1. Supererogation and Sacrifice

In this section I will outline the view that supererogation always involves sacrifice. I will start by exploring what is meant by sacrifice before going on to explain the intuitive appeal of the view.

We should start by noting that those who claim that supererogation always involves sacrifice are appealing to a notion of sacrifice that is less rich than the everyday use of the term. The use of sacrifice in this context is restricted to a cost to the agent performing the act. Jonathan Dancy, for example, claims that supererogation always involves sacrifice, which he defines as “cost to the agent”.Footnote ¹⁰ We might think that the everyday meaning of the term ‘sacrifice’ involves some further conditions, such as being performed intentionally or voluntarily.Footnote ¹¹ I take it, though, that whatever else we mean by ‘sacrifice’, in order for an act to count as a sacrifice it must involve some cost to the agent's interests. Given this, in order to show that acts of supererogation always involve sacrifice it will be enough to show that acts of supererogation do not always involve the relevant form of cost to the agent's interests.

The next point to make about sacrifice is that it is a comparative concept. An act that involves a sacrifice is one that makes the agent worse off in some way. The question we must now ask is what the relevant comparison is here. There are two options. The first option is that the relevant cost to an agent's interests is in comparison to her position before performing the act. The alternative is that we take the relevant comparison to be the position the agent would be in if she performed one of the other acts available to her. It is worth noting that there are different ways in which ‘available’ can be understood in this context. For example, an agent may be physically able to perform some act but psychologically incapable of doing so. We will return to this point in Section 3. For now though, we should note that it is some form of the second, counter-factual, option that picks out the relevant form of cost. Cases of ‘cutting one's losses’, where an agent chooses the least costly option from a range of costly alternatives, should not count as cases of sacrifice.Footnote ¹² Suppose the victim of a mugging is faced with the choice of handing over his money or being beaten and having his money taken from him. The victim recognizes that his assailants mean what they say and that their threat is credible. While the victim appreciates the gravity of the situation, he is not so overcome with fear that he is incapable of making a rational decision. Clearly, in this case choosing to hand over the money without a fight does not count as a sacrifice. While choosing the first option will result in his being worse off than he was before, this is an example of minimizing one's losses rather than making a sacrifice. On the other hand, someone can make a sacrifice while making herself better off than she was before. Someone who chooses not to receive the full value of a cash prize but to take part of it and leave the rest to charity, counts as making a sacrifice despite the fact that she is better off than she was before. What makes this a sacrifice is the availability of an alternative option that would have made her better off. The relevant comparison then is to the position the agent would have been in had she chosen to perform one of the other available acts.

It is worth emphasizing that sacrifice requires the existence of alternative acts that are, in some sense of the word, available to the agent. An act cannot count as involving sacrifice if there were no other acts that the agent could have performed that would have made her better off.

Why, then, would anyone think that performing a supererogatory act will always make an agent worse off than she would have been had she performed some alternative act? The answer is that this view is motivated by the acceptance of a first order view about how moral goodness and moral obligation are connected. According to this view, we are always morally required to perform the morally best act available except when the costs to the agent are too high. Dale Dorsey endorses this point in the following:

According to Dorsey the reason that supererogatory acts are not obligatory is because they involve a significant cost to the agent's welfare. Without this cost the act would be obligatory. Henry Sidgwick makes a similar point when he claims that it is part of common sense morality that people have, “a positive duty to render, when occasion offers, such services as require either no sacrifice one our part, or at least one very much less in importance than the service rendered.”Footnote ¹⁴ Again, the thought here is that it is only permissible to perform a suboptimal act if doing what is best would involve a cost to the agent's self-interest. If we accept this then all permissible acts that are morally better than the minimum morality requires will involve a cost to the agent that doing the minimum does not. It follows from this that all acts of supererogation will involve sacrifice.

