This post contains Part 7 (“Resultant Luck”) of my dissertation prospectus, written in pursuit of my Ph.D in philosophy at the University of Colorado at Boulder and submitted to my committee in early December 2007. The full prospectus is available in PDF format and as an MS Word file. Comments and questions are welcome. While they won’t change the prospectus, they might be of use as I write the dissertation over the next year.
Resultant Luck
Resultant moral luck is “luck in the way one’s actions and projects turn out.”[85] Nagel’s basic claim is that a person’s moral record is influenced by the outcomes of his actions, yet those outcomes are not wholly of his own doing but often substantially influenced by factors outside his control.[86] The proposed cases of resultant luck fall into three broad categories: attempted wrongdoing, decision under uncertainty, and negligent action.
In attempt cases, a person is blamed and punished more severely for the successful completion of some wrongful action than for a mere attempt–even when the difference between success and failure is wholly due to luck.[87] This form of resultant moral luck is most easily found in the standard legal practice of punishing attempted crimes less severely than completed crimes. So as Nagel observes, “the penalty for attempted murder is less than that for successful murder–however similar the intentions and motivations of the assailant may be in the two cases.”[88] In such cases, the assailant’s culpability might depend on “whether the victim happened to be wearing a bullet-proof vest, or whether a bird flew into the path of the bullet–matters beyond his control.”[89] Conversely, virtuous actions may be praised and rewarded more if successful (e.g., if John rescues the baby from the burning building) than if thwarted by luck (e.g., if John drops the baby from a fourth story window due to an explosion behind him).[90]
In uncertainty cases, the agent knowingly takes some inherently risky action, such that the outcome cannot be predicted with any reasonable confidence in advance, and the agent is morally judged based on that outcome.[91] For example, “someone who launches a violent revolution against an authoritarian regime knows that if he fails he will be responsible for much suffering that is in vain, but if he succeeds he will be justified by the outcome.”[92] Bernard Williams’ famous case of Gauguin involves similar moral uncertainty: Gauguin’s abandonment of his family seems justified if he succeeds in painting in Tahiti but not if he discovers his talents to be inadequate.[93] According to the advocates of moral luck, the only moral judgment possible at the moment of decision in such cases is that the agent will be blamed if he fails and praised if he succeeds–because the outcome determines what the agent did, e.g., launching a glorious revolution or a failed bloody coup.[94]
In negligence cases, a person is blamed and punished more when his careless action causes a worse outcome, even though forces beyond his control determine that particular outcome.[95] Imagine, for example, two identical truck drivers, both of whom are long overdue for a brake inspection.[96] One drives safely home through uneventful traffic. The other is forced to brake suddenly to avoid a child darting across the street; when his brakes fail, he kills the child. The second driver blames himself far more than does the first, as do other people. Yet, Nagel observes, “the negligence is the same in both cases, and the driver has no control over whether a child will run into his path.”[97] Similarly, “if one negligently leaves the bath running with the baby in it, one will realize, as one bounds up the stairs toward the bathroom, that if the baby has drowned one has done something awful, whereas if it has not one has merely been careless.”[98] So the negligent person is blamed more or less based on an outcome beyond his control.
To resolve the apparent conflict between luck and responsibility in these puzzling cases of resultant moral luck, the control and epistemic conditions for moral responsibility must be extended beyond actions to outcomes. Moral responsibility for the outcome of some action requires more than just that the action be voluntary. The action also must be the salient cause of the outcome, and the outcome must be voluntary too. The application of those three conditions to cases of resultant moral luck yields surprising results; the agent is sometimes but not always responsible for the actual outcomes of his actions in those cases. So let us first briefly examine the three conditions for moral responsibility for outcomes, then apply them to cases of resultant moral luck.
First, to be morally responsible some outcome, the agent must be morally responsible for his original action: he must have acted voluntarily according to the control and epistemic conditions.[99] So Joe cannot be blamed for missing a critical meeting at work if he’s rushed to the hospital with a heart attack, as the action that caused his absence was wholly involuntary.
