The Non-Political Foundations of the Problem of Dirty Hands

Michael Walzer’s “Political Action: The Problem of Dirty Hands” (Walzer 1973) is truly memorable. Brimming with the type of intelligence and the substantive saturation that is rare in ordinary contributions, Walzer’s article continues to reward, rereading after rereading. Whether or not commemorating single articles is a frequent phenomenon, I am convinced that this article deserves such commemoration, and I am honored to be a part of it. With both erudition and insight, Walzer’s article has spawned a justifiably large secondary literature, and it has promoted reflection on terribly important – and traditionally neglected – topics. Above all, I will focus on Walzer’s central preoccupation: “whether or not a man can ever face, or ever has to face, a moral dilemma, a situation where he must choose between two courses of action both of which it would be wrong for him to undertake” (Walzer 1973: 160). This, in (very) short, is the problem of dirty hands.

Walzer’s insights have influenced my thinking profoundly, and I will try to describe here aspects of that influence. And yet, that I consider myself indebted to Walzer’s article is in a sense odd. After all, and as his very title announces, Walzer emphasizes a distinction between the political and other normative realms. As the title of my article indicates, I will suggest that this is a problematic emphasis, and that we should recognize that the problem of dirty hands is not essentially political. To be clear, I deny neither that there exist important distinctions between the political and other normative realms nor that problems of dirty hands can arise within the political realm. My point rather is that we are better off unmooring dirty hands from its presupposed political confines.¹

Since, as I shall show in due course, there is an important connection between dirty hands and punishment, I will begin by briefly summarizing my views on the distinction between the political and other normative realms within the specialized literature on punishment. In section II, turning specifically to Walzer’s article, I will suggest that the reasons why Walzer may privilege the political context are similar to the reasons why punishment theorists have tended to privilege state punishment. And I will suggest that arguments parallel to those I have put forth against this privileging within the context of the specialized literature on punishment theory can be profitably advanced, too, within the context of the problem of dirty hands. In section III I will attempt to show why this discussion is about much more than mere line-drawing, and will suggest that important, substantive matters are at stake. Finally, in a brief concluding section, I circle back to the central role that punishment has for Walzer’s views on dirty hands – a centrality that is not often – or sufficiently – acknowledged.

1 Punishment with and without the State

I would like to register at the outset my greatest debt to Walzer: one cherished lesson I have drawn from his article is that the presentation and the moral evaluation of human action is much more complicated than ethicists typically assume – and I have applied that lesson to my work on punishment. It was largely as a result of my reading Walzer’s article that I became convinced that punishment, in general, is best seen as an instance of dirty hands (Zaibert 2018). In that regard, it is hard for my indebtedness to Walzer to be more obvious.

Another aspect of my work on punishment, however, is my insistence on blurring or de-emphasizing precisely the distinction between the political and other normative realms upon which, in contrast, Walzer insists. (That is why, in the preceding paragraph I italicized “human action” – a locution I suggest is better than Walzer’s “political action”.) To focus on political (i.e., state, criminal) punishment – as the vast majority of punishment theorists do – risks muddying the waters, at least regarding the theoretical analysis of punishment qua phenomenon. For the discussion of state punishment is almost necessarily tangled up in all sorts of other considerations pertaining to political theory, public policy, economics, and more. These other considerations are no doubt very important in their own right, but they can hinder our efforts to understand punishment as such. And I will suggest that something similar happens in the context of the problem of dirty hands: if we really want to understand what this problem is, we are better off articulating the problem of dirty hands as a fundamental moral problem.

Very roughly, punishment is a treatment which we deem painful (unpleasant, unwelcome, etc.), which we inflict on a perceived wrongdoer, in response to what we perceive was her wrongdoing. It has always seemed to me obvious that we sometimes perceive wrongdoing outside the criminal law, outside the political context: we sometimes find wrong in what our friends, our loved ones, and even perfect strangers, do (Zaibert 2005) (Zaibert 2023) (see also (Gardner 2008) and (Husak 2022). In fact, instances of non-political wrongdoing are arguably more numerous than instances of political wrongdoing: in any case, they clearly exist: we often punish each other for our wrongdoings. Interestingly, even scholars who claim to restrict their attention to state punishment, often (unwittingly?) illustrate their views by presenting examples from non-state contexts (see, e.g. (Duff 2007) amongst many others).

Perhaps aware of the narrowness in his definition of punishment (whereby punishment is essentially just a political/legal phenomenon), Hart almost grudgingly – and at any rate halfheartedly – admitted to the existence of “punishment for breaches of non-legal rules” (Hart 2008: 5). Hart dubbed these cases of punishment “sub-standard”, thus setting the tone for subsequent discussions. Following in Hart’s footsteps, the vast majority of punishment theorists typically treat these other manifestations of punishment as, at best, insignificant marginalia – this is in those cases in which they even acknowledge their existence.

