1. Background

According to the standard definition, moral luck occurs when a factor beyond one’s control affects one’s moral responsibility.Footnote ³ By “moral responsibility,” one might mean at least two things. One, roughly, is to have a moral duty. For instance, parents are responsible for taking care of their children. And two, one might mean being morally blameworthy or praiseworthy. For instance, one might be blameworthy for breaking a window or praiseworthy for saving another person’s life. The sense of responsibility that is relevant in the moral luck debate, and what I will henceforth refer to by “responsibility” unless I indicate otherwise, is the latter one. So, the question of moral luck is whether luck can affect blameworthiness or praiseworthiness.Footnote ⁴

Here are the three types of luck and three types of moral luck that are standardly distinguished.Footnote ⁵ Resultant luck occurs when the consequences of one’s action are partly out of one’s control. Resultant moral luck occurs when one is more or less responsible due to resultant luck. Consider an assassin who is set out to kill her victim. She shoots and the victim dies. Consider another scenario in which when she shoots, a random bird intercepts the bullet and changes its trajectory, and the victim does not die. Whether a bird intercepts the bullet is a matter of luck—that is, beyond the assassin’s control. Nonetheless, if she is less blameworthy in the latter case than she is in the former, there is resultant moral luck.

Circumstantial luck occurs depending on whether one faces a moral challenge partly due to factors beyond one’s control. Circumstantial moral luck occurs when one is more or less responsible due to circumstantial luck. Notice that we do not always choose our moral battles. Sometimes, we walk into the kind of circumstances that present us with moral challenges. Imagine that someone offers a bribe to a judge, and she takes it. Imagine also a second scenario where the same judge is not offered a bribe, and hence, she does not take any bribe (Thomson, Reference Thomson1989). It is a matter of luck whether someone offers her a bribe. Nonetheless, if she is less blameworthy in the latter case than she is in the former, there is circumstantial moral luck.

Constitutive luck occurs when one’s constitution (i.e., one’s dispositions or capacities) is partly due to factors outside one’s control. Constitutive moral luck occurs when one is more or less responsible due to constitutive luck. Notice that our constitution is formed at least partly by factors beyond our control such as our genes, environment, and upbringing. Imagine that Jimmy is stingy partly due to his upbringing. Due partly to his stinginess, Jimmy does not help his friend who needs financial help. Imagine another scenario where Jimmy’s upbringing is different, and he ends up charitable. Due partly to being charitable, Jimmy helps his friend. If Jimmy is blameworthy in the former case but not in the latter, there is constitutive moral luck.Footnote ⁶

To further clarify the challenge for Sifters, consider the following reaction to resultant moral luck. Suppose one thinks that the assassin above is equally blameworthy in both cases. After all, one might think, something as random as a bird flight cannot make a difference between one’s being blameworthy as much as a murderer versus being blameworthy to a lesser degree. But then one would have to say why the judge above should be less blameworthy, if at all, in one case compared to the other. After all, the reason why she is not offered a bribe in the latter case might have been a random bird injuring the would-be bribe giver by flying into his head while he was on his way to offer the bribe. With a further elaborate scenario, we could raise an equivalent of this concern also for charitable versus stingy Jimmy as to why he should be less blameworthy, if at all, in one case as opposed to the other. One way to put the question then is as follows: If a random bird flight is morally irrelevant in the assassin’s case, how could it be relevant in the judge’s or Jimmy’s case?

So, after rejecting resultant moral luck, one might be pushed to reject the other sorts of moral luck. After accepting the other sorts of moral luck, one might be pushed to accept resultant moral luck. It might then seem that Moore’s contention in support of Embracers that “luck is luck”—that if it is relevant to responsibility anywhere, it is relevant everywhere and vice versa—is on the right track. Hence, drawing a line to reject resultant moral luck alone commits one to an unstable or implausible position. Below I argue that rejecting resultant moral luck alone (i.e., the Sifter position) is a stable position. Moreover, if I’m right, it is the Embracers who face a stability concern instead.

