Beyond Consequentialism

Abstract

The most familiar objections to consequentialism claim that the theory gives counterintuitive answers to certain moral questions: requiring us to make massive donations to distant impoverished persons, for example, or permitting us to kill someone and harvest his vital organs to save others. But some consequentialists respond that our intuitions about such cases are themselves mistaken. Paul Hurley's ambitious book challenges consequentialism in a more fundamental way, arguing that the theory is internally inconsistent. The book addresses central issues in normative ethics. It is organized, lucid, and enjoyable to read. Hurley's arguments combine into an elegant whole. Much of the book is a kind of second-order project that consolidates, generalizes, and reframes critiques of consequentialism that others have made: arguments from “demandingness, alienation, confinement, and integrity” (21). Hurley poses questions that consequentialists must answer. Perhaps consequentialists can answer them, but doing so will help them to define their view and clarify their commitments. All consequentialists should therefore read this book.Hurley claims that there are “tensions within the consequentialist approach itself” (3). Most consequentialists, he asserts, accept three propositions:1)Rational Authority of Moral Standards (RAMS): agents have decisive reasons to do what they are morally required to do.2)Nonimpersonal Reasons (NIR): agents have some fundamentally nonimpersonal reasons that sometimes provide them with sufficient reasons not to bring about the best overall state of affairs.3)Consequentialist Moral Standards (CMS): agents are morally required to bring about the best overall state of affairs.In chapters 1–4, Hurley argues that these propositions form an inconsistent triad and that there is a conclusive case for dropping consequentialism (CMS), rather than RAMS or NIR. Chapter 2 introduces the challenge. Chapter 3 develops it by inverting the well-known objection that consequentialism is too demanding. Hurley argues that, on the contrary, consequentialism makes no demands at all because it is a theory of moral standards, not a theory of reasons for action. Consequentialists espouse RAMS but provide no rationale for it (62). RAMS is so widely accepted that critics of consequentialism have not demanded a rationale, nor have they noticed that consequentialists who espouse RAMS and NIR are being inconsistent. Agents have decisive reasons not to conform to such “demanding” moral standards.In chapter 4, Hurley intensifies the challenge, making novel use of Bernard Williams's objections to utilitarianism. Hurley argues that consequentialists typically presuppose NIR in their very conception of “the best overall state of affairs.” The best states of affairs are, by definition, ones in which many individuals act on various nonimpersonal reasons (pursuing personal projects, giving priority to loved ones). So the very formulation of consequentialism presupposes NIR. Consequentialists cannot, therefore, coherently sustain their essential claim that impersonal reasons always trump nonimpersonal ones.In chapter 5, Hurley interprets an intuition that is often thought to support consequentialism—that it is always wrong to fail “to do what is best.” He argues that consequentialists “misappropriate” this intuition, interpreting it tendentiously to mean that it is always wrong to fail to promote the best state of affairs, when in fact our intuition is merely that it is always wrong to fail to perform the best action—an intuition that does not support consequentialism.Chapter 6 proceeds to challenge the consequentialist claim that the impartiality of morality should be understood as impersonality. Inspired by T. M. Scanlon and Stephen Darwall, Hurley articulates a conception of impartiality that is interpersonal, but not impersonal. This standpoint, he suggests, supports agent-centered permissions and restrictions and represents the moral standpoint. The interpersonal standpoint also supports the rational authority of moral standards (RAMS) just where consequentialism fails to do so (chapter 7). Chapter 8 shows how Hurley's foundational challenge to act consequentialism extends to the most prominent recent versions of indirect, rule, and nonfoundational consequentialism, as defended in books by Brad Hooker (2000), David Cummiskey (1996), and Derek Parfit (2011).Hurley makes many important points in these chapters. Consequentialists have not explained how anyone could have decisive reasons to be moral if consequentialism is the correct theory of moral standards. They have not converged on a position in the internalism/externalism debate(s). They have not adequately justified their interpretation of the intuition that it is always right to do what is best. Nor have they adequately defended impersonality as the uniquely correct conception of impartiality. Hurley rightly pressures consequentialists to defend positions on these topics. In the rest of this review, I shall suggest how consequentialists