Abstract
This paper defends the possibility of dirty hands against the longstanding skepticism that an action cannot be simultaneously right and wrong and that dirty hands cases are therefore impossible. While skeptics are right to recognize that prima facie reasons against violating moral duties may be overridden, they are wrong to deny that actions required by necessity may nevertheless remain wrong. Dirty hands cases capture the simultaneous necessity of disregarding moral duties in certain circumstances and the reprehensibility of wronging people even in cases in which this is all-things-considered permissible. Rather than indicating any deficiency in our moral reasoning, the capacity to accommodate the possibility of dirty hands, along with the reality of moral conflict, is a strength of a moral theory.
1 Introduction
It is a truth universally acknowledged that an individual wishing to succeed in politics must be willing to lie, cheat, deceive, be ruthless, and break all kinds of moral rules. This rendering of ‘Austen meets Machiavelli’ captures the pervasiveness of the assumption that certain people, politicians leading the way, must do what is morally reprehensible to achieve what is desirable. This applies not to those who are selfish and corrupt, but to those who genuinely wish to do good.
During 9/11 crisis, after the first plane had crashed into the World Trade Center in New York, President Bush ordered any further suspected hijacked planes to be shot down.1 In September 2002, Wolfgang Daschner, the deputy police chief of Frankfurt in Germany, ordered subordinates to threaten to torture a man named Magnus Gäfgen, who, four days earlier, had kidnapped an 11-year-old boy. Gäfgen was refusing to reveal the child’s whereabouts even after hours of interrogation, and Daschner feared that the child’s life was in imminent danger.2 Finally, at the time President Barack Obama was awarded the Nobel Peace Prize in 2009, he was the Commander-in-Chief of a military embroiled in two wars.
May leaders who claim to strive for peace respond to terrorism with terror, safeguard liberal democracy through violent means, or surrender civil liberties for the sake of public health and safety?3 The problem of dirty hands describes the paradoxical idea that, in politics, it may be right to do what is morally wrong.4 The perhaps best-known contemporary formulation of this phenomenon is Michael Walzer’s. His claim that ‘no one succeeds in politics without getting his hands dirty’ reverberates through a voluminous body of literature grappling with the question of why it may be right to violate moral rules in politics.5 As Bernard Williams put the point, ‘[w]e may want – we may morally want – politicians who on some occasions ignore [moral] problems.’6
Such claims have attracted a great deal of criticism, with opponents questioning the coherence of dirty hands.7 While one action may plausibly be right in one context and wrong in another, dirty hands cases purport to be both right and wrong simpliciter, which – so the criticism goes – is logically impossible. As a result, a considerable mass of scholarship is dedicated to untangling the paradoxical nature of the problem – attempting to show, for instance, that there is no real value conflict to begin with, or that the problem’s paradoxical nature is, in the end, just another simple case of some values or duties overriding others.8
This paper defends the possibility of dirty hands cases against longstanding critiques that take issue with dirty hands’ paradoxical nature. I argue that the paradoxical heart of the problem – the wrongness of doing something that is justified, perhaps even required – is neither captured by the notion that the end justifies the means nor the claim that certain political values simply override individual moral ones. This, I contend, holds even for committed consequentialists. Against the view that dirty hands cannot exist because it is logically impossible for something to be both right and wrong simpliciter, I defend a conception of dirty hands cases that is sensitive to the weight of consequentialist demands in the face of conflicting deontological duties. On this view, committing pro tanto wrongs is compatible with acting all-things-considered permissibly; in certain cases, committing pro tanto wrongs may even be required. Dirty hands cases, it follows, are neither logical impossibilities nor reducible to simple value conflicts. Rather, they demand that we view moral theory as considerably more multidimensional than opponents would have it. Accommodating the intractability of certain moral conflicts neither renders moral theory impractical nor does it reveal it to be incomplete. Far from being tantamount to denouncing moral theory as fatally deficient, acknowledging genuine, intractable moral conflicts reveals a richer, more accurate, and perfectly coherent picture of the moral universe.
The discussion proceeds in four steps: Section ii outlines and dissects the problem of dirty hands, to lay bare the fulcra giving rise to its paradoxical nature. Section iii addresses three principal objections to the possibility of dirty hands cases. The first rejects the possibility of genuine moral conflict altogether, by insisting that there is a plurality not just of values but of entire moral systems. The second assumes that political duties are by nature consequentialist, and claims that we can reduce moral conflicts to simple cases in which one duty overrides another, thus dismantling potential conflicts by providing a straightforward solution. The third presumes a specificationist view of rights that specifies rights so narrowly that they pose no obstacle to consequentialist demands; as a result, dirty hands cases never arise in the first place. Section iv defends the possibility of dirty hands – in particular, their defining paradoxical nature. The pdh does not merely acknowledge that an otherwise wrong, and hence impermissible, action may in certain circumstances be permissible; more importantly, the pdh insists that the act in question remains wrong, despite having been justified, or even required. The question of ‘what ought one to do, all things considered’ is hardly the whole story. Besides the aptness of reactive attitudes such as regret, guilt, and remorse, we should acknowledge the weight of rectificatory duties that may arise from doing even what is all-things-considered permissible. The world is messy, and morality is in no way lacking for reflecting this.
Before we proceed, a few preliminary remarks, beginning with the type of conflict that underlies the problem of dirty hands. Moral conflicts involve two conflicting moral duties that are actual (that is, not merely prima facie) duties.9 A moral conflict arises when (i) there is a duty to perform act a, (ii) there is a duty to perform act b, and (iii) it is only possible to do either a or b, but not both.10 The moral dilemma is a particular kind of moral conflict in which neither duty overrides the other, so that no matter which course of action the agent chooses, she will commit a moral wrong by violating one of the conflicting duties.11
Second, whether a moral duty obtains, roughly speaking, depends on a combination of certain moral reasons and facts. There is a great deal of disagreement about the extent to which moral judgments reflect facts, and which facts have the normative force to constitute moral reasons. My premise that moral conflicts recognize competing normative claims is neutral amongst those metaethical theories which acknowledge that we think and act as though there are moral facts, regardless of how our judgments or attitudes relate to them.
