SOCIAL MORAL EPISTEMOLOGY

SOCIAL MORAL EPISTEMOLOGY* By Allen Buchanan I. Introduction The distinctive aim of applied ethics is to provide guidance as to how we ought to act, as individuals and as shapers of social policies.1 In this essay, I argue that applied ethics as currently practiced is inadequate and ought to be transformed to incorporate what I shall call social moral epistemology. This is a branch of social epistemology, the study of the social practices and institutions that promote (or impede) the formation, preservation, and transmission of true beliefs. For example, social epistemologists critically evaluate the comparative advantages of adversarial versus inquisitorial criminal proceedings as mechanisms for the discovery of truth.2 Social moral epistemology, at its most inclusive, is the study of the social practices and institutions that promote (or impede) the formation, preservation, and transmission of true beliefs so far as true beliefs facilitate right action or reduce the incidence of wrong action. A special department of social moral epistemology is the study of the social practices and institutions that promote (or impede) the functioning of moral virtues so far as their functioning depends upon true beliefs. For example, the virtue of respect does not function properly in an individual if he believes that blacks or Jews are not truly human and are hence unworthy of the respect owed to human beings. (Whether one calls what this person has a truncated form of respect or denies that it is the virtue of respect at all is irrelevant for present purposes.) Without social moral epistemology, I shall argue, applied ethics cannot provide sufficient guidance for right action. Social moral epistemology is epistemology because it focuses on the role of true beliefs in moral action. It is a normative enterprise, not a merely sociological one, because it evaluates the epistemic performance of social * I am grateful to Stefan Sciaraffa and Todd Stewart for their helpful comments on an earlier version of this essay. 1 The term “applied ethics” is usually employed to include bioethics, business ethics, journalism ethics, engineering ethics, and so on. The term suggests that these endeavors consist of the application of an ethical theory or theories to concrete ethical problems in order to guide choice and practice. This is somewhat misleading, however, since in most cases the work in these fields that is said to be applied ethics is not the application of anything that could qualify as a theory. 2 Alvin Goldman, Knowledge in a Social World (Oxford: Oxford University Press, 1999), 289–92.

126

© 2002 Social Philosophy & Policy Foundation. Printed in the USA.

SOCIAL MORAL EPISTEMOLOGY

127

practices and institutions. It is social because it regards the moral agent as socially embedded, and focuses on the role of social practices and institutions in the proper functioning of moral virtues and in right action generally. It is a species of moral epistemology because its concern is not with true belief generally, but rather with true beliefs so far as they play a role in right action. Social moral epistemology, as I will show, provides guidance for the design of social institutions. More specifically, the applications of social epistemology I sketch in this essay support the conclusion that one attraction of liberal institutions is that they encourage beliefs and epistemic habits that tend to enable the moral virtues to function properly. (Thus, the common wisdom that liberal institutions are not designed to produce virtuous people turns out to be somewhat misleading.) I shall also argue that social moral epistemology has important implications for what is probably the dominant method in normative ethics, the method of reflective equilibrium. To employ this method is to strive for an optimal match between, on the one hand, one’s most confident considered moral judgments about particular actions, policies, and institutions, and, on the other, a relatively small set of general moral principles that can serve to ground the particular judgments. Revision can occur in either direction: a hitherto accepted general principle may be qualified or even abandoned altogether if it is found to be inconsistent with particular considered moral judgments that are especially stable and in which we have exceptional confidence, and particular judgments can be reevaluated if they clash with general principles that otherwise appear to yield plausible implications in a wide range of cases.3 Without a theory of social moral epistemology, advocates of the method of reflective equilibrium are ill-equipped to answer the obvious objection to that method — namely, that the moral intuitions with which it begins may be systematically distorted and may in turn systematically distort the outcome of the process. Finally, I will show that the concept of social moral epistemology provides the basis for a critical evaluation of “metaethical communitarianism,” understood as the view that we come to know our moral obligations by locating ourselves within the social practices and traditions of a community. Social moral epistemology is crucial for distinguishing between those traditions that are reliable sources of moral obligations and those that are not. Without social moral epistemology, reliance on a tradition as a guide to our moral obligations is morally risky business. II. The Limitations of Contemporary Applied Ethics Applied ethics as usually practiced suffers from a limitation that greatly impairs its attaining the distinctive goal of increasing right action and 3 On reflective equilibrium, see John Rawls, A Theory of Justice (Cambridge, MA: Harvard University Press, 1971), 48–51.

128

ALLEN BUCHANAN

reducing the incidence of wrong action. Applied ethicists tend to focus exclusively on two tasks: identifying valid moral principles and constructing arguments in support of them. This constructive enterprise is often preceded by the critique of rival principles, either by showing that the rival principles are incompatible with widely shared considered judgments (moral intuitions) or by exposing the unsoundness of the arguments offered in support of the principles. Three examples will demonstrate the inadequacy of this sort of applied ethics and the need for social moral epistemology. A. Medical paternalism First, consider the standard applied-ethics critique of medical paternalism. Medical paternalism is usually defined as the view that physicians may withhold information (e.g., of a diagnosis of terminal cancer) or even lie to patients, or otherwise preempt their making a free, informed choice, when this is done for the patient’s own good. As recently as the mid1970s, medical-paternalist behavior was quite common in the United States, and in some cases physicians even subscribed to the view that paternalism toward normal adult patients is essential to the role of the physician because it is required by the physician’s role-defining fiduciary obligation to act in the best interest of the patient.4 In the early years of modern applied ethics, philosophers advanced conclusive objections against medical paternalism.5 The critique of medical paternalism proceeded along standard applied-ethics lines: The medical-paternalist view was first criticized by exposing the weakness of the arguments that medical paternalists invoked to justify it. Then, on the basis of these criticisms, applied ethicists concluded that the proper principle for physicians is not merely to act in the best interest of the patient, but rather to act in the best interest of the patient subject to the proviso that a competent patient has a right of informed consent to treatment. Regard for the patient’s welfare is to be constrained by respect for patient autonomy. In retrospect, what is striking about the criticism of medical paternalism is that it was too easy. The arguments that advocates of paternalism marshaled to support their view were transparently inadequate. In what follows, I will consider two pro-paternalist arguments, the criticism of which may have played an important role in the emerging consensus among bioethicists —though not necessarily among physicians —that medical paternalism was unjustifiable. The first argument is based on a claim 4 See Allen Buchanan, “Medical Paternalism,” Philosophy and Public Affairs 7, no. 4 (1978): 370–90. 5 See ibid. and Alan Goldman, “The Refutation of Medical Paternalism,” in John D. Arras and Bonnie Steinbock, eds., Ethical Issues in Modern Medicine, 5th ed. (Mountain View, CA: Mayfield, 1998), 58–66.

