INTRODUCTION
It is plausible to think that there exist acts of supererogation (acts that are morally optional and morally better than the minimum that morality demands). It also seems plausible that there is a close connection between what we are morally required to do and what it would be morally good to do. Despite being independently plausible these two claims are hard to reconcile. This problem has been called ‘the paradox of supererogation’ and continues to attract the attention of moral philosophers.Footnote 1 My aim in this article will be to respond to a recent solution to this puzzle proposed by Dale Dorsey.Footnote 2 I will start, in section I by presenting the paradox. I will then, in section II, present Dorsey's solution to this paradox. I will then explain why this response fails to provide a satisfactory solution to the problem. I will start by outlining, in section III, the reasons why we should make room for the category. I will then, in section IV, show that these reasons provide no support for making room for the category that Dorsey proposes. I will finish, in section V, by considering and responding to an objection that might be raised against my argument.
I. THE PARADOX OF SUPEREROGATION
In order to assess properly Dorsey's solution to the paradox of supererogation we must first get a clear idea of the problem. Let us start with the assumption that acts of supererogation exist.
Existence of Supererogatory Acts: Acts of supererogation exist (or at least their existence is a possibility).
As we have seen, a supererogatory act is one that is morally optional. We can formulate this in the following way:
Permissible not Required: If an act φ is supererogatory, φ is morally permissible, but is not morally required.Footnote 3
As we have seen, a supererogatory act is one that is better than some merely obligatory alternative. We can formulate this in the following way:
Morally Good: If an act φ is supererogatory, φ is especially morally good or meritorious in comparison to acts that fulfil obligations in a minimal way.Footnote 4
The final part of the paradox is the claim that moral obligations ought to be closely tied to our moral values. The thought here is that what we morally ought to do should be closely related to what it would be morally good for us to do. More formally, an act's deontic properties ought to be determined by its evaluative properties. This principle is called ‘the good–ought tie-up’.Footnote 5
The good–ought tie-up: If an act is morally obligatory then there are no other available acts that are morally better to perform.
We can now see how the paradox arises. These four claims seem independently plausible but when taken together they lead to a contradiction. If we accept the good–ought tie-up then we should reject even the possibility of acts that satisfy Morally Good and Permissible Not Required. To provide a solution to this paradox we must reject one of these claims.
II. DORSEY'S SOLUTION
Dorsey's proposed solution to this puzzle is to reject the account of supererogation given by Permissible Not Required and Morally Good. In its place Dorsey proposes the following account of supererogation:
Permissible not Required II: If an act φ is supererogatory, φ is rationally permissible, but is not rationally required.
Morally Good II: If an act φ is supererogatory, φ is especially morally good or meritorious in comparison to other rationally permissible actions.Footnote 6
According to this new account, supererogatory acts are morally good but not rationally required. In order to get to grips with this proposal we need to understand what Dorsey means by ‘rational requirement’. Dorsey explains his use of the term in the following way:
Moral requirements are, well, just that: if I fail to conform to a moral requirement, this entails that I will have behaved immorally, or in a morally unjustified way. But there are many different sorts of requirements – not just moral – that I face. I face legal requirements, prudential requirements, requirements of etiquette, requirements of my neighbourhood association. Sometimes these requirements will conflict. But in cases of conflict, it seems natural to ask ourselves what we ought to do really, or all-things-considered. More generally, in the case of conflicting requirements, how should I live? For the sake of brevity, I will refer to this ‘all-things-considered’ requirement, which is distinct from, e.g., moral, legal or prudential requirements, as the ‘rational’ requirement, or rational ‘ought’.Footnote 7
In Dorsey's usage, then, rational requirements are all-things-considered normative requirements.Footnote 8
This account is compatible with the good–ought tie-up, as morally obligatory acts are those that are best supported by moral reasons. It is the rationally obligatory acts that can be morally surpassed. Dorsey also claims that this response to the problem does better than its rivals at handling cases of supererogation that would have been obligatory were it not for the fact that they require a non-trivial sacrifice on the part of the agent.Footnote 9 In the remainder of this article I will argue that Dorsey's account ought to be rejected, regardless of whether it has these advantages. The reason for this is that it is unable to capture the intuitions that push us towards accepting the need for the category of the supererogatory.
III. WHY ACCEPT SUPEREROGATION?
In this section I will investigate the reasons why we should accept that supererogatory acts exist. I will do so by examining J. O. Urmson's discussion of supererogation that provoked the contemporary discussion of the concept. It is worth noting that Urmson is one of the few theorists in the literature to provide arguments in support of the need to make room for supererogation in our deontic scheme. That is not to say that Urmson is unusual in accepting the need to make room for such a category. However, while many take the need for such a category for granted, Urmson offers explicit arguments in defence of this position.