2 The Problem of Dirty Hands
It is a predictable and probable hazard of public life that there will be … situations in which something morally disagreeable is clearly required. To refuse on moral grounds ever to do anything of that sort is more than likely to mean that one cannot seriously pursue even the moral ends of politics.15
Although it is widely accepted, and expected, that certain circumstances will require the perpetration of severe moral wrongs, most people still find something unsettling in the idea of ‘dirty hands.’ How is it that we get our hands dirty by doing what we ought to do?
Michael Walzer’s seminal articulation of the phenomenon of dirty hands in politics suggests that two assumptions characterize the pdh: (1) the wrong action is required by necessity, for instance to avert a greater evil; (2) this action is nevertheless wrong, and the agent is consequently haunted by guilt. This is the sense in which her hands are ‘dirty.’
Let us consider these claims in turn. On Walzer’s view, politics constitutes a ‘world of necessity’ in which certain actions are required.16 This necessity arises from ‘a conflict between collective survival and human rights,’ and results from the politician’s professional duty to prioritize the former in adherence to a radical ‘utilitarianism of extremity.’17 The premise, in other words, is that political leaders’ professional responsibility requires them to prioritize the wellbeing of the collective over individuals’ rights. The starkly consequentialist implications for cases like ‘ticking bomb’ scenarios are familiar.18
But Walzer’s worry is not simply that political ends justify immoral means. The agent can only break out of the dilemma as a reluctant, remorseful quasi-consequentialist, all the while acknowledging that the overridden moral principles were not altogether eliminated by political necessity, but remained intact. This is why it is essential that the politician act as she must ‘with a heavy heart’: we do not, Walzer thinks, want to be ruled by people ‘who have lost their souls.’19 On Walzer’s account, then, it is paramount that the agent recognizes the severity of committing a wrong. If she did not feel remorse and had no scruples about violating moral duties, she would not be a moral politician.20 And a moral politician is what we want.
Part of the assumption is that agents must feel remorse, recognizing that their concessions of moral integrity retain their reprehensibility – even if political necessity requires doing wrong. In this sense, the justification of dirty hands, somewhat paradoxically, lies in their very unacceptability. The phd insists on the validity of deontological convictions even though they must be superseded by a consequentialism that is born out of political necessity.
The phd thus arises in cases in which agents are morally obliged, but can only fail, to reconcile competing duties. Acknowledging the possibility of dirty hands means acknowledging both that there is no way around the fact that political imperatives must sometimes override moral principles and that this is fundamentally problematic, as immoral action is not simply justified by achieving political ends. The dirtiness of hands, and the agent’s appropriate experience of guilt, on which more later, reflect the fact that the action in question, though necessary, remains a moral evil.21
We have now identified the two fulcra of the pdh. Dirty hands (i) result from wrongdoing that is required, for instance by political necessity, and (ii) appropriately imbue a genuine feeling of remorse about the wrong perpetrated. Even in cases in which it is right to do what is wrong, this wrong nevertheless remains wrong.
3 Skepticism about Dirty Hands
Many scholars have questioned the coherence and possibility of dirty hands cases. Actions, one prominent worry goes, cannot be simultaneously right and wrong simpliciter.22 Common criticisms consequently are that accounts of dirty hands suffer from a logical impossibility, moral inaccuracy, or otherwise unwelcome theoretical implications to the effect that morality is impractical, incomplete, or simply unrealistic. To wit, while I stipulated that the problem of dirty hands arises from a moral conflict between two duties, which cannot both be fulfilled, some scholars argue that moral conflicts between duties are altogether impossible; others recognize that conflicts between moral duties are possible in general, but hold that one of the duties must always override the other, leaving no possibility for actual conflict.23 Whereas the phd presupposes that (i) it can be right to do wrong, but (ii) the wrong nevertheless remains wrong, skeptics insist that the wrongness of the wrong effectively loses its purchase, so that the required action is ultimately right, and there remains no meaningful sense in which it is also wrong.
3.1 Plural Moral Systems
that ends equally ultimate, equally sacred, may contradict each other, that entire systems of value may come into collision without possibility of rational arbitration, and that not merely in exceptional circumstances, as a result of abnormality or accident or error … but … as part of the normal human situation.26
The view that ultimate values may be incompatible not only lies at the core of Berlin’s favored account of value pluralism, but also leads him to conclude that there is no one right answer to the question of how people ought to live.27 There are, Berlin claims, disparate moral universes that make different demands on us. They constitute equally ultimate and valid ways of life. But they are irreconcilable, and so one must, in choosing one, forego the other(s).
Different modes of life, on this view, require adherence to different values. To be a successful politician, one must abide by the values that shape political life, not values suited for a private life. If Berlin’s Machiavelli is right, the paradox at the heart of the pdh dissolves. Then all we need to do is acknowledge the existence of different moral systems – ‘exhaustive alternatives between two conflicting systems of value’ – which simply make it impossible to be a morally virtuous individual and, at the same time, a good political leader.28 On this view, the conflict between moral imperatives and political necessity that poses a threat to clean hands disappears as individuals are forced to choose one or the other way of life, and live by corresponding rules. So, if Berlin’s Machiavelli is right, we can avoid the problem by cutting the Gordian knot and choosing one moral system over the other.
