abortion to be the killing of human beings. And 3) that a patient's choice may endanger public health; when the good of the whole is valued above the good of the individual, autonomous choice may be limited; e.g., it may be necessary to override the reluctance of some to be tested for AIDS. Some evidence suggests, however, that these constraints against the exercise of autonomy are losing their strength.

   The loosening of these constraints is most likely more attributable to social factors than to developments within medicine itself. That is, the values of a democratic society contribute to the growing ascendancy of the PRA:

7. The PRA comports well with the values of a democratic society which elevates individualism and freedom over community and authority. The rebelliousness of the sixties catapulted individualism and anti-authoritarianism to new heights.
8. Democratic society is highly tolerant of pluralism. Fear that the doctor and patient may not share the same moral commitments is one of the chief factors undergirding the interest in autonomy.

Recognition of the moral pluralism of modern society drives H. Tristram Englehardt, Jr.'s advocacy of autonomy. His advocacy is based not on the claim that autonomy is a premier value in itself. Rather, he argues that in a pluralistic society in which consensus about moral matters no longer exists and no mechanism exists for adjudicating moral disagreements, the PRA is the best means for ensuring that one group of individuals does not impose its views on another. While it is not right to say that Englehardt's advocacy is solely political, the respect he gives to individuals who adhere to an ethical tradition and who allow that tradition to guide their choices indicates that he is not an unrestricted advocate of the PRA. Moreover, he seems to hold out the possibility that rational individuals could come to some agreement about the morality of actions, even though it hasn't happened. He advocates the PRA as a peace-keeping measure, not as an ideal. Acknowledging that many morally objectionable choices will be made in the name of the PRA, Engelhardt accepts this as the price to be paid for a peaceful coexistence of members of incompatible moral traditions.

Yet, in a context of moral pluralism, the PRA may achieve ascendancy for more than practical reasons. In fact, in a context of moral pluralism, pluralism itself may become a value, and this development may enhance the status of autonomy. The motivations behind the liberalization of laws against abortion may be of such a kind; that is, although the majority of a community may disapprove of abortions, the value of pluralism and the value of choice may supersede the community value.

We can also see autonomy overruling a community value in the well known cases of a Jehovah witness refusing a blood transfusion. The community believes blood transfusions to be normal medical care and the refusal of such to be tantamount to suicide. Yet, the wishes of Jehovah Witnesses have routinely been honored, initially, one suspects, more because of respect for the sovereignty of religious belief, but now, more likely because of the desire to respect autonomy. The respect for the sovereignty of religious belief may have contributed to a climate of moral pluralism and made autonomy more attractive. At any rate, autonomy can clearly supersede community value.

Although moral pluralism does not necessarily entail subjectivism, relativism, and skepticism, the inability to resolve moral disputes that customarily results from debates in a context of moral pluralism, quite readily leads to skepticism.

Daniel Callahan, a vigorous opponent of the preeminence of the PRA in bioethics, argues that skepticism explains the commitment to autonomy 'on the streets':

1. As moral agents, we are essentially independent of each other and isolated; we are not social animals, but morally self-enclosed, self-encompassing animals.
2. There can be no moral truth or wisdom about individual moral goods and goals and few if any about communal ends; morality is inherently subjective and relativistic.
3. The ideal relationship among human beings is the voluntary, contractual relationship of consenting adults; the community has no standing to say what is good or bad in such relationships.
4. In any weighing of the relative interests of individual and community, the burden of proof is always upon the community to prove its case for restricting the liberty of individuals.
5. The only moral obligations I have toward others are those I voluntarily undertake; there can be no such things as in involuntary moral obligation.
6. The only moral obligations that others have toward me are those that autonomously I allow them to have; all I am owed by others is respect for my autonomy.
7. Respect for the autonomy of others is sufficient for overriding my own conscience.

These 'on the street' justifications for preeminence of autonomy have some roots or at least parallels in the philosophic justification of autonomy, a justification that is sometimes difficult to unearth.

The philosophical deliberations of Kant and Mill are frequently invoked to justify the PRA. What moderns likely find appealing about Kant's view of autonomy is 1) his position that rational beings are always to be treated as ends and never as means and 2) his contrasting of autonomy, or the ability to be self-legislating, with heteronomy, or being subject to the laws of another. What is entirely absent from the modern employment of the PRA is Kant's insistence that all choices of rational beings must be subject to the categorical imperative; that all moral choices are subject to universal norms. Freedom, for Kant, is actualized properly only through a life of duty. Moderns, on the other hand, understand 'autonomy' or 'self-legislating' to be largely a principle of 'preference': 'I do what seems good to me to do -- not what is dictated by universal norms, the community, natural law, etc, unless I so choose to make these my moral principles.' The modern view finds some covert support in Kant, however. For, although Kant believed in an universal morality, binding on all, his nonrealistic metaphysics is arguably more compatible with subjectivism. If one can have no certain knowledge of the external, objective world (the noumenal world), the subjective world or the phenomenal world begins to take precedence. While surely the transition to a highly individualistic, subjectivistic ethics has many sources, Kant's philosophy in spite of its inherent incompatibility with such an ethics, contributed to this transition.