Urmson argued that for a normative moral theory to be acceptable it must make room for supererogation.Footnote 10 He gave the following example to support this claim:
We may imagine a squad of soldiers to be practicing the throwing of live hand grenades; a grenade slips from the hand of one of them and rolls on the ground near the squad; one of them sacrifices his life by throwing himself on the grenade and protecting his comrades with his own body.Footnote 11
Urmson claimed that the soldier's action is morally good but not morally required.
That this act is morally good should be fairly uncontroversial, but what is it about this act that leads us think that it is not morally required? Urmson cites two features of this act that give us reason to think that it is not obligatory.
Urmson outlines the first feature in the following:
Though he is clearly superior in some way to his comrades, can we possibly say that they failed in their duty by not trying to be the one who sacrificed himself? If he had not done so, could anyone have said to him, ‘You ought to have thrown yourself on that grenade?’ Could a superior have decently ordered him to do it? The answer to all these questions is plainly negative.Footnote 12
In other words, the soldier's act cannot be morally obligatory because it would be inappropriate to demand that he perform the act.
Clearly, in order for it to follow that the fact that no one could demand this act from the soldier means that it was not obligatory, we need some reason to think that when someone has a duty to act in a certain way it is appropriate to demand that they do so. One reason why we might think this is if we accept that there is a conceptual connection between duty and the appropriateness of demands. This view has a number of supporters. John Stuart Mill, for example, claimed that, ‘It is part of the notion of duty in every one of its forms that a person may rightfully be compelled to fulfill it.’Footnote 13 Similarly, Stephen Darwall claims that, ‘moral obligation and standards of right and wrong are conceptually related to what the moral community, and we as members of it, can demand.’Footnote 14
The second feature of these acts that suggests that they are not morally required is that in both cases it would be inappropriate to blame someone who did not perform the act. Urmson claims that when someone fails to perform an act of supererogation, ‘there could be no question of reproaches for not so acting.’Footnote 15 Urmson's point is that the fact that we could not blame someone for failing to jump onto the grenade gives us reason to think that doing so was not obligatory.
As with the previous point, the absence of a duty only follows from the inappropriateness of blame if we assume that there is a link between failure to fulfil a duty and blameworthiness. Again, one reason we might think such a link exists is that there is a conceptual connection between duty and blameworthiness. This view also has a number of supporters. Darwall, for example, claims that, ‘It is a conceptual truth that an act is morally wrong, if, and only if, it is blameworthy if done without an excuse.’Footnote 16
We can summarize these two arguments in the following way:
The Argument from Demand
P1: If an act is morally obligatory then the agent can legitimately be demanded to perform it.
P2: No one could legitimately demand that the soldier dive on the grenade.
C: Diving on the grenade is not morally obligatory.
The Argument from Blame
P1: If an act is morally obligatory then the agent is blameworthy for failing to perform it (in the absence of an excuse).
P2: The soldier would not be blameworthy for failing to dive on the grenade, even if he lacked an excuse.
C: Diving on the grenade is not morally obligatory.
What we have here are two claims about the appropriateness of demands and blame (P2) which, when combined with two supposed features of the concept of moral duty (P1), provide two arguments in support of Urmson's claim.
These arguments, combined with the assumption that the soldier's act is morally good, show that a tripartite view of deontic categories involving the required, the forbidden and the indifferent is insufficient.Footnote 17 The soldier's act is neither required nor forbidden nor indifferent. In order to accommodate such acts in our deontic framework we must make room for a new category of acts that are morally good but not morally required.
Urmson's arguments, though, do not exhaust the reasons why we might want to make room for the concept of supererogation. Another reason to do so is that this term picks out a recognizable concept in common-sense moral discourse. Most clearly, the phrase ‘beyond the call of duty’ is a familiar part of everyday moral talk and thought. If we want to do justice to this feature of moral discourse then we have good reason to make room for the concept of supererogation.
IV. EVALUATING DORSEY'S PROPOSAL
In the previous section we looked at why there is a need to make room for the category of the supererogatory. By making room for this additional category we are able to reconcile the moral goodness of the soldier's act with the thought that it would be inappropriate to demand this act from him or blame him for failing to act in this way.
The first objection that might be raised against Dorsey's view is that it fails to reconcile the thought that the soldier's act is morally good with the thought that it would be inappropriate for anyone to demand this act from him or to blame him if he had failed to act in this way. Accepting that there are acts that are rationally optional and especially morally good in comparison to other rationally permissible actions fails to explain how it is possible for the soldier's act to be morally good yet for it to be inappropriate to demand that he perform it or to blame him for non-performance. As the discussion in section III demonstrated, if we accept the conceptual claims about moral requirements then the inappropriateness of demands or blame gives us reason to think that the soldier's act is not morally required.
This, though, is not really a problem for Dorsey, as elsewhere he argues against the claim that moral obligations are conceptually connected to blameworthiness and the legitimacy of demands.Footnote 18 Of course Dorsey's position will stand or fall on the arguments that he gives against this conceptual link but unless some reason is given to show that Dorsey's arguments for this position are unsuccessful then this is not a problem for Dorsey's account.