On this view, the political leader’s choice is neither tragic nor difficult. The answer will always be to take whichever measures are necessary for the building of a strong society. If one desires to take up political responsibility, one must be willing to give up the ‘morally good life.’29 As a result, there is no vexing notion that an action may be right in one sense but wrong in another. So long as the action in question accords with the moral system to which the agent is committed, the action is right. Machiavelli’s Prince, then, will not get his hands dirty. He is not a moral politician in Walzer’s terms, which is impossible, but just an ordinary politician.
The kind of pluralism that Berlin finds in the Prince divides morality into different spheres, so that values from one sphere do not come into conflict with those from the other. Machiavelli’s Prince thus escapes conflicts between duties that pertain to different moral systems by committing to just one moral system to begin with.
But moral conflicts do not disappear just because we assume the existence of plural moral universes. And the separation between individual private life and public political life to the point that one is bound by only one set of duties does not preclude value conflicts within the same moral sphere. Even if there were plural moral systems, individual values and corresponding duties could still conflict within the same moral system, so long as fulfilling one duty would violate another sufficiently weighty duty. In politics, values like equality and freedom, or peace and justice, conflict all the time, and we must make difficult decisions about which values to prioritize in different cases.30 Dismissing such conflicts as symptoms of incompatible ways of life does not just miss the truth of agents’ moral experience; it denies the distinctive urgency and ultimacy of moral decisions, relativizing morality to the extreme.31
While relativists effectively treat values as mere arbitrary preferences, Henry Hardy has defended Berlin against the relativist charge by pointing to Berlin’s commitment to the objectivity and universality of values. Still, the question remains: in deciding what to do, how are we to choose between competing values which we acknowledge to be equally valid? Beside a nod to Machiavelli’s concession that, in politics, the end seems to justify the means, Berlin gives us no answer.
One may thus deny the possibility of moral conflicts by advocating a compartmentalization between different systems of morality. If individual morality and politics are two separate spheres, politicians will not get their hands dirty, and there will be no basis for blame or guilt. Indeed, there will be no dirty hands-generating conflict at all, as moral principles will not apply in politics. But, I have begun to suggest, the pdh does not arise from an incongruity between different moral systems. It arises from incompatible duties that inhere in the same moral system.
To see this, consider the distinction between the comprehensiveness and dominance of morality. To say that morality is dominant is to say that it trumps all other considerations; to say that it is comprehensive is to say that it is always relevant, even if it does not always trump other considerations.32 According to Berlin’s Machiavelli, morality cannot be either. By contrast, the idea of dirty hands recognizes the comprehensiveness of morality while remaining sensitive to the possibility that it is not always dominant. Attempts at cordoning off a special section of human activity, where actions are right that would otherwise be wrong, are unsatisfying. For even if they are right that morality is not always dominant, they may still be wrong to deny that morality is nonetheless comprehensive. The argument for a plurality of moral systems, then, does little to eliminate the type of conflict that generates dirty hands. Simply assuming away the comprehensiveness of morality will not do.
3.2 Political Consequentialism
Consider now the notion that politics is in some way ‘special’ in that it requires certain actions that violate moral principles. While the argument for plural moral systems seeks to preclude the possibility of dirty hands-generating conflict altogether, the argument for political consequentialism purports to provide a solution to the conflict that would otherwise give rise to dirty hands. I will say ‘political consequentialism’ to capture the common notion that political agents, more so than others, are bound by distinctively consequentialist demands. Political consequentialism has an air of exceptionalism, in that politics (let’s grant) generally requires people to ascribe greater weight to consequentialist demands than individual, ‘private’ morality.33
Mainstream discussions of dirty hands attach great moral significance to a political agent’s obligations to her community, which are typically taken to imply broadly consequentialist duties.34 One way of putting this is that, while private, individual moral agents may focus on agent-centered considerations of what they do, political agents must prioritize what happens.35 There is a sense in which political action demands a broadly consequentialist outlook – after all, politicians’ primary professional responsibility is to bring about certain beneficial outcomes.
In this, politics is not unique. Dirty hands cases are not linked specifically to political activity. Many professions and associative obligations prioritize outcomes in certain ways. The distinctive susceptibility of politicians to the pdh, I suspect, is intertwined with their commitment to their community, which requires them to prioritize what happens to that community over other, agent-centered considerations, not least since politicians are meant to act in the interests of those whom they represent. If this is right, political life by nature requires politicians, more so than most people, to give special weight to consequentialist considerations.36 Insofar as political actors are primarily representatives, Coady suggests, we may even need to assess them morally differently from other agents.37
One way to understand this critique is by championing a monistic view on which morality is reduced to one moral principle, that of political consequentialism.38 If consequences are paramount, this may eliminate the kinds of moral conflicts that rive rise to dirty hands cases. If politics requires decision makers to prioritize certain outcomes over, say, individual rights, then there is no sense in which they get their hands dirty by violating individual rights, because they are merely fulfilling their duties.
This approach to straightening out the conflict constitutive of dirty hands helps account for why the pdh arises in the first place. However, it also explains away its distinctly troubling nature. All that political consequentialism recognizes is that moral claims must be overridden by political necessity. It does not accommodate the possibility that doing what is all-things-considered permissible may nonetheless include wrongdoing. Hands simply never get dirty. But this obscures part of the story. It only tells us that the action in question is permissible on consequentialist grounds; it does not take seriously the notion that this comes at the cost of committing a wrong.
This is unsatisfying because the cost is real. In the standard trolley case, for example, turning the trolley away from the five and towards the one may be permissible as a lesser evil. But killing the one is still a real moral sacrifice. The fact that one acted all-things-considered permissibly changes nothing about the moral cost inherent in killing the one. After all, one’s lesser-evil justification comes from some external consideration of the greater good of saving the greater number, which has nothing to do with the right of the individual. Even if that person’s right not to be harmed may be permissibly infringed on lesser-evil grounds, this changes nothing about the fact that their right remains intact.
