30 October 2007 Moral Luck: Optional, Not Brute1 Michael Otsuka Abstract: Moral luck is the phenomenon whereby one’s degree of moral blameworthiness or praiseworthiness varies on account of factors beyond one’s control. In Section I, I argue against claims that any or all forms of bad moral luck are ruled out on grounds of fairness. A form of luck that I call ‘option moral luck’ is not so ruled out. This form of moral luck is modelled on Ronald Dworkin’s ‘option luck’ in the domain of distributive justice, where option luck is roughly luck to which one has exposed oneself as the result of one’s voluntary choices and the contrasting category of ‘brute luck’ is luck that is unchosen and unavoidable. In Section II, I offer a non-fairness-based rejection of brute moral luck and defence of option moral luck. These two sections consider the phenomenon of moral luck on account of the way one’s impermissible actions turn out. In Section III, I explore the question of whether one might be subject to bad moral luck on account of the way one’s obligatory or permissible actions turn out.

Moral luck is the phenomenon whereby one’s degree of moral blameworthiness or praiseworthiness, either for what one has done or for the sort of person one is, varies on account of factors beyond one’s control. But what exactly is it to be morally blameworthy? We need to come to a clear understanding of what this is in order to be well placed to assess the plausibility of the claim that there is such a thing as moral luck.2 Some have advanced the claim that the attribution of a vice of character is a form of moral blame and that someone is worthy of such blame if such an attribution is justified. It has also been noted that whether one possesses various vices of character is not always under one’s voluntary control. If both of these claims are true – i.e., if our degree of viciousness is not always within our control, and if someone is worthy of moral blame if the attribution of a vice of character is justified – then we are subject to one form of moral luck.3 Robert Adams has argued for the existence of such moral luck, as he has defended the thesis that one might legitimately be unavoidably morally blameworthy for one’s attitudes,

1

[Acknowledgements to LSE, Oxford, Bled, Pittsburgh, Warwick, UCL, G. A. Cohen, Peter Graham, Alon Harel, Catriona McKinnon, Dana Nelkin, David Owens, Peter Vallentyne, Alex Voorhoeve, and Jonathan Wolff.] 2 A complete inquiry into the plausibility of this claim would also require an understanding of what it is to be morally praiseworthy. In this paper, I shall limit myself to an investigation of the question of whether we can be morally lucky or unlucky in our degree of blameworthiness. 3 We would be subject to what Thomas Nagel calls ‘constitutive luck,’ in which “the kind of person” one is depends on factors beyond one’s control, “where this is not just a question of what [one] deliberately [does], but of [one’s] inclinations, capacities, and temperament” (Thomas Nagel, ‘Moral Luck’, in his Mortal Questions (New York: Cambridge University Press, 1979), pp. 24-38, at p. 28). Nagel’s paper is a reply to a paper by Bernard Williams of the same name, which I discuss in the appendix of this paper.

1

temperament, or character.4 Adams is surely right that a person’s arrogance, callousness, ingratitude, tendency toward Schadenfreude, etc., may have been involuntarily formed and may remain beyond one’s voluntary control. He is also right to maintain that these traits of character are nevertheless properly regarded as vices. According to Adams, the attribution of a vice to somebody can properly be construed as a form of justifiable blame even if the vice is involuntary. Moreover, the blame in question is, as he would describe it, “moral” blame because it is condemnation for a moral failing. I agree with Adams that the attribution of a vice to somebody is the attribution of a bad-making property that is ethical in nature, unlike the non-ethical bad-making properties of stupidity, athletic ineptitude, or ugliness. But the attribution of a vice does not necessarily involve blame. For one can properly regard somebody as nasty and cruel while still leaving open the question of whether that person is blameworthy for being the way she is. One does not withdraw these attributions of vices upon concluding that the person is not to blame for being this way because she is a psychopath who was severely abused as a child and deprived of oxygen in the womb. Even in the light of these findings, she is still nasty and cruel, and these are still vices of character.5 But what is it to deem someone blameworthy for a vice of character above and beyond merely attributing this vice of character to that person? I believe that when one asks whether someone is blameworthy either for a vice of character or for what she has done, one wants to know whether something more than an attitude of horror, loathing, disgust, or pity, or a policy of avoidance, management, quarantine, or elimination, is called for. One wants to know whether, in addition or instead, a “reactive attitude” of a different sort is warranted. This attitude is aptly described as resentment or indignation.6 I believe, and shall assume, that someone is blameworthy for the way she is or for what she has done if and only if resentment or indignation on account of the way she is or what she has done would be warranted. Assuming the soundness of my account of blameworthiness, the question of moral luck that is the focus of this paper comes down to the following question: 4

Robert Adams, “Involuntary Sins,” Philosophical Review 94 (1985): 3-31, 21-4. This paragraph and the following one incorporate material from my “Incompatibilism and the Avoidability of Blame” Ethics 108 (1998): 685-701, at pp. 694-5. 6 Here, of course, I follow Peter Strawson. He regards indignation and resentment as impersonal and personal versions of the same attitude. He describes indignation as the “vicarious analogue of resentment”; it is, in other words, “resentment on behalf of another, where one’s own interest and dignity are not involved.” (See Strawson, “Freedom and Resentment,” Proceedings of the British Academy 48 (1962): 187-211, 199-200.) This description is, Strawson acknowledges, somewhat misleading, since indignation can just as well be on one’s own behalf as on behalf of another. It can be triggered by assaults on one’s own dignity as well as the dignity of others. Therefore, one doesn’t need the concept of resentment in order to fill the role of the first-person blameinvolving reactive attitude. The concept of indignation can fill that role. I don’t think, however, that indignation exhausts the reactive attitudes that are relevant to blameworthiness, since I believe that the concept of resentment captures a distinctive aspect of blameworthiness. 5

2

Is resentment or indignation warranted only to the extent to which that which is the grounds for such resentment or indignation is fully within the control of the individual who is the object of such resentment or indignation?7 If the answer to this question is ‘yes,’ then one denies the existence of moral luck. Why might one answer this question in the affirmative? One might do so because one thinks that things just wouldn’t be fair if we were subject to moral luck. In some remarks in Ethics and the Limits of Philosophy on the “peculiar institution” of morality, Bernard Williams writes that there is a “pressure” from within the perspective of morality “to require a voluntariness that will be total and will cut through character and psychological or social determination, and allocate blame and responsibility on the ultimately fair basis of the agent’s own contribution, no more and no less.”8 Here Williams is articulating, without endorsing, the thought that it would be unfair if how blameworthy one is were not fully within one’s control – if one’s degree of blameworthiness varied as the result of factors that are beyond one’s control.9 In Section I of this paper, I shall argue against claims that any or all forms of moral luck are ruled out on grounds of fairness. A form of luck that I shall call ‘option moral luck’ is not so ruled out. This form of moral luck is modelled on Ronald Dworkin’s ‘option luck’ in the domain of distributive justice, where option luck is roughly luck to which one has exposed oneself as the result of one’s voluntary choices and the contrasting category of ‘brute luck’ is luck that is unchosen and unavoidable.10 In Section II of this paper, I shall offer a non-fairness-based rejection of brute moral luck and defence of option moral luck.

