This file is also available in Adobe Acrobat PDF format
ARGUMENTS FOR ADVERSARIES
ETHICS FOR ADVERSARIES is a philosophical inquiry into arguments that are offered to defend adversary roles, practices, and institutions in public and professional life. The adversary professions in law, business, and government typically claim a moral permission to harm others in ways that, if not for the role, would be wrong. I shall argue that the claims of adversary institutions are weaker than supposed and do not justify much of the harm that professional adversaries inflict. Institutions and the roles they create ordinarily cannot mint moral permissions to do what otherwise would be morally prohibited.
Adversary institutions are pervasive, and the arguments offered to justify such arrangements cut across professional boundaries. The most vivid example is the adversary legal system, in which lawyers are permitted, within its rules, to make the case for what they know to be false and to advance causes they know to be unjust. But many other practices invoke some sort of adversary argument for their justification: competitive markets for goods and services, for labor and capital, and for corporate control; internal competition among managers; electoral politics, interest-group pluralism, constitutionally separated powers, and bureaucratic competition; commercial and political advertising, investigative and advocacy journalism, and the marketplace of ideas. The practice of medicine in a for-profit and managed health care system is becoming an adversary institution too. Though the details of these practices and the nuances of the arguments of practitioners vary, I believe that these adversary settings have more in common than is commonly supposed. Though I shall illustrate my view with specific examples drawn from law, business, government, and medicine, my purpose is to develop a general account of "necessary offices" in politics and the professions.
Adversaries can line up a phalanx of arguments to defend their sharp practices, but not all arguments are good ones. For concreteness, consider the claims that might be made by a political candidate and his campaign strategist who, as James Madison puts it, "practice with success the vicious arts," and willfully distort an opponent's record to smear her reputation in the eyes of voters.1 The claims are presented in pairs. The members of each pair are easily conflated, but that would be a mistake.
EXPECTATION: Political opponents expect to be slandered, and voters expect to be deceived.
CONSENT: Political opponents consent to be slandered, and voters consent to be deceived.
RULES OF THE GAME: The rules of the game of politics permit slandering opponents and deceiving voters.
FAIR PLAY: Fairness to players in the game of politics morally permits slandering opponents and deceiving voters.
INCREASED NET BENEFIT: More benefit than burden is caused by political slander and deception.
PARETO SUPERIORITY: No one on balance is burdened, and some benefit, from political slander and deception.
NO DIFFERENCE: If I don't slander my opponent and deceive the voters, someone else will.
SELF-DEFEAT: If I don't slander my opponent and deceive the voters, someone else will slander and deceive them even more viciously.
ROLE OBLIGATION: The rules of the role of campaign strategist require engaging in the vicious arts by any legal means.
MORAL OBLIGATION: Morality requires that campaign strategists obey the rules of the role of campaign strategists.
SELFLESSNESS: A professional strategist serving a candidate should filter out her own self-interest.
PERSON NEUTRALITY: A professional strategist serving a candidate should filter out her own moral judgments.
Now, in each pair, the first claim might hold, but does not justify much. The second claim would justify much, but does not hold. In no pair does the second proposition follow from the first. Substitute claims about other adversary practices, and the upshot is the same. If you need convincing, this book is for you.
Restricted Reasons and Permissible Violation
Adversaries act for by acting against. The way that adversaries act for poses a problem about good reasons: how to justify adopting the partial aims and point of view of the partisan, thereby restricting the range of moral reasons that count in one's deliberations, so that some good moral reasons are excluded or discounted, and others are given priority or magnified. What justifies such a division of moral labor?2 Call this the problem of restricted reasons. The way that adversaries act against poses a problem about right action: how to justify engaging in adversary tactics that harm others, especially actions that, if performed outside of one's adversary role, would wrongfully violate persons or their rights?3 Call this the problem of permissible violation. Adversaries offer arguments in defense of their practices that appeal to restricted reasons and that assert permissions to violate.
The Problem of Restricted Reasons
Professional and political actors occupy roles that often instruct them to work at cross-purposes, furthering incompatible ends and trying to thwart each other's plans. Prosecuting and defending attorneys, Democratic and Republican candidates, secretaries of state and national security advisers, industrial manufacturers and environmental regulators, investigative journalists and official sources, and physicians and insurance companies often are pitted against one another by their missions, jobs, and callings.