SOCIAL MORAL EPISTEMOLOGY

129

about the scope of the authorization that allegedly occurs when an individual engages a physician; the second is based on a claim about the nature of the fiduciary duty that constitutes the role of medical professional. Both of these pro-paternalist arguments were advanced to show that the general moral prohibition against lying and deception did not apply to physicians in their relations with patients. Accordingly, the critique of medical paternalism proceeded on the assumption that lying and deception are generally wrong, and then endeavored to show that the propaternalist arguments did not establish that the case of physicians lying to or deceiving their patients for the patients’ own good was a justifiable exception to the general rule. According to the first pro-paternalist argument, the patient authorizes the physician to do whatever he thinks is best for the patient simply by virtue of having entered into the physician-patient relationship. Once the implicit premises of this argument are articulated, its unsoundness is glaringly evident. (1) Whenever one engages an agent to serve one’s interests, one thereby authorizes the agent to do whatever he thinks best serves one’s interests. (2) Whenever one enters into the physician-patient relationship as a patient, one is engaging the physician as an agent to serve one’s interests. (3) Therefore, whenever one enters into the physician-patient relationship as a patient, one thereby authorizes the physician to do whatever he thinks best serves one’s interests. The first and perhaps most obvious problem with this argument is that premise (1) is clearly false. If one engages a roofer to fix one’s roof, or a mechanic to fix one’s car, one does not thereby authorize him to do whatever he thinks is necessary to effect the repair, much less whatever is necessary to maximize one’s interests overall. There are always implicit — and sometimes explicit —constraints on the authorization. Moreover, a social practice according to which merely engaging a physician constitutes a blank authorization for the physician to withhold information or preempt patient choice according to his estimate of what would best serve the patient’s interests would be an unreasonable practice. This is the case because of two obvious facts: (a) the distinctive expertise of physicians is medical expertise, not expertise regarding the patient’s overall good; and (b) the very asymmetry of knowledge that makes it attractive for a person to engage a physician creates the possibility of conflicts of interest, that is, opportunities for the physician to take advantage of the patient (for example, by undertaking unnecessary procedures). With respect to (a), bioethicists were quick to point out that it is simply not the case that all (or even most) individuals who engage a physician

130

ALLEN BUCHANAN

have as their only goal or as their top priority the maximization of their health or longevity. People frequently act in ways that are not calculated to maximize health or length of life, and often there is nothing irrational about this. For example, one may prefer a somewhat shorter but more interesting and productive life to a longer one. Even more obviously, some individuals do not value an extension of life if this comes at the price of having a poor quality of life, where this involves suffering and severe disability. So, because health and longevity are not the only values individuals have (or should have), an individual’s good cannot be reduced to his medical good. And because an individual’s good cannot be reduced to his medical good, a physician’s medical expertise does not make him the best judge of a patient’s best interest. In addition, it would be a mistake even to say that whenever one engages a physician one thereby authorizes him to do whatever is in one’s medical interests. In some cases, one engages a physician to gain information, not for treatment. It appears, then, that the medical paternalist is making one or the other of two remarkable blunders: he is failing to see that an individual’s overall good is not reducible to his medical good, or he is assuming that the physician, not the patient, is the best judge of the patient’s overall good. And given (b), the fact that the same asymmetry of knowledge that makes it rational to engage a physician as one’s agent also creates the possibility that the physician will take advantage of his patient, a blanket authorization that includes the right to deceive the patient would make it even more difficult to know when the physician was really acting as he ought. In brief, a social practice according to which simply engaging a physician constitutes an authorization for the physician to withhold or manipulate information would clearly exacerbate the conflict of interest inevitable in any principal/agent relationship. Notice also that fact (b) is hardly a new discovery of bioethicists criticizing medical paternalism. On the contrary, it can be argued that much of the content of codes of formal ethics for physicians can only be understood as admonitions against using the asymmetry of knowledge to which (b) refers to the detriment of patients —for example, by convincing them to pay for treatments that the physician knows are inefficacious. In other words, a major impetus for such codes of ethics is the fact that the asymmetry of knowledge between agent and principal creates the possibility that the agent will act in his own interest at the expense of the principal’s interest.6 But if this is so, then it is puzzling that anyone who understands that this conflict of interest is built into the principal/agent relationship would hold that by merely engaging the physician as one’s agent, one thereby authorizes him to withhold information that may be vital for determining whether one is being taken advantage of. 6 Allen Buchanan, “Toward a Theory of the Ethics of Bureaucratic Organizations,” Business Ethics Quarterly 6, no. 4 (1996): 419–40.

SOCIAL MORAL EPISTEMOLOGY

131

The second chief argument for medical paternalism dispenses with the idea of tacit authorization and appeals directly to the proposition that the overriding principle of the medical professional is the fiduciary obligation to act in the best interests of his patient. According to some medical paternalists, adherence to this principle is the essence of medical ethics. But of course, this principle is utterly indefensible if interpreted in the way it must be interpreted if it is to serve as a justification for medical paternalism —namely, as a license to withhold information or preempt patient choice. For if it is to serve as a justification for medical paternalism, the fiduciary principle must be understood as not being limited by a competent patient’s right of informed consent. But allowing the physician’s pursuit of the patient’s good to override the choice of the patient would only make sense if (i) the physician were more knowledgeable about the patient’s good than the patient, or (ii) the patient were emotionally incapable of making a rational decision about treatment. We have already seen that (i) is extremely implausible unless one makes the mistake of conflating the patient’s overall good with his medical good (understood narrowly as giving absolute priority to maximization of health and/or longevity). (ii) is just as indefensible: it is simply false to say that patients generally, or even desperately ill patients, are not emotionally capable of making decisions about their own treatment. Here it is important to note that the medical paternalist does not restrict the physician’s license to withhold information or preempt choice to the physician’s dealings with incompetent patients —that is, those who lack decisional capacity due to mental illness, cognitive impairment, or immaturity. Indeed, what is striking about medical paternalists is their complete failure to address the question of patient competence. Instead of taking seriously the threshold question of whether the patient has decisional capacity, medical paternalists typically appealed to vast generalizations about the inability of otherwise competent individuals to handle bad medical news. Yet the “evidence” cited to support these generalizations was remarkable flimsy, purely anecdotal, and disconfirmed by studies that showed that most otherwise competent patients can handle bad medical news when it is presented in a sensitive but accurate manner.7 Having easily demolished the transparently unsound arguments invoked to support medical paternalism, applied ethicists thought their work was complete, except for the educational task of informing physicians as to competent patients’ right of informed consent. Yet this limited response to the problem of medical paternalism was inadequate for two reasons. First, the critique of medical-paternalist arguments apparently did not have much direct effect on the behavior of physicians. Applied ethicists were indignant when they continued to encounter hardcore med7 See, for example, Donald Oken, “What to Tell Cancer Patients: A Study of Medical Attitudes,” Journal of the American Medical Association 175 (1961): 1120–28.

132

ALLEN BUCHANAN

ical paternalists who somehow remained unconvinced by the antipaternalist arguments. Second, the standard applied-ethics approach shed no light whatsoever on a disturbing puzzle: Why did otherwise highly educated and intelligent individuals who routinely engaged in sophisticated chains of reasoning regarding scientific matters offer such transparently defective arguments to justify their paternalistic behavior? How could individuals trained in the scientific method espouse such broad generalizations about the medical decision-making incapacity of otherwise cognitively unimpaired persons in the absence of evidence to support them? And how could they mistake medical expertise for expertise in making complex value judgments that require an intimate knowledge of a patient’s life-experience, commitments, and priorities? In other words, what accounts for the selective cognitive disability of paternalistic physicians? (Note that the term “disability” is apt here because what is at issue is a stable defect of reasoning, a disposition to make serious cognitive errors across a wide range of cases in relations with patients.) An initial explanation of this phenomenon of selective cognitive disability might be simply that these physicians exhibited a particular moral vice: they were too quick to accept flimsy arguments that purported to justify a practice that contributed to their domination of the physicianpatient encounter, thus making their work easier by allowing them to avoid situations in which patients might disagree with them. In other words, one might be tempted to assume that to adequately explain the selective cognitive disability of paternalistic physicians, the standard applied-ethics approach of identifying principles and critiquing the arguments offered to support them needs only to be supplemented by a dose of simple virtue ethics: paternalistic physicians experience the disability because they are too quick to believe what it is comfortable for them to believe. They lack the virtue of self-criticism (or impartiality) when dealing with an area of life where some of their important interests are at stake; to put the same point negatively, these physicians exhibit the vice of self-serving rationalization. (By “virtue ethics” here I mean an approach to normative ethics that accords a primary role to the moral virtues, understood as stable dispositions to judge, to act, and to experience various moral emotions such as sympathy, outrage at injustices, and so on. Virtue ethicists define a right action as an action that a morally virtuous agent would perform in the relevant circumstances.) This simple explanation is inadequate, however, and understanding its inadequacy strengthens the case for incorporating social moral epistemology into applied ethics. Simply hypothesizing the vice of self-serving rationalization is insufficient for at least two reasons. First, it does not explain why paternalistic physicians tend to rationalize their paternalistic behavior in the particular way that they do, by appealing to the role of physician. Second, the simple virtue-ethics account does not explain how such a transparently defective rationalization was sustained even after