I. Arguments for the unfairness of moral luck Note that the claim of fairness that Williams articulates is a claim about our degree of blameworthiness (and responsibility) in particular. It is not a claim about our ethical qualities more generally. This claim of fairness has not, for example, been made about our degree of 7

Even if you think that blameworthiness is something other than warranted resentment or indignation, I trust that you will nevertheless regard the question I have just posed as one that is worth pondering. I encourage those who do not endorse my account of blameworthiness simply to substitute the phrase “worthy of resentment or indignation” for every instance of the word “blameworthy” in the remainder of this paper. 8 Bernard Williams, Ethics and the Limits of Philosophy (Cambridge: Harvard University Press, 1985), p. 194. 9 Williams articulates a parallel thought about responsibility. In this paper, I shall not have anything to say about responsibility insofar as it is something distinct from blameworthiness. 10 See Ronald Dworkin, ‘What is Equality? Part 2: Equality of Resources’, Philosophy and Public Affairs 10 (1981): 283-345.

3

virtue and vice. It would, moreover, sound much less convincing if one were to make such a claim about our virtue and vice. It does not sound so convincing to insist that how vicious a person one is must be completely within one’s control because otherwise things would be unfair. As I noted above, we do not withdraw our attribution of a vice to someone when we discover that the vice is involuntary. Rather, we accept the view that some of our ethical qualities – namely, how virtuous or vicious one is – might genuinely be a matter of ineliminable luck. If, therefore, one maintains that one’s degree of blameworthiness must be completely within one’s control because otherwise things would be unfair, it can’t be on the general grounds that it would be unfair if any of one’s ethical qualities (e.g., how virtuous or vicious one is) were a matter of luck. One needs to appeal to something more specific about blame. Why would the norm of fairness under consideration apply to one’s blameworthiness but not one’s viciousness? One thought is that one’s viciousness is just a matter of the way one is. The attribution of a vice to someone is the attribution of a property to that person, where the presence or absence of this property is determined by the way things are, ethically speaking: by whether or not he is, in fact, nasty and cruel, for example. Moreover, the way things are, ethically speaking, is not up to us to stipulate, nor is it always within our control. In contrast to this descriptive picture of vice-attribution, one might paint a more prescriptive picture of blame, whereby regarding somebody as worthy of blame involves something akin to regarding someone as worthy of a penal sentence rather than merely a description of the way things are, ethically speaking. As in the case of punishment, the legitimacy of blame depends on the presence of morally relevant facts that are not up to us to stipulate or even completely under our control. One must in fact have killed somebody intentionally and with malice aforethought if blame of a certain sort is to be appropriate. But the thought is that blame, like punishment, essentially involves reprobation that goes beyond the reporting of an ethical discovery. To deem someone blameworthy, on this picture, is essentially to deem that person worthy of such reprobation. These practices of punishment and blame appear to share criteria of legitimacy that distinguish them from the more descriptive activities. “Well, that’s just the way things are. Life is unfair” might be an appropriate response to someone who complains about the unfairness of some being more nasty and cruel than others as the result of childhood factors beyond their control. But “that’s just the way things are. Life is unfair” sounds inappropriate as a response to an innocent person’s complaint of unfairness upon being convicted by mistake. It may also sound inappropriate as a response to the complaint that one is justifiably blamed for things beyond one’s control. It’s not always up to us how 4

the world is, even insofar as its ethical properties are concerned, but it’s up to us how we go about blaming and punishing others. And since the latter is up to us, we had better conduct ourselves fairly. Let me now distil much of what I have said to this point into the following argument against the existence of moral luck:

An argument for the unfairness of moral luck: P1. Resentment or indignation is warranted only if one is justified in expressing such resentment or indignation by condemning the person in question, where such moral condemnation is a sanction. P2. Being a sanction, such condemnation is subject to a norm of fairness that governs sanctions more generally and that allows for the imposition of a sanction only to the extent to which that which is the grounds for the sanction (i.e., that for which the individual is sanctioned) is fully within the control of the individual. Therefore, resentment or indignation is warranted only to the extent to which that which is the grounds for such resentment or indignation is fully within the control of the individual. I believe that this argument is valid: its conclusion follows from the premises. I believe, however, that this argument can be shown to be unsound by demonstrating the falsity of its second premise (P2). It can be shown that the norm of fairness that P2 articulates is too sweeping. A plausible norm of fairness would, at most, rule out unavoidable blameworthiness. It would not also rule out cases of moral luck in which one’s degree of blameworthiness is a function of factors beyond one’s control but where one could have avoided being blameworthy altogether. We need to differentiate the plausible claim that unavoidable blameworthiness is ruled out on grounds of unfairness from the further claim that it is also unfair that the degree of one’s avoidable blameworthiness vary as the result of factors beyond one’s control. According to the former, plausible claim, it would be unfair if there were nothing a person were capable of doing that would fail to attract blame. It would be unfair, for example, to be “damned if you do, damned if you don’t.” The latter claim amounts to the claim that, if one does something for which one is blameworthy even though one could have behaved decently and blamelessly instead, it would be unfair if the degree of one’s blameworthiness were not fully under one’s control. This latter claim is not plausible. 5

Consider Nagel’s examples of moral luck in which one is deemed more blameworthy and punishable if one’s drunken driving results in the death of a pedestrian rather than not or if one’s attempt at murder succeeds than if it fails. We can imagine that, like you, I was so drunk that I would have run over a pedestrian in the crossing, as you did. But fortunately for me, the crossing was empty. And Nagel has us imagine that the only thing that separates the shooter who kills from the shooter who misses his target is a bird that flies into the path of and deflects the bullet in the latter case. In these cases, one is deemed more blameworthy and punishable even though luck is the only thing that makes the difference between one’s killing someone or not. Note that, in these cases, even though the degree of one’s blameworthiness for one’s actions may differ depending on external circumstances, whether or not one attempts murder or drinks and drives is still up to oneself. Here one could have behaved – and could reasonably have been expected to have behaved – in a manner for which one would have been entirely blameless. Where is the unfairness here? Note that these cases are analogous to cases of what Ronald Dworkin has called “bad option luck” as opposed to “bad brute luck,” where there is nothing unfair about bad option luck. Option luck, as Dworkin defines it, is “a matter of ... whether someone gains or loses through accepting an isolated risk he or she should have anticipated and might have declined.” Brute luck, by contrast, is “a matter of how risks fall out that are not in that sense ... gambles.”11 Dworkin illustrate this distinction as follows: “If I buy a stock on the exchange that rises, then my option luck is good. If I am hit by a falling meteorite whose course could not have been predicted, then my bad luck is brute (even though I could have moved just before it struck if I had any reason to know where it would strike).”12 Now let us suppose that equally wealthy and capable individuals are presented with an equal opportunity to gamble in a casino. Suppose further that nobody is a compulsive gambler, and any wagers they make are fully voluntary. There would be no unfairness here if some grow richer than others because they have made high-stakes gambles at roulette that others could have made but chose not to. Nor would there be any unfairness when two people make equally risky high-stakes gambles at roulette, and one wins and the other loses. This contrasts with the unfairness when one person is unavoidably, and through no choice of his for which he had any evidence to believe he had reason not to make, struck down by a meteorite, whereas the other is spared. 11 12

Dworkin, Sovereign Virtue, chap. 2, p. 73. Ibid.

