Degrazia • the Principles of Biomedical Ethics

The fifth edition of Beauchamp and Childress’s Principles of Biomedical Ethics is distinguished by its emphatic embrace of common morality as the ultimate source of moral norms. This essay critically evaluates the fifth edition’s discussion of common morality and, to a lesser extent, its treatment of coherence (both the model of ethical justification and the associated concept). It is argued that the book is overly accommodating of existing moral beliefs. The paper concludes with three suggestions for improving this leading text. Throughout its five editions (1979–2001), Tom Beauchamp and James Childress’s Principles of Biomedical Ethics has been enormously important to the field of bioethics. One would be hardpressed to find a text that has been more influential and more frequently cited. In addition, the overall achievement of each edition has been very impressive: the integration of large segments of ethical and political theory as well as law and history; the sensitive analysis of critical concepts; the informed discussions of current issues in bioethics; the peppering of real cases throughout the text—in sum, a formidable mastery of salient information, key concepts, and normative reasoning. If the authors had produced no other works in their careers than this one, they would have contributed much. One of the book’s strengths is its authors’ willingness to change. The changing methodology—or at least the way methodology is presented— over the course of the editions offers an interesting example. In the first * This article is a lightly edited version of a paper presented at The State of Bioethics: From Seminal Works to Contemporary Explorations, the thirtieth anniversary celebration of the Kennedy Institute of Ethics, Georgetown University, 6 April 2002. KENNEDY INSTITUTE OF ETHICS JOURNAL • SEPTEMBER 2003 [ 220 ] three editions, there was an emphasis on ethical theories that could support the mid-level principles of bioethics: respect for autonomy, beneficence, nonmaleficence, and justice. One author, Beauchamp, held that rule-utilitarianism was preferable to any form of deontology; the other author, Childress, held that rule-deontology was preferable to any form of utilitarianism. A close reading of the text reveals that neither fully endorsed either theory. Rather, if required to choose, this is how each would choose (see Beauchamp and Childress 1979, p. 40; 1983, pp. 40–41; 1989, p. 44). In other words, each author was pluralistic at the level of ethical theory from the start—contrary to virtually everyone’s interpretation. The mid-level principles, the authors maintained, enjoyed greater certainty than any ethical theory. However, it seems fair to say that the emphasis on theoretical pluralism grew with succeeding editions, as did the stress on reflective equilibrium as a model that held principles, rules, and specific judgments together. The fourth edition (1993) introduced a massive expansion of theoretical perspectives to be considered and, methodologically, integrated specification into the mix. Finally, the fifth edition (2001) is distinguished by its emphatic embrace of common morality as the ultimate source of moral norms. One mark of a book’s importance is the frequency with which it is criticized. In effect, critics say, “Your work is worth troubling over.” In this paper, I “trouble” over the authors’ discussion of common morality—and, to a lesser extent, their treatment of coherence—in the hope of providing some insight into the problem of how to improve an excellent book. Before taking up these topics, I briefly sketch the authors’ so-called “principlism,” a central and enduring component of their methodology. BEAUCHAMP AND CHILDRESS’S VERSION OF “PRINCIPLISM”1 In the authors’ view, four major principles—respect for autonomy, nonmaleficence, beneficence, and justice (the latter being really a cluster of principles)—are general crystallizations of accumulated moral insight. But these principles are notoriously subject to conflict. In cases of conflict, principles should be balanced and, if possible, specified—made more specific—in an effort to resolve the conflict for the case or problem at hand (see Beauchamp and Childress 2001, p. 18).2 It is important to understand, however, that moral reasoning in this model does not always move “downward,” from more general to more specific norms. In the model of reflective equilibrium, also called “the coherence model of ethical justification,” while one often applies general DEGRAZIA • THE PRINCIPLES OF BIOMEDICAL ETHICS [ 221 ] norms to specific cases, and may specify more general norms in cases of conflict, one also may use reflection on concrete cases to revise or even to reject a more general norm.3 A good example of such “upward” moral revision is revising a rule of confidentiality, after considering the famous Tarasoff case, to allow breeches of confidentiality where a patient threatens grave bodily harm to an identifiable third party. What justifies a particular revision of moral