SOCIAL MORAL EPISTEMOLOGY

133

medical-paternalist arguments were publicly shown to be fallacious. To explain both of these facts, it is necessary to locate medical-paternalist action and the rationalizations for it within an institutional context in which a certain occupational group has attained elite status, thereby receiving special privileges that insulate the group’s members from criticism and that systematically impede the functioning of virtues that may operate quite effectively in areas of the elite’s lives in which the professional role is not implicated. The medical profession is a socially constructed inequality, or rather a cluster of inequalities.8 Physicians as a group have typically enjoyed high social status, a sheltered economic position relatively insulated from market competition, and much-higher-than-average incomes (at least for the last one hundred years). This cluster of advantages is not the product of market forces. On the contrary, it is the result of government-supported monopolistic licensure arrangements and, until very recently, legal prohibitions on advertising, and hence on market competition, among physicians; organized medicine lobbied for these arrangements and prohibitions in the name of maintaining the ethical purity of the profession. A plausible explanation of the selective cognitive disability of medical paternalists would presumably include an account of how the culture of elitism that the cluster of socially constructed inequalities supports tends to disable physicians’ capacity for self-criticism in their professional activities, and at the same time insulates them from criticism by those outside the profession. Something much more general than the distinctive elitism of the medical profession may also play a role in the selective cognitive disability evidenced in medical paternalism. I noted earlier that the rationalization of medical paternalism relies in part on a tendency to exaggerate the domain of physicians’ expertise, resulting in a conflation of the patient’s medical interests and his interests overall. This may be only one instance of a much broader phenomenon that I have elsewhere called “expertise imperialism” —the tendency of experts to appeal to their genuine expertise in one area to justify their exercise of control in areas to which their expertise is in fact irrelevant.9 There are at least two distinct though compatible explanations of the tendency toward expertise imperialism. First, the same dedication to a particular discipline or skill that motivates an individual to become an expert in it may lead him to overestimate that discipline’s or skill’s effectiveness in dealing with problems outside its immediate purview. Second, in a society that relies heavily on the division of labor, expertise is power. An expert who stakes a successful claim to a broader area of 8 Allen E. Buchanan, “Is There a Medical Profession in the House?” in Roy G. Spece, Jr., David S. Shimm, and Allen E. Buchanan, eds., Conflicts of Interest in Clinical Practice and Research (New York: Oxford University Press, 1996), 105–36. 9 Buchanan, “Toward a Theory of the Ethics of Bureaucratic Organizations.”

134

ALLEN BUCHANAN

expertise often garners higher status, commands higher remuneration, and secures a wider field for autonomous action. For example, in a society in which many social problems are medicalized, physicians become all the more powerful. Above all, a successful claim to a domain of expertise insulates the claimant from criticism, by disqualifying nonexperts from being recognized as the sort of persons who are in a position to make credible criticisms of the expert’s behavior. Thus not just medical professionals, but socially recognized experts generally, exhibit a combination of characteristics pregnant with the possibility of self-serving bias. On the one hand, the members of such groups have a common interest in sustaining and, if possible, expanding the privileges that they enjoy; this corporate interest can be in conflict with the interests of other individuals and of the community at large. On the other hand, the privileges that experts enjoy tend to insulate them from external criticisms that might serve as checks on the tendency to indulge in self-serving rationalizations. Hence, the more general phenomenon of expertise imperialism and the particular elitist culture of the medical profession have two things in common: both can exist only where a certain group already reaps the advantages of hierarchical social institutions that give it a special status (that is, both are creatures of social inequality); and both facilitate the systematic disabling of experts’ capacity for self-criticism in areas of activity in which the corporate interests of those experts are at stake. A developed social-moral-epistemological analysis of the phenomenon of medical paternalism would delve more deeply into the relationship between the existence of a medical profession as a distinct and privileged subgroup in the population and the liability among physicians to selective malfunctioning of the virtue of self-criticism. In particular, it would be necessary to explore the ideological resources that such a subgroup commands. For example, it is not difficult to see that the higher status and income of physicians as a professional group equips them to project an image of themselves to the general public that tends to disarm the public’s critical faculties and thereby gives physicians freer reign to act on the tendency toward self-serving rationalization to which all human beings are liable. More specifically, physicians traditionally have benefited from what I have elsewhere called “status trust,” a relative relaxation of critical attitudes that is accorded by the public to a physician as a physician, independently of any evidence-based belief in the competence or integrity of the particular physician as an individual.10 As a socially and economically privileged group, the medical profession has considerable resources with which to present a public image that encourages status trust. Al10 Buchanan, “Is There a Medical Profession in the House?”; Allen Buchanan, “Trust in Managed Care Organizations,” Kennedy Institute of Ethics Journal 10, no. 3 (2000): 189–212.

SOCIAL MORAL EPISTEMOLOGY

135

though status trust undoubtedly brings certain advantages for patients (in particular, it reduces anxiety and minimizes decision costs by picking out a certain group as being trustworthy providers of health-care services), it also creates certain disadvantages. Most obviously, surplus status trust — status-based trust that is accorded to the wrong people or is excessive —can lead patients to acquiesce in unnecessary or overpriced treatments. Furthermore, like the other phenomena discussed above, status trust poses ethical risks to physicians by insulating them from criticism. If patients regard physicians qua physicians as exceptionally trustworthy and selfless promoters of the patient’s good, they may be inhibited from criticizing them, with the result that one important check on physicians’ tendency toward self-serving rationalization will be absent. B. The loving S.S. father A second example illustrating the role of social institutions in disabling the virtues —and thereby explaining why we need to engage in social moral epistemology —is inspired by philosopher Jonathan Glover’s discussion of the limits of sympathy.11 Glover explores what he calls the “moral resources,” universal human capacities that can serve as a constraint on immoral behavior. Among the most important moral resources, according to Glover, is sympathy. However, Glover is quick to point out that in many individuals sympathy is limited, extending only to certain subgroups of human beings, not to all. Thus, the same man may be a loving father who cares deeply about his own children and an S.S. guard at a death camp who routinely commits atrocities against other children. Glover is right to insist that great immoralities are often, perhaps usually, committed by persons who are not moral blank slates or sociopaths, but who possess moral virtues that function well in some but not all areas of their lives. He is also correct to suggest that false beliefs play a role in limiting the operation of the moral resources. For example, Glover suggests, unsurprisingly, that Nazi indoctrination was designed to instill the belief that Jews are subhuman and hence not worthy of sympathy, and to create a Nazi “moral identity” that included the belief that it is necessary to override or even extirpate misplaced feelings of sympathy that had been inculcated by Christianity. However, Glover’s analysis is incomplete, because it does not seriously address the role that social practices and institutions play in inculcating the beliefs that systematically limit sympathy and thereby enable immoral behavior. Here I can only sketch some of the main features that such an analysis would include. 11 Jonathan Glover, Humanity: A Moral History of the Twentieth Century (New Haven, CT: Yale University Press, 2000).