6

So why, then, is there any unfairness when two people voluntarily choose to gamble with their moral worth, so to speak, by choosing to try to shoot someone or to drive while drunk? Why isn’t the person who shoots and kills or drives and kills morally unlucky in just the same way in which someone who plays roulette and loses is financially unlucky? These are not cases in which one is unavoidably unlucky. Rather, they are cases in which one is unlucky as the result of one’s voluntary choice to gamble.13 I acknowledge that the analogy between gambling with one’s moral worth and gambling in a casino is imperfect. We cannot make every wager regarding our moral worth (i.e., our degree of blameworthiness) that is akin to an ordinary wager where we place our monetary worth on the line. It is not, for example, possible to take gambles that result in our being less blameworthy if we shoot and kill an innocent person than if the bullets miss. The point is, rather, the more modest one that no unfairness stands in the way of those who shoot and kill being more blameworthy than those who shoot and miss. This is because the risk that you would end up more blameworthy if, as luck would have it, you end up harming someone is a risk to which you voluntarily exposed yourself. The upshot of this discussion is that P2 overstretches in appealing to a norm of fairness “that allows for the imposition of a sanction only to the extent to which that which is the grounds for the sanction (i.e., that for which the individual is sanctioned) is fully within the control of the individual.” At the very most, an opponent of moral luck is only entitled to an invocation of a weaker norm of fairness that allows for the imposition of a sanction only if it was possible for the sanctioned individual to avoid being sanctioned altogether. In the light of this, we might reconstruct a more modest argument from fairness against moral luck, whose first premise is the same as that of the original argument but whose second premise, and therefore whose conclusion, diverge:

13

Note that Kant himself might have no quarrel with the claim that the shooter who kills is guilty of greater wrongdoing, worthy of more severe punishment, and perhaps even more morally blameworthy than the shooter who does not kill. “If we are of good will, for Kant,” according to John Gardner, “our moral agency is invulnerable to luck—that our actions turn out for the worse cannot transform our actions from right to wrong. We do the morally right thing come what may. ...But if we are not of good will, luck in the way our actions turn out can contribute constitutively to our breach of duty. It can be part and parcel of our moral wrongdoing, tarnishing our moral agency.” (Gardner, “The Wrongdoing that Gets Results,” Philosophical Perspectives 18 (2004): 53-88, at p. 63.) Barbara Herman offers a similar Kantian rationale for the more severe punishment of the shooter who kills than the shooter who doesn’t. On this account, the shooter who kills is responsible, and hence liable to punishment, for the bad outcome of his impermissible act because he has “stepped outside the protection of the moral” and made himself author of his own immoral law. Herman maintains, however, that the shooter who kills is nevertheless no more morally blameworthy than the shooter who misses. (Herman, “Feinberg on Luck and Failed Attempts,” Arizona Law Review 37 (1995): 143-49 at pp. 148-49.)

7

An argument for the unfairness of brute (but not of option) moral luck: P1. Resentment or indignation is warranted only if one is justified in expressing such resentment or indignation by condemning the person in question, where such moral condemnation is a sanction. P2*. Being a sanction, such condemnation is subject to a norm of fairness that governs sanctions more generally and that allows for the imposition of a sanction only if it was possible for the sanctioned individual to avoid being sanctioned altogether. Therefore, resentment or indignation is warranted only if it was possible for the sanctioned individual to avoid being sanctioned altogether. This argument is, as I have said, more modest than the first one. This argument rules out unavoidable blameworthiness – i.e., it rules out bad brute moral luck. It does not, however, rule out bad option moral luck. It does not, for example, rule out a drunk driver’s being more blameworthy because he kills, or an assassin’s being more blameworthy because his bullet meets its target, even when luck is the only thing that distinguishes the drunk driver or the assassin who kills from the one who doesn’t. Since this argument sets itself against bad brute moral luck while allowing bad option moral luck, one might be tempted to describe it as a defence of ‘moral luck egalitarianism’ – which is a term I have just coined to describe the view that one’s degree of blameworthiness must be free of the sort of unchosen inequality that is analogous to the unchosen inequality that is condemned by philosophers such as Dworkin, G. A. Cohen, Richard Arneson and Larry Temkin who adhere to an ideal of distributive luck egalitarianism.14 Both forms of luck egalitarianism set themselves against the inequality of some being in a less good position than others through no choice of theirs. In the context of moral luck egalitarianism, being in a less good position is defined in terms of one’s degree of blameworthiness rather than, as in the case of distributive luck egalitarianism, one’s welfare, resources, or capacities. So, according to moral luck egalitarianism, if one is more or less blameworthy than others, this must be on account of one’s choices rather than solely on account of factors beyond one’s control. In rejecting all bad brute moral luck, the position defended in the above argument is in fact stronger than moral luck egalitarianism, as the following example illustrates. If everyone were equally blameworthy at birth because of the irredeemable fact of original sin, such 14

See Ronald Dworkin, ‘What is Equality? Part 2’, G. A. Cohen, ‘On the Currency of Egalitarian Justice’, Ethics 99 (1989): 906-44, Richard Arneson, ‘Equality and Equal Opportunity for Welfare’, Philosophical Studies 56 (1989): 77-93, and Larry Temkin, ‘Equality, Priority, and the Levelling Down Objection’, in M. Clayton and A. Williams (eds.), The Ideal of Equality (Basingstoke: Palgrave, 2002), pp. 126-61.