This view has a certain intuitive appeal. It seems reasonable to think that if we can help others at no cost to ourselves then this is what we ought to do. In addition, when we think about paradigmatic examples of supererogation it seems plausible to think that what is preventing these acts from being obligatory are the costs that performing the act have for the agent. The righteous gentiles, for example, put their lives in danger in order to help save the lives of those persecuted by the Nazis. Wallenberg risked and eventually lost his life as a result of his decision to leave Sweden to help the Jews. Attorney Gideon Hausner, the man who prosecuted Adolf Eichmann described Wallenberg as, “a man who had the choice of remaining in secure, neutral Sweden when Nazism was ruling Europe. Instead he left this haven and went to what was one of the most perilous places in Europe. And for what? To save Jews.”Footnote ¹⁵ It seems obvious then that these acts involved sacrifice. After all they could have chosen not to help and in doing so they would have avoided the risks that accompanied their actions.Footnote ¹⁶

Despite the initial plausibility of this view, I will argue that it is mistaken. While many acts of supererogation do involve sacrifice there are others that do not. It cannot be the case then, that all supererogatory acts are prevented from being obligatory by the level of sacrifice involved in their performance.

There are a number of ways in which we might argue against this view. We might argue that there are cases where it can be permissible to perform a morally suboptimal act even though performing the optimal act would better promote the agent's well-being. This, though, will not be the approach that I will take here. Instead I will argue that the traditional view rests on a view about the range of alternative acts that agents can perform that fails to do justice to the first personal perspective of many of those who perform acts of supererogation. However, in order to be in a position to make this point I must first introduce Bernard Williams’ idea of moral incapacity. It is this task to which I will turn in the next section.

2. Moral Necessity and Moral Incapacity

Let's return to the claims made by the moral exemplars we considered in the introduction. As we saw, moral exemplars often claim to have experienced a sense of internal compulsion. They claim that they felt that they could not have acted otherwise. That they had to act as they did. In this section I will investigate how we should understand these claims. I will argue that these claims should be understood as a form of what Bernard Williams calls ‘practical necessity’.

Williams begins by pointing out that it is a familiar part of practical deliberation that in deliberating an agent may conclude that there is a certain action that she must do or has to do.Footnote ¹⁷ Williams claims that in reaching this conclusion, the agent has discovered that performing that action is a practical necessity for her. The concept of practical necessity is distinct from other practical conclusions we might reach through deliberation. Most importantly, deciding that one must act a certain way is different to deciding that one ought to act in that way. Williams explains the difference between the two by saying that, “Ought is related to must as best is related to only.”Footnote ¹⁸ In other words, to say that one ought to Φ is to say that of all the available acts, Φ is the act that is most favoured, while to say that one must Φ is to say that Φ is the only available option. A further difference between ought and practical necessity is in what it takes to show each to be false. As Williams points out, if someone says that she cannot act in a particular way but then does so intentionally then her claim was false. However, if someone says that she ought not act in a certain way then this can be true even if she does act in that way and does so intentionally.Footnote ¹⁹ In Williams’ view, practical necessities are conceptually linked to incapacities. As Williams puts the point:

Williams develops this thought in his later discussion of moral incapacity in his article of the same name. By moral incapacity Williams is not referring to, “an incapacity to engage or be engaged in moral outlook,”Footnote ²¹ the kind of incapacity we might associate with despair or apathy. Rather, Williams is describing:

An example of this kind of incapacity can be found, according to Williams, in the life of Martin Luther, who at The Diet of Worms in 1521, when asked whether he was willing to revoke his writings criticizing the practice of selling indulgences, famously declared, “Hier steh' ich, ich kann nicht anders,” (Here I stand; I can do no other).

Williams claimed that in order to do justice to this claim we should understand it as a genuine incapacity. A proper understanding of Luther's claim involves appreciating that he was incapable of acting otherwise. Luther's claim, then, cannot be equivalent to the claim that he ‘ought not’ to act differently. Nor is Luther claiming a physical incapacity to do otherwise. Luther was certainly physically capable of doing otherwise and indeed had good prudential reasons to do so. Rather, this is a case of moral necessity, as Luther's practical necessity to act as he did is one that is an expression of his moral outlook.Footnote ²³

It could be objected that moral necessities are just equivalent to moral obligations. However, this would be mistaken for two reasons. The first reason is that the conditions which can make it is false to say that one is morally obliged to act a certain way are different to those that make it false to say that performing that act is a moral necessity. As was the case with the practical ought, the fact that an agent fails to perform some act does not show that he had no moral obligation to perform it. This, though, would show that performing the act was not a moral necessity for him. An act then can be a moral obligation without being a moral necessity. Similarly, an act can be a moral necessity without being a moral obligation. It is after all, perfectly coherent to think that while Luther was acting from a moral necessity to act as he did, he was not morally required to act in this way.Footnote ²⁴

Nor are moral necessities reducible to categorical imperatives. After all, it seems perfectly reasonable to accept that Luther acted from a moral necessity without accepting that everyone should face a moral necessity in the same circumstances. An act can be a moral necessity, then, without being a categorical imperative.