136

ALLEN BUCHANAN

First, as in the case of the social-moral-epistemological analysis of medical paternalism, social inequalities, including those that attach to individuals recognized as scientific experts, undoubtedly played a role in the Nazi atrocities. The Nazi effort to “harden” Germans against the Jews —to place Jews beyond the domain of sympathy —relied heavily upon a pseudoscientific racialism that in turn was promulgated by playing upon ordinary people’s status trust in scientists and physicians. In support of the “euthanasia” program that preceded the Holocaust, propaganda films and school textbooks presented revered scientists “exposing” the false humanitarianism of sympathy toward “defective types” who were at best parasites on the social body and at worst carriers of degeneration.12 The cooptation of the German scientific community and above all of the medical establishment played an important role in the systematic disabling of the virtue of sympathy among Germans.13 Educational institutions and the government-controlled press not only identified certain individuals as authorities, but also prevented counterauthorities from voicing dissent from the official message. Social practices, including preeminently the social recognition of authorities, contributed to the formation of beliefs that systematically disabled the virtues of sympathy and respect, and thereby contributed to grievous moral wrongs. Reflection on the Nazis’ use of hierarchical social institutions to cultivate virtue-disabling beliefs highlights the importance of what might be called “epistemic deference” as well as such deference’s dependence upon hierarchical institutions —in particular, upon the social recognition of authorities. Epistemic deference may be defined as the disposition to regard some other person or group of persons as especially reliable sources of truths. Social institutions that recognize some persons as experts encourage this sort of deference. Indeed, it is difficult to imagine how the Nazis could have succeeded so well in disabling the virtues of so many people had the Nazis not been operating in a society in which there was so much surplus epistemic deference to governmental, educational, and scientific authorities. (By “surplus epistemic deference” I mean either deference that is misplaced, as when a person or a group is regarded as a reliable source of truths about matters on which he or it is not, or excessive, as when there is an overestimation of a person’s or group’s reliability as a source of truths.) And in fact, historians of the Third Reich often emphasize that German society in the decades before Hitler was exceptionally authoritarian, and more specifically, rife with patriarchal authoritarian-

12 “Genocide,” in The World At War documentary series (London: Thames Television, 1975). 13 Robert Jay Lifton, The Nazi Doctors: Medical Killing and the Psychology of Genocide (New York: Basic Books, 1976); David Proctor, Racial Hygiene: Medicine under the Nazis (Cambridge, MA: Harvard University Press, 1989).

SOCIAL MORAL EPISTEMOLOGY

137

ism. In such a society, epistemic deference can become an epistemic vice — that is, a disposition to form, sustain, and transmit false beliefs. Of course, all societies include hierarchical institutions, and all societies that exhibit a division of labor recognize authorities. Moreover, the right sort of deference to the right sorts of authorities is essential for the efficient creation, preservation, and transmission of true beliefs. Because of this, social epistemology must include an account of how reliable authorities are to be identified.14 Yet the social inequality that inevitably attaches to the recognition of authorities also carries with it the possibility of there being surplus status trust and surplus epistemic deference, which lead to the systematic disabling of both the virtues of persons generally and the capacities for self-criticism in those who are designated as authorities. A developed social moral epistemology would delineate the optimal balance between epistemic deference and a “critical epistemic egalitarianism” that challenges putative authorities. More generally, social moral epistemology should explore the ways in which different institutional embodiments of the division of labor either minimize or exacerbate the risks of surplus status trust, surplus epistemic deference, and expertise imperialism. It may turn out, for example, that a fairly widespread attitude of critical epistemic egalitarianism (within certain limits) is extremely important for keeping epistemic deference toward authorities from reaching excessive levels, and that this epistemic attitude can only flourish within certain institutional frameworks. Because it is oblivious to these issues of social moral epistemology, the standard applied-ethics exercise of exposing defective moral principles and unsound arguments fails to capture the full complexity of the sources of moral failure. To say that the problem with the Nazis was that they were mistaken in their moral principles or that they got the wrong principles because they were inept at moral argumentation would be woefully inadequate. In fact, it is not too much of an exaggeration to say that the core moral pathology of Germans in the Third Reich was not that they lacked sound moral principles, but rather that they systematically and grossly misapplied them, in part because their leaders artfully employed existing social institutions to spread a web of false beliefs that excluded whole segments of humanity from the domain of moral subjects. (Of course, this is not to deny that those leaders built upon systematic distortions of belief that existed in the anti-Semitism that was already prominent in German culture.)15 In virtue-ethics terms, then, the flaw of most Germans who participated in or supported the Holocaust was not that they were amoralists bereft of virtues, but that the virtues they possessed were truncated. A web of false 14

For one such account, see Goldman, Knowledge in a Social World, 267–71, 304–11. On this point, see Daniel Jonah Goldhagen, Hitler’s Willing Executioners: Ordinary Germans and the Holocaust (New York: Knopf, 1997). 15

138

ALLEN BUCHANAN

beliefs, inculcated and sustained by social practices and institutions, systematically disabled the proper functioning of the virtues (or, if you prefer, resulted in people having truncated facsimiles of the virtues). The case of medical paternalism and that of the loving S.S. father both illustrate the importance for social moral epistemology —and for applied ethics —of the concept of epistemic vices, ingrained dispositions to form, sustain, and transmit false beliefs, particularly false beliefs that play a role in disabling the moral virtues. Social moral epistemology aims to uncover the connections between (1) immoral behavior, (2) the malfunctioning of the virtues, (3) the role of epistemic vices in disabling the moral virtues, and (4) the social practices and institutions that nurture epistemic vices or fail to provide checks on them. What the medical-paternalism case and the S.S. father case suggest is that (a) the epistemic vices include selfserving rationalization and surplus epistemic deference, and that (b) social inequalities —in particular, those that support the synergism between status trust, surplus epistemic deference, and expertise imperialism —can play a significant role in the disabling of the moral virtues. C. Sexist education and racism In the past two decades, studies have shown that false generalizations about natural differences between males and females not only contribute directly to sexist discrimination by providing justifications for it, but also contribute indirectly by helping to create discriminatory practices and institutions that produce experiences that appear to confirm the generalizations themselves.16 Teachers’ attitudes toward the mathematical abilities of males and females illustrate the point vividly. Teachers have tended to believe that boys naturally have greater mathematical ability than girls do. This belief often leads teachers to spend less time with girls in math classes, with the result that girls’ math skills are developed less than boys’ are.17 Then, when girls score lower on math tests, the generalization that they have less aptitude for math appears to be confirmed. (It may also be the case that in like fashion, teachers in some cases spend more time teaching literature to girls, because they believe girls are naturally better suited to this pursuit than boys are.) Sexist educational practices produce experiences that reinforce the beliefs invoked to justify those practices; these experiences thereby lend legitimacy to the practices themselves. As with the cases discussed above, the inadequacy of the standard applied-ethics approach is starkly evident here. It would be misleading, or at least seriously incomplete, to say that the problem with a sexist teacher is that she violates the principle of equal treatment for equals. 16 See, for example, Gilah C. Leder, “Equity in the Mathematics Classroom: Beyond the Rhetoric,” in Lesley H. Parker, Léonie J. Rennie, and Barry J. Fraser, eds., Gender, Science, and Mathematics: Shortening the Shadow (Dordrecht, The Netherlands: Kluwer, 1996), 95–104. 17 Ibid.