might interpret RAMS and NIR so as to avoid Hurley's central incoherence argument.Hurley's argument that the RAMS/NIR/CMS triad is inconsistent assumes that, if you have a decisive reason to Φ, then you do not have a sufficient reason not to Φ. This conditional is true only if ‘reason’ has the same denotation in the antecedent and the consequent. So it matters whether ‘reasons’ in RAMS has the same denotation as the second appearance of ‘reasons’ in NIR (in the phrase ‘sufficient reasons’). Must it? In some contexts, ‘reasons’ denotes a subset of reasons (such as prudential reasons), not reasons, simpliciter. In other contexts, ‘reasons’ denotes a broader set of reasons (at the limit: all reasons). As Hurley emphasizes, RAMS is implausible as applied to prudential reasons. On any plausible theory of prudential reasons and any plausible theory of moral standards, morality can require an agent to act against his or her own interests, as when he or she could benefit by making a lying promise without suffering any adverse consequences. RAMS becomes plausible only if ‘reasons’ is understood in a broader, more moralized sense, such that the agent can be said to have ‘decisive reasons’ not to make the beneficial lying promise.Given this interpretation of RAMS, I think there is a reading of NIR that is compatible with both consequentialism (CMS) and RAMS. We can understand ‘sufficient reasons’ in NIR to denote reasons in the narrower sense. A devoted father might have sufficient reasons to spend thousands of dollars on music lessons for his daughter and decisive reasons not to spend the money on a racehorse for himself (although he has a pro tanto prudential reason to buy the racehorse if it will bring him joy). These are reasons only in the narrower sense. If he wrongly embezzled the money, then he has decisive reasons, in the broader sense, to return it to its rightful owner, rather than spending it on anyone. It is not inconsistent to assert that he has decisive reasons, in the narrower sense, to buy the music lessons, and decisive reasons, in the broader sense, not to do so.A decisive reason, in the narrower sense, could simultaneously constitute a pro tanto reason in the broader sense. For example, the father's decisive reason to buy the music lessons could constitute a pro tanto reason, in the broader sense, to do so, but not a sufficient reason, because the money was embezzled. A consequentialist can similarly hold that the father's decisive reason to buy the music lessons constitutes a pro tanto reason, in the broader sense, to do so, but not a sufficient reason, because impoverished children will benefit more from the money than his daughter will benefit from the lessons. Consequentialists can accept that decisive reasons, in the narrower sense, always constitute pro tanto reasons in the broader sense. They need only insist that such pro tanto reasons to Φ do not constitute sufficient reasons to Φ, in the broader sense, if Φ-ing fails to bring about the best overall state of affairs because in that case there are stronger pro tanto reasons not to Φ.I have not proven, of course, that these stronger pro tanto reasons always exist. That would require proving that consequentialism is true. I am just questioning Hurley's charge that consequentialism is incoherent. Nor have I shown that any actual consequentialist endorses my interpretations of RAMS and NIR. I have simply argued that there is a plausible interpretation of each claim such that, contra Hurley, they do not jointly conflict with consequentialism. However, Hurley succeeds in demonstrating that consequentialists cannot consistently endorse RAMS and NIR if they interpret the claims as he does.I have also said nothing about what reasons, in the broader sense, are. Here I can offer only an analogy. Consider an agent who is indifferent to her own future welfare. She knowingly ingests a drug that produces immediate pleasure but will kill her next week. She stubbornly denies that she acts irrationally. There is a narrow sense of ‘rational’ in which she acts rationally—she takes effective means to her dominant end of the moment (obtaining immediate pleasure). But there is a broader (although still prudential) sense of ‘rational’ in which she acts irrationally—she acts against her long-term self-interest. She is rationally required, in this broader sense, to care about her future interests, even if she does not. Analogously, there is a narrow sense of ‘rational’ in which agents are often rationally justified in not bringing about the best overall consequences because doing so will not promote their own welfare. But there is a broader sense of ‘rational’ in which they would be rationally required to do so. Hurley writes as though there were a straightforward understanding of ‘rational,’ simpliciter (50–53), to which moral theories must answer, but I am not sure what sense he has in mind.The views expressed in this review are the author's. They do not represent the views of the National Institutes of Health or the United States Department of Health and Human Services.