belief, whether “upward” or “downward,” is the overall coherence of the set of norms that results from introducing the new judgment. That, of course, raises a question: What is coherence? CRITIQUE OF BEAUCHAMP AND CHILDRESS’S TREATMENT OF COHERENCE AND COMMON MORALITY Surprisingly, Beauchamp and Childress (2001, p. 400) say very little about what coherence is. What they could and should say, I think, is that the criteria developed in an earlier chapter for the evaluation of ethical theories—consistency, comprehensiveness, justificatory power, and so forth (pp. 338–40)4 —are precisely the relevant criteria for coherence, even if this means amplifying the ordinary meaning of “coherence.” In other words, coherence for the purposes of ethical theory and evaluation is the holistic embodiment of theoretical virtues, the characteristics one expects of any good theory.5 But one theoretical virtue, or criterion for evaluating theories, that is not on their list is plausibility. A set of moral norms should be largely believable, at least upon reflection. Rather than include a criterion of plausibility, the authors accomplish roughly the same thing by stressing the importance of “considered judgments,” which they characterize, quoting Rawls, as “judgments in which our moral capabilities are most likely to be displayed without distortion” (p. 398), for example, “Slavery is wrong.” I think it would help to say much more about what should count as a considered judgment.6 Among the judgments of common morality, we will need to distinguish those that are credible and those that are not. I suggest that considered judgments, which have an initial credibility (though not infallibility), are moral judgments that are adequately informed in terms of both factual information and relevant moral alternatives, stable even when one is not under pressure to accept them, free from personal conflict of interest, and the like. We know from experience that people may accept particular moral judgments for any number of reasons. Sometimes we believe a moral judgment because it is reasonable and defensible, even in the face of the KENNEDY INSTITUTE OF ETHICS JOURNAL • SEPTEMBER 2003 [ 222 ] most thoughtful criticism. But sometimes we believe a moral judgment because believing it promotes our own interests, or avoids the need for unpleasant questioning or even a change in lifestyle; or because believing it conforms to our political ideology or religious dogma; or simply because we are misinformed about the relevant facts. A vivid awareness of our fallibility suggests the importance of being positioned to recognize when our moral beliefs have a credibility problem (unlike considered judgments). In my view, Beauchamp and Childress’s enthusiastic embrace of common morality in the fifth edition does far too little to address this concern. What is common morality? In one place the authors define it as “the set of norms that all morally serious persons share” (Beauchamp and Childress 2001, p. 3). Now that is a small set—one that arguably includes, at a very general level, only nonmaleficence. Note that libertarians reject beneficence as a principle of obligation, while perhaps some morally serious non-Westerners do not embrace respect for autonomy; and it is quite clear that there is no substantive principle of justice that all morally serious people accept. Elsewhere, however, the authors appear to use the term “common morality” more inclusively, describing it as “a heap of obligations and values unconnected by a first principle” (2001, p. 407). And their frequent appeals to common morality—seemingly appeals to widely shared intuitions, such as the judgment that obligatory beneficence is limited (2001, p. 169)—suggest the more inclusive reading. Beauchamp and Childress provide, as examples of common-morality theorists, Frankena (1970) and Ross (1930). Ross, they say, has influenced them more than any other twentieth-century author (2001, p. 402). That may be a reason for concern. Although Ross overcame some of the limitations of utilitarianism and Kantian theory with his list of prima facie duties, his view, like a statement of mid-level principles, represents only a start. And, importantly, Ross offers no analytical tools for determining when claims of self-evidence or specific judgments are unreliable. The authors next state that, while common-morality theory begins with ordinary, shared moral beliefs, it may reach conclusions that are not commonly shared (2001, p. 403). This is a crucial point, but the authors do not really develop it beyond stating that they wish to unite common morality with the coherence model of ethical justification (2001, p. 403). And sometimes they do not seem to follow their own good point. At this juncture, it will be helpful to distinguish two senses of common morality. “Common morality 1,” as I will define it, is the set of widely shared moral beliefs. “Common morality 2” is the set of moral beliefs DEGRAZIA • THE PRINCIPLES OF BIOMEDICAL ETHICS [ 223 ] that probably would be widely shared among morally serious