Sometimes, when adversaries further conflicting moral ends, one is thought to be right and the other wrong; or, one is thought to act for the better, and the other for the worse. But sometimes the actions of both actors are thought to be for the good; indeed, observers often believe that both actors ought to act as they do, though what one ought to do conflicts with what the other ought to do. But how can two political or professional actors facing the same situation be required to act in opposing ways? How can two adversaries who act to further conflicting purposes both have good enough reasons to do so? Why are the reasons the one has to act not reasons, or not good enough reasons, for the other?
One reply is pervasive in both casual and considered talk in support of adversary institutions. Although the form of the argument varies from practice to practice, the heart of it looks something like this: actors occupying professional or public roles are not to make all-things-considered evaluations about the goodness or rightness of their actions, but rather, they are to act on restricted reasons for action, taking into account only a limited or partial set of values, interests, or facts. Reasons for action are restricted in two ways. First, the professional is exempted from the broadest, most inclusive of deliberative concerns, allowing for specialized moral aims across roles--role relativity. Second, when acting on behalf of others, the professional is precluded from counting the most local of deliberative concerns, requiring uniformity within role--person neutrality. Each adversary actor ought to do so (or, more modestly, is permitted to do so) because, in the aggregate, the institution of multiple actors acting from restricted reasons properly takes into account the expansive set of reasons, values, interests, and facts. The competitive market, the system of legal representation, and electoral politics--each turns for justification to a version of this division-of-moral-labor argument. Adam Smith and James Madison, in different ways, appeal to such a division. Some arguments in support of freedom of expression--notably, that of John Stuart Mill--take a similar form.
These arguments rely on some notion of a favorable equilibrium that will result if adversaries restrict their concerns to narrow aims. The mechanism by which a system is to arrive at this equilibrium varies, and at least three evocative images are used to describe it. The division-of-moral-labor image itself suggests not necessarily conflict between professionals or professions, but rather, selective attention to interests, values, and reasons by specialized actors, whose efforts result in the efficient and possibly harmonious manufacture of social value. The adversary-equipoise image adds to this specialization a contest between identified opponents, in which the aims and efforts of one are poised against the contrary aims and efforts of the other in careful balance, so that if one shirks her part, the favorable equilibrium will be upset. In contrast, Adam Smith's famous invisible-hand image, though it too seeks to justify restricted reasons, suggests the opposite of both specialization and individual importance of actors. In the face of competitive pressure and market reaction, actors are interchangeable, no one has room for successful discretion, and no one's actions make a difference to the outcome.
The favorable claims made for the resulting equilibria also vary. The strongest versions of equilibrium arguments in the professions unconditionally claim that it is better that all actors narrow their reasons for action than if all actors tried to take the broadest range of reasons for action into account. Such first-best claims are often made for competitive markets and for the moral doctrine known as ethical egoism. The conditional division-of-labor argument claims that if some actors pursue restricted or partisan aims, then it is for the best that all do (though it might be better still if none did). The conditional argument often is invoked to justify the training and deployment of soldiers and lawyers. The weakest version claims only that adversary institutions adequately anticipate and neutralize the bad effects of partisan action. Madison, whose arguments for separation of powers and representative government are often misread as championing the strong argument, actually subscribes to the weakest form.4
Arguments for divided moral labor in public and professional life face both a factual and a conceptual challenge. Factually, the equilibrium mechanism by which partial actions are said to serve impartial goods must be specified, and the conditions necessary for a good equilibrium outcome must be shown to hold. What precisely is the route by which manipulative and misleading commercial or political advertising is supposed to lead to market efficiency or legitimate government? Conceptually, it must be shown why a prescription to act from filtered and partial reasons follows from the evaluation that it would be a good state of affairs if practitioners did so. It may be good that diverse, conflicting beliefs are held in the marketplace of ideas, but that doesn't give any actor a good reason to adopt a false belief, and it doesn't justify circulating information one believes to be false. Good forms of social organization do not by themselves dictate the forms of moral reasoning particular actors within institutions ought to employ. The gap between what an institution may allow and what an actor within an institution may do is especially great when the action in question deceives, coerces, or violates persons in other ways.