Second, moral responsibility for some outcome requires that the action be the salient cause of the outcome, i.e., the unusual factor operating against the background of ordinary causes.[100] So the salient cause of the explosion in a machinist’s shop would not be the commonplace sparks from welding that ignited it but rather the gas leak. The person responsible for the explosion would not be the welder but rather the person who cut the gas line. On a repair job for a leaky gas line, however, the gas would be the to-be-expected background condition, such that the careless worker who created the spark would be morally responsible for the resulting explosion. In general, a person ought to cultivate his knowledge of the ordinary causes that operate in his environment and affect his endeavors, as well as protect and expand his capacities to act on that knowledge. If he fails to do so, a person is properly blamed for voluntarily placing his projects, his flourishing, and even his life in jeopardy.[101]
Third, moral responsibility for some outcome requires the outcome to be voluntary, even if not desired or intended. The agent must (1) be able to produce the outcome or not and (2) know that his action might plausibly produce such an outcome–or be incapable and/or ignorant voluntarily. In other words, for some outcome of an action to be voluntary, the agent cannot be involuntarily incapable or ignorant with respect to that outcome. So if a professor gives all his students Bs regardless of the quality of their work, then those students might be praised and blamed for their work in the class but not for the outcome thereof, namely their grades. Since they could not have done better or worse by their own actions, they were involuntarily incapable with respect to that outcome. Similarly, a woman cannot be blamed for her family’s food poisoning absent some reason for her to suspect the ordinary-looking bag of spinach used for the dinner salad to be infected with E. coli. Her ignorance is involuntary–unlike the ignorance of the person who dismisses news reports of tainted spinach as mere scare-mongering by the carrot lobby. In short, this third condition precludes moral responsibility for outcomes that the agent cannot reasonably avoid or predict.
So what do these conditions for moral responsibility for outcomes tell us about moral responsibility in the cases of attempt, uncertainty, and negligence? As we shall see, the first condition (voluntary action) is satisfied easily, but the second condition (salient cause) and third condition (voluntary outcome) are only sometimes satisfied.
Applying the first condition, the action in cases of attempt, uncertainty, and negligence is clearly voluntary. The agent satisfies the control condition since he has the power to act or not: the hit man can squeeze the trigger or not, Gauguin can abandon his family for Tahiti or not, the mother with the child in the bath can leave the room or not. The agent also satisfies the epistemic condition since he’s aware of the basic character of his action, whether malicious, risky, or negligent.[102] So in the cases of resultant luck, the voluntary action satisfies the first condition for moral responsibility for outcomes. That shows that the person can be praised or blamed for his original action, even if ultimately not for the outcome thereof. Speaking generally, since an action is distinct from its ultimate outcome, it can and ought to be morally judged on its own merits, considering the information and alternatives available to the person at the time, regardless of the outcome.[103]
The second and third conditions for moral responsibility for outcomes yield far more complex and interesting results than the first. In some but not all cases of resultant moral luck, the action is the salient cause of the outcome and the outcome is voluntary. In those cases, the outcome corresponds to the moral character of the action. To simplify the analysis of these cases somewhat, these two conditions will be applied together for each of the three kinds of resultant moral luck: first attempts, then negligence, and finally decision under uncertainty.
In cases of attempt, the second and third conditions are only satisfied when the attempt produces its intended outcome, not when the attempt fails. When a person deliberately pursues and successfully achieves some end, his actions are the salient cause of the outcome. For example, when the hit man kills his intended victim, his actions in pursuit of that end are the salient cause of her death. Similarly, when the firefighter rescues the child from the burning building, his efforts are the salient cause of that life’s being saved. The outcome in such cases is also more than voluntary: it’s actively, deliberately intended. So in cases of successful attempt, the agent is clearly responsible for the outcome of his action. In contrast, when the attempt fails due to some intervening factor, that factor is the salient cause of the actual outcome. So the bird that flies into the path of the hit man’s bullet is the salient cause of the continued existence of the intended victim. And the explosion that rocks the burning building is the salient cause of the death of the child. Those outcomes are also not voluntary but rather directly contrary to the wishes of the agent; he will regret the outcome. Consequently, the hit man is not responsible for the good outcome of his attempted murder and the firefighter is not responsible for the bad outcome of his attempted rescue. The hit man can be thoroughly blamed for his depraved action, for putting his intended victim at great risk, for putting him in fear for his life, and more–but not for killing him (since that didn’t happen) and not for failing to kill him (since he didn’t cause that). The same basic analysis applies to praise for the firefighter.
