This post is a summary of chapter 1 of George Sher’s Desert, keyed to session 1 of the MT Seminar on Philosophy. For background, read this post.
The Preface of George Sher’s Desert begins with a statement of the obvious: the concept of moral desert is central to moral judgment and deliberation, and yet (at least as of 1987, the date of the book’s publication), we lack a systematic account of the nature and justification of moral desert. Indeed, ever since Rawls’s “critique” of moral desert in A Theory of Justice (if it is one), doubt has been cast on the coherence or viability of the concept. So the book sets out to provide a justification: “one of my central aims is to display the underlying justification of desert claims (George Sher, Desert, p. xi).
Sher calls his approach to ethical theorizing “pluralistic”: instead of beginning with some higher-order account or principle that entails desert claims, he starts with the desert claims “we all make,” and works to systematize them with the proviso that “desert need not have any single normative basis” (Sher, Desert, p. xii). So the approach is both “pluralistic” and in a sense, “bottom up.”
Chapter 1, “The Dimensions of the Problem,” begins by laying out four distinctive puzzles about moral desert which “seem especially perplexing,” (p. 3) which clearly require resolution if an account of desert is to get off the ground, and by implication, whose resolution serves the larger end of displaying the underlying justification of desert claims.
1. The first puzzle concerns the apparently close connection between moral desert and moral obligation. In some cases, it is morally obligatory to provide or withhold what is morally deserved or undeserved. What makes this puzzle particularly puzzling is that in some cases, moral desert reverses ordinary moral obligations. For instance, while we typically have a duty of non-maleficence, if someone deserves punishment, we’re obligated to harm them, a reversal of the usual state of affairs. On the positive side, we typically have a right or permission to mind our own business, but a hiring officer arguably has the obligation to hire the best qualified application from a pool of applicants. So the first puzzle is why moral desert has so much normative power.
2. The second puzzle is in effect a double puzzle: a puzzle in itself, and a puzzle in relation to (1). Many things are deserved or undeserved, and yet obligate no one to do anything in particular. This is puzzling in itself: how can there be a moral norm that requires nothing of anyone in particular? There’s a puzzle here in relation to puzzle (1), as well: how can one and the same norm obligate in some contexts, and fail to do so in others?
3. A third puzzle concerns desert’s orientation to the past. We might, for instance, say that a political candidate does not deserve to be elected to office because of his past misdeeds, where saying this competes with or overrides considerations about the beneficial policies he might implement in the present and future. Why should considerations about the past be thought to override considerations about the future? Why not treat the past either as a moral equivalent of a sunk cost, or as morally relevant only when information about the past is required to realize some present or future aim?
4. A fourth puzzle concerns the very adequacy of the concept: how can anyone deserve anything? Claims about desert are often grounded in actions: Smith performs an action, X, and is thought to deserve something in virtue of having performed X. But Smith’s capacity to perform X is arguably based on capacities that Smith did nothing to deserve. Every desert claim involves the performance of actions that presuppose certain capacities for that action, and many if not most of these capacities are inherited (whether by genetics or upbringing or both), rather than being the object of intentional action. If the capacities for desert-conferring actions are undeserved, how do the actions confer desert? How can what is undeserved be the basis for what is deserved?
Consideration of these four puzzles raises the possibility that there is no single, unified conception of moral desert. Only (4) applies to all and only claims of desert. Claims (1) and (2) are disjoint, hence might well have disjoint (and conceptually unrelated) solutions. The solution to (3) may be unrelated to the solution to (1): if (1) requires an account of why the best qualified ought to be hired, and that account is fundamentally forward-looking, then the solution to (1) is not only different but potentially in conflict with the account of (3). (Example: if a politician ought to be voted-for on the basis of his qualifications to bring about the best policies in the future, such an account may well conflict with one that emphasizes the irreducible backward temporal orientation involved in .) So it’s entirely possible that the solution to (1), (2), and (3) are conceptually distinct from (even fundamentally unrelated to) each other, leaving us with a gerrymandered account of moral desert.