SOCIAL MORAL EPISTEMOLOGY

139

Instead, the problem is that she fails to apply the principle properly because she applies it in an environment in which sexist practices have systematically distorted her experience in such a way that this experience seems to support the belief that girls and boys are not equals when it comes to mathematical ability. Similarly, in a racist society in which blacks are systematically deprived of opportunities for developing white-collar work skills or sophisticated mechanical skills, experience will appear to confirm the generalization that they are only fit for menial labor. In the case of medical paternalism and that of the loving S.S. father, social practices and institutions contribute to morally wrong behavior by nurturing epistemic vices that disable the moral virtues. In the case of sexist educational discrimination or racism, wrong acts may not result from epistemic vices (such as self-serving rationalization or excessive epistemic deference), but from one’s distorted experience. This distorted experience, in turn, results from the fact that false beliefs about natural differences between boys and girls or between races shape educational practices. Though the case of sexism or racism differs from the other cases in this way, in all three cases social practices and institutions disable moral virtues by adversely affecting the formation, preservation, and transmission of true beliefs. It is important to emphasize that the connection between social practices and institutions and impairment of the virtues is often mediated not just by false beliefs, but by the dispositions I referred to earlier as epistemic vices. Consider again the case of racist or sexist beliefs. A racist believes that blacks are naturally inferior to whites in certain respects — less intelligent, less able to control their passions, and so on. A racist culture, through the social practices and institutions by which people become members of the cultural group, does more than just help to instill these false beliefs in individuals; it also inculcates or at least supports certain epistemic vices that enable those individuals to sustain those beliefs in the face of evidence that seems to contradict them. For example, I grew up in the American South at a time when it was literally an apartheid society: blacks had separate (and grossly inferior) schools, separate restrooms, and so on, and interracial marriages were prohibited by law. Like others in this society, I was taught by various authority figures (parents, pastors, teachers, etc.) that blacks are mentally inferior to whites. I have already noted how, in a racist society, the systematic exclusion of blacks from white-collar jobs reinforces the belief that blacks are fit only for menial tasks. Such systematic exclusion was clearly present in the American South. As a result of this structural inequality, which greatly reduced the chances that whites would have experiences that revealed the full mental capacities of blacks, the belief in the mental inferiority of blacks was deeply entrenched. As a result, when a particular black person clearly exhibited extraordinary mental abilities, instead of abandoning their belief in blacks’ natural inferiority, whites simply appealed to an-

140

ALLEN BUCHANAN

other strand in their complex web of racist beliefs: the notion that the presence of some “white blood” in a black person could raise him above the low position of blacks generally. Thus, instead of regarding the experience of seeing extraordinary mental abilities in a particular black person as disconfirming the universalization that blacks are mentally inferior, whites “saved the hypothesis” by assuming that that black person “must have some white blood.” Similarly, women who excel at sports formerly reserved for men have sometimes been explained away as being androgynous. These examples illustrate a crucial point. The false beliefs that disable virtues are typically elements of a web of mutually supporting false beliefs. Because of the existence of epistemic vices —including the willingness to preserve a belief, in the face of evidence that appears to clearly contradict it, by relying on another belief that is equally unwarranted — rehabilitating truncated virtues will typically require more than just providing good evidence that a particular belief is false. Instead, the process will require attacking a large set of mutually supporting beliefs, and this in turn may require discrediting the putative authorities (whether they are persons or sacred texts) that are invoked to justify those beliefs. It follows that applied ethics will have to encompass social criticism, not just in the sense of critically evaluating social policies, but by scrutinizing the social practices and institutions that foster mutually supportive false beliefs.

III. The Limitations of Virtue Ethics To their credit, virtue ethicists have shown that exclusive preoccupation with moral principles and arguments to support them misses the central role of stable character traits in the moral life. However, efforts to develop a virtue-ethics approach in contemporary applied ethics have been sporadic and incomplete, and have failed to achieve a fundamental transformation of the field. Although the contemporary virtue-ethics approach to applied ethics is a great improvement over the standard applied-ethics focus on principles and arguments, it still represents an impoverished conception of applied ethics because it fails to take seriously the role that social practices and institutions play in generating and sustaining beliefs that either enable or disable the virtues. In other words, because the goal of applied ethics is right action, we cannot ignore the relationship between the virtues and social practices and institutions. Even if the virtue ethicist distinguishes between having a genuine moral virtue (say, sympathy) and having a truncated or disabled pseudovirtue, and emphasizes, in her account of the genuine virtues, the role of true belief and of dispositions to form and sustain true beliefs, this is not enough. It is also necessary to provide an account of the various ways in

SOCIAL MORAL EPISTEMOLOGY

141

which social practices and institutions support or undermine the virtues in their genuine or fully developed form. At this point, the virtue ethicist might object that by emphasizing the role of moral education, virtue ethics does accommodate the notion that social practices and institutions influence the functioning of the virtues. After all, moral education does not occur in a vacuum: it takes place within a web of social practices and institutions. Two facts make this reply less than convincing. First, although contemporary virtue ethicists typically do note the importance of moral education in cultivating the virtues, they have little or nothing to say about how different social practices and institutions impede or promote the formation of the virtues. Second, what is needed is not merely an account of how a person comes to have a virtue —a theory of primary moral education, as it were. What is also required is an explanation of how social practices and institutions can disable the virtues in individuals who have undergone largely successful primary moral education; one way they can do this, for example, is by producing selective cognitive failures among elites, as in the case of medical paternalism. It is true that in a thoroughly sexist or racist society, even primary moral education can be impaired by distorted experience. But as the medical-paternalist case indicates, even those who have developed their basic moral powers without initial distortions can later suffer impairments as they move into areas of the social structure that make them especially vulnerable to certain epistemic vices, or that make them prone to epistemic vices in certain areas of their activities. IV. The Social-Moral-Epistemological Argument for Liberal Institutions John Stuart Mill famously argued that liberal institutions have important epistemic advantages.18 Freedom of information and expression, along with legal and social permission to engage in diverse lifestyles, enables individuals to form true beliefs about what is most conducive to their own good. What might be called Mill’s social value epistemology rests on his empiricist, individualistic view of the good. We must, he thinks, discover through experience (broadly interpreted to include the testimony and example of others) what is best for us, and liberal institutions provide the most reliable framework for such discovery. Moreover, what is good for one may not be good for another (even if there are certain conditions that are necessary if each of us is to be able to find out what our distinctive good is). Social moral epistemology, as I conceive it, differs from Mill’s empiricist social value epistemology in three respects. First, social moral epis18

See John Stuart Mill, On Liberty (Indianapolis, IN: Bobbs-Merrill, 1956).