A monistic theory that reduces morality to achieving desirable consequences cannot make sense of moral sacrifices inherent in permissible rights infringements or compromises of one’s moral integrity.39 If there is only one end that matters, only duties that serve to achieve desirable outcomes will have any force. As a result, agents will not feel the pull of distinct, conflicting duties; they will not experience the type of conflict that results in the notion of dirty hands. If there are cases in which it is appropriate that agents feel remorse for wrongful actions, even if they are required, political consequentialism will not do.40 It just provides a clear formula as to what ought to be done, all the while denying the moral cost of duties not acted upon.
3.3 Specificationism
Specificationism seeks to define the constraints imposed by particular rights as narrowly as possible. On this view, rights, properly understood, do not conflict with other demands. Once rights are delineated sufficiently specifically, they are, in Hillel Steiner’s term, ‘compossible.’41 When rights are specified exactly, as John Oberdiek puts the idea, ‘what is justifiable is not opposed to any right in any way.’42 The notion of justified rights transgressions, and with it that of dirty hands, thus becomes obsolete.
Accordingly, if X is justified in doing Y, then Z has no right that X not do Y. That is, Z might have a right that people not do Y, except in cases in which a justification for doing Y obtains. If people only have rights not to be treated in highly specific ways, no rights are transgressed in situations in which agents otherwise would get their hands dirty.43
On the specificationist approach, rights play no role in determining what constitutes wrongful conduct. Rather than functioning as primary considerations in moral deliberation, rights become conclusions at which we arrive as a result of already having decided what it is permissible, prohibited, or obligatory to do. Specificationism effectively demotes rights to features of the moral universe that are completely descriptive.
On my view, by contrast, rights make it the case that there are certain things that we must not do to others, for no other reason than that they have certain rights. Even if, in tragic situations, morality permits the transgressing of certain rights, this does not change anything about what people’s rights prohibit. Value conflicts and dirty hands scenarios are troubling precisely because some rights transgressions are all-things-considered permissible despite the fact that victims have rights not to suffer the harms that may permissibly be inflicted on them. To say, even if only for the sake of argument, that the fictional mayor rushing to determine the location of a bomb justifiably tortured a potentially innocent suspect, end of story, would be deeply unsatisfying. When people’s rights are infringed, having been justified in doing so on lesser-evil grounds is decidedly not the whole story.
As a result of specifying away any possibility of conflicts between rights, specificationism is unable to make sense of the fact that innocent victims’ rights not to be harmed continue to exert normative power, even though agents acted all-things-considered permissibly in contravening those rights. It simply denies that this kind of tension is possible. While the consequentialist approach to denying dirty hands cases denies the abiding relevance of rights that may be permissibly infringed, the specificationist approach distorts their very purpose.
Time to take stock. We have considered three different ways to resist the notion of dirty hands. The first claimed that there is a plurality of moral systems, so that moral conflicts do not arise in the first place; the second claimed that politics is by nature consequentialist, suggesting that dirty hands conflicts are straightforwardly resolvable because there always is a consequentialist solution. The third approach defined rights so narrowly that they pose no real obstacle to consequentialist demands, and so never give rise to dirty hands scenarios in the first place. I have begun to suggest that we should resist all three objections, because none satisfactorily address precisely what it is that renders dirty hands distinctly and paradoxically troubling. This claim requires further defense.
4 Defending the Problem
Let us begin with the structure of the kind of normative view that recognizes the possibility of dirty hands. To what extent, and in what sense, is the problem of dirty hands precisely that – a problem?
In Section ii, we identified as one of the constitutive elements of the pdh the somewhat paradoxical claim that an agent may be committing a wrong in doing what is required by political necessity. The arguments for plural moral systems and political consequentialism remove this sense of wronging from the equation, implying that, if an agent acts in accord with some previously chosen moral system or with political consequentialism, because this is what is professionally required of her, there is no sense in which her action is wrong, and no sense in which she becomes blameworthy. These critiques thus deny the defining, peculiar feature of the pdh, that doing what is required under certain conditions may remain wrong – not wrong, all things considered, but wrong nonetheless.
The pdh arises from conflicting moral duties, neither of which naturally overrides or outweighs the other. The pdh at once attempts to sanction a path out of the conflict for the agent, without rejecting the validity of each of the conflicting duties. The fact that one duty is made to override the other, the pdh signals, does not ‘resolve’ the actual conflict for good. That is why the problem does not dissolve once we work out what we ought to do, all things considered. Whichever course of action one takes, one commits a wrong. Hence, what the pdh tells us is not that an action is both right and wrong. What it in fact tells us is that a wrong action, such as the transgression of an actual, intact right, may be permissible under certain circumstances, without this rendering it any less wrong.
Consider the appropriateness of reactive attitudes, such as regret, in dirty hands cases. As Williams puts it, ‘whatever he does, the agent will have reason to feel regret at the deepest level.’44 This is the case not least if one infringes another person’s right on lesser-evil grounds, such as in the hypothetical trolley case. When consequentialist demands override individual rights, this may generate a moral ‘remainder,’ which may manifest itself as regret. The person who must get her hands dirty, through no fault of her own, finds herself in a situation in which she has no choice but to act in a way that subsequently renders it appropriate for her to feel regret for what she has done.45
The pdh does not merely acknowledge that an otherwise wrong action may in certain circumstances be right; the pdh also insists that the act in question remains wrong, despite having been permissible.46 This is why Walzer’s moral politician suffers from her actions, and is plagued by a sense of guilt for having gotten her hands dirty. This has no practical impact on the world, but it makes all the difference for our understanding of the problem.