In cases of negligence, the third condition (voluntary outcome) is satisfied whatever the outcome, whereas the second condition (salient cause) is satisfied only when the negligence produces its expected kind of harm. As concerns the third condition of voluntary outcome, by acting negligently, a person fails to act with the care required to ensure some particular outcome. Instead, he permits the outcome of his action to be determined by the random forces in his environment. Genuinely negligent action is voluntary: the person must know, even if only dimly, that he is acting carelessly instead of taking due care. As such, the outcome of negligent action is clearly voluntary, even if greatly regretted thereafter. The mother who leaves her child in the bath alone could ensure his safety by remaining in the room, so she’s not involuntarily incapable with respect to the outcome. She’s also aware that an unattended child might drown, so she’s not involuntarily ignorant with respect to that outcome. In fact, the negligent person renders himself incapable voluntary: he willfully relinquishes control over the outcome of his action, allowing luck to determine what happens. As concerns the second condition of salient cause, the causal influence of luck in such cases does not automatically diminish the moral responsibility of the negligent person. When the negligence produces no disasters, as when the mother finds her baby alive and well in the tub, she deserves no credit for that outcome because her negligence is not the salient cause of the baby’s safety. Then she can only be blamed for acting negligently, i.e., for needlessly risking harm to her child. However, if the baby does drown, then her negligence is the salient cause of that outcome. In that case, she would be culpable not just for acting negligently but also for the death of her child. For these reasons, the negligent person is responsible for the to-be-expected harms caused by his actions but not for the lucky avoidance thereof.
In cases of decision under uncertainty, the person is responsible for the outcome when he can succeed or fail and does so by his own efforts. The person satisfies the second condition when he is the salient cause of his own success or failure. So Gauguin could be responsible for becoming a world-class painter (or not) when due to his own choices and actions, but not when due to some external circumstances thrust on him.[104] So if his move to Tahiti was critical to his development as a painter, then Gauguin can take credit for his ultimate success. Yet if he failed in Tahiti due to a devastating injury to his dominant hand during a random criminal assault, then he cannot be blamed. In either case, however, he could be blamed for the negative effects of his departure on his family. The application of the third condition of voluntary outcome is somewhat more complicated. The person acting in uncertainty satisfies the epistemic half of that condition: he knows the range of possible outcomes, even if unable to determine their respective probabilities. However, the agent may or may not be able to satisfy the control half of that condition: that depends on the capacities of the agent. Gauguin may well have the talent to make himself into a world-class painter, but he spends his time in Tahiti drinking his days away. In that case, he would be culpable for the outcome, i.e., for his failure to become a world-class painter. Yet imagine that Gauguin’s talents are truly inadequate, such that the move to Tahiti can and does only marginally improve the quality of his work. In that case, he would not be responsible for his failure to become a world-class painter because he lacked the requisite capacity. That result is surprising but correct: he should not be blamed for failing to do the impossible. That does not render him exempt from blame, however. Depending on the wisdom of his decision, he might be culpable for a lack of good judgment concerning his own talents.[105] Then Gauguin could be blamed for trying to become a world-class painter, particularly for the harms he inflicts on himself and others in the process of that pursuit, but not for failing to achieve that goal, since he’s not capable of success.