142

ALLEN BUCHANAN

temology is concerned with morality, not with value generally. Second, and more important, social moral epistemology is neutral on both the question of value-empiricism and the question of the individuality of the good. The value of the social-moral-epistemology enterprise does not depend upon the assumption that our knowledge of moral values is radically empirical in the way Mill thought our knowledge of the good is, but it does not deny radical value-empiricism either. Nor does the value of social moral epistemology depend upon the assumption that there is great diversity as to the good for persons. Third, social moral epistemology, unlike Mill’s view, focuses not only on the role that social practices and institutions play in generating truth or exposing falsehoods, but also on the role that truth and falsehood play in the promotion or disabling of the moral virtues or the moral powers generally. In spite of these differences, there is a fundamental similarity between Mill’s empirical value epistemology and social moral epistemology: both provide materials for a powerful argument for liberal institutions and attitudes. Mill argues that the distinctive liberal institutions provide the best conditions for the reliable formation of true beliefs about individual good. The sample social-moral-epistemological analyses sketched in Section III suggest that those same institutions tend to promote the proper functioning of the moral virtues. Freedom of information and expression, for instance, makes it more difficult for those in positions of authority to use social institutions to foster beliefs that disable the moral virtues. For example, had Germany had a free press, it would have been more difficult for Nazi propagandists to foster the belief that Jews were dangerous, subhuman vermin not entitled to respect or sympathy; a free press would also have made it more difficult for entrepreneurs of ethnonational conflict in Croatia to instill the belief that Serbs were beasts bent on exterminating Croatians. Similarly, characteristic liberal attitudes, including that fundamental egalitarianism that is suspicious of deference to authority and stingy in bestowing status trust, also provide significant protections against the disabling of the virtues by surplus epistemic deference and expertise imperialism. It is something of a truism that liberal societies allow people the freedom to pursue diverse conceptions of human excellence, and that this precludes the use of common institutions to instill virtue in citizens. However, this statement is misleading so far as it suggests that liberal institutions do not themselves foster moral virtue. Once the social-moralepistemological argument for liberal institutions is appreciated, it becomes clear that liberal society nurtures virtue in important though indirect ways —not by enforcing or privileging a particular conception of what the virtuous life is, but by reducing the risk that the moral virtues will be systematically disabled by false beliefs. The social moral epistemology sketched in this essay also suggests another quite general conclusion about the relationship between the proper

SOCIAL MORAL EPISTEMOLOGY

143

functioning of the moral powers and the character of the social institutions within which they are exercised: other things being equal, the more inegalitarian a society is, the greater the risk of there being surplus status trust and surplus epistemic deference generally, and of there being selective cognitive disabilities among those at the top of the social hierarchy. In other words, as the foregoing analysis of medical paternalism indicated, institutionalized inequalities may foster epistemic vices that can impair moral judgment. This is not an argument for equality. It is, however, an argument to show that in addition to the other costs of inequality, there may be epistemic costs that have serious moral implications. Interestingly enough, the connection between social inequality and the impaired functioning of the moral virtues was explicitly identified by de Tocqueville 170 years ago. In Democracy in America, he quotes at length a letter written by an aristocratic woman in prerevolutionary France to illustrate how extreme social inequalities result in what I have called “truncated” moral virtues. As a digression in a warm and caring letter to a person of her own class, Madame de Sevigne describes in a horribly insensitive, almost frivolous way the barbaric punishments meted out to townspeople who protested against a tax increase: My God, my daughter, how amusing is your letter from Aix! At least re-read your letters before sending them. Allow yourself to be surprised by their charm. . . . Would you like to know the news from Rennes? . . . The day before yesterday they broke on the wheel the violinist who started the dance [of resistance] . . . ; he was quartered, and his four quarters exposed in the four corners of the town. Sixty bourgeois were seized, and they will begin hanging them tomorrow. This province is a fine example for the others, and especially to [encourage people to] respect the governors and their wives, and not to throw stones in their garden. Mme de Tarente was in her woods yesterday during enchanting weather.19 Commenting on this letter, de Tocqueville states: One would be wrong to believe that Mme de Sevigne, who wrote these lines, was a selfish and barbaric creature: she loved her children with passion and showed herself very sensitive to the sorrows of her friends. . . . But Mme de Sevigne had no clear conception of what it was to suffer if one was not a noble.20 19 Alexis de Tocqueville, Democracy in America, ed. Sanford Kessler, trans. Stephen Grant (Indianapolis, IN: Hackett, 2000), 250. 20 Ibid., 251.

144

ALLEN BUCHANAN

He then offers an explanation of this truncation of the virtue of sympathy: extreme inequality, the vast social distance between Madame de Sevigne and the persons whose agonies she so lightly describes, limits the reach of her sympathy. De Tocqueville goes on to remark that things are quite different in an egalitarian society: In our day, the hardest man, writing to the most unfeeling person, would not dare to indulge cold-bloodedly in the cruel jesting that I have just quoted. . . . What does this [relative restraint] come from? Have we more sensitivity than our forefathers? I do not know, but undoubtedly, our sensitivity is brought to bear on more objects. When ranks are almost equal among a people, with all men having more or less the same manner of thinking and feeling, each of them can judge in an instant the feelings of all the others: he casts a rapid glance at himself; that suffices for him. There is thus no misery that he cannot easily conceive of and whose dimensions are not revealed to him by a secret instinct. It does not matter whether it is a question of strangers or enemies: his imagination puts him immediately in their place. It mixes something personal with his pity and makes him suffer himself while the body of someone like him is torn apart.21 On de Tocqueville’s view, one is capable of feeling sympathy only for those whose suffering one can vividly imagine, and one can only vividly imagine the suffering of those whom one recognizes as being like oneself. Extreme social inequality makes members of different classes alien to each other, thereby making them unable to imagine vividly each other’s suffering: this truncates the virtue of sympathy. Clearly, not all social inequalities have this effect. Social moral epistemology should explore the effects of various forms and degrees of inequality on the functioning of the moral powers. V. Social Moral Epistemology and the Reflective-Equilibrium Method Proponents of the reflective-equilibrium method (which includes just about everyone working in normative ethics today) say little about the social framework within which the method is inevitably employed. But if that social framework includes practices or institutions that foster epistemic vices or in other ways produce or sustain false beliefs that disable the moral virtues, then there will be systematic distortions in the moral in21

Ibid.

SOCIAL MORAL EPISTEMOLOGY

145

tuitions of persons existing within that framework. Consequently, the results of the reflective-equilibrium method may be unreliable. The central point here is that the moral intuitions employed in the reflective-equilibrium method are not free-floating. They are embedded in those complex dispositions to feel, to judge, and to act that are called the virtues. If the virtues are disabled as a result of epistemic vices or selective cognitive disabilities, the moral intuitions of those with disabled virtues will not provide reliable data for the process of attempting to reach a reflective equilibrium between principles and particular considered moral judgments. Critics of the reflective-equilibrium method have often complained that the method is not reliable because our intuitions are “culturally bound” — the implication being that “our” reflective equilibrium may well be parochial, at best a coherent representation of the views of the particular group we happen to belong to. This objection is far from lethal, however. Even though it is true that our moral powers are developed within a particular society, it does not follow from this that our moral intuitions must be “culturally bound” in the sense of being parochial, for the simple reason that modern human life is rife with examples of individuals from one culture coming to have new beliefs, attitudes, and even basic values as a result of exposure to other cultures. The more serious challenge the reflective-equilibrium method must meet is this: Are the social practices and institutions that influence one’s moral intuitions epistemically reliable, at least so far as the beliefs that influence those intuitions are concerned? The better our social framework withstands the critical scrutiny of social-moral-epistemological analysis, the more confidence we should have in our own use of the reflectiveequilibrium method. The less adequate the social framework of another culture is from the standpoint of social-moral-epistemological analysis, the less we should worry when intuitions fostered by that framework conflict with our own. It might be argued, however, that social practices and institutions that foster epistemic vices or otherwise disable the virtues will also impair the ability of the social moral epistemologist to recognize these problems in his own social framework. In other words, a defective social framework may make it impossible for a social moral epistemologist working within it to identify its defects. My sketch of a social-moral-epistemological analysis of medical paternalism suggests that this need not be so. The moral disabilities fostered by social practices and institutions are often selective, not only in the sense that they might operate only in one area of an individual’s life, but also in the sense that they are not distributed equally among all members of society. Even if it may be especially difficult for members of a medical elite or socially recognized authorities in some other area to detect the flaws in their self-serving rationalizations or to recognize their own ex-