Finally, moral necessities and incapacities must be distinguished from other forms of necessity and incapacity. We can start by noting the similarity between all three. If someone is unable to perform an act because of any of these forms of incapacity then performing that act is not a possibility for her. However, moral incapacities are similar only to a specific form of these incapacities. Many physical and psychological incapacities are such that those who possess them cannot act in certain ways either intentionally or unintentionally.Footnote ²⁵ If I say that I cannot lift a five hundred pound weight or do long division, yet it turns out that I can when under hypnosis then it is not true that I cannot do these things.

Other forms of incapacity, however, are such that those who possess them are unable to act in a certain way if they know that that is what they are doing. For example, having a relaxed and friendly conversation with one's spouse's lover.Footnote ²⁶ Someone might be quite capable of doing this for as long as she is unaware of the stranger's full identity. To possess a moral incapacity is to be incapable of acting in such a way intentionally. These incapacities are, in other words, such that they make it impossible for their possessors to wilfully perform these actions.Footnote ²⁷ This does not rule out the possibility that the agent could perform these acts unintentionally.Footnote ²⁸ It is clearly this kind of incapacity that Luther possessed. Presumably, Luther would have been quite capable of rejecting his writings if he could be shown why they were incorrect. What he was incapable of was rejecting his writings solely in order to spare himself.

Before we are able to relate this discussion to our cases of admirable supererogatory acts we must first make a distinction between moral incapacities and other forms of incapacity to perform an act intentionally. Unlike these other forms of incapacity, moral incapacities are grounded in the agent's character.Footnote ²⁹ Of course, some psychological incapacities will also be grounded in an agent's character. The incapacity to converse with one's spouse's lover, for example, is plausibly an incapacity that is rooted in an agent's character. However, what is distinctive about moral incapacities is that these incapacities involve a self-endorsement of this aspect of the one's character. Someone with a psychological or physical incapacity to perform an act intentionally could try to overcome or remove these incapacities. Someone who is physically incapable of lifting a 500 lb weight could begin a weightlifting programme in the hope of one day being capable of doing so. Similarly, someone who is psychologically incapable of flying may take a course that could help her to deal with her fears in the hope of one day being able to fly. Someone with a moral incapacity, on the other hand, cannot coherently try to remove this incapacity from her character. The reason for this is that as soon as she ceases to identity with this incapacity, it ceases to be a moral incapacity, though the agent may remain psychologically incapable of performing the acts in question. This follows from the fact that moral incapacities are defined as those incapacities that are expressive of the agent's moral outlook.

For example, someone who is incapable of murder cannot be said to be morally incapable of murder if he is actively trying to overcome this incapacity. The reason for this is that as soon as the agent undertakes to overcome this incapacity it can no longer be said to be one that is expressive of his moral outlook. Moral incapacities then are not simply incapacities to act in certain ways, they are also incapacities to try to act in those ways.Footnote ³⁰

Another distinctive feature that Williams claims for moral incapacities is that these incapacities are, at least implicitly, the result of a deliberative process.Footnote ³¹ While physical and psychological incapacities limit the range of choices that an agent can decide between, moral incapacities are the conclusion of a piece of deliberation. Of course, moral incapacities can limit the range of deliberation too. When someone discovers she has a moral incapacity this limits the range of options that are open to her to consider. However, these limitations are always the result of some prior process of deliberation rather than limitations that exist prior to any deliberations taking place.

Craig Taylor has challenged this aspect of Williams’ account of moral incapacity.Footnote ³² According to Taylor there are cases of moral incapacity where no moral deliberation is involved. For example, suppose a revolutionary, R, discovers that one his comrades is a spy. R decides that the only course of action is to kill the traitor. However, when he comes to pull the trigger he finds that he is unable to do so – and this is not because he is squeamish, or afraid of the consequences, but because he thinks it would not be morally right. As Taylor points out, one way in which it seems reasonable to interpret this case is to say that R did not undergo a further period of deliberation. Rather he made the discovery that despite concluding that he must kill the spy, he was not morally capable of doing so.Footnote ³³ Moreover, it could be the case that having made this discovery R endorses this incapacity and would make no attempt to change it. Taylor claims that in such a case it seems reasonable to describe this as a moral incapacity. The problem with Williams’ claim that moral incapacities are always the result of a piece of deliberation is that it leaves us unable to say this.