2. Rejecting Resultant Moral Luck Alone

I want to begin by highlighting that—contrary to the appearances above and in the literature overall—there is nothing puzzling about the motivation to draw some line and sift out some things that are due to luck as irrelevant to responsibility. I will discuss this further in the next section. For now, consider (un)lucky, unforeseeable results of an action far down the causal chain—say, a few millennia after the action. It should seem implausible, even to Embracers, that such results affect one’s responsibility. If so, Embracers should draw a line to sift out some luck as irrelevant to responsibility.Footnote ⁷ But then, Moore’s contention that “luck is luck” seems dubious: Both Sifters and Embracers should draw a line and treat some luck differently than others. That is, Sifters are not the only ones in the “line-drawing” business. So, the Sifter position should not seem unmotivated or without any initial plausibility.

Indeed, much more must be said to establish that the Sifter view is plausible, and here is my argument:

By “what grounds responsibility,” I refer to the fundamental factors in virtue of which one is responsible. That is, I refer to what makes one responsible or what explains one’s responsibility. So, P1 says that factors other than the grounds for responsibility can affect responsibility only if they affect what makes one responsible. This seems highly plausible. By analogy, consider a consequentialist theory according to which what makes an action morally right or wrong is its consequences. If this is correct, it is unclear how a factor can affect the rightness or wrongness of an action without affecting the consequences of that action.

So, the controversial premise should be P2. Let me defend it. As many philosophers agree, responsibility depends on two conditions: the control condition and the epistemic condition.Footnote ⁹ The idea is that responsibility requires acting freely and being, or being reasonably expected to be, aware of various morally relevant factors.Footnote ¹⁰ To illustrate, consider this. Suzy pushes the main door to a busy building open and thereby hits Billy, who has been idly standing on the other side, causing him to fall and get hurt. If, through no fault of her own, Suzy is sleepwalking when she pushes the door open, she is neither in control of her action nor is aware of what she is doing. Hence, intuitively, she is not blameworthy. Suppose, however, she is not sleepwalking, and she is in control of her action, but she is not aware of Billy standing on the other side. Suppose also that it is a busy building, and people go in and out all the time. Now she is not only not aware of it that Billy is standing on the other side, it is also not reasonable to expect that she be aware of it. No one has any business idly standing on the other side of the door to a busy building. Intuitively, she is again not blameworthy. Imagine instead that Suzy is aware of it that Billy is standing on the other side, that she will hurt him if she opens the door, and that it is wrong to so hurt him, but she freely pushes the door anyway. Now it seems that she is blameworthy.

Given the two conditions above, it is helpful to see P2 as involving two separate claims. (P2.1) The moral challenges that one faces, and one’s dispositions or capacities can, but the consequences of one’s action cannot, affect whether the agent satisfies the epistemic condition. (P2.2) The moral challenges that one faces, and one’s dispositions or capacities can, but the consequences of one’s action cannot, affect whether the agent satisfies the control condition. Below I defend these claims, starting with P2.1.Footnote ¹¹

If there is an epistemic requirement for responsibility, it must be satisfied at the time of action or prior to it. It is implausible that what one is (reasonably expected to be) aware of after one performs an action is relevant to responsibility. Consider Suzy’s pushing the door open and realizing afterwards that Billy was standing on the other side. It seems implausible that this postact awareness retroactively makes her blameworthy in the past. Or imagine that right after Suzy pushes the door open, the door miraculously becomes transparent. Hence, now it is reasonable to expect, if Suzy were to push that door open again, that she be aware of someone standing on the other side. Yet again, it seems implausible that this newly obtaining fact retroactively makes her blameworthy in the past. So, the epistemic requirement for responsibility must be satisfied at the time of or earlier than the act. But the consequences of an action occur after the action.Footnote ¹² So, the consequences cannot affect satisfying the epistemic requirement for responsibility.

Notice, however, that moral challenges, and one’s dispositions or capacities, can affect satisfying the epistemic requirement at the time of or prior to an action. If it is a matter of circumstantial luck that one happens to be passing by a pond in which a child drowning, it is partly a matter of circumstantial luck whether one is properly aware of the morally relevant factors of the situation. To the extent that one is aware of those factors, one can be praiseworthy or blameworthy depending on how one behaves upon this realization. So, a moral challenge that one faces, even if it is due to luck, can result in one’s being responsible because that challenge can affect what makes one responsible—that is, one’s awareness. The same is true of constitutive luck—that is, one’s “lucky” dispositions or capacities. Most notably one’s epistemic capacities such as perception, memory, and reasoning affect what one is (reasonably expected to be) aware of. Hence, these capacities can affect responsibility.