people who give initial credence to considered judgments (as just explained) and expand their moral thinking in conformity with the criteria for evaluating moral theories—e.g., consistency and argumentative support. In other words, common morality 2 is a set of moral beliefs in reflective equilibrium. Beauchamp and Childress sometimes intimate that they are committed to common morality 2. But some of their remarks and discussions seem to favor common morality 1. This is suggested, for example, in their frequent appeals to social consensus. Here it is instructive to consider why they favor common morality over any specific ethical theory. One reason they offer is that there is far more consensus on common-morality principles and rules than on any theory (2001, p. 404). (Note the circularity of justifying common morality in terms of consensus.) A stronger argument is pragmatic: that appeal to norms of common morality will work better for practical decision making and policy development (2001, p. 404). Although this is probably often the case, there is reason for concern about the authors’ heavy reliance on consensus or what most people actually believe. As already noted, people hold moral beliefs for a variety of reasons, not all of which are conducive to reliability. In my view, there is nothing paradoxical in the Socratic thought that 90, or even 99, percent of all people might share a moral belief yet be wrong; experience reveals that moral judgment can be distorted in many ways. Consider these moral judgments: (1) Homosexuality is morally problematic. (2) It is more incumbent on women than on men to change their names upon getting married and to sacrifice some of their own interests, such as their professional development, for the sake of family. (3) Property rights are sacred. (4) If a wealthy couple sends their child to private school, though the public school in their district is very good, and they contribute nothing to charity, they have not thereby failed in an obligation of beneficence. (5) Meat eating is morally unproblematic; so is going to a circus that features animal acts. (6) Animals exist for human use. These very widely shared judgments seem to reflect common morality 1. I happen to believe all six are false. From a progressive perspective that does not kowtow to religious or social tradition, capitalist ideology, uppermiddle-class priorities, or speciesism, we need to watch out for common KENNEDY INSTITUTE OF ETHICS JOURNAL • SEPTEMBER 2003 [ 224 ] moral prejudices and use critical tools very vigorously. Fortunately, common morality offers not only widely shared moral judgments, but also methods of questioning moral judgments—that is, critical tools—though we tend not to use them enough. When we do so, we move toward common morality 2. Near the end of the last chapter, Beauchamp and Childress apparently gravitate toward common morality 1. Confronting the question of whether they can construct a theory in any rich sense, they seem unambitious: “Perhaps mid-level principles, polished analyses of the moral virtues, and coherent statements of transnational human rights are all that we should attempt, rather than a theory that conforms to the criteria delineated [earlier]” (2001, p. 407). Why give up so easily? Throughout the book they defend a highly integrated set of norms, of different levels of generality and of different types. That already involves more structure than the previous statement aspires to. Why not keep seeking the most coherent and plausible set of interconnected norms possible? Their reason: “. . . attempts to bring the common morality into greater coherence through specification risk decreasing rather than increasing moral agreement in society” (2001, p. 407). This statement suggests prioritizing common morality 1 over common morality 2, not to mention abandoning the coherence model of justification—to which they had earlier pledged their allegiance. The importance of this move should not go unnoticed: To embrace common morality 1 is to reduce normative ethics to descriptive ethics, eliminating the logical space needed for the very concept of moral progress. The quoted statement also reveals, I think, excessive aversion to disagreement. Although this may seem obvious, standing up for what is right—for what makes the most moral sense, all things considered—sometimes means accepting disagreement and not pleasing everyone. Although the deeply democratic instincts that guide the authors in these passages are themselves morally admirable, they can be theoretically paralyzing—and they do not get us closer to truth and justification. Beauchamp and Childress (2001, p. 408) conclude rather pessimistically: “A common-morality theory will certainly not satisfy the full set of criteria we delineated at the beginning of Chapter 8.” That is a major concession. If those criteria are irrelevant or pointless, why were they employed in evaluating ethical theories? Alternatively, if the criteria are valid, then why favor a common-morality approach that apparently does no better in