There is not space to adjudicate on this debate here. The important point to take from this discussion, however, is that the reason that it seems plausible to call this incapacity a moral incapacity is that R endorses it. If R were to discover that he was incapable of killing the spy and feel disgusted at his weakness then we could not say that this incapacity was expressive of R's moral outlook. The key difference, then, between moral incapacities and other forms of incapacity is that moral incapacities express the moral outlook of the agent.

This concept of moral incapacity provides the perfect way of understanding the sense of ‘inner compulsion’ reported by many moral exemplars. When Wallenberg claimed to have “no choice” and that he would “never be able to go back to Stockholm without knowing inside myself I'd done all a man could do to save as many Jews as possible,” the most straightforward way of understanding this is as a claim of moral incapacity. Similarly, the concept of moral incapacity allows us to take seriously the righteous gentile's claim that she “could not sit there and do nothing.” Taking these claims seriously involves accepting that those who performed these acts were morally incapable of choosing another course of action.

One final way in which one might attempt to explain away the concept of moral necessity is by claiming that the rescuers’ reports are best understood as claims of moral commitment rather than moral incapacity. In his response to Williams’ argument Tod van den Beld claims that purported claims of moral incapacity are best understood as reducible to, “I ought not and I shall not.”Footnote ³⁴ Perhaps, then, these are not incapacities but simply expressions of moral commitments.

The problem with this reductionist account of moral incapacity is that it fails to do justice to the differences between lacking a capacity to act in a certain way and making a commitment not to act in a certain way. If someone makes a commitment not to act in a certain way and then acts in that way then she has either been insincere, weak willed or ignorant of the difficulties involved in keeping the commitment. However, someone who fails to fulfill her commitments does not thereby show that she did not have any commitments in the first place. On the other hand, someone who claims to be morally incapable of acting a certain way and then does so has not displayed weakness of will or a lack of sincerity. Rather, she has shown that her claim was mistaken.Footnote ³⁵ The conditions under which claims of commitments or incapacities are shown to be false are different. This reductionist account cannot do justice, then, to the differences between commitments and incapacities and so ought to be rejected.

Moreover, while the claim that the rescuers’ reports are reducible to moral commitments is not wildly implausible when applied to Wallenberg, whose claim was made while he was still performing the actions he is describing, it appears bizarre when applied to the other rescuers who were talking decades after they had acted. It is hard to see how the reductionist approach could make sense of this. After all, these rescuers are not reporting a commitment that they made at the time they are describing why they acted as they did. Of course, it would be possible for a rescuer to say that they acted because they had made a commitment and were determined to stick to it. Importantly, though, this is not what the rescuers we are looking at did say. Rather, they talked about incapacity. So while such a case is possible it does not look like this is how we should understand the cases we are discussing here. It seems far more plausible to interpret these remarks, as Oliner and Oliner do, as describing a sense of “inner compulsion,”Footnote ³⁶ or as Colby and Damon do as, “not […] a matter of choice.”Footnote ³⁷

Of course, showing that these acts were viewed as cases of necessity by the rescuers does not by itself show that these acts were cases of moral necessity. In order to show this I must also show that these incapacities were expressive of the moral outlooks of these agents. It seems clear, at least for the majority of these cases, that these necessities are of this sort. After all, accompanying most of the statements of necessity are explanations that appeal to the moral reasons that count in favour of these actions. Wallenberg, for example, talks about the need to, “rescue as many Jews as possible,” while other rescuers point to, “the daily misery that was occurring,’ and, “The Germans shooting people in the street.” In all these cases, these moral reasons are being offered as explanations for the agents' moral incapacities. This gives us good reason to think that these incapacities are expressive of the rescuers' moral lives.