We need not rely on reactive attitudes like guilt and remorse, as evidence of genuine moral conflict. The claim is not simply that our intuitions about ex post duties, such as duties of compensation, offer useful insights into which ex ante duties we expect there to be – though this may be true.47 The point, rather, is that the moral realm is not exhausted by what agents ought to do, all things considered. The notion of moral remainders and the resulting aptness of certain reactive attitudes such as regret are also part of the story, as are certain rectificatory duties that may arise from doing what is, all things considered, permissible. Moral theory is about more than what to do at a particular point in time.48
Moral duties that are overridden are not automatically eliminated. If I promised to meet you for tea, but come across a drowning child on my way, my obligation to meet you is overridden. Rescuing the drowning child instead of fulfilling my duty to meet you is not a case of getting my hands dirty. I might owe you an explanation, and perhaps an apology for not showing up. But for all practical purposes, my duty to meet you for lunch is eliminated by the exceptional circumstances. What the pdh captures is that the same is not true if weighty moral requirements, like those implied by individual rights, are overridden by political necessity or other consequentialist demands. If X violates Y’s right not to be tortured because X believes that this is the only way to get Y to tell her where Y hid the bomb that will extinguish the entire city, the moral duty not to torture Y does not just disappear. Even assuming, if only for the sake of argument, that X acted permissibly, all things considered, this is not all that matters. What also matters is that X must, for the rest of her life, live with the moral remainder of the wrong she committed.49
Even if a duty is overridden by overwhelming necessity, it is not necessarily eliminated. This is why the agent’s hands are dirty, why it would be apt for her to experience remorse, and why she may owe a duty to apologize and make efforts to otherwise address the wrong done.
The ‘problem’ inherent in dirty hands, then, is the real moral cost involved in failing in some duty, even if this is justified, all things considered. This cost is what mainstream challenges to the pdh are mistaken to discount. Dirty hands are a problem in that they result from doing wrong.50
Principal critiques of the phd, I suggested, attempt to resolve its paradoxical nature at too high a price. The first objection, the argument for plural moral systems, seeks to avert the presumed moral conflict by rejecting its possibility from the outset. This sacrifices morality’s comprehensiveness. The second objection, the argument from political consequentialism, leaves us no choice besides consequentialist monism. This sacrifices the reality of the moral wrong concomitant with permissibly prioritizing consequentialist demands in certain cases. The third objection, the specificationist approach, denies the possibility of dirty hands altogether. Its price is the complete distortion of the nature and purpose of rights. None of these options are appealing.
What is more, such drastic measures are unnecessary. In what follows, I argue that accommodating the paradoxical aspect of the pdh, far from being a weakness in our moral theory, reflects an unsatisfying, but nonetheless real, part of moral experience.
5 Moral Theory and the Happenings of the World
The mere fact that an action may be all-things-considered required by certain circumstances is compatible with it being a pro tanto wrong that must be addressed and rectified. Even a duty that is overridden does not lose its grip on the agent. This is why she has dirty hands. There is, as I have begun to suggest, nothing contradictory about the recognition that certain acts can be both pro tanto wrong and all-things-considered required. Some actions that are required by political necessity are regrettable because they fail in a moral duty and thus constitute a wrong. This is why the pdh is a problem in the literal sense: it describes a conflict between two actual duties, both of which retain their validity, even if one is made to override the other. We know that the overridden duty retains its moral force because there is nothing that made it go away; hence, the aptness of certain reactive attitudes such as regret. Victims have a justified complaint that they have been wronged, and they may be owed compensation.
There is a pervasive sense in the literature that there had better not be real moral conflicts. The worry is that, if actual moral conflicts exist, this implies something about our theorizing that ought to make us uneasy. For example, it might mean that our moral theory is (a) impractical, demanding that we pursue conflicting values, which is impossible; (b) incomplete, leaving us with no answer as to which of the conflicting values to pursue; or (c) non-realistic, dismissive of what is right as a matter of fact.51 While some views therefore reject the possibility of genuine moral conflict, others conclude that moral theory as a whole is deficient if it accommodates what appears like a logical inconsistency.
But there is no need to try to eliminate conflict in moral theory. In recognizing the type of conflict that gives rise to dirty hands, we, in the first instance, recognize something about the world, rather than making a claim about morality itself. Moral conflicts are not necessarily a symptom of defective thought; they are simply a result of circumstances in the world.52 Acknowledging the reality of actual moral conflicts need not entail existential anxieties for our theory, because conflicts between duties arise from external circumstances. Duties conflict as a result of things that happen, including other people’s actions. To borrow Martha Nussbaum’s phrase, it is ‘under pressures endemic to life in a world where choice is constrained by necessity’ that moral conflicts occur and the pdh arises.53 This is why, in acknowledging the reality of moral conflict, in the first instance, we learn something about the world. We in no way compromise the coherence of our moral theory.
There is nothing wrong with a moral theory that identifies conflicting duties. The problem is not with how the agent perceives the situation; rather, as Williams puts it, ‘what is wrong lies in his situation itself.’54 Conflicts between values, and between their corresponding moral claims, are, in Nussbaum’s language, simply ‘a correct description of the way in which natural circumstances restrict the possibilities for choice.’55 A plurality of conflicting values is a reflection of the world as it is, not a logical contradiction that needs to be straightened out in our theory.56
In the end, morality is about more than what we ought to do at a particular point in time. What matters about moral acts, Stocker was right to insist, goes beyond whether they are to be done. Recognizing that the notion of moral remainders captures something real, such as rectificatory duties, opens the door to questions about what is to be done about those remainders. Once we accept the possibility of dirty hands cases – and, more generally, genuine moral conflicts – we face the question what morality requires when one cannot do everything one ought to do.57
Accommodating the possibility that there is something both right and wrong about an agent’s potential courses of action in dirty hands cases is not a weakness of our moral theory. Moral conflicts are not indicative of morality’s limits, but part of our human existence in the world.