Punishment theorists’ focus on political punishment is understandable. After all, it is precisely this manifestation of punishment that can cause more widespread harm. While perfectly legitimate, this consideration is explicitly based on ameliorative aspirations, on protecting human beings. It is not a theoretical consideration. In other words, this focus is not meant to advance our understanding of punishment, qua phenomenon. As it turns out, it is not simply that these ameliorative efforts do not always overlap with theoretical endeavors: a much greater problem is that they can actually hinder those endeavors. Such is the case of the excessive – and widespread – focus on state punishment. This focus naturally leads scholars toward issues of criminalization, political legitimacy, of the liberal concern with protecting the weak (from the awesomely powerful state), and so on. These are, again, terribly important matters. But they take attention away from the essence of punishment: the fact that when we punish someone, we are seeking to make that person suffer. This purely conceptual point is linked to a normative one: making people suffer is prima facie a bad thing to do. And yet, (I stipulate) sometimes we are justified in punishing wrongdoers, in making people suffer. How is that possible?

These points, and the question(s) they invite, have as much applicability outside the political context as they have within it. Imagine you discover that your best friend has done something horrible, and that you are debating whether (or to what extent) to punish her. The usual conflict between the different rationales typically mobilized within the context of state punishment apply here as well. You would wonder whether you should punish your friend because she (and/or others) will learn a lesson, or because you want her (and/or others) to think twice before she behaves like this again, or because you want to incapacitate her (and/or others) from ever behaving like this again, or because you think she deserves to be punished.

In other words, something very much like the debate which in the political sphere opposes consequentialists (by which I mean utilitarians: thinkers who believe that punishment can only be justified by its potentially good consequences) and retributivists (thinkers who believe that punishment is justified by its being deserved) will ensue. And the debate is, in a sense, purer in the non-political context, since the important (but ameliorative) considerations (relating to special protections of the weak vis-à-vis the powerful state, etc.) that matter in the political sphere either do not obtain in the non-political sphere or obtain in a mitigated way. For example, limitations on the permissible ways of obtaining evidence – crucial in the political context of a criminal trial – have very little currency in non-political contexts, if any. You may have discovered your friend’s horrible deed because you read her private diary; inappropriate as reading her diary without her permission is, this would have no bearing on the dilemma that you face regarding your reasons for punishing her. She cannot ask for a mistrial or avail herself of other protections as a result of your violation of her privacy.²

These extraneous factors can make the theoretical tasks at hand more complicated. What exactly is punishment? What can justify it? What does it mean to say that punishment is justified? How do we reconcile the values in conflict whenever we punish anyone? These are difficult questions, whose answers become even more elusive if we also have to juggle considerations concerning political legitimacy and the reach of the state’s power when considering them.

This is the parallel I want to establish between the specialized literature on punishment theory and the specialized literature on dirty hands. Just like punishment theorists have excessively privileged the political, so have dirty hands theorists. And just as this privileging has detrimental effects within the context of punishment theory, so it has within the context of the discussion of dirty hands.

2 Shifting Frames

The very title of Walzer’s article seems to leave no room for doubt about the fact that he focuses on the problem of dirty hands understood as a political problem – a focus that is visible throughout the rest of the article as well. But it is not Walzer’s only view. Walzer actually oscillates between this ostensibly political understanding of dirty hands and a much broader moral understanding of it, such as the one I recommend.

Walzer began his article incisively, arguing that the symposium on the rules of war that appeared in Philosophy and Public Affairs to which he was reacting was not centrally about the rules of war. It was, in Walzer’s estimation, “actually (or at least more importantly) a symposium on another topic” (Walzer 1973: 160). I think that Walzer is profoundly correct in this assessment. In fact, appropriating here Walzer’s own overture, I will suggest that, despite ample textual evidence – including Walzer’s very title – his article is not centrally about the problem of political action. It is, in my estimation, actually (or at least more importantly) an article on another topic.

Just as I take it that Walzer would not deny that what he thought was the actual topic of the symposium to which he was reacting had implications for the narrower topic of the rules of war, I do not deny that what I think is the actual topic of Walzer’s article has implications for the narrower topic of political action – such as those he illuminatingly discusses. Still, the topic of Walzer’s article is, at its core, much more fundamental, and it transcends matters of political action, just as the topic of the symposium to which Walzer reacted transcends the issues of the rules of war. The actual (proper, central, etc.) topic of Walzer’s article is precisely what he himself identifies as the topic of the symposium: “the actual topic [of the symposium] was whether or not a man can ever face, or ever has to face, a moral dilemma, a situation where he must choose between two courses of action both of which it would be wrong for him to undertake” (Walzer 1973: 160, emphasis added). This passage conspicuously contains what at the outset I suggested is the essence of dirty hands.