8

blameworthiness would be unavoidable and hence a matter of bad brute moral luck. But there would be no inequality in bad brute moral luck. Such equality of bad brute moral luck would therefore be consistent with the realization of moral luck egalitarianism. It would, however, be condemned by the position defended in the above argument. Because P2* of this argument condemns all bad brute moral luck as unfair without also condemning bad option moral luck as unfair, P2* is more plausible than P2. What, however, is the ground for the norm of fairness articulated in P2*? I shall argue that it is not clear that there are any convincing grounds. Is it that sanctions involve the infliction of harm on the sanctioned person, and it is unjustifiable because unfair for an individual to have harm he cannot avoid inflicted upon him by another? This would be too sweeping a claim. For one thing, it would rule out various cases in which it is thought to be justifiable foreseeably to inflict harm on the innocent if that is necessary to prevent more individuals from being harmed – e.g., the trolley case. It would also rule out various justifiable cases in which people competitively harm innocent individuals. Suppose, to borrow an example from Nozick, that your coming to town and successfully courting an individual breaks the heart of an unsuccessful rival suitor and drives that rival to despair and loneliness.15 If you hadn’t been around, this other person would have won this individual’s hand in marriage. Moreover, you succeed in winning this person’s hand only because this other person is naturally less charming and attractive than you. It is a piece of bad luck for this other person that you came to town before this other person was able to secure this individual’s hand in marriage. Life is unfair, as they say. This sort of unfairness does not, however, support a norm that would decisively condemn your coming to town and courting this individual.16 The claim that it is unjustifiable, because unfair, for an individual unavoidably to have harm inflicted on him by another would also rule out being harmed in cases such as the following that are more akin that of moral blame: Suppose that someone is a nasty piece of work through no fault of his. He is vicious and cruel, but he cannot help being that way because he is the victim of fetal alcohol syndrome and a horrible upbringing in which he was constantly and severely beaten and sexually molested as a child. We might justifiably shun 15

See Robert Nozick, Anarchy, State, and Utopia (New York: Basic Books, 1974), p. 237. One might query whether the unlucky suitor is really harmed by the other individual, any more than the loser of a foot race is harmed by the successful efforts of the first-place runner to run faster than he. There are, however, other cases that it seems clearly appropriate to describe as instances of justifiable competitive harm: I might be justified on opening a restaurant in a given neighbourhood even though, in so doing, I clearly harm the owner of the restaurant across the street by depriving him of customers and causing him to go out of business.

16

9

this person and advise others to avoid his company because he is such bad news – so nasty and vicious. We may do so even if we know that such social isolation will make him miserable. In extreme cases, we may even lock this person up for the safety of others if he’s so nasty and vicious that he cannot help but lash out dangerously against others. We may lock him up even though this is clearly harmful to him. There is a clear sense in which it is unfair that such an individual be shunned or locked up, since he would be treated in this unpleasant fashion on account of factors beyond his control. We do not, however, affirm a norm of fairness that renders such shunning or locking up unjustifiable.17 When it comes to criminal penalties, we take a very different attitude. Take the case of strict liability to criminal punishment. H.L.A. Hart writes: There are now [criminal] offences (known as offences of ‘strict liability’) where it is not necessary for conviction to show that the accused either intentionally did what the law forbids or could have avoided doing it by use of care.... ‘[S]trict liability’ is generally viewed with great odium and admitted as an exception to the general rule, with the sense that an important principle has been sacrificed to secure a higher measure of conformity and conviction of offenders.18 Why do we reject strict liability for criminal punishment even though we do not reject strict liability to confinement of the dangerously insane? Joel Feinberg has argued that our objection to strict liability to criminal punishment is largely explained by the element of moral blaming involved in such punishment. We are far more willing to tolerate strict liability to stiff but non-criminal fines and penalties than we are willing to tolerate strict liability to what would otherwise be equally harmful criminal punishment through incarceration precisely because the penalties lack the expressions of indignation and reprobation that characterize punishment through incarceration.19 Strict liability to criminal punishment is disturbing because we especially abhor being unavoidably blamed. Feinberg’s explanation of our aversion to strict criminal liability points away from a more general norm of fairness that P2* posits. A more general norm of fairness, such as one 17

See T. M. Scanlon, What We Owe to Each Other (Cambridge: Harvard University Press, 1998), p. 285. Punishment and Responsibility (New York: Oxford University Press, 1968), p. 20. 19 See Joel Feinberg, “The Expressive Function of Punishment,” in Doing and Deserving (Princeton, N.J.: Princeton University Press, 1970), pp. 95-118. See also Scanlon, What We Owe to Each Other, p. 266, where he notes that punishment involves a blaming component. 18

10

that rules out the infliction of harm on an individual on account of things that are outside of his control, cannot differentiate between criminal punishment and the confinement of the dangerously insane. Rather, Feinberg’s explanation of our objection to strict criminal liability rests on an appeal to the very thing that the argument in question is trying to establish as sound: namely, the conviction that unavoidable blame is unjustifiable. So it will not offer non-question-begging support of this argument against moral luck. There is, I think, a deeper problem with P2* than the problems I have been describing to this point. That problem has to do with the questionable nature of a presupposition of P2*: that presupposition being that blameworthiness involves justifiable sanction. Recall that earlier we entertained the thought that how blameworthy one is might be subject to norms of fairness that do not apply to one’s degree of viciousness, because one’s blameworthiness involves the justifiable imposition of a sanction, whereas one’s viciousness is a matter of how things are. It is now time to consider the soundness of this link between blameworthiness and sanction, which is to say that it is now time to consider P1, which the two arguments against moral luck share in common. I shall now explain why I think P1 should be rejected. As Thomas Scanlon and Gary Watson have pointed out, one needs to draw a distinction between someone’s being blameworthy and the justifiability of blaming practices involving express condemnation or sanction.20 One might coherently conclude that someone was blameworthy even if one thought it unjustifiable for anyone to expressly condemn him or to sanction him. The drawing of such a distinction is supported by the following difference in meaning between “I think he is worthy of blame for φ-ing” and “I think he is worthy of punishment for φ-ing.” The latter implies the justifiability of imposing a penalty in a way that the former does not. The former might simply indicate that one thinks the individual worthy of resentment or indignation on account of what he has done without also indicating the justifiability of any penalty.21 Here, once again, I invoke an account of blameworthiness in terms of the justifiability of the reactive attitudes of resentment and indignation.22 Such

20

See Scanlon, What We Owe to Each Other, esp. pp. 276 and 285, but also p. 267, and Gary Watson, “Two Faces of Responsibility,” Philosophical Topics 24 (1996): 227-48, at p. 239. 21 As R. Jay Wallace writes: ‘Though blame often and naturally finds expression in sanctioning behavior, it is not necessarily so expressed – thus I can blame a person “privately” without expressing my response to anyone at all, much less sanctioning the person whom I blame (who may, anyway, be outside my sphere of causal influence).’ (Responsibility and the Moral Sentiments (Cambridge, Mass.: Harvard University Press, 1996), p. 56.) 22 Scanlon understands one’s blameworthiness differently as it’s being the case that a form of moral criticism applies to one because one has governed oneself in a manner that cannot be justified to others. According to