Moreover, for many rescuers, these incapacities appear to have originated from well-entrenched character traits rather than from a result of a particular set of circumstances. In their comparison of the two groups Oliner and Oliner claim that the differences between rescuers and non-rescuers cannot fully be explained by the circumstances. A full explanation must also make reference to the moral characters of the members of both groups. Of particular importance were two aspects of the character of the rescuers. The first was recognition that Jews were worthy of assistance. For some this was a result of a general recognition that all those in need should be helped, regardless of who they are. As a rescuer named Louisa, who features in Oliner and Oliner's study, explained, “It wasn't that we were especially fond of Jewish people. We felt we wanted to help everybody who was in trouble.”Footnote ³⁸ For other rescuers it is less clear whether their recognition that Jews deserved the same rights as themselves extended to other persecuted groups. Wallenberg saw his mission to be to save as many Jews as possible, rather than to save as many people as possible.Footnote ³⁹ Either way, this compared favourably to non-rescuers who Oliner and Oliner found to be more inclined towards “detachment and exclusiveness,”Footnote ⁴⁰ with several claiming explicitly that they, “did not help them because they were Jews.”Footnote ⁴¹ The second key feature of the characters' of rescuers was a particularly strong sense of responsibility towards others. Oliner and Oliner claim that rescuers were significantly more likely than non-rescuers to feel a sense of responsibility for those being persecuted.Footnote ⁴² Wallenberg certainly seemed to have a strong sense of responsibility, as Bierman describes, as soon as stories of the persecution of the Jews began to emerge Wallenberg felt that he needed to help in some way and until leaving for Hungary felt a deep sense of frustration at being unable to help.Footnote ⁴³

Finally, it should be noted that these incapacities reflect positively on the rescuers' moral lives. Williams mentions in passing that moral incapacities might not always be a virtue of an agent and could instead be a failing.Footnote ⁴⁴ In the cases we are looking at, though, it is clear that these incapacities are positive parts of these agents' moral characters. Not only did these incapacities lead the rescuers to perform morally good acts but they appear to have led the rescuers to do so for the right reasons. As Oliner and Oliner point out, while there were some rescuers who seem to be motivated by self-interest, the subjects of their study appear to have been motivated by a humanitarian desire to help those affected by Nazi persecution.Footnote ⁴⁵ Similarly, Wallenberg appears to have been motivated not by a desire for personal gain but from a feeling that he had a, “mission to save the Jewish nation.”Footnote ⁴⁶

In this section I have argued that a plausible interpretation of the claims made by those, like Wallenberg, who risked their lives to rescue the victims of Nazi persecution are plausibly viewed as claims about moral incapacity. Moreover, the source of this moral incapacity is plausibly seen as originating from a deep connection to the moral values that supported becoming a rescuer. In the remainder of this essay I will argue that this creates problems for the view that these acts involved sacrifice.

3. Sacrifice and Moral Incapacity

In this section I will argue that accepting that the rescuers were morally incapable of acting otherwise provides good reason to reject the view that such an alternative existed.

As seen in §2, in order for an act to involve sacrifice there must be an alternative act that the agent could perform that would better promote her welfare. This means that assessing the relevance of moral incapacity to whether or not an act involves sacrifice will involve investigating the concept of sacrifice more closely. In particular, what alternatives are being ruled in or out by the clause ‘available to the agent’. Clearly, in order for an act to be one that an agent could perform it will need to be an act that she is capable of performing. However, it is far from clear what kinds of capacity are relevant here.

I will make a start at answering this question by looking at the clearest case of incapacity, physical incapacity. Clearly, physical incapacity limits the range of options that are being considered when we discuss sacrifice. To suggest that Beethoven's life would have been improved had he been able to hear a performance of his ninth symphony, composed after he had become completely deaf, is perfectly comprehensible. To suggest that this shows that he made a sacrifice by not doing so is ludicrous. The fact that he was physically incapable of hearing this symphony prevents this act from being counted as one of the acts that are available to him.

The same is true for psychological incapacity. To suggest that Emily Dickinson would have been better off had she led a less reclusive life is a perfectly reasonable, though contestable, statement to make. To suggest that Dickinson made a sacrifice by leading a reclusive life is quite bizarre, at least if we accept the theory that Dickinson suffered from a form of agoraphobia. Dickinson's psychological incapacity to lead a less reclusive life makes it absurd to suggest that she made a sacrifice. Given her incapacity she had no choice but to live in this way. As with physical incapacity, when determining if an act involves sacrifice acts that an agent is psychologically incapable of performing should not be included in the comparison class.