Accepting that conflicting duties are generated by real-world circumstances rather than faulty mechanisms in moral theory also has significant practical implications. If the pdh is a real moral conflict arising from a particular type of situation, we should work to reduce and eliminate such situations in public policy. After all, getting one’s hands dirty is an emergency solution. As Berlin put it, the only justification for employing immoral measures is that they are exceptional and necessary for the preservation of a system whose purpose ‘is precisely to preclude the need for such odious measures.’58 Indeed, it would be a mistake to assume that difficult moral problems can be solved before they arise: getting one’s hands dirty is an emergency measure, not an actual solution to the problem. It is not a blanket permission to employ all means necessary for the achievement of certain ends. This, incidentally, is something that the argument from political consequentialism neglects.
Accepting the moral burden of having to get one’s hands dirty when necessary is not tantamount to a carte blanche for overriding all moral constraints. That is not the point of substantiating the possibility of the pdh. Acknowledging that a wrong must be committed should lead us to consider what is required to change the circumstances, including our institutional background, that might have made it necessary to commit a moral wrong in the first place.59
This raises questions about the extent to which dirty hands cases are the result of someone else’s wrongdoing, though a more detailed discussion must wait for another day. Paradigmatic examples of dirty hands involve someone else creating a situation in which the agent must do wrong, such as in ticking bomb scenarios. Indeed, Machiavelli argues that the Prince must learn how to be ‘not good’ only and precisely because other people are not virtuous.60 One question this raises is whether dirty hands cases are always generated by others’ wrongdoing, or whether they may also arise from unfortunate circumstances, in the absence of any wrongdoing.61
Before we conclude, let us briefly revisit our starting point. While Machiavelli suggests that one can be either morally virtuous or an efficient politician – decrying any attempt at reconciling the two as pointless, foolish, and doomed – Walzer hinges leaders’ moral legitimacy on the attempt to be both: genuinely regretting the violation of moral duties and the inevitable failure to retain one’s moral integrity make all the difference for the very legitimacy of political authority. Incidentally, Max Weber similarly suggested that the politician who has a sense of vocation must strive to reconcile a deontological ethic of conviction with the consequentialist ethic of responsibility.62 The politician who has a sense of vocation must both believe in transcendent moral ideals and remain sensitive to the constraints that political reality sets. What is required is the willingness to accept dirty hands as simultaneously necessary and lamentable.
6 Conclusion
The aim of this paper was to defend the possibility of dirty hands and the paradoxical nature that distinguishes it. I began with the assumption that the pdh arises from a conflict between actual duties, neither of which cancels out or naturally overrides the other. One feature of the pdh that I stipulated is its simultaneous acknowledgement of an ex ante necessity to commit a moral wrong and the ex post appropriateness of certain reactive attitudes such as regret. I then addressed three objections to the possibility of dirty hands cases: the argument from a plurality of moral systems, which denies the possibility of moral conflict altogether; the argument from political consequentialism, which attempts to turn the dilemma into an ordinary case of overriding duties; and the specificationist approach, which defines rights so narrowly that they pose no real obstacle to consequentialist demands, so that dirty hands scenarios do not arise in the first place. I argued that none of these approaches successfully dismantles the pdh, as none can convincingly account for the fact that, although an action may be required by certain consequentialist demands, it may nevertheless remain wrong. Principal attempts at straightening out the moral conflict, I argued, fail to satisfactorily address the fact that the duties that are violated nonetheless retain their force.
What this discussion revealed is that we should not take the pdh to claim simply that an action is both right and wrong, but that a wrong action, such as the violation of an actual, intact duty, may be required by certain circumstances, without this rendering it any less wrong. This is no indication of a deficiency in our theory. It just reflects the world as it is.
By insisting that the agent’s hands are dirty, the pdh captures the realization that it is not enough to recognize that prima facie reasons against violating moral duties may be overridden. A duty is violated, people are wronged, and this matters even when one has a very good reason for one’s action. This, we saw, may have various practical implications if we entertain the possibility that dirty hands might appropriately entail not only certain reactive attitudes but also rectificatory duties of some sort, such as apologies or compensation.
a remarkable and characteristic phenomenon in philosophical investigation: the difficulty … is not that of finding the solution but rather of recognising as the solution something that looks as if it were only a preliminary to it … . This is connected … with our wrongly expecting an explanation, whereas the solution of the difficulty is a description … The difficulty here is: to stop.’63
Acknowledgements
For helpful comments and discussion, I am grateful to Joel Rosenthal, Michael Weinman, Thomas Norgaard, and an anonymous reviewer for this Journal. Earlier versions were presented at the Warwick Graduate Conference in Political and Legal Theory in 2015 and at the European College of Liberal Arts/Bard College Berlin in 2014, where this paper originated as an undergraduate thesis.
Biographical Note
Linda Eggert is an Early Career Fellow at Balliol College and in the Faculty of Philosophy at the University of Oxford.
Dana Milbank, ‘Cheney Authorized Shooting Down Planes,’ The Washington Post (18 June 2004), A01.
Richard Bernstein, ‘Kidnapping Has Germans Debating Police Torture,’ The New York Times (10 April 2003).
For discussion, see Michael Ignatieff, The Lesser Evil: Political Ethics in an Age of Terror (Princeton, NJ: Princeton University Press, 2005).
While dirty hands cases are paradigmatically at home in politics, they may occur in any sphere of human conduct in which moral conflicts arise. See Section iii.2.