Let us turn to P2.2. If there is a control requirement for responsibility, it must be satisfied either at the time of action or prior to it. It is implausible that whether one controls one’s behavior after one performs an action is relevant to one’s responsibility. Imagine that, in the case where Suzy is sleepwalking, she wakes up right after she pushes the door. She can control her behavior now, but it is implausible that this newly obtaining fact retroactively makes her blameworthy for what she previously did. So, the control requirement for responsibility must be satisfied at the time of or earlier than the act. But the consequences of an action occur after the action and hence cannot affect satisfying the control requirement for responsibility.Footnote ¹³

Notice again, however, that moral challenges, and one’s dispositions or capacities, can affect satisfying the control requirement at the time of or earlier than the act. A moral challenge can be more/less difficult and hence can result in it being more/less difficult to handle one’s actions in response. Consider an adult human drowning in a pool. An inexperienced teenager witnessing the event might panic and not easily be able to control his actions in response. An experienced adult may succeed in keeping calmer and being more in charge of his actions in response. In extreme cases, one may even “freeze” and totally lose control over one’s actions. Consider also that a courageous person is more in charge of her actions in scary situations than a less courageous person and that better epistemic capacities enable one to be more strongly in control of reacting appropriately to one’s circumstances.

So, P2.1 and P2.2 seem true. Below I respond to three objections to my argument.

3. Objections

Objection 1. It Is Implausible to Deny that Actual Consequences Can Be Part of the Ground. Some hold that, in addition to control and awareness, actual consequences of an action can be part of what grounds responsibility. Embracers, or proponents of resultant moral luck, in particular are committed to this view since they hold that one can be more or less responsible because of actual consequences of one’s action. Notice, for instance, that if there is a difference between the successful assassin’s responsibility and the unsuccessful assassin’s responsibility, this can only be due to actual consequences of the successful assassin’s action. This is because they are on a par regarding all other factors. One might then object that denying that actual consequences can also be part of the ground for responsibility is implausible or leads to an unstable position.

In response, notice that we are not told yet which actual consequences specifically do ground responsibility. So, first, the idea might be that (AC) just any consequence of an action makes one responsible. But (AC) seems false. Consider (i) random consequences and (ii) numerous distant consequences (e.g., consequences that occur many years later) that can result from an action. Suppose you touch the switch to turn on the lights in your room, but unbeknown to you, the switch is connected to a torture machine in the next room, which results in an innocent victim’s being tortured. As even Embracers would accept, it is implausible that the victim’s being tortured makes you blameworthy.

Second, the idea might be that (EC) only the expected consequences make one responsible—either the consequences expected by the agent or the consequences that the agent is reasonably expected to foresee. To my mind, (EC) is also the most plausible way to avoid the above objection to (AC): Since you neither expect nor is it reasonable to expect that touching the switch will result in an innocent person being tortured, you are not blameworthy. The problem is that if (EC) is true, it is false that actual consequences of an action can be part of what grounds responsibility. Expected consequences are not actual consequences. One’s beliefs about the potential consequences of an action, or even the beliefs that are reasonable to expect one to have regarding them, can be false. It is reasonable for you to believe that if you touch the switch in your room, the light will turn on. Nonetheless, in the case above, your act leads instead to someone’s being tortured. Notice moreover that (EC) collapses back into the epistemic condition. This is because the idea that one can be responsible because of expected consequences of one’s action is included in the idea that one can be responsible because of what one is (reasonably expected to be) aware of.