What then of moral incapacities? At first these might seem different to other forms of incapacity. After all, the thought that Wallenberg made a great sacrifice by acting as he did appears to be a plausible one.

However, there are two responses to make to this suggestion. First, there is reason to doubt how seriously these intuitions should be taken. The problem with giving much weight to these intuitions is that it seems to ignore the viewpoint of the agent. To observers it may seem as if Wallenberg had the choice to stay or leave Sweden, and that he freely and knowingly chose to sacrifice his comfortable and safe life. However, the reason that this appears plausible to many observers is that many people when faced with the same circumstances would have been capable of acting differently and staying in Sweden. According to Wallenberg, though, this was simply not an option. It is instructive that Wallenberg made his claim about having no choice in response to being asked whether he was scared of the risks he was exposing himself to. If someone were to suggest to Wallenberg that he had performed a sacrifice it would be perfectly reasonable for him to deny this by simply repeating that there was no other way he could have acted.

More importantly, though, the view that acts that an agent is morally incapable of performing can be included in the comparison class for sacrifice but not acts that she is physically or psychologically incapable of performing appears bizarre when considered in light of the difference between moral and psychological incapacities. As I explained in §2, the key difference between moral incapacities and other forms of incapacity is that the agent morally endorses this incapacity. If an agent were to cease to endorse this incapacity and try to overcome it then it would cease to be a moral incapacity. While an agent may remain unable to perform the act, as soon as she attempts to overcome the incapacity the incapacity can only be considered a psychological incapacity and not a moral incapacity. Note, though, that this means that the difference between whether an incapacity is considered to be a moral incapacity or a psychological incapacity can come down to whether the agent endorses the incapacity. Given this, it would be odd if psychological incapacities prevent certain alternatives from being included in the comparison class but moral incapacities did not. After all, if this were the case then the endorsement of a moral incapacity would be sufficient in these cases to turn an act that is not part of the relevant comparison class into one that is. This, though, seems like an odd view. There seems no reason to think that the endorsement of incapacities should alter whether the acts the incapacity renders unavailable to the agent should be included in the comparison class for sacrifice. It seems much more plausible to think that, just like physical and psychological incapacities, when determining if an act involves sacrifice acts that an agent is morally incapable of performing should not be included in the comparison class.

4. Supererogation and Sacrifice Reconsidered

I am now in a position to raise the challenge against the traditional view of supererogation. I started this essay by introducing the example of Raoul Wallenberg who risked, and eventually lost, his life in order to save the lives of those facing persecution by the Nazis. This act seemed like a paradigmatic example of supererogation. Interestingly, Wallenberg, along with others who performed similar acts, claimed to have had no choice but to act as they did. In §2 I argued that if we take these statements seriously then we should accept that Wallenberg and those like him were morally incapable of acting otherwise. I then argued, in the previous section that acts that an agent is morally incapable of performing should not be included in the comparison class for sacrifice. Clearly, this presents a problem for the traditional view of supererogation, outlined in §1, that all acts of supererogation involve agential sacrifice. If we accept that Wallenberg's acts were supererogatory, that he was morally incapable of acting in any other way and that this means that his act cannot be considered a sacrifice then we have to reject the traditional view of supererogation.

The problem with the traditional view is that it fails to take the first personal accounts of moral exemplars into account and, as a result, wrongly assumes that the agent acts by choice in performing the acts that appear from an observer's point of view to be available alternatives. As a result, it wrongly rules that acts like Wallenberg's cannot be supererogatory. However, the problems for the traditional view do not end there. Given that Wallenberg's acts cannot be considered supererogatory on the traditional view, the next question to ask is how it is that they should be classified. There are two responses a defender of the traditional view might provide here. First, she could claim that these acts are obligatory. Alternatively, she could claim that these acts have to be classified in some other way. In the remainder of this section I will argue that neither option is acceptable and that, as a result, the defender of the traditional view faces a dilemma in her attempt to classify acts like Wallenberg's.

I will start by explaining what is wrong with classing these acts as obligatory. It might be thought that categorizing Wallenberg's acts in this way does not create any new problems for the view. Rather, it simply faces the objection that this clashes with the intuitively plausible thought that these acts are supererogatory.