Michael Walzer, ‘Political Action: Dirty Hands in Politics,’ in Marshall Cohen, Thomas Nagel, and Tim Scanlon (eds), War and Moral Responsibility (Princeton, NJ: Princeton University Press, 1974), 66. See also Michael Ignatieff, Fire and Ashes (Cambridge, MA: Harvard University Press, 2013).
Bernard Williams, Moral Luck: Philosophical Papers 1973–1980 (Cambridge: Cambridge University Press, 1981), 54.
For example, Kai Nielsen, ‘There is No Dilemma of Dirty Hands,’ in Igor Primoratz (ed), Politics and Morality (Palgrave-Macmillan, 2007), 1–7. See also Michael Yeo, ‘On the One Hand and On the Other’; Michael McDonald, ‘Hands: Clean and Tied or Dirty and Bloody’; David P. Shugarman, ‘Democratic Dirty Hands?’ all in Paul Rynard and David P. Shugarman (eds), Cruelty and Deception: The Controversy over Dirty Hands in Politics (Peterborough, Ont.: Broadview Press, 2000; Sydney: Pluto Press, 2000).
Michael Stocker, Plural and Conflicting Values (Oxford: Clarendon Press, 1990), 85–126; Walter Sinnot-Armstrong, Moral Dilemmas (Oxford: Blackwell Publishers, 1988), Chs. 1–3. For a recent discussion of the relationship between the phd and moral conflicts, see Christina Nick, ‘Dirty Hands and Moral Conflict – Lessons from the Philosophy of Evil,’ Philosophia 50 (2022): 183–200.
For an argument that moral conflicts (of dirty hands) involve conflicting duties, see Michael Stocker’s ‘Dirty Hands and Ordinary Life,’ in Plural and Conflicting Values (Oxford: Clarendon Press, 1990), 9–36; Stocker, ‘Dirty Hands and Conflicts of Values and of Desires in Aristotle’s Ethics,’ Pacific Philosophical Quarterly 67 (1986): 36.
Stocker, Plural and Conflicting Values, 86; see also Bernard Williams, ‘Ethical Consistency,’ in Problems of the Self: Philosophical Papers 1956–1972 (Cambridge: Cambridge University Press, 1973), 167. On the nature of moral conflicts, see also Thomas Hurka, ‘Monism, Pluralism, and Rational Regret,’ Ethics 106 (1996): 555–575.
Sinnot-Armstrong, Moral Dilemmas, 58; Martha Nussbaum, The Fragility of Goodness, 2nd ed. (Cambridge: Cambridge University Press, 2001), 30; see also Williams, ‘Ethical Consistency’; Stocker, ‘Moral Conflicts: What They Are and What They Show,’ in Plural and Conflicting Values.
C. A. J. Coady, ‘Politics and the Problem of Dirty Hands,’ in Peter Singer (ed), A Companion to Ethics, (Oxford: Blackwell Publishers, 1993), 373.
In the relevant cases, the people who are harmed retain rights against being thus harmed; they have neither forfeited nor waived them. What nonetheless renders it permissible for agents to inflict these harms is a lesser-evil justification.
Susan Mendus, Politics and Morality (Cambridge: Polity Press, 2009), 8. As noted earlier, dirty hands are not necessarily limited to politics, but may result in any sphere of human activity in which moral conflicts arise. See Section iii.2.
Williams, Moral Luck, 60.
Michael Walzer, Just and Unjust Wars (New York: Basic Books, 1996), 325.
Ibid.
Walzer, ‘Political Action: Dirty Hands in Politics,’ 68–69.
Ibid., 79.
Ibid., 73.
Ibid., 63.
See Kai Nielsen, ‘There is No Dilemma of Dirty Hands’; Michael Yeo, ‘On the One Hand and On the Other’; Michael McDonald, ‘Hands: Clean and Tied or Dirty and Bloody’; David P. Shugarman, ‘Democratic Dirty Hands?’ all in Paul Rynard and David P. Shugarman (eds), Cruelty and Deception: The Controversy over Dirty Hands in Politics. For defenses of the pdh, see Ronald Beiner, ‘Missionaries and Mercenaries’; John Allett, ‘Bernard Shaw and Dirty-Hands Politics’; Tom Sorrell, ‘Politics, Power, and Partisanship’; Evan Simpson, ‘Justice, Expedience, and Practices of Thinking,’ in the same volume.
Even if we rank duties according to priority, moral dilemmas are still possible. See Sinnot-Armstrong, Moral Dilemmas, 31, 72, 95–97.
Isaiah Berlin, ‘The Originality of Machiavelli,’ in H. Hardy (ed), The Proper Study of Mankind (Berlin: Farrar, Straus and Giroux, 2000), 320.
Ibid., 289.
Ibid., 320. See also Hannah Arendt, ‘Truth and Politics,’ in Between Past and Future, revised edition (Penguin Classics, 2006), 223–259; Reinhold Niebuhr, Moral Man & Immoral Society (Westminster: John Knox Press, 2002); Max Weber, ‘Politics as a Vocation,’ in D. Owen, T. B. Strong (ed), The Vocation Lectures (Hacket Publishing Company, 2004), 92.
Berlin, ‘The Originality of Machiavelli,’ 312, 316.
Ibid., 303.
Ibid.
We may also question whether values are really incommensurable. For discussion, see, for example, Ruth Chang’s work.
Incidentally, assuming an infinite number of ultimate ethical codes also makes it difficult for Berlin to identify what it means to exercise good moral leadership in politics. See, for example, his analyses of Winston Churchill and Franklin D. Roosevelt in The Proper Study of Mankind.
C. A. J. Coady, ‘The Problem of Dirty Hands,’ in Edward N. Zalta (ed), The Stanford Encyclopedia of Philosophy (Spring 2014 Edition).