The Problem of Permissible Violation
Imagine a society, Badland, where people are motivated by self-interest alone, and where everyone pursues his or her own advantage in every interaction with intense vigor. In those pursuits, no one avoids harming another unless there are penalties discouraging such harm, and all craftily engage in manipulation and deception if doing so will advance their ends. In Badland, Avarice talks Bully into buying a worthless painting, Bully dumps toxic waste near Cutthroat's backyard, and Cutthroat refuses to repay the loan borrowed from Avarice. Now, if there are stringent enough rules in place to govern the pursuit of self-interest, Badland might be a just society. Kant held that just laws could be written even for a nation of devils. But Badland would not be a good society--its inhabitants would be vicious, not virtuous, and we would not admire such people or their character.
Across the border, in Roland, people have the same motivations, but do not pursue their own advantage. Rather, each appoints a trustee who pledges to advance the trustor's interests through a blind trust, and each trustor also is someone else's trustee. Exactly the same conflicts are fought, the same manipulations occur, the same harms inflicted, but each actor is acting as a faithful professional in fulfilling obligations to a client. In Roland, Comity sells Arista's worthless painting to Bono, and thinks that she has a duty to Arista to get the best price; Arista lobbies the legislature to pass a law permitting the expeditious disposal of Bono's poisons despite the risk to Comity's health, and believes that it would be wrong not to seek a rule most favorable to the polluter; Bono is required by the rules of his profession to extricate Comity from her debt to Arista through the skillful manipulation of Roland's legal system. The people of Roland believe that it would be wrong not to meet their fiduciary responsibilities, distasteful as they are. Because they devote their days to fulfilling their professional obligations, they pride themselves on their virtue.
The puzzling self-understanding of the inhabitants of Roland (which perhaps is no more puzzling than the self-understanding of our marketers, lobbyists, and litigators) raises what we might call the problem of hired hands: how can a professional have an obligation to do on a client's behalf what would be wrong if done on the professional's own behalf? The answer cannot simply be that the professional has promised. Whether the promissor is a contract killer or a contract liar, a promise to wrong another has no moral force. One response is to redescribe the doing, so that the action of the professional is said to be "fulfilling professional responsibilities" or "realizing social values served by the division of moral labor," rather than "lying," "poisoning," or "stealing." Another is to redescribe the actor so that it is the professional role that performs the nasty acts, not the person who occupies the professional role--a response to the hired-hands problem we might call the no-hands solution.
The problem of hired hands is an instance of a more general problem. How can acts that ordinarily are morally forbidden--violence, deception, coercion--be rendered morally permissible when performed by one who occupies a professional or public role? Occupants of adversary roles claim such a moral permission when the rules of their profession permit, and claim to be morally required to exercise these permissions when the rules of their profession require. But why do the conventional rules of a practice have the power to create moral permissions and requirements? True, adversary roles direct practitioners to filter out moral reasons that count against harming others, but why are practitioners morally allowed, let alone morally required, to follow such directions?
To begin, I examine in some detail a professional role that Montaigne counts among his necessary offices: the executioner of Paris.5 Charles-Henri Sanson is appointed by Louis XVI, and serves the punitive needs of the ancien régime for decades. What becomes of the King's Executioner come the French Revolution? He becomes Citizen Sanson, the king's executioner. Sanson adapts seamlessly to the Revolution and its new technology, the guillotine, and ministers with professional detachment to each defeated political faction throughout the Terror and its aftermath. First, the historical record of what is known about Sanson, how he was viewed, and how he wished to be viewed is reconstructed. Then, the most plausible arguments that might have been made in his defense are constructed. The claims that can be made on Sanson's behalf strikingly resemble the claims made by politicians, bureaucrats, lawyers, business executives, and journalists to justify their commitments to their professional roles when the roles ask them to act in ways that ordinarily would be wrong. By exploring one extraordinary professional career and the arguments from the morality of roles that can be offered in its defense, unsettling doubts are raised about arguments in defense of less sanguinary professions and their practices. These doubts are explored more systematically in subsequent chapters.
The three chapters of Part II, "Roles and Reasons," assess arguments in defense of adversary practices that invoke role as an important moral concept. Roles claim to change the morally apt descriptions of actions (lawyering isn't lying) and to restrict the moral reasons that properly enter into a roleplayer's deliberations (defense attorneys are not to consider the consequences of setting a dangerous offender free). Do these bids to filter descriptions and reasons succeed?