Third, the idea might be that (AEC) only the actual consequences among the expected consequences make one responsible. (AEC) can also avoid the above objection to (AC). The victim’s being tortured is an actual but unexpected consequence of your action. Hence, you are not blameworthy. Notice that (AEC) draws a line between the actual consequences that are, and those that are not, among the expected consequences. The question is “why.” Among all the actual consequences, why should only the expected ones matter for responsibility? First, one might think that it is because they are expected. But then what ultimately matters is the expectancy, not the actuality, of the consequences. Thus, (AEC) collapses back to (EC). Second, one might think that they matter because control matters, and one cannot control what is unexpected—that is, what one is not (reasonably expected to be) aware of.Footnote ¹⁴ But is it the expectancy of the consequences that ground control or (also) the actuality of them? If it is the former, this is consistent with (EC). If it is the latter, the view seems dubious. It implies that what comes after you do all that is in your power can make it that you had more or less control. Suppose you want to throw a rock and break a glass bottle. You do all that is in your power—for example, consider what sort of rock to throw, how fast you should throw, at what angle to throw it, and so on—and throw the rock accordingly. Now, consider two variations of what happens next. One, the rock hits the bottle. Two, someone randomly knocks the bottle down, and the rock flies past. The view implies that your level of control was less in the latter. But this seems odd. In both the cases, you used all the power you have and exerted all the effort you could. It seems implausible that what someone else does, after you do all you could, can make it that you had more or less control. Moreover, if awareness in question is necessary for control, and the awareness condition is better captured by (EC), this is further reason to hold that responsibility relevant control is grounded in the expectancy, and not the actuality, of consequences (see also fn.#13). Third, one might think that they—that is, actualized expected consequences—matter because they are expected and actual. Thus, (AEC) is a brute fact. But this does not answer the “why” question above and amounts to mere insistence that actuality of the expected consequences must somehow matter.

Indeed, it does not follow that (AEC) as a brute fact is false. The following seems coherent: While mere actuality of consequences does not matter, when the actuality is combined with the expectancy, the combination matters in a way that is not reducible to the expectancy alone. However, if the debate between Sifters and Embracers comes to this—that is, choosing between (EC) versus (AEC)—the objection that the Sifter position is unstable loses its force. There is nothing unstable in holding (EC) as opposed to (AEC), and (EC) is at least as plausible as (AEC) if not more. Consider this. Billy flips the switch while fully and reasonably expecting that it will electrocute an innocent victim but, due to a faulty wiring, it does not. Intuitively, Billy is blameworthy even though the expected consequence is not actualized.Footnote ¹⁵ (EC) can account for this but (AEC) cannot. To account for this, Embracers can hold that the expectancy of consequences alone can make one blameworthy just as much as the expectancy and the actuality of consequences. But this makes (AEC) redundant. Alternatively, Embracers can hold that the expectancy alone can make one blameworthy, but the expectancy and the actuality can make one more blameworthy. That is, Billy is blameworthy, but he would have been more blameworthy had the victim got electrocuted. Notice that this just asserts that resultant moral luck exists. But if the debate between Sifters and Embracers comes down to this, Embracers must give up the objection that the Sifter position is unstable in arguing that resultant moral luck exists. Once the former boils down to the latter, it cannot be evidence for the latter. Additionally, first, empirical studies—especially when executed carefully—show that most ordinary people (up to nearly 90%) hold that the actuality of consequences does not increase blameworthiness.Footnote ¹⁶ This speaks for (EC) and against (AEC). Second, (EC) is theoretically simpler since it involves only the expectancy of the consequences and not the actuality of them. Third, as discussed in the previous paragraph, (EC) answers the “why” question above and does so straightforwardly. These are further good reasons to choose (EC) over (AEC).Footnote ¹⁷

Here’s another significant point that follows. Embracers must draw a line between some actual consequences and others. The discussion above suggests that this task is not easy and that they might be pushed to reject all actual consequences as relevant to responsibility. This raises the question about whether the Embracer position is stable. The discussion above suggests that it might not be. Hence, Objection 1 not only is ineffective against Sifters but thorough consideration of it raises a stability concern for Embracers instead.

Objection 2. Other Accounts of Responsibility. My argument relies on the view that responsibility is grounded in control and awareness. One might then worry whether I can secure my conclusion if this view is rejected.

In response, first, the view in question is intuitively plausible. Moreover, philosophers in the moral luck debate typically accept it. Many philosophers outside the moral luck debate also accept it. So, even if one rejects this view, it is not too troubling. I will still have given an argument that the Sifter position is plausible.