This last consideration suggests that we take a look at concrete instances of moral desert with a view to identifying patterns of similarity and difference between them. Sher offers a list of fifteen paradigm instances, noting both their heterogenity and their “common structure” (Sher, pp. 6-7). As for the common structure, all fifteen desert claims “display the form ‘M deserves X for A,'” where M is a moral agent (or patient?), X is something deserved, and A is a “desert basis,” usually an act, an omission, or a trait/characteristic of or by M. The heterogeneity arises in part through variation in the values of the variables of the common structure: different agents deserve different things in different contexts, on different bases. The question is how to make sense of both the heterogeneity and the common structure. “Making sense” is both an explanatory and a justificatory matter: how could we rationally come to hold the beliefs we do about desert?
Three prominent (or then-prominent) attempts to answer that question all seem to fail: act-utilitarianism, Rawlsian institutionalism (considered as a form or variant of rule-utilitarianism), and intuitionism.
Act-utilitarianism holds that an agent’s every act ought to maximize greatest expected overall utility, given the options available to the agent. Since reward and punishment have incentive effects that serve that aim, act-utilitarians are obliged to reduce moral desert to the incentive-producing considerations that promote overall expected utility.
While some utilitarians (e.g., R.M. Hare) have gone a long way toward accommodating act-utilitarianism to moral desert (and/or vice versa), the fact remains that intuitive judgments about moral desert conflict with the demands of maximizing utility. Intuitively, it’s thought that people deserve reward or punishment regardless of whether that promotes overall utility. But if so, that intuition by definition cannot be captured by act-utilitarianism, no matter how sophisticated.
Rule-utilitarianism holds that the set of rules we adopt ought to aim to maximize greatest overall utility, given the rule-adopting options collectively available to us. One natural and plausible variant on rule-utilitarianism is what might be called Rawlsian institutionalism, after the view Rawls took and developed on the subject between roughly 1955 and 1971 (“Two Concepts of Rules,” Philosophical Review 64:1 ; A Theory of Justice , e.g., section 48).
Rawlsian institutionalism drops the utilitarian focus on utility maximization, offers non-utilitarian criteria for the justness of our institutions, and turns moral desert into a norm internal to institutions. The justness of just institutions is not a matter of promoting moral desert, but of other, unrelated considerations (irrelevant here); claims about moral desert are reducible to the legitimate demands and expectations of just institutions, but are not applicable to or beyond those institutions. “Thus the concept of moral worth is secondary to those of right and justice, and it plays no role in the substantive definition of distributive shares” (Rawls, Theory of Justice, rev. ed., p. 275).
Once again, Sher takes this view to conflict with our bedrock intuitions about moral desert. Considerations of moral desert ought to apply to and beyond institutions, not just within them. If we assume this (a big if), then Rawlsian institutionalism fails.
Intuitionism takes claims about moral desert to be both epistemically and morally basic, known directly by intuition, and not susceptible to further justification. This view has a natural affinity with Sher’s own “bottom up” approach to the subject. Both Sher and the intuitionist seem to take probative force of first-order claims of moral desert for granted, differing only on the implications of this first move. Intuitionism stops with those first-order judgments, regarding them as justified insofar as we understand their meaning. Sher seeks to use such first-order claims “as a prologue to reconstructing justificatory arguments” grounding moral desert in deeper moral principles.
Taken literally and at face value, intuitionism fails, though in a somewhat weaker sense of failure than the preceding two theories. Where utilitarian accounts of desert fail through their failure to account for salient moral phenomena, intuitionism fails by acknowledging the reality of all of the relevant moral phenomena, but failing to provide a second-order justification of them. Though moral intuitions about desert have some probative force, this force by itself is insufficient to justify the claims about desert thatan adequate account of moral desert would make. Hence a further account is required, one that goes beyond what intuitionists offer, and that the book aims to provide.