Notice that this other topic, as described by Walzer, which we both find immensely important, has, at least on first approximation, no particular connection to anything political. What Walzer describes is a predicament that “a man” – that is, any human being, not just a politician – can face. This predicament flows naturally and unavoidably from the complexity of moral life. Walzer suggests that this complexity can stand in the way of our efforts to find “coherence and harmony in the moral universe”, and even to lead “a moral life” (Walzer 1973: 161, emphases added). In at least this much, I place Walzer’s article in close proximity to other relatively recent contributions which invite us to engage seriously with the often-overlooked intricacies of the moral landscape.³

Exhibiting a sort of “self-incurred minority” (Kant 1996: 17),⁴ contemporary moral theorists often underestimate (or altogether deny) the complexity of moral life. Consider how out of place Walzer’s references to the harmony of the moral universe, or the difficulties in leading a moral life appear within the ubiquitous context of “trolley problems” and similarly fashionable schemas in contemporary moral philosophy. Wondering how human beings ought to feel about what they sometimes ought to do appears to be unimportant to many contemporary moral philosophers. In marked and much welcome contrast, Walzer’s article invites us to discuss subtle and deeply human distinctions. Thus, and focusing only on the emotional context, Walzer distinguishes generic sadness and melancholy from more robust and complicated forms of guilt arising from the realization that, when we have faced dirty hands situations, we have done wrong. In a particularly trenchant passage, Walzer quips that “surely we have a right to expect more than melancholy” from the politician that, having faced a dirty hands situation, is aware that she has done wrong (Walzer 1973: 167).

As it happens, part of the symposium to which Walzer reacted concerns whether utilitarianism – qua comprehensive moral doctrine – has theoretical space for (let alone interest in, or curiosity towards) dirty hands situations.⁵ As Walzer pithily put it, the problem of dirty hands “may appear to utilitarian philosophers to pile confusion upon confusion” (Walzer 1973: 162). Surely pointing out to the complex moral life that generates the very possibility of dirty hands problems will also appear confused to the utilitarian. As I have argued elsewhere, it is not only utilitarian thinkers who may dismiss these sorts of themes and considerations: much of what Williams aptly called “the morality system” is in fact committed to versions of this dismissal (Williams 2006; Zaibert 2018). And I cannot agree more emphatically with Walzer’s claim that the view that moral dilemmas such as dirty hands are somehow not “real” (as, R.M. Hare, in the context of that seminal symposium, would have it) is not “at all comforting” (Walzer 1973: 161). Neither is there any daylight between Walzer and myself regarding his admission that the problem of dirty hands is not “only a political dilemma”, or about his further concession that “we can get our hands dirty in private life also, and sometimes, no doubt, we should” (Walzer 1973: 174).

Oddly, however, Walzer’s capacious initial frame – properly articulated in terms of general moral philosophy, as I have shown – quickly gives way to a secondary frame, articulated in much narrower political terms. Walzer’s article thus quickly veers in a different direction, and the initial frame that I find so valuable is replaced by – or at least straitjacketed within (or besides) – an altogether different frame. Concessions such as the ones quoted in the previous paragraph cede the stage to an approach that not only presupposes a certain sharp distinction between the political and other normative realms, but that it largely circumscribes dirty hands problems to the political realm. In fact, immediately after making those concessions, Walzer hastens to assert that the problem of dirty hands is “posed most dramatically in politics” (Walzer 1973: 174). This tense duality – that for Walzer the problem of dirty hands is a fundamental moral problem, though also somehow (“more dramatically”, etc.) a specifically political problem presents significant interpretative difficulties for understanding Walzer’s article. And, more importantly, it obscures the fundamental nature and importance of the problem of dirty hands.⁶

Consider how, for example, although Walzer concedes that we all – politicians and non-politicians alike – can face dirty hands dilemmas, he still insists that there is something special about the way in which politicians face these (or about what is at stake when it is politicians who face these dilemmas, etc.). This something special is, in Walzer’s opinion, the result of the fact that “the men who face the dilemma of dirty hands [upon whom he focuses] have in an important sense chosen to do so” (Walzer 1973: 164). Furthermore – and rather astonishingly – Walzer claims that the cases of dirty hands situations on which he is studiously focused “tell us nothing about what it would be like, so to speak, to fall into the dilemma” (Walzer 1973: 165, emphasis added). “Nothing”, really? I do not think that it takes any particular intellectual agility – or that it betrays any undue naivete – to expect that some of what may be true of some cases of dirty hands (the political cases – and even the very “thin” or “highly stylized” cases Walzer discusses) may tell us something about other cases of dirty hands (the non-political cases).

There are problems here that go beyond Walzer’s humility. First, “falling into the dilemma” is precisely what flows from the normal predicament of human life. Any human being with a modicum of sensibility will notice how she cannot escape sometimes falling into the dilemma. If it were true that Walzer’s article said “nothing” about this situation, Walzer’s article would be in this regard hard to distinguish from ordinary contributions that flatten moral experience.

Second, and more importantly, Walzer’s view that politicians have “chosen” to face dirty hands situations is not very persuasive. The sense in which for Walzer the men who face the dirty hands dilemma have chosen to do so is the result of their having decided to enter politics. Walzer thinks that deciding to enter politics somehow entails deciding to face dirty hands situations, and eventually deciding to at least sometimes actually getting our hands dirty: someone who entered politics committed to never getting her hands dirty will have a short career, and will accomplish little. Indeed, Walzer points out that “politicians often argue that they have no right to keep their hands clean” (Walzer 1973: 165). And here a remarkable alleged difference between politicians and non-politicians becomes quite prominent. For while Walzer admits that this is true of politicians, “it is not so clearly true of the rest of us”; after all, “probably we do have a right to avoid, if we possibly can, those positions in which we might be forced to do terrible things” (165). Tellingly, Walzer suggests that this alleged right that non-politicians may have (and that politicians, regarding their political activity, certainly lack) “might be regarded as the moral equivalent of our legal right not to incriminate ourselves” (Walzer 1973: 165).

