Breakdown of Moral Judgment (Ethics, April 2014)

%UHDNGRZQRI0RUDO-XGJPHQW $XWKRU V (ULF&DPSEHOO 6RXUFH(WKLFV9RO1R $SULO SS 3XEOLVKHGE\7KH8QLYHUVLW\RI&KLFDJR3UHVV 6WDEOH85/http://www.jstor.org/stable/10.1086/674845 . $FFHVVHG Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp . JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org. . The University of Chicago Press is collaborating with JSTOR to digitize, preserve and extend access to Ethics. http://www.jstor.org This content downloaded from 132.239.1.231 on Fri, 28 Mar 2014 10:55:31 AM All use subject to JSTOR Terms and Conditions ARTICLES Breakdown of Moral Judgment* Eric Campbell I argue that moral judgments function as commitment strategies that rely on a deflection of attention from our motivations and values. Revealing the hidden workings of these strategies allows me to illustrate and explain some of the widely unrecognized practical downsides of moral discourse. I recommend a departure from moral discourse in favor of paying more and better attention to our actual concerns. Important strengths of my approach over contemporary forms of moral abolitionism lie in my ability to sidestep moral error theories, my acknowledgment of the significant value of moral discourse, and thus the restricted target of my recommendation. According to Justin D’Arms and Daniel Jacobson, “The central naturalist thought in ethics is that the normative force of values must ultimately be located in motivational force, if it’s not to be convicted of systematic reification error. The puzzle, of course, is how then to make sense of the seeming lack of contingency on moral motive that moral judgments possess.”1 And in a recent paper, Richard Joyce claims that “it is something of a travesty in moral philosophy that . . . philosophers have largely contented themselves with the unexamined assumption that morality is a Good Thing without which we’d all be worse off. But intuitions on this matter, even widespread ones, will not stand in for concrete data.”2 This * Thanks to David Brink for comments on a distant ancestor of this article, referees and editors at Ethics for several helpful comments and suggestions ðincluding a change in titleÞ, and George Ainslie for his pioneering work on hyperbolic discounting and its ramifications. 1. Justin D’Arms and Daniel Jacobson, “Expressivism, Morality, and the Emotions,” Ethics 104 ð1994Þ: 739–63, 762. 2. Richard Joyce, “Metaethics and the Empirical Sciences,” Philosophical Explorations 9 ð2006Þ: 133 – 48, 143. Ethics 124 (April 2014): 447–480 © 2014 by The University of Chicago. All rights reserved. 0014-1704/2014/12403-0001$10.00 447 This content downloaded from 132.239.1.231 on Fri, 28 Mar 2014 10:55:31 AM All use subject to JSTOR Terms and Conditions 448 Ethics April 2014 essay provides a solution to the naturalist’s puzzle and uses that solution to suggest an answer to the question of the value of moral discourse.3 The solution to the puzzle is that the central moral concepts that figure in moral judgments are commitment strategies that rely on a deflection of attention from the motivations to which they are in fact related. That is, the seeming lack of contingency on moral motive is a design feature of moral judgments. Revealing the hidden workings of these commitment strategies allows me to describe and explain some of the potentially severe practical pathologies that predictably arise in moral discourse. Both because these largely invisible pathologies are widespread and often severe, and because moral discourse is in any case threatened by an awareness of its motivational function, I recommend ðat least for some of usÞ a departure from moral discourse in favor of paying much more and better attention to our actual motivations, concerns, and commitments. In offering a normative critique of moral discourse, my approach has affinities with moral abolitionism, understood as the view that we ought to abolish all talk of morality. However, I think it has significant advantages over contemporary forms of abolitionism. First, my recommendation to depart from peculiarly moral concepts is not a consequence of my commitment to naturalism combined with a belief that we cannot locate normative force in motivational force—though I am a naturalist and do believe no such location will ever succeed. Second, my critique of moral judgment does not rest on any traditional ði.e., semantic or epistemologicalÞ moral error theory. In relying on the claim that moral judgments are all false, abolitionists ðand other error theoristsÞ have allowed discussion of the value of moral discourse to get sidelined in favor of seemingly intractable semantic disagreements. I take it as a significant virtue of my critique that it neatly sidesteps all such disagreements. Finally, unlike contemporary abolitionists, my critique can and must recognize the very significant value of moral discourse and that this value will be different for different peopleðsÞ. It is highly unlikely that such a central practice would have developed and been sustained so long if it were almost uniformly destructive, as contemporary abolitionists claim.4 However, one can recognize the value of moral or religious discourse while nevertheless seeking to move beyond them, and 3. I will treat “moral discourse,” “moral judgment,” and “moral cognition” as roughly interchangeable; I vary my usage depending on context and connotation, but they may all be intersubstituted without altering anything important in my argument. 