Coalitions of Reasons and Reasons To Be Moral

H.A. Prichard famously argued that philosophers who aim to answer the question "why are we bound, or why ought we, to do what is right?" are pursuing a vain quest.² Prichard advertized this finding in a provocative way, asserting that moral philosophy is the mistaken enterprise of trying to reply to a question to which no replies are possible. Should we be skeptical about the resources of philosophy for addressing issues about morality's authority? In my view, Prichard was mistaken. But identifying his error requires getting clear about how reasons of autonomy contribute to a person's normative reasons for action.

I. Prichard's Argument and the Initial Statement of the Case Against It

Prichard's reasoning is sometimes opaque. But here is what I think he had in mind. If a person is asking why she should to do what she believes is her moral duty, then she is asking a question about morality's authority, or what I will call the authority question. A suitable answer to the authority question must be normative. It must identify an 'ought.'

There are two kinds of reply to the authority question that Prichard considers and goes on to reject. The first appeals to the goodness of acting on a moral duty. Prichard is willing to entertain the possibility that good consequences determine whether some action is right.³ He is willing to allow, in other words, that an action would not be a moral duty (but perhaps some other kind of duty, like a duty of etiquette, or a religious, or a legal duty) unless that action caused good consequences. He claims, however, that even if philosophy can answer questions about the characteristics that render actions right, that enterprise leaves the authority question unresolved. This is because answers of that kind are not normative. A person could agree with the utilitarian, for example, that the production of good consequences is what makes right actions right and sensibly conclude that the question relating to why she ought to do what is good has not been addressed. In this vein, Prichard complains that if the utilitarian argument "is to restore the sense of obligation to act, [it] must presuppose an intermediate link, viz., the further thesis that what is good ought to be."⁴ This complaint allegedly generalizes to all proposals regarding morality's constitution.

According to Prichard, only one other kind of reply to the authority question is possible. It holds that a person ought to do her duty because moral compliance is the best means – whether causal or constitutive – to her advantage, interests, or happiness. I refer to this as a teleological theory of obligation since it identifies a person's reason to be moral with the attainment of a non-moral goal. Prichard attributes arguments of this kind to Plato, Butler, Hutcheson, Green, Paley, Mill, and Sidgwick.⁵ His general complaint against teleological theories of moral obligation is that, even if philosophers succeed in demonstrating that doing our duty is conducive to realizing our ends, this finding does nothing to resolve questions about morality's authority.

There are two ways to interpret Prichard's complaint. On one reading he is critical of teleological theories of obligation because they are not genuinely normative. Discussing attempts to justify moral requirements by showing that compliance is a necessary means to our ends, Prichard writes: "The answer is, of course, not an answer, for it fails to convince us that we ought to keep our engagements; even if successful on its own lines, it only makes us want to keep them."⁶ The complaint is that teleological theories of obligation have in fact opted out of the business of identifying objective reasons to be moral. They resolve obligation into motivation or incentives.

But if that is Prichard's argument, it naturally invites a reply: why not reformulate a teleological theory to hold that a person ought to adopt the necessary means to her ends?⁷ This reply can be developed in different ways. One possibility holds that a principle of instrumental reason is correct...