Walzer’s analogy between the legal right protecting us against self-incrimination and the posited moral right not to face dirty hands dilemmas is remarkably weak. The posited moral right of which Walzer speaks simply does not exist in anything like the required (or even a useful) sense. There are two options worth considering here. We (regular people) can frivolously (recklessly, imprudently, selfishly, etc.) – in short, without good reason – choose to place ourselves in a position in which we then may be “forced to do terrible things”. Or we can so place ourselves for good reasons: a particular terrible thing is in fact necessary to avoid even more terrible things, etc.⁷ We clearly have no moral right to risk causing harm without good reason. But we also lack a moral right to refuse to cause harm when that is (ex hypothesi) the best, or likeliest, (or only) way of preventing greater harm – when we have a good reason.

Thus, we, as much as the politician, lack the moral right of which Walzer speaks. In fact, the Walzerian picture of moral (non-political) life that emerges from this passage is uninspiring, and at odds with the themes visible in the early pages of his article. Sometimes life puts us (non-politicians) in difficult situations – in situations in which it would be too self-servingly easy to invoke something like the imagined moral right not to get involved so as to ensure that we do not do “terrible things”. In those situations, our refusal to get involved could be criticized as much (albeit perhaps differently) as the refusal of the politician to get involved. The fact that it may be the politician’s “job” to get involved in some difficult situations, does not entail that us non-politicians can glibly sidestep involvement in these sorts of situations, by (almost magically) invoking this imagined moral right. So, here too, we, as much as the politician, lack the “moral right” of which Walzer speaks.

Consider the first case first; that is, the case of our frivolous (etc.) decision to put ourselves in a situation whereby we are forced to do terrible things. The talk about a “right” here is understated to a fault: it would not be quite a right that we would have not to place ourselves in such position: what we would have seems more like a duty to not so place ourselves. These “terrible things” are presumably terrible because (inter alia) they harm people – and we clearly have duties (both prima facie and otherwise; both legal and moral) not to harm people. If we turn to the second type of case – in which we do have a good reason to do this terrible thing (in order to avoid even more terrible things, etc.), it also strikes me as at least misleading to talk about a right. The good reason of which we speak here would, again, suggest that our intervention is less optional than our typical exercise of rights. I have a right to sell my car, or to grow a beard, say, and it is entirely up to me whether to exercise those rights or not; this seems quite unlike the case in which my risking doing a terrible thing is a way of avoiding more terrible things. In either of these cases, rather than a right, we appear to have something more akin to a duty: in the first case we – politicians or not – seem to have something like a duty to not intervene; and in the second case something like a duty to intervene.

It is important to point out that, for ease of exposition, I am following Walzer in talking about “rights” (and “duties”). More nuanced terms would better capture Walzer’s point, and would avoid the impression that we are somehow limited by the narrow deontic strictures of a morality of rights and duties: there is much more to morality than rights and duties. In fact, it is precisely the conviction that morality is indeed very broad and complex – so as to permit dirty hands situations – what I think animates, at bottom, Walzer’s position. It is not that the politician “has no right” to avoid dirt; it is, rather, that her avoidance of dirt can be objectionable, whether or not there was a “right” in play. But these reformulations would leave my main point unaffected: the non-politician’s refusal to get her hands dirty can be as objectionable (in whatever terms such objection is to be expressed) as the politician’s.

In contrast, and here whatever may be left of Walzer’s analogy between action and incrimination finally collapses, we clearly lack any duty not to incriminate ourselves. Notwithstanding the havoc that guilty pleas have wreaked in contemporary criminal justice systems, incriminating ourselves, in abstract, can be an admirable thing to do – a form of owning up to our past misdeeds, a form of apologizing, or even as a form of initiating a process of reconciliation.⁸ Yet, in the concrete context of the criminal justice system – where this legal right lives – and in light of necessary checks to the state’s awesome punitive power, we do need protection against being forced to incriminate ourselves. Unlike the problem of dirty hands, this right, predicated on systematic and institutionalized (if unavoidable) power asymmetries, emanates from fundamentally political phenomena. Nothing of the sort applies to the posited moral right not to place ourselves in positions in which we may do terrible things, in either of the two possible cases we have discussed. Except for contrived examples, there is little redemption in harming others (simpliciter); in contrast, confessing our misdeeds by way of incriminating ourselves has intrinsic redeeming potential. Again, nothing like the differential power between state and individual (and its institutionalization, etc.) necessarily obtains between random individuals.