11

justifiability does not, I think, necessarily imply the justifiably of expressly condemning or otherwise sanctioning the object of these attitudes.23 In other words: on the account of blameworthiness that I embrace, to deem someone blameworthy for φ-ing is to deem him worthy of resentment or indignation for φ-ing. One can, moreover, deem someone worthy of resentment or indignation without expressly condemning that person, shunning him, penalizing him, or outwardly manifesting this blame in any other way.24 Someone who objects to moral luck on grounds of unfairness might try, at this point, to perform an end-run around my criticisms of the arguments under consideration for the unfairness of moral luck by advancing the claim that unavoidable blameworthiness is ruled out simply because it would be unfair if one were the object of resentment or indignation even though one could not have behaved in a manner that would not have elicited such attitudes. On this view, this would be unfair even if being the object of resentment or indignation were not accompanied by sanction or express condemnation. One who maintains that being the unavoidable object of resentment or indignation in itself, even apart from any sanction or express condemnation, is ruled out on grounds of unfairness must meet my earlier challenge of explaining why it is not similarly unfair to regard a person as vicious if he could not help being this way. One might try to distinguish blameworthiness from viciousness on the grounds that a person’s blameworthiness entails the warrant of certain reactive attitudes (even if not their outward expression), whereas a person’s viciousness is simply a matter of the way he is. One who draws this distinction might claim that the reactive attitudes are governed by a norm of fairness that condemns as unfair a person’s unavoidably being the object of a negative reactive attitude. One might hold this view while at the same time maintaining that there is no analogous norm of fairness that governs our mere descriptive judgments of the way someone is. Even if a person profoundly dislikes the fact that others believe that he is vicious, and his viciousness was unavoidable,

Scanlon, a person may be blameworthy in this sense even if there are decisive reasons against other people’s voicing this moral criticism. 23 I should note that Strawson does not himself draw such a sharp distinction between the reactive attitudes of resentment and indignation on the one hand and express condemnation or sanctioning on the other hand. {He speaks of the “partial withdrawal of goodwill which these attitudes [of disapprobation and indignation] entail, the modification they entail of the general demand that another should, if possible, be spared suffering” and writes that “the preparedness to acquiesce in that infliction of suffering on the offender which is an essential part of punishment is all of a piece with this whole range of attitudes [such as resentment and indignation] of which I have been speaking” (op. cit, p. 90 in Watson, 2nd ed.).} 24 A lack of outward manifestation might, however, involve a certain amount of dishonesty about one’s true feelings.

12

that doesn’t render their beliefs in his viciousness unfair. There may be no unfairness about these descriptive judgements because they are well-grounded in the evidence.25 As stated, this proposal is unsound because it is clear that not all of the negative reactive attitudes are governed by a norm of fairness that decisively rules out one’s unavoidably being the object of the attitude in question. The reactive attitude of distrust, for example, is not governed by such a norm, as Pamela Hieronymi points out.26 We do not regard it as unwarranted because unfair to continue to harbour an attitude of distrust towards some dishonest scoundrel upon discovering that he cannot help but behave in the way he does because of his formative circumstances. The above proposal for distinguishing between unavoidable viciousness and unavoidable blameworthiness fails because it fails to discriminate between reactive attitudes such as distrust and those of resentment and indignation. One might try to offer a principled means of distinguishing reactive attitudes such as distrust, whose unavoidability is clearly not ruled out on grounds of unfairness, from those such as resentment and indignation, whose unavoidability is allegedly rule out on such grounds. I think, however, that reflection on the following case should cast doubt on the prospects of unearthing a sound rationale for the claim that it is unjustifiable unavoidably to be the object of resentment or indignation because this is unfair.27 Suppose that you are filled with indignation upon reading an historical account of the deeds of a Nazi war criminal. As you read further, however, you become convinced that the man was a sociopath who could not have behaved decently instead. As I shall try to explain in the next section, having been so-convinced, you should acknowledge that indignation is unwarranted. But note that this transformation in your attitude would not plausibly be explained by the thought that it would be unfair to him for you to harbour this attitude of indignation now that you have discovered that he could not have behaved decently instead. How is this Nazi the victim of unfairness on account of the fact that some stranger in his distant, posthumous future feels indignant? We can suppose that, had he been told that a Nazi-hater such as you would be filled with indignation upon reading of his deeds, he either would not have given a damn or else have positively relished the prospect of your distress. In casting doubt on the claim that unavoidably being the object of resentment or indignation is unjustified because unfair, I do not commit myself to the sweeping claim that 25

I shall return to this point below. See Pamela Hieronymi, “The Force and Fairness of Blame,” Philosophical Perspectives 18 (2004): 115-148, at pp. 119-120, 136. 27 I am indebted to David Owens for this case. 26

13

resentment or indignation can never be unfair. We might properly condemn a person for unfairly concluding that someone is blameworthy if this conclusion is based on a culpably inaccurate assessment of the evidence.28 As I have indicated above, the very same accuracybased norm of fairness governs our judgments of a person’s viciousness. An accuracy-based charge of unfairness might be levelled against an irresponsibly hasty and ill-informed judgement that somebody is untrustworthy or possesses any other vice of character that he in fact lacks. This norm does not, however, rule out unavoidable viciousness. One might accurately conclude that a person is vicious even if his viciousness is unavoidable. Moreover, this particular norm of fairness does not, in itself, rule out unavoidable blameworthiness as unfair any more than it rules out unavoidable viciousness as unfair.

II. A Non-Fairness-Based Account of Moral Luck I would now like to conclude by presenting my own different approach to moral luck. I reach the conclusion that there is such a thing as option moral luck while denying the existence of brute moral luck. Yet my defence of this view that moral luck is optional but not brute does not make appeal to any norm of fairness. Rather, it rests on the claim that what determines whether someone is blameworthy, and how blameworthy he is, are simply facts having to do with the quality of his will, the state of his knowledge, the injury he has done, and the gratuitousness of his behaviour. My defence of the view that moral luck is optional but not brute draws on the account of blameworthiness in terms of the justifiability of the reactive attitudes of resentment or indignation that I have just sketched above, and which I first presented in “Incompatibilism and the Avoidability of Blame”. There I defended a ‘Principle of Avoidable Blame’, according to which “One is blameworthy for performing an act of a given type only if one could instead have behaved in a manner for which one would have been entirely blameless.”29 On behalf of this principle, I argued that resentment or indignation directed at someone on account of what he has done should dissipate upon discovery that the person could not have behaved any better. This is because what gives rise to resentment and indignation is the gratuitousness of bad behaviour, where such behaviour is “gratuitous,” as I shall define the term, when the person indulges in it even though he knows that he could have

28 29

See Hieronymi, “The Force and Fairness of Blame,” p. 129. ‘Incompatibilism and the Avoidability of Blame’, p. 688.