Mill's ethics exhibited nearly no concern that ethical norms be universal. Mill largely wanted to give man free reign to actualize his individuality within the limits of harming others or harming his own ability to make free choices. In fact, for Mill, one major source of man's happiness is his autonomous pursuit of self-defined goals. Mill's exaltation of freedom was based largely on skepticism, on the view that since certainty about the nature of reality and about morality is impossible, individuals should be free to shape their lives in accord with their own views.

The fact that two such disparate philosophies such as those of Kant and Mill are rather willy-nilly invoked to support commitment to the PRA, suggests that the justification of the PRA does not rest upon the strength of the philosophical argumentation of Kant, Mill or any other philosopher. It is not their philosophies that validate commitment to the PRA; it is the compatibility of their philosophies with skepticism.

The close alliance between the recent ascendancy of the PRA in bioethics and skepticism can be seen in the work of Beauchamp and Childress. They somewhat gingerly propose that truth is not possible in the moral realm:

...it is far from settled that this treatment of moral truth is adequate. It is doubtful that a successful body of coherent beliefs, no matter how stable, yields truth. For one reason, it is doubtful that moral statements have truth values and that truth is a category that should appear in moral theory. For another reason, we would need a theory of truth, itself a complicated and controversial subject. We are content to conclude here that justification successfully occurs in ethics and that the right approach to justification is the coherence account...

While Beauchamp and Childress do not straightforwardly deny that truth is possible in the moral realm, their firm expression of doubt about this possibility and their approach to bioethics, qualifies them, in effect, as skeptics.

The 'coherence theory of justification' advocated by Beauchamp and Childress requires that we start 'with considered judgments that are settled moral convictions in a broad expanse of ethics, and then cast... the net more broadly in specifying, testing, and revising those convictions.' All convictions are ultimately revisable; they depend upon no fundamental unchangeable principles nor is one seeking to establish any unchangeable principles.

Although Beauchamp and Childress cite Kant and Mill as possible sources for a philosophical justification of the PRA, they ostensibly ground their commitment to the PRA in common morality: 'Respect for the autonomous choices of other persons runs as deep in common morality as any other principle....' Furthermore, they find the grounding of this (and their other principles) in common morality to be more secure than a grounding in philosophical theory:

If we could be confident that some abstract moral theory was a better source for codes and policies than the common morality, we could work constructively on practical and policy questions by progressive specification of the norms in that theory. But fully analyzed norms in ethical theories are invariably more contestable than the norms in the common morality. We cannot reasonably expect that a contested moral theory will be better for practical decisionmaking and policy development than the morality that serves as our common denominator. Far more social consensus exists about principles and rules drawn from the common morality (for example, our four principles) than about theories.

Beauchamp and Childress do not explore the reasons why 'common morality' values autonomy so highly. While in the understanding of Beauchamp and Childress, 'common morality' is not identical with public opinion, it seems fair to note that, as Callahan observes, the reason that the public values the PRA is largely because of skepticism. Thus, in my view, it is largely skepticism, their own and that which seems to characterize our age, that undergirds the commitment of Beauchamp and Childress to the PRA.

Beauchamp and Childress have claimed that the PRA is a prima facie principle that in some circumstances must yield to other values. Yet, the shift in Beauchamp's and Childress' treatment of assisted suicide reveals that in spite of their denials, the PRA has begun to assume status as an absolute value in their account. The question of assisted suicide is a good test of the moral limits of the PRA since assisted suicide seems to violate what we identified earlier as generally recognized limits the PRA: 1) that the patient not be doing clear and serious harm to himself; 2) community values; and 3) the common good.

In the third edition of their Principles of Biomedical Ethics, Beauchamp and Childress seem to agree that assisted suicide did not fall within the limits of what is permitted by the PRA. They certainly saw it as a reversal of the values of the medical community ('do no harm') and of the larger community, and that it threatened the common good, as well:

...we need to ask which side in the debate [on assisted suicide] has the burden of proof -- the proponents or the opponents of a practice of selective killing. One prominent view is that the opponents of a practice of prohibiting killing bear the burden of proof because the prohibition of voluntary euthanasia infringes liberty and autonomy. However, a policy of voluntary euthanasia, based on either a negative right to die (a right to noninterference) or a positive right to die (a right to be killed), would involve such a change in society's vision of the medical profession and in medical attitudes that a shift in the burden of proof to the proponents of change seems to us essential. We have argued that the prohibition of killing expresses important moral principles and attitudes whose loss, or serious alteration, could have major negative consequences. Because the current practice of prohibiting killing while accepting some 'allowed deaths' has served us well, if not perfectly, it should be altered only with the utmost caution. Lines are not easy to draw and maintain, but in general we have been able to respect the line between killing and letting die in medical practice. Before we undertake any major changes, we need more evidence than we now have that the changes are needed in order to avoid important harms or secure important benefits and that the good effects will outweigh the bad effects.