Second, rejecting the view in question does not entail rejecting the conclusion of my argument. Suppose an objector holds another view of responsibility. To reject the conclusion, she must show that the consequences of an action (in addition to the moral challenges that one faces, and one’s constitution) can affect what grounds responsibility on that view. Assume that she holds what is sometimes considered as the main rival of the view of responsibility above: a Quality of Will (QOW) account of responsibility according to which what fundamentally matters for responsibility is one’s quality of will in performing an action.Footnote ¹⁸ There are various ways of fleshing out what “quality of will” amounts to. On one account, one is praiseworthy or blameworthy for an action just in case that action arises from a proper concern or lack thereof for another person’s morally significant interests. To illustrate, imagine that an assassin shoots to kill an innocent person, and the victim dies. The QOW view says that what makes the assassin blameworthy is her lack of concern for the victim. Notice that whether the assassin’s action arises from a lack of proper concern for the victim cannot be affected by the consequences of her action—for example, whether the victim dies. This is because the consequences of her action occur after what gives rise to her action. That is, on the QOW view, consequences of an action cannot affect what makes one responsible.Footnote ¹⁹ However, as should be clear given the discussion above, moral challenges that one faces, and one’s dispositions or capacities, can affect the kind of concern that gives rise to one’s action. Hence, the objector not only does not avoid my conclusion but arrives at the same conclusion. Since the same conclusion is secured on two different and widely accepted accounts of responsibility, my argument is even more plausible now.

Objection 3. Consequences Can Affect Awareness. Suppose Jane pushes a random button and consequently gets teleported next to a pond where a kid is drowning. She looks around and notices the kid. Immediately, she gets into the water and saves him. She seems praiseworthy. But notice that her praiseworthiness is partly due to her awareness of the kid in the pond, which is partly due to a “lucky” consequence (i.e., being near the pond) of her action (i.e., pushing the button). Hence, consequences of an action can affect satisfying the epistemic requirement for responsibility. If so, one, a conjunct of P2.1 is false. And two, (un)lucky consequences of an action can affect one’s responsibility which implies that resultant moral luck exists.Footnote ²⁰

In response, recall that the relevant conjunct of P2.1 claims that consequences of one’s action cannot affect satisfying the epistemic condition on responsibility in the context of that action. And notice that the consequence (i.e., being near the pond) of Jane’s action (i.e., pushing the button) does not affect Jane’s awareness relevant to that same action. What happens is that a consequence of Jane’s previous action affects Jane’s awareness relevant to the next action. Hence, P2.1 remains unscathed.

Here are some further considerations if this seems too quick. First, notice that Jane’s case is no different than paradigm examples of circumstantial moral luck—Jane faces a moral challenge partly due to factors beyond her control. One might worry that Jane faces the moral challenge partly because of her own action. But this is true also of virtually all instances of real-life circumstantial luck cases. We do not just pop into existence in circumstances where we face moral challenges. We end up in those circumstances partly because of our previous actions. Indeed, even granting that Jane’s case is a circumstantial moral luck case, it might not be logically inconsistent to categorize it also as a resultant moral luck case. But I do not see a compelling reason to do so—given especially the reasons below.

Second, the paradigm examples of resultant moral luck involve cases where the consequences of an action make one more or less responsible in the context of that same action. Consider the comparisons between lucky versus unlucky reckless drivers or successful versus unsuccessful assassins. Moreover, among Sifters, I am not aware of anyone rejecting the idea that one can be responsible for a future action, A₂, partly due to consequences of a past action, A₁. All they reject is the idea that one can be more or less responsible in the context of an action, A₁, due to consequences of that same action, A₁. So, even if one considers the kind of cases like Jane’s as cases of resultant moral luck and hence holds that there is resultant moral luck, one would be holding a view in which moral luck literature is overwhelmingly uninterested. Those who reject resultant moral luck in particular will not be moved. This is also because the dispute then would be merely verbal without any substantive disagreement.

4. The Argument Extended

In this section, I show how my argument applies in famous cases in the moral luck literature and respond to further potential objections along the way. Consider the following case:

Embracers hold that Attempter is less blameworthy than Killer. Sifters hold that they are equally blameworthy. My argument also tells us that they are equally blameworthy because the only difference between the two—that is, whether the victim dies—does not affect what grounds their responsibility.

Imagine another person, Rester. Rester is currently resting at home because a random bird hit her head yesterday causing her to have a headache. She is not incapacitated but takes the day off. If not for the fortuitous event yesterday, she would be out and about today and Mystery Man would have approached her, asking her to kill another innocent person, Victim3, for money. Imagine that Rester would then take the job and kill Victim3. Assuming that Killer, Attempter, and Rester are on a par regarding all other relevant facts, if Rester is less blameworthy than Killer or Attempter, there is circumstantial moral luck.