14

behaved decently instead. It is this gratuitousness that makes the behaviour in question so galling and worthy of resentment and indignation.30 Take any imagined pair of individuals who have behaved badly (e.g., who have maliciously injured another) and hold everything constant except for the fact that the one could have behaved less badly, and knew that she could have, whereas the other could not have behaved less badly though she had the justified but false belief that she could have. Against the first person, the victim of the injury would be warranted in thinking the following: “How dare you treat me this badly when you didn’t have to, and you knew you didn’t have to.” This objection carries force and provides grounds for resentment and indignation. But a victim is not entitled to such an objection when she knows that the aggressor could not have behaved less badly. Rather, the most she is entitled to say is: “I realize that you could not have behaved any less badly. But how dare you treat me this badly when you had the justified but false belief that you didn’t have to treat me this badly.” This accusation lacks force.31 On my account it is the existence of a coherent ‘how dare you’ complaint that grounds the warrant of resentment and indignation. Moreover, the lack of a coherent ‘how dare you’ complaint against the person who couldn’t have behaved less badly is not to be explained by the thought that such a complaint would be unfair, where the unfairness in question consists of an individual’s being burdened on account of factors beyond his control. Recall the Nazi war criminal example, which reveals that the appropriateness of resentment or indignation doesn’t vary depending on whether the object of such resentment or indignation is in any way burdened by being such an object. The Principle of Avoidable Blame that I endorse requires that whether or not one is blameworthy at all for what one has done be under one’s control. But it does not require that the degree of one’s blameworthiness be completely under one’s control. It is consistent with this principle that one be more blameworthy if one’s attempt at murder succeeds than if it fails, or if one’s drunken driving results in the death of a pedestrian rather than not, even if luck is the only thing that makes the difference between one’s killing someone or not. In these cases, even though the degree of one’s blameworthiness for one’s actions may differ depending on external circumstances, whether or not one attempts murder or drinks and

30 31

Ibid., p. 696. This paragraph incorporates material from ‘Incompatibilism and the Avoidability of Blame,’ pp. 696-7.

15

drives is still up to oneself. Hence, one could have behaved in a manner for which one would have been entirely blameless.32 What follows is a sketch of a defence of the positive claim that one’s degree of genuine blameworthiness is, in fact, a function of luck in the way things turn out in cases such as the drunk drivers or the people who shoot to kill. In these cases, blameworthiness was avoidable and one freely chose to do wrong. Here, one’s responsibility for killing someone adds to one’s degree of blameworthiness: one is worthy of resentment or indignation for harming someone even though one did not have to harm this person and could instead have behaved entirely decently. This is above and beyond the resentment or indignation both the successful and the unsuccessful agent are worthy of for their bad intentions and actions.33 The successful murderer or the drunk driver who kills is worthy of resentment and indignation for killing, and the failed murderer or the lucky drunk driver is not.34 It is crucial to the justification of resentment and indignation in the above cases that it is a reasonably foreseeable consequence of pulling the trigger or driving drunk that one might end up killing someone. In cases in which harm is an unforeseeable consequence of wrongdoing, by contrast, one is not worthy of resentment for the harm that results. Suppose, for example, that one’s drunkenness while driving results in one’s skidding off a mountain road and, by a series of flukes, this sets off an avalanche that causes a landslide, which buries a village, killing hundreds.35 Given the unforeseeability of these consequences, the driver would not be worthy of resentment for the deaths he has caused through his wrongdoing.36 32

This paragraph incorporates material from ‘Incompatibilism and the Avoidability of Blame,’ p. 700. It is not clear to me whether Strawson would agree. On his account of resentment and indignation are reactive attitudes “to the quality of others’ wills.” (“Freedom and Resentment,” p. 99 and/or 100 [check]) Presumably, the quality of a person’s will doesn’t vary depending on whether, as luck would have it, he kills someone. [Check to see whether Strawson limits resentment and indignation to such reactions to the quality of the will.] 34 Robert Kane has offered a different defence of the assassin’s responsibility for killing that appeals to the fact that he did what he intended to do and to the related claim that he endorses the outcome. (See his ‘Responsibility, Luck, and Chance,’ Journal of Philosophy. See also his The Significance of Free Will, pp. 5556 and 111-12. See Barbara Herman’s related account on p. 147 of her Feinberg piece.) But Kane’s explanation of option moral luck in this case fails to account for the fact that the drunk driver – who does not intend or endorse his killing of the one but bitterly regrets it – is also blameworthy for killing someone. 35 I’m grateful for Paul Snowdon for presenting an essentially similar case in conversation. [Someone at Pittsburgh might also have raised this issue.] There is an actual case which approaches this one in which a driver in Selby, England, fell asleep at the wheel because he was driving on too little sleep, which resulted in his vehicle’s going off the motorway and through a narrow gap in safety barriers. The vehicle ended up on a railway, whereupon a train collided with it and derailed. Six passengers and four rail crew were killed. 36 Peter Graham raises a related difficulty in which one knows that there’s a very small chance that one’s trivially wrongful behaviour will result in enormous harm. He thinks it wouldn’t be justifiable to blame the person for the harm in this case. I would say, in reply, that when an act is wrong in part because of the known risk that it involves, then one is blameworthy for the foreseeable harm to which the risk gives rise. If, however, the wrongness of the act has nothing to do with the risk – the risk, rather, is the sort of thing that one would be justified in imposing by so-acting were this act not wrong for other reasons – then one isn’t blameworthy for the resulting harm. 33

16

One might raise the following challenge to my claim that drunk drivers or assassins who kill are worthy of greater resentment for their foreseeable killings than those who did not kill. Suppose that each of two people grabs a gun which they justifiably believe is loaded, but, unbeknownst to them, one of them is empty of bullets. They each aim at the same victim head’s at point blank range and simultaneously pull the trigger.37 Should we really regard the one as more worthy of resentment than the other simply by virtue of the fact that the one person pulled the trigger of a gun whose chamber was empty, as luck would have it? No doubt many would intuit that we should instead regard the two as equally blameworthy. I would nevertheless maintain that the person who kills is more blameworthy. I believe that the apparent intuitive force of the claim that the two are equally blameworthy traces to a lack of vividness regarding which of the wrongdoers actually caused the death. To illustrate this point, consider another case in which each of two expert fencers grabs a sword and lunges at the victim’s heart simultaneously. They both justifiably believe that these swords are lethal instruments. Yet unbeknownst to them, one of them is a theatrical prop that is designed to harmlessly collapse upon impact. In this case it will be clear which of them drew blood. Moreover, I think it justifiable for those close to the victim to resent that person more than the other for actually piercing the victim’s flesh and grievously, mortally injuring him. He’s the murderer – the one who deprived them of the person they love. Now in saying that those who kill are worthy of a greater degree of resentment or indignation than those who do not kill in the cases of the drunk drivers and the shooters, I do not thereby commit myself to the claim that the former are worse people than the latter. Judith Thomson maintains that, when the difference between the lethality of their behaviour is simply down to luck, the person who kills is no worse a person than the person who doesn’t kill. Thomson also maintains we have no good grounds to regard the person who kills with more moral indignation than the person who does not.38 I agree with Thomson’s former claim but deny her latter claim. Though the killer is no worse a person than the non-killer, we have good grounds to resent him, or regard him with indignation, for killing that person. He is more blameworthy than the non-killer – because worthy of a greater degree of resentment, on account of what he has done – even though he is no worse a person. How bad a person is might depend only on factors internal to him such as his intentions, motives, beliefs, and volitions, rather than on any consequences of his actions that lie beyond his skin. But it does 37

This case is inspired by a slightly different case that Elizabeth Cripps presented in conversation. Judith Jarvis Thomson, “Morality and Bad Luck”, reprinted in Daniel Statman, Moral Luck (Albany, New York: State University of New York Press), pp. 195-215, at pp. 205-06.