However, a further problem remains. If we reject the arguments given in section III then there seems little reason to accept Dorsey's account of the supererogatory. By denying these arguments Dorsey would also remove the need to make room for the concept. If we reject the arguments given in section III then the next move would not be to find some other way to understand the supererogatory. Rather, it would be simply to abandon this category, as it would not be doing any work. By denying that moral obligations are conceptually linked to blameworthiness and the legitimacy of demands, Dorsey removes the need to make room for any category of the supererogatory and so his proposed account is redundant.
Of course, there is another way for Dorsey to show the need to make room for this additional deontic category. As we saw in section III another reason why we might want to make room for the concept of supererogation is that it captures a feature of folk moral discourse. Someone sympathetic to Dorsey's view might be tempted to suggest that this account should be seen as an attempt to make sense of this feature of ordinary moral discourse.Footnote 19
The problem with this response, though, is that if this is what Dorsey's account of the supererogatory is setting out to achieve then it is unsuccessful. In order for Dorsey's account to be a plausible account of the concept being picked out by ‘beyond the call of duty’ we need to accept both that requirements of practical reasons are a feature of ordinary normative discourse and that these requirements are picked out by the term ‘duty’. It could plausibly be argued that the concept of a rational requirement plays some role in ordinary talk and thought. It seems reasonable enough to think that the concept of a requirement of practical reason is being appealed to when people make utterances like, ‘You must do your homework,’ or ‘You must take your medicine.’ However, it is implausible to claim that it is this concept that is being picked out by the term ‘duty’. This can be seen if we substitute ‘have a duty to’ for ‘must’ in the previous utterances to read: ‘You have a duty to do your homework’ and ‘You have a duty to take your medicine.’ It is clear that this substitution creates a change in meaning between the two sentences, at least if we took the original utterances to refer to rational requirements. These new utterances suggest a moral requirement not a rational requirement. Dorsey's account then is not a plausible articulation of what is meant by the ordinary language phrase ‘beyond the call of duty’. If Dorsey's account is supposed to articulate what is meant by the ordinary language phrase ‘beyond the call of duty’ then it is unsuccessful. If this is not what Dorsey's account is designed to achieve then we are again left without a motivation for making room for this concept in the first place.
V. OBJECTIONS AND REPLIES
We might think that there is a way of responding to this objection. Perhaps Dorsey could claim that although the arguments given in section III should be rejected there are similar arguments that could be given to show the need for the category he proposes. If we reinterpret the claims made in the premises of both arguments to refer to rational requirements rather than moral requirements then this would provide us with the following two arguments to show that Urmson's soldier is not rationally required to act as he did:
The Revised Argument from Demand
P1: If an act is rationally required then the agent can legitimately be demanded to perform it.
P2: No one could legitimately demand that the soldier dive on the grenade.
C: Diving on the grenade is not rationally required.
The Revised Argument from Blame
P1: If an act is rationally required then the agent is blameworthy for failing to perform it (in the absence of an excuse).
P2: The soldier would not be blameworthy for failing to dive on the grenade, even if he lacked an excuse.
C: Diving on the grenade is not rationally required.
Of course for this response to work it would need to be shown that there is good reason to accept the first premise of these arguments. This is problematic, as it is far from clear that there is good reason to accept the first premise of either argument. To see why, suppose I have a prudential requirement to consume a healthy diet. Suppose further that this prudential requirement is sufficiently strong to generate an all-things-considered normative requirement. It is not obvious from this that it follows that others can legitimately demand that I consume a healthy diet or blame me if I fail to do so.
Even if we think that it can be shown that blame and demands are always appropriate for rational requirements a further problem with this approach remains. The problem is that even if the revised arguments are effective they fail to provide any reason to create a new deontic category that fits the account that Dorsey provides. The original arguments may have shown that we could not divide morality into the required, the indifferent and the forbidden but pose no threat to this division. The fact that an act can be morally better than a rational requirement does not give us any reason to propose a new deontic category; it simply tells us that moral requirements are not rational requirements. While this result is interesting, it does not provide any reason to create a new moral category. The tripartite division of morality is unthreatened by either argument. Even if we accept the revised versions of these arguments, then, there is no reason to accept Dorsey's account of the supererogatory.
CONCLUSION
I have argued that Dorsey's attempt to solve The Paradox of Supererogation by offering a new account of supererogation is unsuccessful. I started by looking at the reasons why we might be tempted to make room for supererogation in our ethical scheme. I then argued that these considerations provide no support for the existence of a deontic category fitting Dorsey's account. This creates a problem for Dorsey's account, as there is no longer a need to make room for any category of the supererogatory and, as a result, the proposed account is redundant.Footnote 20