This, though, would be a mistake. Classing Wallenberg's acts as obligatory would create a new problem for the traditional view. The problem is that it is unclear whether or not it would also be obligatory for someone morally capable of acting otherwise to act as Wallenberg did if they were in a similar situation. As I pointed out in the introduction, it seems very plausible to think that in general, acts like Wallenberg's are supererogatory. It can now be asked what effect one agent's moral incapacity has for others who are capable of either performing or not performing that act. There are two options here.

First, the defender of the traditional view could claim that an agent's moral incapacity can have an effect on what other agents who lack the incapacity are obliged to do in similar circumstances. This, though, is a bizarre view of moral obligations. The fact that one agent acts a certain way because they are morally incapable of acting otherwise is not the kind of consideration that can make it morally obligatory for others to act in this way.

The alternative for the supporter of the traditional view is to claim that the fact that one agent is morally incapable of performing some act has no influence on whether others are obliged to act in similar way. This seems more plausible but leads to odd consequences. If this is the case then while the person who would be morally incapable of doing otherwise would face this obligation, those who could act otherwise would not. This is no less problematic. As I argued in §2 the explanation for the moral incapacities of Wallenberg, and other rescuers, is down to a deep appreciation of the moral reasons that count in favour of their actions. To say that these moral incapacities create extra obligations then would be to subject those with a deeper appreciation of moral values to higher standards of moral obligation than the rest of the moral community.

I have argued that the view that the rescuers' acts are obligatory faces a dilemma when it comes to the issue of whether this affects whether it would be obligatory for those who are capable of acting otherwise. However, the supporter of the traditional view is not committed to claiming that Wallenberg's acts are obligatory. Perhaps, a supporter of the traditional view might wish to claim that his acts were neither obligatory nor supererogatory. The problem with this move is that it is under motivated. Given that these acts appear to be paradigmatic examples of supererogation, good reason needs to be given in order to make a persuasive case that they cannot be categorized under our existing range of deontic concepts. The fact that on closer inspection categorizing these acts in this way is incompatible with the traditional view gives us little reason to do so. It is worth bearing in mind at this point that the traditional view is not supposed to be a conceptual analysis of supererogation. Rather, it is supposed to be an explanation as to how it is that supererogation is possible. This is important. If the traditional view was a conceptual claim then it would be more plausible to suggest that the existence of acts that are beyond duty that do not involve sacrifice shows that there are acts that cannot be categorized by the existing deontic scheme. However, given that this view is supposed to explain how it is that supererogation is possible, it is ad hoc to suggest that a new category is needed for any acts that do not fit this explanation.

Concluding Remarks

I have argued that the traditional view of supererogation as involving agential sacrifice faces a significant problem. As I argued in Section 1, in order for an act to be said to involve sacrifice there must be an alternative act that the agent could perform that would better promote her well-being. In Section 2 I argued that it is plausible to think that many of The Righteous Gentiles were morally incapable of acting otherwise. In Section 3 I argued that acts that an agent is morally incapable of performing should not be included in the comparison class for sacrifice. This creates a problem for the traditional view, as it is committed to the claim that the acts performed by Righteous Gentiles who were morally incapable of acting otherwise cannot be considered supererogatory. As I argued in Section 4, this is problematic both because it is an implausible claim to make and because it raises the further question of how these acts should be characterized. In answering this question the supporter of the traditional view faces a dilemma. The traditional view of supererogation is unable to categorize these acts in a way that is not either ad hoc or deeply problematic

This discussion has important implications. As I pointed out in the introduction, the traditional view is often advocated by those seeking to defend mainstream views in Normative Ethics and Metaethics. It is, for example, often appealed to by those seeking to reconcile consequentialism with the existence of supererogation. Similarly, those who argue that there is a close connection between moral reasons and moral obligation often appeal to a version of this view in order to explain why the morally best act is not always morally obligatory. Given that the traditional view is flawed, supporters of these other theories cannot appeal to it as a way of reconciling their theories with the existence of supererogatory acts.

Finally, this discussion raises an interesting question about what the appropriate form of response to those who perform admirable supererogatory acts as a result of a moral incapacity. I have already argued that these moral incapacities do not alter what moral obligations agents possess. However, we might wonder whether we should find agents who act as a result of a moral incapacity any more or less admirable or praiseworthy than those who perform the act but could have acted otherwise. This is not a question that I have the space to address here but is one that is worthy of further consideration in the future.Footnote ⁴⁷