146

ALLEN BUCHANAN

pertise imperialism, these moral-epistemic defects may be quite transparent to those who occupy another location in the social structure. Though I cannot pursue the suggestion here, I surmise that a developed social moral epistemology would give much greater plausibility to —yet at the same time strongly qualify —the view of critical race theorists and feminist social epistemologists that individuals who occupy positions of subordination in racist or sexist social orders are epistemically privileged. Suppose, as I have argued, that social privileges and higher status can make one more vulnerable to certain epistemic vices, by creating a corporate interest that promotes expertise imperialism and self-serving rationalization, and, at the same time, by insulating one from criticism by others. If this is the case, then the idea that those who are socially disadvantaged by hierarchical institutions are epistemically privileged makes perfectly good sense. Epistemic privilege, on the view I have put forth, consists largely in the absence of those selective epistemic disabilities to which persons occupying certain elevated social positions are liable. Notice, however, that the term “epistemic privilege” here may be misleading. A member of an oppressed group may be free of the particular epistemic defects to which her oppressors are prone, but this is not to say that she is free of epistemic defects. It may turn out that those at the bottom of the social hierarchy merely suffer from different epistemic defects that are equally debilitating. Hence, it is misleading to say that those in disadvantaged or oppressed conditions are epistemically privileged simpliciter. It might be more accurate to say that such individuals enjoy a local epistemic privilege. In Section IV I offered the broad outlines of a social-moral-epistemological argument for liberal institutions and attitudes. That argument has an interesting implication when it is combined with the social-moralepistemological account of local epistemic privilege as being freedom from those selective epistemic disabilities that are due to a special social status. Individuals who are excluded from positions of power and authority will not be liable, in any society, to the epistemic vices that afflict their “superiors,” but only in liberal societies are their views likely to have an impact. In a society in which there are formal or economic restrictions on freedom of expression, the (locally) epistemically privileged members of disadvantaged groups will not be able to express their views. Furthermore, in a society in which the attitude of epistemic egalitarianism is not widespread, those who are (locally) epistemically privileged by virtue of their inferior social position will not be treated as credible sources of truth, even if they are able to express their views. In societies with these features, then, these individuals’ views will not serve as a corrective to systematically false beliefs in the way they may in a society that allows for freedom of expression and in which a substantial degree of epistemic egalitarianism is embedded in the culture.

SOCIAL MORAL EPISTEMOLOGY

147

VI. The Limits of Metaethical Communitarianism “Metaethical communitarianism” may be defined as the view that we come to know what our obligations (and rights) are by locating ourselves within a framework of social practices and institutions, a framework understood as the embodiment of the tradition of a community.22 Thus, I come to know what I ought to do in a certain situation by recognizing, for example, that I am a father. Under metaethical communitarianism, social knowledge about oneself —roughly, about the nature of one’s roles as defined in an institutionally embodied tradition —carries with it (or perhaps even constitutes) moral knowledge. This view might be confused with a much weaker and more plausible thesis —namely, that in general, knowing one’s social roles and their presumptive obligations is necessary for determining what one ought to do. On this view, a responsible moral agent, in determining what he ought to do in a particular situation, will generally have to take into account the distinctive obligations that are imputed to the various roles he occupies. This is a much weaker thesis than metaethical communitarianism, for two reasons. First, it is compatible with a proper knowledge of one’s social roles not always yielding an answer to the question of what one ought to do. (Thus Achilles, situated in social institutions that embody a “heroic” moral tradition, may know that he must show prideful indignation toward his treatment by Agamemnon, and that he should provide leadership for the Greek forces, but he may not be able to determine how to balance these role-constituted obligations. “How long should I sulk in my tent?” he may ask himself.) 23 Second, the weaker thesis about the relation between moral knowledge and knowledge of one’s social roles is compatible with it being the case that in some societies some of one’s social roles require one to do what is immoral. The weaker thesis is also compatible with, but does not entail, the psychological claim that an individual will sometimes not be able to determine what to do without recourse to socially defined roles. Social moral epistemology provides support for the weaker thesis, and at the same time supplies a fundamental criticism of metaethical communitarianism. Philosopher Philip Kitcher has suggested that social epistemology should encompass the idea that traditions can serve as “reliable suppliers of epistemic short-cuts,” just as individual authorities can.24 Kitcher applies this idea to scientific traditions. I want to suggest that it can be extended to traditions as putative sources of moral knowledge. 22 For examples of this view, see Alasdair MacIntyre, “The Virtues, the Unity of a Human Life, and the Concept of a Tradition,” in Michael Sandel, ed., Liberalism and Its Critics (New York: New York University Press, 1984), 125–48; and Michael Sandel, “Justice and the Good,” in Sandel, ed., Liberalism and Its Critics, 159–76. 23 This example was given to me by Julia Annas. 24 Philip Kitcher, “Knowledge and Tradition” (paper presented at conference in honor of Alvin Goldman, Tucson, AZ, February 2000).

148

ALLEN BUCHANAN

Social moral epistemology should explore the features of moral traditions that either promote or hinder their functioning as reliable suppliers of morally relevant truths, and in particular truths that enable the proper functioning of the virtues. To the extent that it is possible to distinguish between a tradition and its current institutional embodiment, social moral epistemology must provide a critical analysis of both. Thus, for example, it may be that some traditions have tended to identify certain individuals (e.g., priests) as preeminent moral authorities, but that different ways of institutionally embedding the role of priest have different costs and benefits from the standpoint of maximizing the morally relevant epistemic virtues of priests and minimizing their morally relevant epistemic vices. A social-moral-epistemological evaluation of a tradition must not confine itself to evaluating the tradition as a set of ideas, attitudes, and abstract processes; it must also examine the concrete institutional embodiments of those various elements. Notice that the applicability of social moral epistemology to moral traditions does not presuppose any sort of metaethical cognitivism, the view that ethical judgments themselves are true or false. A noncognitivist can still explore the role of a tradition (or of social practices and institutions generally) in the formation and preservation of whatever true beliefs are relevant to the proper functioning of the moral virtues. More generally, even if the moral virtues are not themselves sources of moral truths, on the ground that there are no moral truths, it is still the case that the virtues’ proper functioning, on any reasonable account, can be disrupted by false beliefs about nonmoral matters. Unless social-moral-epistemological analysis shows that a particular tradition is a reliable supplier of truths relevant to moral judgment, there is no more reason to regard that tradition as a reliable source of moral knowledge than there is to regard any particular putative scientific tradition as a reliable source of knowledge about the natural world. Therefore, the metaethical-communitarian thesis requires a radical reformulation: traditions can serve as reliable suppliers of knowledge relevant to moral decision-making if they include, or are embedded in, practices that facilitate the formation, preservation, and transmission of beliefs that enable the proper functioning of the moral virtues. It may also be true that something stronger can be said —namely, that without recourse to a moral tradition, an individual is seriously impaired or even fully debilitated as a moral decision-maker, just as it might be said that an individual cannot develop a comprehensive and accurate set of beliefs about the natural world without recourse to a scientific tradition. But this, of course, falls far short of the metaethical-communitarian view that reliance upon a tradition’s conception of one’s roles is a reliable source of moral knowledge. Everything will depend upon whether the tradition in question is a reliable supplier of truths relevant to moral judgment, and this in turn will depend in part upon whether the tradi-