meeting these criteria than the ethical theories that were judged inadequate on their basis? I believe the authors’ discussion of common DEGRAZIA • THE PRINCIPLES OF BIOMEDICAL ETHICS [ 225 ] morality—while admirable in reflecting both deeply democratic instincts and a willingness to innovate—goes much too far in the direction of antitheory and a tacit embrace of the moral status quo. This is, I suggest, the chief difficulty in the fifth edition’s stated methodology. A secondary difficulty, to my mind, is an overzealous effort in particular discussions to accommodate multiple “constituencies.” The same democratic instincts may be at work here. The result, I think, is reduced coherence and justification. I offer here a few brief examples in which one or both of these difficulties appear. EXCESSIVE ACCOMMODATION OF EXISTING MORAL BELIEFS AND/OR COMPETING CONSTITUENCIES: THREE EXAMPLES The first example concerns obligations of beneficence. In discussing whether and to what extent we have such obligations, the authors contend that common morality recognizes “significant limits to the demands of obligatory beneficence” (2001, p. 169). Descriptively, that seems right. They go on to argue that common morality does not generate a clear line determining the extent of any such obligations (2001, p. 173). That also seems correct (although some of the reasoning leading to this conclusion is very puzzling7 ). In a later chapter, however, they suggest that there are no general obligations of beneficence8 at all: “If such norms of charity express what we ‘ought to do,’ they do so not from obligation but from personal ideals that exceed obligation” (2001, p. 359). This conclusion strikes me as insufficiently respectful of human beings’ moral status. If it is part of common morality at all, it belongs to common morality 1; certainly, this thesis could not survive in reflective equilibrium. For one thing, we generally recognize that universalizability and putting oneself in another’s shoes are important critical tools in examining moral beliefs. Now ask yourself what you would think if you were a starving child who knew there were many millions of advantaged people who could take steps to help others in dire need, with no significant costs to themselves: You would doubtless favor a system in which people of means recognized some general obligations of beneficence.9 Obligatory beneficence is tied to distributive justice, leading to the second example. Beauchamp and Childress’s chapter on justice starts promisingly, condemning the American status quo of health care delivery (2001, p. 240) and employing a rule of fair opportunity as a basis of criticism and reform. If the discussion had continued to develop this conception of justice, it might have proved more cogent than it actually turns out. InKENNEDY INSTITUTE OF ETHICS JOURNAL • SEPTEMBER 2003 [ 226 ] stead, the desire to accommodate a wide range of views, including libertarianism, apparently takes over, and the authors recommend universal access to a decent minimum of health care as the appropriate goal (2001, p. 245). Presumably libertarianism, which recognizes no right to health care, exerts downward pressure on the extent of the right crafted by the committee of theories. But considering that 20-some nations provide universal access to a set of benefits much more comprehensive than the term “decent minimum” suggests, while spending less per capita than do Americans (including the uninsured), I believe there is a very strong case that the decent-minimum standard is too low (see DeGrazia 1996b).10 A third, and final, example involves the chapter on nonmaleficence, which includes excellent discussions of the concept of harm, the killing/ letting die distinction, the doctrine of double effect, and other topics. After arguing that some individual acts of physician-assisted suicide are justified, the authors shy away from recommending legalization. They cite slippery slope concerns despite admitting that there is “scant evidence” to support such empirical claims (2001, p. 146). Later, citing the U.S.’s success in holding the line between justified and unjustified cases of passive euthanasia, they are optimistic: “[W]e should be able to hold the line between justified and unjustified assistance in suicide. We appreciate that this observation conflicts somewhat with our earlier comments on . . . slippery slope arguments, but we believe that these two points of view can be reconciled . . .” (2001, p. 152). How? Beauchamp and Childress’s commentary on assisted suicide now appears incoherent. It is difficult to avoid the impression that in the earlier discussion they are trying to accommodate conservative participants in this debate and later they are trying to accommodate liberal commentators. Although the effort to forge a compromise between major competing views sometimes generates morally compromised—that is, less than fully defensible—results, as I believe to be the case in the previous two examples, here the effort to compromise apparently leads to contradiction.