38

17

not follow that he cannot ever be worthy of resentment or indignation for the bad consequences of his actions that are not entirely up to him. I think he can.39

III. Option Luck and the Obligatory and the Permissible In the previous section, I defended the claim that one might be subject to bad option luck on account of the way one’s impermissible actions turn out. In this section, I explore the question of whether one might be subject to bad option luck on account of the way one’s obligatory or permissible actions turn out. In particular, I want to consider whether one might be blameworthy if one’s obligatory or permissible action turns out badly, even though one would not have been worthy of any blame if it had turned out well. Consider the following case in which someone is morally compelled to take a risky course of action yet it appears that he would be worthy of blame if this gamble fails. This person is the head of a government and has reached the point in an international crisis where no permissible course of action is open to him apart from sending in the commandos to try to free some of his innocent countrymen who are being held captive in a foreign country. He has exhausted all diplomatic channels and would embolden the enemy if he did nothing at this point. Might he nevertheless be worthy of blame if the raid fails, his soldiers are shot, and all the captives are brutally executed as a result? As shall become clear in a moment, the letter of the Principle of Avoidable Blame does not itself settle the matter of whether a head of government might be worthy of blame whether or not he tries to rescue the captives. One needs to examine the underlying rationale for this principle in order to settle this. Consider a fairness-based rationale for the Principle of Avoidable Blame – i.e., a rationale according to which unavoidable blameworthiness is ruled out because it would be unfair to the individual if he were blameworthy no matter what. On such a rationale, it appears that there is nothing obviously wrong with a politician’s being unavoidably blameworthy whatever he does in such circumstances: since no individual is duty-bound to become a politician, and hence each politician was morally and well as literally free to choose not to become a politician, this is, from a temporally extended perspective, a case of bad option moral luck rather than bad brute moral luck. One could have 39

Kant might be on my side here. According to John Gardner: “...Kant tolerated the possibility that what contributes to the moral deficiency of one’s actions need not be something that also contributes to one’s status as a morally deficient person. ...The moral deficiency in her action, i.e. its being a breach of duty, could be constituted partly by the deficiency in her character, her ill will, and partly by the way that action turned out. ...The wrongdoing involved in breaching [the duty of care, for example] is only partly a matter of the wrongdoer’s carelessness, a deficiency that may be traced to her character. It is also partly a matter of the damage she does by her carelessness.” (Gardner, “The Wrongdoing that Gets Results,” p. 63.)

18

avoided being blameworthy altogether by refraining from becoming a politician. This case is, in one respect, similar to the cases of the drunk drivers and the assassins: they too could have avoided being blameworthy altogether by refraining from driving while drunk or pulling the trigger. But in another respect this case is quite different from these others. It is different insofar as here the action in question is obligatory rather than impermissible. Given what I say in Section I of this paper, I would reject a fairness-based rationale for the Principle of Avoidable Blame. I also resist the apparent consequence of a fairnessbased rationale that, having chosen to be a politician, a person might find himself in a predicament in which he will be blameworthy, as a matter of bad moral option luck, no matter what he does in executing the duties of his office. I think that the office itself – rather than the particular individual who freely fills it – is relevant in the following respect: we need a head of government, who is likely to find himself in predicaments in which (given the way the world is) he will need to take risky actions with potentially very bad consequences (e.g., sending in the military to try to rescue the captives).40 Since we need there to be someone to fill the role of head of government who will take such action – we would rightly blame a president or prime minister who was paralyzed by the prospect of failure and never took gambles – it seems unreasonable to blame a leader for the bad consequences of taking a gamble that we think he ought to have taken. If we believe that, given the facts available at the time, he ought to have taken a certain gamble, then the fact that this gamble is obligatory immunizes the gambler from blame for any bad consequences that ensue.41 Does the same holds for morally permissible but non-obligatory gambles? Suppose that what someone does is neither contrary to duty nor required by duty but merely permissible. Might this person be worthy of blame if things turn out badly? Consider a variation of the above example in which, given the knowledge a leader has at the time, the case for launching the rescue mission and not launching it is equally strong, and he has sufficient reason to do either. Suppose, moreover, that he knows for certain how things will turn out if he refrains from the rescue mission.42 In this example, even if we set aside

40

Cf. Williams: ‘But it is a predictable and probable hazard of public life that there will be these 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.’ (Bernard Williams, ‘Politics and Moral Character’, reprinted in his Moral Luck (Cambridge: Cambridge University Press, 1981), p. 60.) 41 Here and in the following paragraph I offer a more extended defence of conclusions that I reached in ‘Incompatibilism and the Avoidability of Blame’, pp. 700 -701. 42 Suppose, for example, that the captives have been unjustly convicted in a show trial of trumped-up espionage charges and sentenced to twenty year prison terms in a foreign land, and he knows that they will be held for the

19

complexities that arise from the fact that he could have chosen not to be a politician, and we instead assume that he had no choice but to become head of government, the Principle of Avoidable Blame does not settle the matter of whether this person is blameworthy if things turn out badly. This is because here the agent has an option (of refraining from attempting the rescue) that he knows that he will not be blameworthy for choosing. Therefore, his being blameworthy if the permissible gamble turns out badly is consistent with the Principle of Avoidable Blame, given that he had a known non-blameworthy alternative.43 Nevertheless, one might argue that even if he launches the rescue mission, and, as luck would have it, it fails disastrously, he is not blameworthy for this failure. He is not blameworthy because neither resentment nor indignation is appropriate, since he was fully justified in choosing as he did, given the information available at the time of his choice.

precise term of their prison sentence if he do not rescue them and that no harm above and beyond the loss of liberty consequent upon a twenty year confinement will come their way. 43 Consider, by way of contrast, a case in which one gamble is merely permissible rather than obligatory because there is another gamble that is also permissible. Here, as luck might have it, the person might be destined to lose either gamble he takes. If he would be blameworthy for losing a gamble in such a predicament, and if we further assume that the person has no choice but to be head of government, then blameworthiness would be unavoidable, contrary to the Principle of Avoidable Blame. Such cases are common. In predicaments such as the rescue mission under consideration, it is realistic to assume that great harm might befall the captives whatever one chooses: if one refrains from trying to rescue them, their captors might eventually kill them or they might eventually set them free; but if one tries to rescue them, this attempt might fail and trigger a slaughter of the captives by their captors who would otherwise have set them free.