The problem is that Walzer’s threadbare description “doing terrible things” in these passages is at odds with the pregnant description with which the article starts: a situation in which “doing terrible things” is wrong, but somehow (and deeply interestingly) also the right thing to do – a situation in which we cannot avoid getting our hands dirty. Properly seen, this is the predicament of moral life, when appreciated in its bewildering complexity – a complexity that Walzer (bucking contemporary trends) masterfully explores, but that at times oddly seems to dismiss or underestimate. This complexity simultaneously exposes the profundity of Walzer’s initial description, and the inadequacy of Walzer’s later, bald description. If the doing of terrible things is something we choose (choose “to be forced into doing”) for no good reason (frivolously, recklessly, etc.), then we lack a right not to do that, and we may have a duty to not do it. But if the doing of terrible things is part and parcel of the predicament of moral life which informs Walzer’s initial frame, then we may in fact be praised for our willingness to get our hands dirty, and we may have something resembling a duty to so doing.⁹

Even if I have succeeded in arguing that Walzer’s oscillation between the political and the moral is problematic and his analogy between a legal right and an elusive moral right is weak, perhaps some of Walzer’s other arguments for the alleged importance of privileging the political may fare better. I thus turn to this possibility next.

3 Difficult Choices and the Pervasiveness of Dirt

Walzer highlights three characteristics of politicians that bolster the alleged uniqueness of the political: that politicians act “on our behalf, even in our name” (Walzer 1973: 162), that they are the potential victims of “the pleasures of ruling”, which (we are told) “are much greater than the pleasures of being ruled” (Walzer 1973: 163), and that they are likely to turn to “violence and the threat of violence” (Walzer 1973: 163). I find Walzer’s treatment of these characteristics too quick, but he thinks that they – or rather: their combination – highlight both the peculiar station of the politician and the peculiar sense in which Walzer believes that deciding to devote one’s life to politics is itself a difficult choice. After all, before deciding to enter politics, we must be aware of the risks involved in the combination that our susceptibility to vanity and its (concomitant?) dangerousness generate as we act on others’ behalf. And yet, we forge ahead all the same: despite these risks “good and decent people still enter political life” (Walzer 1973: 164). Why?

Perhaps good and decent people discount the risks because, as Walzer eloquently puts it, “the politician has, or pretends to have, a confidence in his own judgment that the rest of us know to be presumptuous in any man” (Walzer 1973: 163). But this is not terribly helpful, and actually highlights the fragility of the distinction between the political and the non-political, for an unstable, dilemmatic situation looms. If, on the one hand, the confidence of which Walzer speaks is present before entering politics, then it simply cannot help us draw the distinction he wishes to draw. If, on the other hand, the confidence arises later (once one has entered politics) then it does not help explain how good and decent people overcome the risks of political life – that is, of course, unless we assume that these good and decent people are also short-sighted (unable to foresee how the confidence will eventually arise). In other words, problems arise either if this confidence is taken to be present before one decides to enter politics or if it is taken to arise only after we entered politics.

Now, there may of course be other ways of explaining how good and decent people look past the risks inherent in the decision to enter political life. But I focus on the instability of this confidence in order to highlight what I take to be an undeniable point: we can act on others’ behalf (or in their own name), we can be vain, and we can be dangerous, outside the political realm. We can be over-confident outside politics. It may even be granted that these features are typical of political life. But they are neither necessary aspects of political life, nor necessarily absent from other (non-political) forms of life. Thus, I do not think that they suffice in establishing the requisite distinction between the political and the non-political.

Similarly, for all I know Walzer is correct in that “no one succeeds in politics without getting his hands dirty” (Walzer 1973: 164). He may, perhaps, also be correct in that, as a matter of statistics, politicians face dirty hands situations more often than ordinary folk, or in that when politicians face these dilemmas, they face truly momentous situations, potentially affecting many people. And yet these facts (if that is what they are) are theoretically unimportant: non-politicians can find themselves in these sorts of situations as well – even if less frequently, or less momentously, than politicians do. Non-politicians, too, may need to get their hands dirty in order to “succeed” in life, and they may find themselves in situations in which their actions may affect many people. The broader point that I am trying to emphasize is that, independently of politics, no one can lead a recognizably human life without getting their hands dirty.

The most surprising point is that I do not think that a sharp distinction between the political and other normative dimensions is either crucial or even useful to Walzer’s deeper point. In a way this is already visibly through passages in Walzer’s paper. The fact that he oscillates between the two frames discussed above already suggest this much. In a way, the oscillation climaxes in the last section of the paper, when Walzer fruitfully identifies “three broad traditions of explanation” concerning dirty hands, which “derive in some very general way from neoclassical [Machiavelli], Protestant [Weber], and Catholic [Camus] perspectives in politics and morality” (Walzer 1973: 174, emphasis added). Take Machiavelli’s neoclassicism: notwithstanding the fact that Machiavelli, too, appears to focus on the political, Machiavellian action is evidently possible outside the political context. Non-politicians regularly invoke the mantra that “the end justifies the means”, and their invocations are as cogent as the politicians’. In fact, Walzer admits that Machiavelli’s “paradox” (i.e., the problem of dirty hands) “depends on [a] commitment” to moral (i.e. non-political) standards (Walzer 1973: 175). On a similar vein, Isaiah Berlin concludes his tour de force on Machiavelli suggesting that his “cardinal achievement” is, in the final analysis, “his uncovering of an insoluble dilemma”; grosso modo, this dilemma consists in the recognition that “entire of systems of value may come into collision without the possibility of rational arbitration” – and these values are unquestionably moral (Berlin 1979: 74).¹⁰ For Berlin, Machiavelli’s genius was, in part, that he did not conceive of two “autonomous realms of morals and politics” (Berlin 1974: 54).