4. Ian Hinckfuss, “The Moral Society: Its Structure and Effects,” Preprint Series in Environmental Philosophy, Department of Philosophy, 16 ðAustralasian National University, 1987Þ; Richard Garner, Beyond Morality ðPhiladelphia: Temple University Press, 1994Þ, and “Abolishing Morality,” Ethical Theory and Moral Practice 10 ð2007Þ: 499–513; Joshua Greene, “The Terrible, Horrible, No Good, Very Bad Truth about Morality and What to Do This content downloaded from 132.239.1.231 on Fri, 28 Mar 2014 10:55:31 AM All use subject to JSTOR Terms and Conditions Campbell Breakdown of Moral Judgment 449 one can seek to move beyond them without thinking that everyone ought to move beyond them. In my view, recognizing these two related points is the best way to do so successfully and honestly. Here’s how I proceed. In Section I, I discuss the most prominent general form of moral error theory, which is based on the idea that moral discourse is committed to some kind of end- or motivationindependent authority, together with the claim that either no such authority exists or we have no good reason for thinking it does. I then cite a diverse collection of prominent theorists over several centuries who have felt that there is something frustratingly mysterious or opaque about our moral discourse5—something that renders it fictional, self-deceptive, hypocritical, manipulative, or otherwise involves bad faith or an obstacle to human flourishing—and that several such theorists have explicitly connected these problems with the mysterious end-independent authority of moral judgments. In Section II, I provide an explanation for this mysterious sense of authority, one that is centered on the motivational importance of subjective commitments and their reliance on the deflection of attention from our own motives. Paradigmatic moral judgments involve moral sentiments and moral concepts. I argue that we have very good reasons to think that these sentiments and concepts developed as commitment strategies to combat our deep tendency for preference switching over time, or hyperbolic discounting of the future. Crucially, it is functionally central to the commitment strategies manifested in our moral concepts that they deflect attention from their status as commitment strategies as well as the nature of the motivations or ends they express or promote. In Section III, I address the question of the value of moral discourse by critiquing Joyce’s normative moral fictionalism, which argues that we ought to pretend to believe our moral judgments, precisely due to their value as subjective commitment devices. I contend that Joyce’s arguments fail primarily because they ignore the potential for severe pathologies that accompany such commitment strategies, pathologies that are greatly exacerbated by the deflection of attention from our motives. This deflection of attention is rightly but not explicitly at the functional heart of Joyce’s normative fictionalism, which comprises an attempt to retain the motivational benefits of the subjective commitment strategies that we experience as a sense of motivation-transcendent authority. It’s crucial to understand that my critique is not of commitments in general, or even of subjective commitment strategies in general. Rather, the critique about It” ðPhD diss., Department of Philosophy, Princeton University, 2002Þ; Joel Marks, Ethics without Morals ðNew York: Routledge, 2012Þ. 5. By “theorist” I mean to include all those who investigate the nature of moral judgment, in or outside philosophy departments. This content downloaded from 132.239.1.231 on Fri, 28 Mar 2014 10:55:31 AM All use subject to JSTOR Terms and Conditions 450 Ethics April 2014 is focused on the aspect of moral judgments that requires a sustained lack of awareness of the nature of our own commitments and the kinds of concerns they serve to promote. Nothing in my critique of moral judgment entails that moral judgments are systematically false or involve any reification error. In the final section I briefly describe some important advantages to critiquing moral discourse without going through any traditional error theory and explain how my critique succeeds in doing so. I then conclude that the tendency of moral discourse to deflect attention and inquiry from our own values and concerns gives at least some of us excellent reason to consider moving beyond it, as part of a broader attempt to gain normatively important forms of self-knowledge. I. MOR AL JUDGM ENT, MOR A L ER ROR THEORY, AND PRAC T ICAL OOMPH By “moral judgment” I mean to include any judgment about how it is good or right to act or feel that—in a sense it has proven difficult to cash out—purports to have a kind of normative authority or privilege that transcends any person’s ðor institution’sÞ contingent ends, standards, attitudes, values, drives, interests, or ðotherÞ motivations.6 This inclusive notion of moral judgment no doubt