20

Appendix on Williams on Moral Luck Bernard Williams say that it would not be irrational for a lorry driver to feel ‘agent-regret’ for having involuntarily run over a child ‘through no fault of his’. There would, in fact, be something amiss if he did not feel such regret. (p. 28)44 So-called agent-regret is akin to remorse, where ‘“remorse” is what we have called “agent-regret”, but under the restriction that it applies only to the voluntary’ (p. 30). Warranted agent-regret does not imply that the person is blameworthy. That it does not is revealed by Williams’s assertion that the lorry driver’s agent-regret is grounded in ‘something special about his relation to the happening, something which cannot merely be eliminated by the consideration that it was not his fault’ (p. 28). Though the lorry driver is not blameworthy, he may be liable to recompense or restitution: ‘The lorry-driver may act in some way which he hopes will constitute or at least symbolise some kind of recompense or restitution, and this will be an expression of his agentregret.’ (p. 28) Two comments: (i) Such liability might appear to be an unfair instance of bad brute moral luck, though it might instead be described as bad option luck, since it is voluntarily-incurred cost of engaging in a dangerous activity. (See my above discussion on moral luck for the way things turn out when one acts permissibly.) (ii) Suppose that, by some extremely rare chaos-theoretical cascading that nobody could have predicted, my opening my window one evening caused Hurricane Katrina to form. When informed of the effects of my agency, I would not feel any agent-regret (which is, recall, analogous to remorse). What’s the difference between this case and the case of the lorry driver? I think it’s that one knows that lorry driving is a dangerous activity, whereas it is impossible to know that window opening is a dangerous activity. Williams says of Paul Gauguin that: even if Gauguin can be ultimately justified [in his choice to abandon his wife and children in order to devote himself to a life as a painter], that need not prove him with any way of justifying himself to others, or at least to all others. Thus he may have no 44

Williams also says that there is all the more reason to feel agent-regret ‘in cases where agency is fuller than in such an accident, though still involuntary through ignorance’. (p. 28) (All parenthetical page references are to Williams’s essay ‘Moral Luck’ in his Moral Luck (Cambridge: Cambridge University Press, 1981).)

21

way of bringing it about that those who suffer from his decision will have no justified ground of reproach. Even if he succeeds, he will not acquire the right that they accept what he has to say . . . (pp. 23-24) If the justifiability of reproach implies that blame is justified, then this case differs from the lorry driver case, where blame is not justified. Williams claims that Nagel holds that ‘Gauguin may not be able to justify himself to others, in the sense that they will have no justified grievance’ and therefore that ‘Gauguin’s most basic retrospective feelings do not have to be moral’ (p. 37). Williams denies this inference, since he rejects a ‘strong assumption about the nature of ethical consistency, to the effect that, if someone has acted justifiably from a moral point of view, then no-one can justifiably complain, from that point of view, of his so acting’ (p. 37). Here are his grounds rejecting this ‘strong assumption’: But this as a general requirement is unrealistically strong, as can be seen from political cases, for instance, in which we can have reason to approve of the outcome, and of the agent’s choice to produce that outcome, and of his being an agent who is able to make that choice, while conscious that there has been a ‘moral cost’. It is not reasonable, in such a case, to expect those particular people who have been cheated, used or injured to approve of the agent’s action, nor should they be subjected to the patronizing thought that, while their complaints are not justified in terms of the whole picture, they are too closely involved to be able to see that truth. Their complaints are, indeed, justified, and they may quite properly refuse to accept the agent’s justification which the rest of us may properly accept. The idea that there has been a moral cost itself implies that something bad has been done, and, very often, that someone has been wronged, and if the people who have been wronged do not accept the justification, then no-one can demand that they should. It is for them to decide how far they are prepared to adopt the perspective within which the justification counts. (p. 37) Williams notes that there are important differences between Gauguin and politicians but he nevertheless concludes that we cannot infer from the fact that Gauguin’s justification ‘would not properly silence all complaints’ that this justification is not a moral justification. (pp. 389)

22

Assuming that neither Gauguin nor a politician would be blameworthy for not abandoning his family or for not going into politics, neither case qualifies as a counterexample to the Principle of Avoidable Blame. One might, however, draw on Williams’s reasoning from the above passage in order to try to defend the possibility of unavoidable blameworthiness. One might invoke cases in which one is obliged to do, and would be blameworthy for refraining from doing, something that nevertheless gives rise to a justified complaint from the perspective of the victim of one’s agency. One might go on to argue that, since he has a justified complaint, the victim is also justified in blaming his victimizer. I would resist this last claim. It is important to draw a distinction between having a justified complaint and being justified in reproaching (i.e., blaming) the agent. We can acknowledge the former without acknowledging the latter. The justified complaint might, for example, track nothing other than the fact that one’s rights having been infringed, from which it follows that one is owed compensation. A rights-infringement counts as a ‘moral cost’ since ‘something bad has been done’ when there is such an infringement. It does not follow, however, from the fact that one’s rights have been infringed that one is justified in blaming the rights-infringer, since such infringement might have been justified.45 What if the victim is justified in complaining that he has been wronged? Does it follow from this that the agent has done wrong, all things considered? Arguably this does not follow, as the justified complaint that one has been wronged might track nothing other than the fact that one’s rights have been infringed. Yet it does not follow from the fact that one’s rights have been infringed (as opposed to violated) that the infringer has done wrong, all things considered (as opposed to pro tanto). Suppose that the agent has done wrong, all things considered. Does it follow from this that the agent is blameworthy? Contrary to those who affirm the possibility of blameless wrongdoing, I think it does. I therefore deny the possibility of unavoidable wrongdoing as well as unavoidable blameworthiness. I believe that unavoidable wrongdoing is impossible for substantive normative reasons: a set of moral requirements whereby everything is prohibited, while conceptually possible, would be an unreasonable set of requirements. Moreover, moral requirements are genuine (i.e., actually binding) only if reasonable.

45

Here I track Judith Thomson’s distinction between and infringement and a violation of a right. A rightsviolation is, by definition, an unjustifiable rights-infringement. Some rights-infringements are, however, justifiable. (See Thomson, The Realm of Rights (Cambridge, Mass.: Harvard University Press, 1990), ch. 4.)

23

Consider the following analogy between morality and legality. An ideally just legal system cannot be one in which illegality (i.e., legal impermissibility) is unavoidable. If our current legal system implies that someone might flout the criminal law no matter what, we should take that to be a defect in need of remedy. Perhaps, given our limitations, the best legal system we can devise will be such that someone will break the law no matter what he does. But the fact that we can’t do better doesn’t make this other than a defect of this system. It would amount to a reification of a system of laws as something other than a human construct that we can reconstruct to think that an ideally just system of laws might be one in which illegality is unavoidable. In expressing the related possibility that moral wrongdoing might be unavoidable, Nagel famously says that the world might not just be bad; it might also be evil. But if, to follow Kant, morality, like laws, is legislated, then this is to make the mistake of thinking that there is an arbitrary and capricious lawmaker who holds sway over us from on high. Rather it is we, according to Kant, who legislate morality autonomously. Of course, if we are to legislate well, we must be guided by mind-independent values. In this respect I depart from Kant. Nevertheless, it is still up to us to promulgate the most reasonable set of laws in the light of these values. And the most reasonable set of laws will not be one in which impermissibility (whether it be legal or moral) is unavoidable.

24