A COMPLETE MORAL THEORY MUST PROVIDE an account of what is often called "moral motivation," the job of which is to explain "what reason one has to be moral." This much tends to be agreed upon. Less clear, however, are the details concerning what is being demanded. The term "moral motivation" seems to imply that what is sought is primarily a psychological account, concerning, perhaps, what does or could motivate an agent to moral action. The language of "the reason to be moral," on the other hand, has a normative ring to it: The explanation being sought here concerns what reasons there are. Thus while there is largely an agreement on the need to provide some kind of account concerning the link between morality and motivation, the details are unclear in the abstract, and anyone hoping to answer the call must do some clarificatory work at the beginning.
T. M. Scanlon, in making a case for his unique brand of contractualism, does just this. In his words, a satisfactory moral theory must "explain the reason-giving and motivating force of judgments of right and wrong" (1998: 147)--a challenge that clearly has both motivating and normative components. Rather than explaining "why one ought to be moral," or "why one has reason to be moral," Scanlon interprets the moral motivational question as asking "how the fact that an act is wrong provides a reason not to do it" (1998: 147-49). And this question can be read in two different ways. On the one hand, we want to know the answer to an empirical question, namely: What do we care about when we care about right and wrong? And on the other hand, we are asking a normative question: Why is right and wrong something we must care about (1998: 148)? The question of moral motivation, then, is actually two questions. For this reason, Scanlon suggests that a better name for the challenge of "moral motivation" is "the motivational basis of morality," as this name wears its dual character more clearly on its face. (1)
In What We Owe to Each Other,; Scanlon outlines the central problem facing any account of the motivational basis of morality, providing his own, contractualist solution. The problem, Scanlon thinks, is to navigate "Prichard's Dilemma" (PD), which arises from the demand that explanations of the reason to act morally be both (a) helpfully explanatory and (b) relevant to morality. According to PD, all moral theories must explain the reason to be moral by reference either to a moral consideration or a nonmoral consideration. Explanations by reference to moral considerations, however, are trivial and unhelpful (thereby violating (a)), while explanations by reference to nonmoral considerations offer implausibly external incentives to be moral (thereby violating (b)). Scanlon's solution to this dilemma is to explain the reason to be moral by reference to a moral-but-still-helpful value--namely, the value of living with others on terms that all can accept.
In this paper, I will accept many of Scanlon's philosophical commitments. Despite this theoretical friendliness, however, I will suggest that Scanlon's own solution fails to navigate PD. I then attempt to derive an alternative solution from his framework, but argue that it, too, fails. In the end, I take the failure of these promising views to indicate that PD is unlikely to be solved by a traditional account of moral motivation, and so suggest a change of strategy.
According to Scanlon, any attempt to provide an account of the motivational basis of morality faces a difficult challenge, which he calls "Prichard's Dilemma," after H. A. Prichard's description of a similar dilemma (1912). The challenge here is that the question--"Why be moral?"--seems to require, on the one hand, a moral answer. One has reason not to act immorally because the fact of an action's being wrong is a reason not to do it. But this, of course, is not much of an answer at all; it takes the reason-giving force of morality for granted, when the reason-giving force of morality is precisely what we want explained. The same is true of other, closely related answers to the moral-motivational question. In a particularly Kantian moment, one might be tempted to say that one has reason to be moral "because it is one's duty," (2) but this answer fares no better. It simply assumes that duties have reason-giving force, which is to say that morality has reason-giving force. On this horn of the dilemma, the answers are what Scanlon calls "trivial," or what I call "unhelpful," as the explanations take for granted precisely the phenomenon that we are trying to explain.
On the other hand, one might attempt to provide a clearly nontrivial, obviously helpful explanation of the reason to be moral by reference to a nonmoral consideration. For example, one might think that the most satisfying way to account for the reason to be moral would be to show that being moral would make one happy, or would otherwise be in one's interest. (3) But on this horn, we face a different challenge, as such an answer seems to provide one with the wrong kind of reason to be moral. If I ask why I have reason to donate to famine relief, there seems to be something inappropriate about citing how good it will make me feel in response. To do so, Scanlon says, would be to offer an "implausibly external incentive" for acting rightly. And the whole Kantian tradition seems to recognize this danger, as it tells us that such an act would lack true moral worth; what one ought to do is act rightly out of respect for the moral law. While considerations of one's happiness might in fact motivate some people to perform some right actions, such considerations are not what we would expect the paradigmatically moral person to cite as her reason for acting. According to this set of intuitions, then, an account of the motivational basis of morality must also be relevant to morality.
PD, then, demands of any account of moral motivation that it be both helpfully explanatory and relevant to morality. While Scanlon makes this challenge fairly quickly, I take it that his intuitions are widely shared. Accounts that explain our reason to be moral in terms of our happiness or interests do, in fact, seem to have a problem explaining the special value of acting out of duty, while those that explain our reasons by reference to very closely related concepts like "duty" are unsatisfying.
Although Scanlon's analysis of the two, competing desiderata for a theory of moral motivation is compelling, the challenge is not a true dilemma. Instead, candidate moral motivational accounts seem to be able to succeed more or less with regard to the criteria of relevance and helpfulness. This can be seen even in the two "trivial" candidates, as the Kantian view that one has a reason to act rightly because it is one's duty is slightly less trivial (and therefore more explanatorily helpful) than the view that one has a reason not to act wrongly because the fact of an action's being wrong just is a reason. Similarly, although many bristle at the idea that one has a reason to act rightly because it would make her happy, it is plausible that this is a more morally relevant answer than "one has a reason to act rightly because it would result in more facial muscles being used" (supposing that were true). And so it in fact looks like PD is not a genuine dilemma, but a challenge of meeting two, seemingly contrasting desiderata, solutions to which satisfy these criteria to more or less a degree. Scanlon says, similarly:
Answers [to PD] can thus be arrayed along one dimension according to their evident moral content, ranging from those that appeal to what seem most obviously to be moral considerations (thus running the risk of triviality) to those having the least connection with moral notions (thus running the risk of seeming to offer implausibly external incentives for being moral)" (1998: 150). Scanlon's setup of the moral motivational problem does, however, help us to see why PD--although not a true dilemma--is so challenging. The moral motivational question, remember, has two parts: A successful theory must explain both "what we care about when we care about right and wrong," as well as "why this is something we must care about" (1998 148). (4) Or, more positively, a successful view will ground the reason to be moral in "a substantive value which seems at the same time to be clearly connected to morality and, when looked at from outside morality, to be something which is of obvious importance and value, capable of explaining the great importance that morality claims for itself' (1998: 151). Although Scanlon uses slightly different language in different places to characterize the second half of this burden, the general idea throughout is that a solution to PD requires identifying a reason to be moral that both seems to be of the right kind (and so morally relevant), and that is appealing in itself, even when viewed apart from morality. When we focus on the latter half of the challenge, we are pushed in the direction of looking for incentives: What kind of reason can we come up with that seems obviously compelling, in itself, apart from any relationship to morality? Such nonmoral incentives, however, remind us of the former half of the challenge, and we find ourselves dissatisfied with a view that has paradigmatically moral actors acting for nonmoral reasons. Thus, we simultaneously want the answer to have moral content, to satisfy the former half of the challenge, while being devoid of moral content in order to satisfy the latter. A satisfying solution to PD must somehow convince us either that a moral explanation can be helpful and nontrivial, or that a nonmoral explanation can still be morally relevant. (5) On my understanding of Scanlon's account, he opts for the former strategy, to which I now turn.
Scanlon's Account: The Good of Mutual Recognition
According to Scanlon's well-known criterion of wrongness, an act is wrong just...