SOCIAL MORAL EPISTEMOLOGY

149

tion, embodied as it actually is in a particular set of practices and institutions, enables or disables the virtues (and fosters or combats various epistemic vices). My purpose here is not to develop a social-moral-epistemological critique of traditions as sources of moral knowledge, but only to indicate that without such an analysis, the claim that any particular tradition is a reliable supplier of moral truths or morally relevant truths is groundless. However, if what I have suggested regarding the moral-epistemic virtues of liberal institutions and attitudes is correct, this much can be said: traditions that are reliable suppliers of moral knowledge or of morally relevant truths will be, other things being equal, those that best approximate an optimal balance between, on the one hand, the efficiencies provided by epistemic deference to the tradition’s authority and, on the other, a critical epistemic egalitarianism that serves to check sources of unreliability in the tradition. In addition, given that those individuals that a tradition identifies as sources of moral authority (or as authoritative interpreters of the tradition) can suffer from the epistemic vices to which authorities generally are liable, some significant degree of freedom of information and expression will be needed for the effective functioning of the attitude of critical epistemic egalitarianism. Thus, rigidly hierarchical, thoroughly illiberal traditions —or traditions embodied in rigidly hierarchical, thoroughly illiberal institutions —are, other things being equal, less reliable sources of moral knowledge than more liberal traditions or traditions situated in more liberal institutions.

VII. The Relationship between Social Moral Epistemology and Moral Theory Social moral epistemology, as I have presented it, is not a substitute for moral theory. It does not attempt to provide a justification for basic moral principles or for a list of the virtues. Instead, it augments moral theorizing, as the latter is usually conceived, in two ways. First, by examining the ways in which different traditions may either facilitate or impede the formation, preservation, and transmission of morally relevant beliefs, social moral epistemology provides a critical tool for increasing the reliability of the reflective-equilibrium method. Second, it furthers applied ethics’ goal of promoting moral behavior by exposing the role that social practices and institutions play in fostering beliefs that impair the functioning of the moral virtues. Clearly, both of these uses assume, but are not themselves intended to justify, the claim that there are such things as right and wrong or that there are genuine moral virtues. All of the examples I have employed in this essay have featured the role of social practices and institutions in fostering false empirical (that is,

150

ALLEN BUCHANAN

factual) beliefs —for example, that ordinary adults are not competent to accept or refuse treatment, that girls naturally have less capacity for mathematics, or that blacks are naturally inferior. My main point has been that wrong behavior often occurs not because people are utterly lacking in the moral virtues or because their most basic moral principles are defective, but because they possess truncated versions of the moral virtues or systematically misapply unexceptionable moral principles because they hold certain false empirical beliefs, beliefs that certain kinds of social practices and institutions are more likely to foster than others. The type of enterprise I am engaged in here might be called “modest social moral epistemology.” It does not address the most fundamental questions of ethical theory —in particular, the question of whether what we commonly take to be basic moral truths, as opposed to empirical ones, are truths. A more radical social moral epistemology would explore the role of social practices and institutions in the formation, preservation, and transmission of the most basic moral beliefs themselves, including those that concern our conceptions of both fundamental moral principles and the virtues. This more ambitious project would presumably have to be preceded by, and rely upon, a cognitivist metaethical theory that provided a justification for a set of basic moral principles and for the claim that some particular set of stable character traits are in fact the moral virtues. Why is this the case? What both the modest and the radical socialmoral-epistemological enterprises have in common is the critical investigation of how social practices and institutions shape morally relevant beliefs. The modest version restricts itself to a comparative evaluation of various social practices and institutions to the extent that they shape the empirical beliefs that are relevant to the functioning of what we commonly take to be the moral powers. The radical version focuses on the social shaping of our most basic beliefs about the fundamentals of morality. For radical social moral epistemology to be a normative enterprise, as opposed to a merely descriptive (i.e., sociological) one, it must rest on the assumption that moral beliefs are true or false; this is because radical social moral epistemology attempts to evaluate social practices and institutions as to how well they facilitate the formation, preservation, and transmission of moral beliefs. Unless moral beliefs are true or false, no sense can be made of evaluating different social arrangements as to how well or badly they foster true moral beliefs. Because the radical socialmoral-epistemological enterprise raises separate and more complex issues than its more moderate counterpart, I have not pursued it in this essay. It is worth emphasizing, however, that the modest social moral epistemology I have sketched in this essay is a valuable enterprise quite apart from the more radical project. Even if it remained restricted to an investigation of the role of social practices and institutions in fostering empirical

SOCIAL MORAL EPISTEMOLOGY

151

beliefs that influence moral performance, social moral epistemology would transform the enterprise of applied ethics. VIII. Conclusions Social epistemology evaluates social practices and institutions in their capacity as sources, preservers, and transmitters of truths. Social moral epistemology, in its most inclusive form, evaluates social practices and institutions in their capacity as sources, preservers, and transmitters of truths that are of special relevance to right action. In this essay, I have focused on the role of social practices and institutions in fostering empirical beliefs that are relevant to moral performance. A more radical social moral epistemology would delve deeper, examining the role of social practices and institutions in the creation, preservation, and transmission of the most basic moral beliefs, including beliefs about what the moral virtues are. One important aspect of the social-moral-epistemological enterprise is examination of the effects of social practices and institutions on the functioning of the virtues. Features of social practices and institutions that encourage epistemic vices and can systematically disable the virtues include surplus status trust, surplus epistemic deference to putative authorities, and expertise imperialism. Other features of social practices and institutions foster the proper functioning of the virtues —for example, by promoting an optimal balance between epistemic deference to authorities or tradition and a critical epistemic egalitarianism that encourages opinions from persons of all social classes and roles and that insulates no person or group from criticism. In the absence of a social-moral-epistemological analysis of the social framework in which the reflective-equilibrium method of doing normative ethics is practiced, one cannot conclusively rebut the objection that this method is unreliable because the intuitions upon which it relies are parochial. Social moral epistemology also provides an indispensable tool for evaluating the claim that socially embedded moral traditions are sources of moral knowledge. At least to the extent that social moral epistemology can determine which sorts of traditions do a better job of creating, preserving, and transmitting empirical beliefs that are relevant to the proper functioning of the moral virtues, it provides an important qualification on the metaethical-communitarian thesis that we learn what is morally required of us by locating ourselves in a tradition. Thus, social moral epistemology is relevant not just to normative ethics, to the extent that that field relies upon the reflective-equilibrium method, but also to the metaethical enterprise. The value of social moral epistemology is perhaps clearest, however, in the case of applied ethics. Applied ethicists have typically confined their efforts to criticizing defective principles and unsound arguments, and to

152

ALLEN BUCHANAN

proposing alternative principles and arguments to support them. If the goal of applied ethics is to promote right action and reduce the incidence of wrong action, then the standard applied-ethics approach is woefully inadequate. In many cases, immoral action results from false beliefs disabling the moral virtues (for example, by truncating the virtue of sympathy or of respect so that it is not extended to some groups of persons). As a result, promoting right action and reducing the incidence of wrong action requires study of the ways in which people come to have and maintain those false beliefs. Virtue ethics commendably stresses the role of the virtues in right action, but neglects the question of which sorts of social practices and institutions facilitate the proper functioning of the virtues. An applied ethics that takes seriously the goal of promoting right action and reducing the incidence of wrong action must develop an account of the relationship between the principles upon which individuals act, the role of the virtues in determining the scope and meaning of those principles, and the effects of social practices and institutions on the formation, preservation, and transmission of the beliefs that are required for the proper functioning of the virtues. Philosophy, University of Arizona