Walzer’s treatment of the other two traditions for approaching dirty hands problems he identifies similarly reveals porous boundaries “separating” the political from the moral. Weber’s “good man with dirty hands” (Walzer 1973: 176) is forced to do wrong when he acts as a politician – but the wrongs are assuredly moral wrongs. Finally, in the context of his analysis of Camus Walzer is again explicit about the interplay between politics and morality: “[i]n most cases of dirty hands, moral rules are broken for reasons of state” (Walzer 1973: 179). But clearly they can be broken by other moral rules (and not just by “reasons of state”; reasons of state can also be broken by other reasons of state, etc.) As it turns out, it is the Catholic approach (exemplified, in Walzer’s mind, by Camus) that Walzer finds most attractive, and the explanation for this is surprisingly connected to punishment.

Before addressing this typically overlooked connection between dirty hands and punishment (even in Walzer’s own article), I wish to conclude this section by emphasizing that whether dirty hands is best seen as a political or a moral problem is not a minor terminological or methodological matter. To this end, I will contrast Walzer’s two hypothetical examples around which much of his article revolves against two actual (albeit very schematically presented) historical examples. The consideration of these historical figures shall allow us to see deep problems facing the effort to insist on a bright line distinction between the moral and the political.

Walzer’s first hypothetical is that of a politician who, in order to attain power and then advance an ex hypothesi morally respectable agenda, needs to strike shady deals and engage in suspect or downright immoral behavior – he needs to be willing to learn how to get her hands dirty – or, in Machiavellian terms, how to be bad. Walzer’s second hypothetical involves a politician who is already in power, but who, in the context of exercising said power, faces the prospect of achieving an ex hypothesi morally admirable end but only as a result of ordering the torturing of a human being. (I have just italicized “attain” and “exercising” because Walzer seems to attach some importance to this difference, but the difference is largely irrelevant both to my purposes here and to Walzer’s in his article: the second hypothetical can be seen simply as a ratcheted-up version of the first.) There is no denying that difficult choices such as those sketched in Walzer’s hypotheticals can indeed obtain. The problem, again, is that these are situations which non-politicians can also face.

Consider, in contrast, the (in)famous case of Chaim Mordechai Rumkowski, the Aelteste der Juden in the Nazi ghetto in Lodz (Litzmannstadt). Historians differ widely in their assessment of Rumkowski’s behavior during the holocaust, ranging from seeing him as a collaborationist to seeing him as a hero (Trunk 1972; Unger 2004). But one undeniable fact – crucial for current purposes – is that Rumkowski faced incredibly difficult dilemmas. If his behavior during the holocaust is to be in any way defended (a matter I am here expressly not adjudicating), the defense has to include the fact that he was willing to get his hands dirty. Yes, he had to work closely with the Nazis; yes, he had to crack-down on the general ghetto population; yes, he had to deport people to what he had reason to believe were places even worse than the already hellish ghetto, yes, he had to “rule” the ghetto tyrannically – but he did all these things in order to (in his mind) avoid greater evils.

Although assuredly not a philosopher, Rumkowski himself appeared to defend his behavior by invoking something like dirty hands. Upon being ordered by the Nazis to assist in the deportation (i.e., in the murdering) of Jewish children, Adam Czerniakow, Rumkowski’s counterpart in the Warsaw ghetto, committed suicide – after confessing in a note to his wife “I can no longer bear all this” (Hillberg et al. 1979: 23). Rumkowski boasted that, in contrast, he had the wherewithal to do whatever was necessary to save as many innocent Jews as possible. Unlike Czerniakow – whose actions Rumkowski saw as utterly ineffectual – he himself was willing to get his hands dirty indeed (Horwitz 2008: 295 ff.). Probably partly as a result of Rumkowski’s actions, the Lodz ghetto was the last to be liquidated by the Nazis – had the war ended only a few months earlier, Rumkowski could have been credited with saving thousands upon thousands of innocent victims.

Again, these admittedly rudimentary descriptions of Rumkowski’s and Czerniakow’s different reactions to the increased horror of the Nazi’s genocidal plans merely seek to illustrate the terrible dirty hands situations that these men faced. Where Czerniakow chose to end his life rather than dirtying his hands in certain ways, Rumkowski chose to cover himself in dirt. And it is not clear that Rumkowski’s dirt renders his actions worse than Czerniakow’s. One last time: I am not here adjudicating this matter; I am simply stressing that it is not at all clear whose actions are less bad.