Chapter 3, "Doctor, Schmoctor," introduces the concept of a social and professional role and discusses the various ways that roles can generate moral obligations and permissions. I argue for practice positivism, the idea that the rules of practices, roles, and institutions do not have any necessary moral content--they simply are what they are, not what they morally ought to be. There is nothing incoherent about a role that can be performed well only by being bad, as Sanson demonstrates. Practice positivism has two upshots: first, roles can demand too little, because a practitioner might not have acquired a moral obligation to comply with a good role. This, I argue, is the challenge new institutional arrangements for the delivery of medical expertise, such as health maintenance organizations, present to the traditional doctor-patient relationship. Second, roles can permit too much, because a practitioner can acquire a moral obligation to comply with a bad role--or so it might seem.
Roles characteristically claim to generate moral prescriptions that vary from professional role to role (role-relative prescriptions) but that do not vary by the personal attributes of those who occupy the role (person-neutral prescriptions). Chapter 4, "The Remains of the Role," explores the argument, often heard from British civil servants, that the demands of person neutrality properly filter out the substantive moral objections that persons occupying roles might have to what the role requires. The standard view of the civil service is compared with a butler's view of private service in Ishiguro's novel, The Remains of the Day. I argue that public servants must make political philosophical judgments about both the justice and legitimacy of public policies, and that sometimes those judgments will justify disloyalty and disobedience. The political actor must not defer to the authority of his role obligations without exercising judgment about the legitimacy of the role or of the content of the actions it prescribes.
Chapter 5, "Are Lawyers Liars?" examines the claim that professional practices create new ways of acting that can be judged only by the rules of the practice. If this strategy of redescription succeeds, then the central question of this book is misposed. To the question, How can a professional role morally permit actions that otherwise would be morally wrong? the response is, There is no otherwise. The argument of redescription seeks to short-circuit the moral evaluation of both actions and actors by redescribing them in practice-defined terms. But the argument does not work, because act and actor descriptions persist. Whichever way a practice describes an action, preconventional descriptions do not disappear, and so actions can always be evaluated under multiple descriptions. The rules of the practice of business might claim to redescribe the breaking of a promise so that it is no longer the breaking of a promise, but merely the nonperformance of a contract, and the practice of lawyering might claim to redescribe lying so that willfully causing beliefs one knows to be false is no longer a lie, but merely zealous advocacy. Still, these claims fail.
The harms that adversaries cause cannot be redescribed away or filtered out--the claims of the target not to be mistreated pierce the masks that role players wear. The three chapters of Part III, "Games and Violations," consider arguments that address directly the complaints of those harmed by adversary practices and institutions.
In Chapter 6, "Rules of the Game and Fair Play," I assess several arguments claiming that presumptively wrong actions, if permitted by the rules of a game, for that reason are no longer wrong, and so become morally permitted. Arguments from consent and tacit consent are explored, and shown either to fail to justify sharp tactics or to require stringent conditions that are unlikely to be met in practice. The most promising argument in support of at least some deceptive, coercive, and violent practices is grounded, surprisingly, on the principle of fair play, which obligates us to do our fair share in schemes of social cooperation from which we willingly benefit, and not to free-ride on the burdens shouldered by others. The fair-play argument is here employed in a new way, to establish a moral permission that otherwise would not exist, rather than to establish a moral obligation that otherwise would not exist. Necessary and sufficient conditions for the fair-play argument to work in establishing a permission are developed, and the question of whether various games in business, legal practice, and politics meet these conditions is explored.
If some ways of treating others never are morally permissible, however, then arguments in defense of some adversary permissions cannot get off the ground. Chapter 7, "Are Violations of Rights Ever Right?" establishes the conceptual possibility of morally permissible violations. I distinguish violating persons from violating the rights of persons not to be violated. Though it may be a contradiction to violate a right in order to express the inviolable status of rights, persons, not rights, have that status, so it is no contradiction to violate a right in order to express the inviolable status of persons. I explore a number of conditions under which the violation of persons could meet a test of reasonable acceptance--for example, when constraints against violation are self-defeating or Pareto-inferior. But adversary professions and practices typically do not meet these conditions, and so the violations they inflict typically do not meet the test of reasonable acceptance.