In this passage Beauchamp and Childress place the burden of proof on supporters of assisted suicide, since the widespread acceptance of this practice would require undesirable changes in society's views of the medical profession and would be harmful to society. In elaboration of their opposition to assisted suicide, they state 'a policy that authorizes killing medicine -- even in a few cases -- stands to violate the obligation of nonmaleficence by creating a grave risk of harm in many cases.' They speak of a likely erosion of trust between patient and doctor: 'The prohibition of killing is an attempt to promote a solid basis for trust in the role of caring for patients and protecting them from harm. This prohibition is both instrumentally and symbolically important, and its removal could weaken a set of practices and restraints that we cannot easily replace.' They invoke the principle of the slippery slope and summarize their position in this way:

The main reservation expressed in this argument is the following. If rules permitting mercy killing were once introduced, society might gradually move in the direction of nonvoluntary and perhaps involuntary euthanasia -- for example, in the form of killing handicapped newborns to avoid social and familial burdens. There could be a general reduction of respect for human life as a result of the official removal of barriers to killing. Rules against killing in a moral code are not isolated fragments; they are threads in a fabric of rules, drawn in part from nonmaleficence, that support respect for human life. The more threads we remove, the weaker the fabric becomes. ....'

By the fourth edition of their text, published only five years later, Beauchamp and Childress argue that the burden of proof against assisted suicide now lies with opponents. Their shift in view does not seem to be based on an argument that there has been a change in community values outlined in their third edition. Now the harm that comes to a patient through denying the patient's autonomous choice outweighs all other goods:

... If a person desires death rather than life's more typical goods and projects, then causing that person's death at his or her autonomous request does not either harm or wrong the person (though it might still harm others -- or society -- by setting back their interests, which might be a reason against the practice). To the contrary, not to help such persons in their dying will frustrate their plans and cause them a loss, thereby harming them. It can also bring them indignity and despair. Furthermore, if passive allowing to die does not harm or wrong a patient because it does not violate the patient's rights, then assisted suicide and voluntary active euthanasia similarly do not harm or wrong the person who dies. Those who believe it is sometimes morally acceptable to let people die but not to take active steps to help them die must therefore give a different account of the wrongfulness of killing persons than the one we have suggested. The burden of justification, then, seems to rest on those who would refuse assistance to those who wish to die, rather than on those who would help them.

Note, again, that the reason offered for shifting the burden of proof to the opponents of assisted suicide is not the potential harm to the community, but rather the harm that refusing wishes does to the moral agent. In another statement justifying placing the burden of proof on opponents of assisted suicide, Childress combines the concern for the community with a concern for autonomy:

Other communitarian concerns may also outweigh the principle [of respect for autonomy] in some contexts. Consider the debate about relaxing the societal and professional rules against physician-assisted suicide and active euthanasia. The principle clearly establishes a prima facie case for changes in these rules as a way to respect patient autonomy. However, the debate rightly focuses on whether over time the community as a whole, including its most vulnerable members, would be best served by a change in the rules -- for example, would such a change promote or threaten patient autonomy under conditions of serious illness? (my emphasis)

Here the community value that would need to offset the goodness of honoring a patient's wish to die is no longer the need to protect doctor/patient trust or the life of the handicapped infant, but the value of patient autonomy.

The implications of the growing preeminence of the principle of autonomy and the shifting grounds for its justification, have serious implications for bioethics. Deontological argument and utilitarian arguments seem now without place in bioethics. Neither bioethicists, doctors, or patients need to justify their choices in terms of moral principles: the overriding concern is whether or not a patient's autonomy is being respected. This turn from considering the moral acceptability of issues to ensuring the autonomy of choices renders the discipline of bioethics rather vacuous.

It would require another essay of this length to argue that moral pluralism need not lead to the unfettered primacy of autonomy or to the evacuation of the substance of bioethical discussions. Bioethicists could be helping patients understand the implications for their decisions of their own moral traditions; they could be helping patients without a tradition identify what principles of moral discernment are acceptable to them; they could be helping patients learn how to respect their own principles, the values of the communities in which they live, and the values of the doctors who serve them. A respect for patient's autonomy would clearly be a driving force behind such activity but would retain bioethics as a discipline that is concerned with evaluating the moral status of medical practices. It is time for bioethics to reconsider the value it places on the PRA.