Embracers hold that Rester is less responsible than Killer or Attempter, and Sifters can agree. Here is how my argument suggests that they can. Let us assume for simplicity that all three of them are equally in charge of their actions. Notice that Rester’s state of awareness is entirely different from that of Killer’s or Attempter’s, and this is due in part to their circumstances. The latter are aware that they are setting out to kill an innocent person while the former is not. So, what can make Rester as responsible as Killer or Attempter is not in place to begin with. Since they are in different states regarding what grounds responsibility, they are differently responsible. Put differently and more simply, the difference is explained by the fact that while Rester does not at all set out or will to kill anyone, Killer and Attempter do (cf. also Nyman, Reference Nyman2022, 961, 966–968).

If one finds the judgment above implausible, here are further things to note. One, it is open to both Embracers and Sifters to hold that all three of them have equally morally undesirable characters. Indeed, this does not mean that they are all equally blameworthy for their actions. Consider psychopaths. One might hold that they have morally repugnant characters while holding—as it is often accepted—that they are morally neither blameworthy nor praiseworthy since they do not recognize moral reasons. But if Killer, Attempter, and Rester are responsible for their characters, they can also be equally blameworthy for their characters—that is, being such that they would freely and knowingly kill an innocent person. This is also open to Embracers and Sifters. And notice that what is in question in the foregoing paragraph is not responsibility for a character but for an act.Footnote ²¹ Two, notice that the differential moral judgment above is no different than the judgment in the following case. Suppose you are walking by a pond where a kid is drowning. You can be morally praiseworthy or blameworthy depending on how you react. Suppose alternatively that you were at home, watching a movie, totally unaware that a kid is drowning in the pond. It is perfectly plausible that in the latter scenario you are neither as praiseworthy nor as blameworthy as you could be in the former scenario. If the differential judgment between these two scenarios is plausible, it is equally plausible in comparing Killer and Attempter on the one hand, and Rester on the other. Three, if the differential judgment is implausible, this is not a problem only for Sifters but also for Embracers. This is important because recall that the main concern in this paper is the debate between Sifters and Embracers. And what is equally problematic for Embracers cannot be a reason to choose the Embracer position over the Sifter position.

Imagine now another person, Sneezy. Just like Killer and Attempter, he takes the job from Mystery Man to kill an innocent person. He takes his position, his finger is on the trigger, he wills or attempts to pull the trigger, but sneezes before he can pull it, and thereafter misses the opportunity to kill the victim. Two questions arise. One, is he less blameworthy than Killer or Attempter or is he equally blameworthy? Two, is Sneezy’s case a circumstantial luck case?

Here is why these questions are relevant. While some hold that (A) it is a circumstantial luck case and Sneezy is equally blameworthy as Killer or Attempter (Zimmerman, Reference Zimmerman2002), some hold that (B) it is a circumstantial luck case and Sneezy is less blameworthy than Killer or Attempter (Hartman, Reference Hartman2017, 105–111). One might take (A) as a reason to reject circumstantial moral luck along with resultant moral luck. After all, if Sneezy’s circumstantial luck does not lessen his blameworthiness, this is evidence that there is no circumstantial moral luck. Or one might take (B) as a reason to accept resultant moral luck along with circumstantial moral luck. This is because if Sneezy’s circumstantial luck lessens his blameworthiness despite significant similarities among Sneezy, Attempter, and Killer, it is evidence that resultant luck in Attempter’s action lessens his blameworthiness despite significant similarities between Attempter and Killer. Hence, (A) and (B) can generate concerns for Sifters.