Chapter 8, "Ethics in Equilibrium," assesses appeals to the overall good of a system of adversaries pursuing partial and partisan purposes. A few pages ago I sketched the forms that such appeals take--to an invisible hand, to a division of labor, or to equipoise. In this chapter, I examine both the claim that harmful actions taken under an equilibrating mechanism produce good consequences and the claim that such actions pass the test of reasonable acceptance. Adversaries in equilibrium might argue that, because of the institutional structure in which they act, the harms they cause fall on the easier-to-justify side of two distinctions: the difference between intentional and accidental harm and the difference between doing and allowing harm. Both moves fail. The adversary cannot redescribe his aims in a way that makes the violation he inflicts an accidental effect of the overall good aims of the institution. Invisible hands don't violate people--people violate people. And, though the designers and rule makers of an adversary institution can be understood to allow certain activities, practitioners do them, and actions that are not wrong to allow might be wrong to do. Failure to note this asymmetry leads to the conflation of justified forms of social organization with justified forms of moral reasoning by practitioners within institutions.
If institutions may permit or even require actions that practitioners within those institutions ought not to do, then a different sort of adversary relation arises--not one designed by governments, markets, or professions, but one that follows from the conflict between the authority of these institutions and the judgment of practitioners. Part IV, "Authority and Dissent," takes up this conflict.
Chapter 9, "Democratic Legitimacy and Official Discretion," considers the conditions under which the occupant of a political role may take adversary action against superiors or constituents in the face of substantive political disagreements. When may government officials create and exercise effective discretion to pursue policies that dissent from the wishes of superiors or of most citizens? I argue that neither the obedient servant nor the catch-me-if-you-can entrepreneur are proper models of official discretion. Criteria for justified discretion are offered, drawing on accounts of legislative representation and civil disobedience. Conclusion: dissenters ought to make judgments about the common bads or injustices at issue and about the legitimate jurisdiction and the legitimate reasons of the sources of political mandates. These conditions of democratic legitimacy are matched with possible strategies of dissent--persuasive, incentive, or deceptive. In the end, it is not the role-relative prescriptions of divided government institutions that justify adversary action by political actors, but rather, dissenting political judgments about justice and legitimacy.
Chapter 10, "Montaigne's Mistake," illustrates the book's main arguments through an analysis of an event important in shaping American views of authority and dissent under a division of constitutional labor, the conflict between President Richard Nixon and Special Prosecutor Archibald Cox over the Watergate tapes. I assess various ways that a claim for a division of labor in moral reasoning, given an institutional separation of powers, can be pressed, so that both those who brought about Cox's dismissal and those who refused are justified. I conclude that, under plausible assumptions, to obey the president is to violate political liberties in a way that fails to meet any of the tests of morally permissible violation.
I hope that there is something in this book for readers who come to it with a variety of foundational views in moral and political philosophy. Chapter 2 and Part II raise questions and reject answers in a way that travels fairly light. But obviously one cannot be completely ecumenical in one's philosophical commitments and still say something. As will become clear in Parts III and IV, I believe that some version of contractualism owing much to Kant is the right account of moral philosophy, and some version of liberalism owing much to Rawls is the right account of political philosophy. But it is not my project here to demonstrate the correctness of contractualism or liberalism in general. Whatever the correct view is about divisions of moral labor, surely there is room for a division of intellectual labor. I leave those more foundational pursuits to others.6 Nor is it my purpose to enter into intramural discussions about just which formulation of contractualism or liberalism is most promising--on that, except for occasional lapses, I am intentionally noncommittal. I hope that what I have to say about adversary roles and institutions is sufficiently robust to hold water, with some minor plugging, under any plausible version. Rather than argue for, I argue from a contractualist sensibility, in the hope that, if you are not already convinced, you will feel enough of its tug to question whether morality is, at bottom, simply about the summing up of benefits and burdens across persons. If utilitarianism or a straightforward consequentialism is the correct moral theory, then there are no deep moral objections to an adversary profession or institution that, all things considered, breaks even on benefits and burdens.7 If, instead, moral justification is about giving reasons to each person burdened that she, if reasonable, would accept, an appeal to good consequences alone is not likely to meet the test of reasonable acceptance. This book is a search for acceptable reasons adversaries could give to those they deceive, coerce, and otherwise violate. Though there is truth in many arguments for adversaries, these truths are far more limited in scope and setting than is often claimed. The task ahead is to draw these limits.
Return to Book Description
File created: 8/7/2007
Questions and comments to: email@example.com
Princeton University Press