What seems clear to me is that these men were not politicians. I know of no more brutal (if tacit) questioning of Rumkowski’s status as a politician than Primo Levi’s description: “[h]e had a carriage drawn by a skeleton nag in which he rode through the streets of his miniscule kingdom, streets crowded with beggars and mendicants” (Levi 1998: 63).¹¹ But perhaps a case can be made that these men were indeed politicians. At least Rumkowski seems to display the three marks of the politician Walzer highlights: he somehow acted in the name of the rest of the (mostly) Jewish population of the Ghetto, he was (very) vane, and he was (very) dangerous. Both these men held some form of authority over populations “under their control”; both had “cabinets”, appointed “judges” and “police forces”, issued “edicts”, and seemed to have had some of the characteristics of garden-variety political leaders. Of course (and thus the preceding scare quotes), these were mere caricatures of real cabinets, judges, and police forces. These “politicians” were prisoners of the Nazis – whatever power they may have had over other prisoners was utterly limited by the ultimate power that the Nazis ruthlessly wielded over all of them. As prisoners of the Nazis, moreover, they can hardly be said to have chosen this “political” life (if that is what it was).¹² Like the rest of the population over which they “ruled”, these “politicians” (appointed by the Nazis) were destined to be killed too – and the vast majority of them, including Rumkowski, met their horrible ends in Nazi gas chambers.

So, while I do not think that these men were politicians, it appears that whether or not they were is a complicated question. Precisely in light of this complexity a further objection to my decision to discuss them suggests itself: why choose complicated cases at all? If my point is that what Walzer appears to think is particularly significant/typical/obligatory in the political context can occur in non-political contexts, then why not offer examples that are clearly non-political? In order to avoid having to deal with the complicated contours of real political lives (or political stations, or offices, or roles), I could have, for example, tinkered with Walzer’s second hypothetical (involving the torture of a human being) along the following lines. Rather than it being a politician (or a quasi-politician) who is stipulated as engaging in torture, we imagine it is ourselves (stipulated to be non-politicians) who are, with our own hands (say), torturing a person. We can further stipulate that our reasons are as good (or as potentially good, or at least not obviously bad) as the politician’s in Walzer’s second hypothetical.

The problem is that the necessary assumption here – i.e., that torturing another human being in order to save innocent people, etc. can be a non-political act – may itself be questioned. It could be suggested that the moment that I (a putative non-politician) decide to torture a human being in order to save others (etc.), I become a politician, or at least that this particular act of mine is inherently political. The first of Walzer’s marks of the political would seem to apply to me in this example: I would be acting on others’ behalf (or even in their own name). And the two other marks that Walzer mentions may lag not very far behind: I may discover that I enjoy the pleasures of doing this, and I may develop a proclivity for turning to this sort of (violent) expedient in order to attain worthwhile results in the future, thus becoming dangerous.

While I am not sympathetic to this sort of strategy (some cases are clearly political, and others are clearly non-political), there are many complicated cases. And I choose the particularly complicated cases of Rumkowski and Czerniakow in order to highlight two interrelated difficulties facing emphases on the political such as Walzer’s. First, the existence of complicated cases emphasizes the elusiveness of any sharp demarcation between the political and other normative realms. It is hard to see the point of privileging the political, as Walzer does, if one cannot point to a clear distinction between the political and the non-political. In other words, if complicated cases are frequent (as I suspect they are), then assuming that the political is generally easy to distinguish from the non-political seems misguided. Second, we may end up defining the political in precisely those ways that would make it coextensive with the possibility of dirty hands. Self-servingly, whenever anyone faces a dirty hands situation, she would, eo ipso, be acting as a politician, or within a political situation. If such overly stipulative move is admitted, however, then the whole point of distinguishing the political from the non-political in the case of dirty hands becomes viciously circular: any case of dirty hands would be, by definitional fiat, political.

But there are more problems here. Suppose that, unlike me, you think that Rumkowski – and his “cabinet” members, and the leaders of his “police” force, etc. – were indeed politicians. Where do you stop? How about the “ordinary” prisoners who were not (or not directly) related to Rumkowski’s “administration”. Think, for example, of a relative of a leader in Rumkowski’s “police” force, utterly disconnected from the ghetto “administration” and in fact morally opposed to his relative’s decision to join said force. Still, it is easy to imagine how this person may find himself facing dirty hands situations: his very survival (or that of his loved ones) may have depended on accepting a favor from his better-connected relative. It strikes me as implausible to suggest that the moment in which this prima facie non-politician accepts (or requests) help from his relative, he (by definitional fiat) becomes a politician. Or think of another regular prisoner in the ghetto, who has no relatives in Rumkowski’s “administration” but who somehow schemes in ways that allow her to save those closest to her. Again, it strikes me as implausible to suggest that she thereby becomes a politician. Was Captain Vere a politician? Was Lieutenant Barney Greenwald?

This foray into hellish Nazi ghettos highlights the myriad ways in which the problem of dirty hands can manifest itself. Luckily, most of us have never had to face dirty hands situations whose consequences could even approximate those that Rumkowski and other prisoners in the Nazi ghettos faced. But that does not change that fact that human beings often find themselves in situations in which there is no completely clean course of action, even if the stakes are low. Paradoxically, perhaps, the extreme scenario of the Nazi ghettos helps make explicit how widespread such moral dilemmas are: virtually everyone there was susceptible to get her hands dirty. One of the points on which I have insisted here is that we face those dilemmas in non-extreme scenarios as well, even if less conspicuously or less dramatically.