However, (A) and (B) are both problematic. First, it is unclear that Sneezy’s case is a circumstantial luck case.Footnote ²² Circumstantial luck is about whether one faces a moral challenge. But Sneezy, Killer, and Attempter all face the same kind of moral challenge—that is, the Mystery Man’s offer to kill an innocent person or to attempt to kill an innocent person. It is Rester that does not face a similar moral challenge partly due to factors beyond her control. And pairing Rester’s case with Killer or Attempter gives us a clear case of circumstantial luck. But it is far from clear that what is in question is circumstantial luck when we pair Sneezy’s case with Killer or Attempter. Second, it is consistent with rejecting resultant moral luck that Sneezy is equally blameworthy as Attempter and Killer. One plausible way of saying why Attempter is equally blameworthy as Killer is that they are on a par regarding what they tried or attempted to do—that is, to kill an innocent victim, and the results of their acts of trying to do what they tried to do cannot make them more/less blameworthy. Indeed, some hold that this is precisely what denying resultant moral luck entails (Enoch, Reference Enoch, Heuer and Lang2012, 101; Khoury, Reference Khoury2018; Wolf, Reference Wolf2001, 13–18). But then, Sneezy must also be equally blameworthy as Attempter or Killer since they are all on a par regarding what they tried to do. Third, my argument also tells us that Sneezy is just as blameworthy as Attempter and Killer. This is because all of them have equal control and awareness, which remain unaffected by the victim’s death, the bird flight, or the sneeze.Footnote ²³

Imagine one last person, Timid. Partly because he is genetically predisposed to be timid, he is always stayed away from guns. Knowing this, Mystery Man does not even approach Timid. Hence, Timid does not kill anyone. However, if not for his genetic predisposition, he would have had a different character, ended up taking the job from Mystery Man, and killed an innocent person. Nonetheless, if he is less blameworthy than, say, Killer, there is constitutive moral luck.

Embracers hold that Timid is less blameworthy than Killer or Attempter. And again, Sifters can agree. Like the previous example, here is how my argument suggests that they can. Assume for simplicity that Killer and Timid have equal control over their respective actions. Notice that Timid’s state of awareness is entirely different from that of Killer’s. While the latter is aware that she is attempting to kill an innocent person, the former is not. Hence, what can make Timid as responsible as Killer is not in place. And since they are in different states regarding what grounds responsibility, they are differently responsible.

One might object that the assumption above—that Timid and Killer have equal control over their respective actions—is questionable. If Timid’s behavior is partly due to his genetic predisposition, which is beyond his control, he cannot be in charge of his behavior. The same is true of Killer—if his act of killing is partly due to his “lucky” genetic predisposition, he cannot appropriately be in control of his action. So, neither Timid nor Killer satisfies the control condition, and hence, neither of them are blameworthy, which entails that they are equally blameworthy.

In response, first, notice that the comparison between Timid and Killer is relevantly similar to the following. Suppose Timid is not genetically predisposed to be timid. Instead, his genetic predisposition results in poor eyesight or blindness. Partly due to this predisposition, he is always stayed away from guns. But Killer has no such predisposition, and her eyesight is clear. Assume that the rest of the story is the same. The objector is committed to holding that Timid and Killer, in this new scenario, do not control their actions which implies that blind people cannot control their actions and having a clear eyesight is an impediment to being in control of one’s behavior. But this seems implausible.

Second, on most contemporary accounts of free will, an act’s being partly due to genetic predispositions is consistent with one’s exercising responsibility relevant control over that action. On one compatibilist account (i.e., the deep-self view), one has control just in case one’s behavior is governed by one’s deep self. One’s deep self is defined in terms of one’s second-order volitions, cares, values, or judgments of the good.Footnote ²⁴ And notice that an act’s being partly influenced by a genetic predisposition is consistent with that act’s being governed by what one cares about or values deep down. On another compatibilist account (i.e., the guidance control view), one has control just in case one’s action issues from a moderately reasons-responsive mechanism and one takes ownership of that mechanism (Fischer, Reference Fischer and Fischer2007; Fischer & Ravizza, Reference Fischer and Ravizza1998). A mechanism is moderately reasons-responsive just in case it is overall receptive and reactive to reasons for action. Notice again that one can recognize reasons for an action and react appropriately to those reasons while the action is also influenced by a genetic predisposition. Finally, on many libertarian accounts, one has control only if one is the actual source of one’s action a prerequisite of which is that the act be causally undetermined by prior events.Footnote ²⁵ On these accounts, a prior event’s merely partial contribution to an act is consistent with that act’s being genuinely up to one. That is, an act’s being partly due the influence of a genetic predisposition is consistent with one’s being the genuine source of that act.Footnote ²⁶