For example, Coady, ‘Politics and the Problem of Dirty Hands,’ 378; Williams, Moral Luck, 61.
For discussion of a leader’s special duty to her community, see Walzer, ‘Political Action: Dirty Hands in Politics,’ 81; Plato, The Republic, Allan Bloom, trans., 2nd ed. (Basic Books, 1991), 412 d-e; Coady, ‘Politics and the Problem of Dirty Hands,’ 378; Williams, Moral Luck, 64; Niebuhr, Moral Man & Immoral Society, 257.
For the distinction between what we do and what happens, see Thomas Nagel, The View From Nowhere, revised ed. (New York: Oxford University Press, 1989).
For what it’s worth, I think what I call political consequentialism here is ultimately most plausibly explained by an appeal to associative obligations rather than some claim that consequentialism somehow acquires special validity in politics, but this has no bearing on my argument.
Coady, ‘Politics and the Problem of Dirty Hands,’ 376.
See Sinnott-Armstrong, Moral Dilemmas, 32.
For example, Williams, Moral Luck, 63.
It would be a mistake to rely on moral emotions as an indication of moral permissibility; after all, our emotions might simply be mistaken. The point, rather, is about the one-dimensional structure of morality that denies the kind of permissible wrongdoing that might appropriately give rise to such emotions in the first place. I wish to thank an anonymous reviewer for pressing me on this.
Hillel Steiner, An Essay on Rights (Blackwell, 1994).
John Oberdiek, ‘Lost in Moral Space: On the Infringing/Violating Distinction and Its Place in the Theory of Rights,’ Law and Philosophy 23 (2004): 325–346.
Oberdiek, ‘Lost in Moral Space,’ 346; see also Leif Wenar, ‘Rights,’ in Edward N. Zalta (ed.), The Stanford Encyclopedia of Philosophy (Spring 2020 Edition).
Williams, Moral Luck, 31, 74; see also Williams, ‘Ethical Consistency,’ in Problems of the Self, 166–186. Stephen de Wijze terms the relevant reactive attitude ‘tragic remorse.’ See Stephen de Wijze, ‘Tragic Remorse – The Anguish of Dirty Hands,’ Ethical Theory and Moral Practice 7 (2005): 453–471.
For example, Nussbaum, The Fragility of Goodness, 27, 51.
This also explains why dirty hands cases sit some awkwardly with the distinction between justifications and excuses, as the action in question is justified, wrong, and the agent’s regret is appropriate all at the same time. For a discussion, see Sinnot-Armstrong, Moral Dilemmas, 42; J. L. Austin, ‘A Plea for Excuses,’ Proceedings of the Aristotelian Society 57 (1956–1957): 1–30. Not everyone believes that those with dirty hands are blameworthy. For an example, see Neil Levy, ‘Punishing the Dirty,’ in Igor Primoratz (ed), Politics and Morality. Palgrave-Macmillan (2007), 38–53. Thanks for an anonymous reviewer pressing this point, on which I remain agnostic here.
Judith Jarvis Thomson, for instance, regards ex post compensatory duties as evidence of certain ex ante moral duties. See Thomson, Rights, Restitution, and Risk (Cambridge, MA: Harvard University Press, 1986), 40–1, 66–68; Joel Feinberg, ‘Voluntary Euthanasia and the Inalienable Right to Life,’ Philosophy & Public Affairs 7 (1987): 102. See also Sinnot-Armstrong, Moral Dilemmas, 37, 44.
See Stocker, Plural and Conflicting Values.
For example, Sinnot-Armstrong, Moral Dilemmas, 92–95.; Williams, Moral Luck, 63; Stocker, Plural and Conflicting Values, 123.
For a defense of the view that there are moral conflicts in which wrongdoing is unavoidable, see Christopher W. Gowans, Innocence Lost: An Examination of Inescapable Moral Wrongdoing (Oxford: Oxford University Press, 1994).
Stocker, Plural and Conflicting Values, 85–86.
For the kind of view I have in mind, see Philippa Foot, ‘Moral Realism and Moral Dilemma,’ The Journal of Philosophy 80 (1983): 379–398.
Nussbaum, The Fragility of Goodness, 38. For related defenses of the pdh as a part of moral reality, see Stuart Hampshire, ‘Public and Private Morality,’ in Stuart Hampshire (ed), Public and Private Morality (Cambridge: Cambridge University Press, 1978), 23–53; see also Gowans, Innocence Lost.
Williams, Moral Luck, 75.
Nussbaum, The Fragility of Goodness, 46.
As an anonymous reviewer points out, a consequentialist might agree with Nussbaum that natural circumstances restrict our choices, but still deny that these choices leave a moral remainder. My view is that such an account would be missing something.
Stocker, Plural and Conflicting Values, 103.
Berlin, ‘The Originality of Machiavelli,’ 311.
A notable example of this was set by the British diplomat, Carne Ross, who quit his profession over the Iraq War, as his job demanded moral sacrifices with such frequency that dirty hands were the norm rather than the exception. See Carne Ross, Independent Diplomat: Dispatches from an Unaccountable Elite (Cornell University Press, 2007).
Niccolo Machiavelli, The Prince, George Bull, trans. (London: Harmondsworth: Penguin Books, 1961), 91.
For the view that dirty hands generally result from ‘immoral circumstances created by other persons,’ see Stephen de Wijze, ‘Dirty Hands: Doing Wrong to do Right,’ South African Journal of Philosophy 13 (1994): 27–33.
Weber, ‘Politics as a Vocation,’ 92.
Ludwig Wittgenstein, Zettel, Section 314, cited in Nussbaum, The Fragility of Goodness, 23.