Paying attention to the problem of dirty hands is one fruitful and mature way in which moral philosophers can avoid their self-incurred minority and its concomitant narrow approaches built around overly schematic thought-experiments or presupposing overly mechanistic solutions to moral problems. But assuming that dirty hands is essentially or predominantly (or most interestingly, etc.) political risks thwarting that progress. At bottom, dirty hands is a fundamental moral problem, and one that is often overlooked in moral theory.

4 Conclusion

Walzer credits Machiavelli with being “the first man” to discuss the dirty hands problem (even if Machiavelli did not use those very words) (Walzer 1973: 175). In a very interesting passage Walzer identifies a possible misreading of Machiavelli: it is a mistake to take the famous Machiavellian phrase “when the act accuses, the result excuses” to mean that the politician’s bad actions are “justified by the good results he brings about” (Walzer 1973: 175). For if that were so, then “it wouldn’t be necessary to learn what Machiavelli claims to teach: how not to be good” (Walzer 1973: 175). After all, in standard use to justify an action is to render it “not bad”: the justified actor is not “not being good”. Walzer’s point – that the dirty act is not justified (in this standard sense) – again strikes me as profoundly correct: if we are justified in doing something, then there is no need to be instructed in how not to be good.¹³ Failing to heed Walzer’s point would drastically water-down Machiavelli’s point – and it would make the problem of dirty hands vanish into thin air.

But there is another option available here: that the very process of justifying an action is immensely more complicated than typically assumed. On this interpretation, Machiavelli’s dirty actor’s act may be after all justified – except that to justify something is not quite to render it “not bad”. Typically, some reasons may render a certain action prima facie wrong; but those initial reasons may be defeated by other reasons that may ultimately justify the action. The crucial point is that the defeated reasons do not just disappear: they remain. While Walzer does not mention moral remainders in his article, he clearly has them in mind: “when rules [or reasons] are overridden [or defeated], we do not act or talk as if they have been set aside, cancelled, or annulled. They still stand [or remain] […]” (Walzer 1973: 171).) And this – the complexity of justification – may be (related to) the instruction Machiavelli thought he was imparting.

I do not adjudicate here between these two options, if indeed two options. Perhaps the lessons as to how not to be good are coextensive with (or otherwise very similar to) the lessons involved in rethinking the meaning of “justification”. The larger point is that Walzer’s searching views on the problem of dirty hands are revealed in their deep fundamentality only if we recognize them in their full splendor – namely, not as part of (narrow) politics, but broadly: as necessary parts of any life worth living.

This allows me to complete the circle that brings us back to punishment, since it is in that context that I have emphasized a certain naivete regarding justification. Punishment is an excellent example of the complexity we are discussing: the suffering we inflict is not “set aside, cancelled, or annulled” because we were (ex hypothesi) justified in inflicting it. To think otherwise is naïve and pernicious.¹⁴ And this may explain why punishment plays such an important (yet overlooked) role in Walzer’s article.¹⁵ Walzer prefers the “Catholic” approach to dirty hands (as illustrated by Camus) because, unlike the other two approaches, it “requires us at least to imagine a punishment or penance that fits the crime” (Walzer 1973: 179). The Catholic approach demands some “determinate” (Walzer 1973: 178) and “socially expressed” (Walzer 1973: 177) punishment – and this is the reason why Walzer prefers it.

Walzer concludes his article with a disquieting warning about the infinite regress that lurks in the background of his favored approach centered around punishment. Since the dirty actor’s action is not justified (or in my preferred terms: if it being justified leaves so many moral matters unresolved), we “must make sure he pays the price” of his action: he must be punished (Walzer 1973: 180). But we cannot ensure that he pays “without getting our hands dirty, and then we must find some way of paying the price ourselves”, a way of punishing ourselves (Walzer 1973: 180). I cannot possibly deal here with the formidable regress which Walzer identifies,¹⁶ but I just want to note a potential risk in Walzer’s take on it.

The protestant approach exemplified by Weber also prescribes a punishment, but in Walzer’s estimation, such punishment is too private and indeterminate, entirely left to the conscience of the tragic hero, thus giving us reason to suspect the politician with dirty hands of “either masochism or hypocrisy or both” (Walzer 1973: 177).¹⁷ Walzer prefers the Catholic approach over the Protestant approach – Camus over Weber – because only the former prescribes a “determinate penalty” for a “determinate crime” (Walzer 1973: 178). Emphasizing determinateness in this way seems in tension with the emphasis on the complexity of moral life that otherwise pervades Walzer’s article. Just as the notion of “justification” may be more complicated – less “determinate” – than typically assumed, so may be the notions of “crime” and “penalty”. Whether or not it is quite a regress, the problem with which Walzer ends his article, is real, important, and difficult. But it is most definitely not by simply enacting very precise penal codes, or by consulting (narrow) treatises in penal or political theory that we will find tools for dealing with it. Rather, it is in decidedly moral theorizing at its most general and abstract level, where, if we are lucky, we may find some of those tools.¹⁸