So far, the development of the three conditions for moral responsibility for outcomes and their application to cases of attempt, negligence, and uncertainty seems to have confirmed that the problem of resultant moral luck is genuine. Indeed, the analysis shows that a person’s responsibility for the actual outcomes of his actions may depend on the influence of luck. In fact, however, a person’s moral responsibility for the actual outcome of negligent, attempted, and uncertain actions is wholly distinct from the proper judgment of that person’s actions or character. Because the lucky person is not any less dangerous to our projects and interests just because the disaster that his actions might have caused was forestalled by luck, he cannot be judged morally better than the unlucky person. The significance of moral responsibility for outcomes, at least in these cases involving luck, is that it determines that for which the person must atone (if he acted wrongly) or that for which he may reap the rewards (if he acted rightly). In particular, the person who acted wrongly yet avoided disaster by luck has less damage to repair than his unlucky counterpart.[106] That influence of luck is not morally problematic: no principle of morality dictates that people who do morally equivalent acts will have exactly the same effects in the world to remedy or enjoy. In addition, this distinction between blame and atonement promises to explain the seemingly problematic influence of luck in legal judgments, in that damages (e.g. for negligence) and punishments (e.g. for attempts) are (and ought to be) determined not only by the wrongness of the original act but also by the actual harms done by the defendant for which he must atone. As with moral judgments, the influence of luck in such legal judgments would not be unjust.
Notes
[85] Nagel 1993, p. 60.
[86] Nagel 1993, pp. 61-3.
[87] Nagel 1993, p. 61.
[88] Nagel 1993, p. 61.
[89] Nagel 1993, p. 61.
[90] Nagel 1993, p. 58.
[91] Nagel 1993, pp. 61-3.
[92] Nagel 1993, p. 63.
[93] Williams 1993, pp. 37-41.
[94] Nagel 1993, pp. 62-3; Williams 1993, pp. 38-9.
[95] Negligent action involves some willful indifference of the agent to the risks of his actions. In contrast, the agent in cases of decision under uncertainty knows the risks but lacks the power to substantially mitigate them. So the original action in cases of negligence is always wrong; that’s not true of decisions under uncertainty.
[96] Nagel 1993, p. 61.
[97] Nagel 1993, p. 61.
[98] Nagel 1993, p. 63.
[99] Sartorio 2004, p. 329. Sartorio argues that a person is not responsible for some outcome just because he caused it. Rather, the person must be morally responsible for the action that caused the outcome.
[100] For similar views of causation, see Moore 1994, p. 255 and Feinberg 1970, p. 166. To its credit, this view of causation seems to avoid the problems of the too-broad and too-narrow “but for” test often used in law. Still, many details remain to be worked out.
[101] Such is why we can speak of what a person “knew or ought to have known.” He ought to have known in the sense that the information was available to him and clearly relevant to his endeavors, yet he chose to ignore it.
[102] Malicious actions require conscious intent to do some wrong. Negligence requires some awareness of the care that could and should be taken in that situation. Decisions under uncertainty require knowledge of the risks of the action.
[103] Nagel and Williams deny that the action is distinct from its outcome (Nagel 1993, p. 62; Williams 1993, pp. 38-9). That is commonly disputed, e.g., by Athanassoulis 2005, pp. 272-3; Rosebury 1995, pp. 517-9; and Latus 2005, pp. 1-4. Those critics also argue that, at least in some cases, a better or worse outcome might suggest some virtue or vice in the agent’s reasoning in retrospect. Of course, proper moral judgment of risky actions may be exceedingly difficult in practice. When information about the decision is scarce, people’s actual judgments may be prone to hindsight bias, meaning that success or failure is wrongly regarded as proof of right or wrong choice, respectively (Royzman and Kumar 2004, p. 338).
[104] This analysis mirrors the distinction between intrinsic and extrinsic luck made in Williams 1993, p. 40.
[105] Kenny 1988, p. 110.
[106] In serious cases, such as when a drunk driver kills a pedestrian, the damage may be irreparable.
Go to Works Cited or the Proposed Bibliography.