includes at least some judgments of prudence or well-being that we would not ordinarily regard as moral. I treat moral judgment this inclusively because it is this central feature of moral judgments I wish to explain and critique in this essay.7 It is also the feature most targeted by moral error theories, and so those theories as well as my critiques extend to judgments of prudence or wellbeing insofar as they share this feature. I will generally speak of moral judgment or discourse rather than “morality” for several reasons. One is that I think “morality” is a far more ambiguous, contested, and loaded term than “moral judgment,” though the latter has its share of controversy as well. Another, related reason is that I mean to exclude from my discussion any of the practical commitments associated with folk morality or normative moral theories, to the extent that this is possible. The most odious and ignorant, along with the most noble and wise judgments, are moral judgments just in case ðor perhaps to the extent thatÞ they fit the description above. Therefore, when critiquing moral judgment, I wish to remain neutral on the 6. For simplicity’s sake I will often just use “ends” or “motivations” to stand for this disjunction. 7. I employ the term “moral judgment” to capture all such judgments because I think that typical judgments of prudence and well-being are far more sensitive to considerations of a person’s own desires, values, and concerns than what we would ordinarily consider moral judgments. This content downloaded from 132.239.1.231 on Fri, 28 Mar 2014 10:55:31 AM All use subject to JSTOR Terms and Conditions Campbell Breakdown of Moral Judgment 451 question whether and to what extent any substantive practical commitments associated with “morality” could or should survive a greater awareness of the motivations underlying our moral judgments. Traditional moral error theorists claim that moral judgments are systematically incoherent, false, untrue, or epistemically unjustified.8 By far the greatest proportion of moral error theorists target the ostensible “objectivity,” or end-independent practical authority of moral judgments.9 The arguments are generally variations on the following theme.10 1. Moral concepts presuppose some kind of end- or motivationtranscendent practical authority. ðConceptual claimÞ 2. No such authority exists, or we are not justified in thinking it does. ðOntological or epistemological claimÞ 3. No moral judgments are true, or we are not justified in believing any of them. ðDifferent versions of error-theoretic conclusionÞ. I have left the first premise in the argument vague. That is because it has proven quite difficult to analyze or otherwise elucidate this ostensibly problematic conceptual presupposition of moral judgments, and so different theorists have made different attempts to cash it out. I’ll follow Joyce in calling it practical oomph.11 Practical oomph is a usefully and deliberately vague term meant to point to the apparent motivationtranscendent normative force of moral judgments. The term implies no philosophical theory about the socio-anthropological phenomenon it points to. I employ this term partly because I want to be clear that in explaining the practical oomph of moral judgments in the next section, I make no commitment as to whether or not this oomph involves a semantic commitment in ordinary moral discourse. The most developed and plausible attempt to cash out such a semantic commitment is to hold that moral judgments semantically presuppose inescapable and authoritative reasons for action that are in8. I will typically only say “false” to stand for this disjunction. 9. John Mackie, Ethics: Inventing Right and Wrong ðNew York: Penguin, 1977Þ; Michael Ruse, Taking Darwin Seriously ðOxford: Blackwell, 1986Þ, and Evolutionary Naturalism ðLondon: Routledge, 1995Þ; Richard Alexander, The Biolog y of Moral Systems ðNew York: Aldine De Gruyter, 1987Þ; Hinckfuss, The Moral Society; Garner, Be yond Morality ; Robert Wright, The Moral Animal ðLondon: Little, Brown, 1994Þ; Richard Joyce, The Myth of Morality ðCambridge: Cambridge University Press, 2001Þ, and The Evolution of Morality ðCambridge, MA: MIT Press, 2006Þ; Greene, “The Terrible Truth”; Jonas Olson, “In Defence of Moral Error Theory,” in New Waves in Metaethics, ed. M. Brady ðBasingstoke: Palgrave Macmillan, 2010Þ, 62–84; Marks, Ethics without Morals. 10. Other moral error theories could target different apparent ostensible presuppositions. Further, this argumentative theme is meant to only roughly characterize the several error theories that do target moral judgments’ apparent commitment to end-independent practical authority. 11. Joyce, The Evolution of Morality. This content downloaded from 132.239.1.231 on Fri, 28 Mar 2014 10:55:31 AM All use subject to JSTOR Terms and Conditions 452 Ethics April 2014 dependent of anyone’s contingent desires or institutional affiliations.12 The idea that moral judgments presuppose such reasons is called conceptual rationalism. The denial that there are any such reasons is called substantive antirationalism. Here I want to use that latter term a bit more loosely, to represent the denial that there can be any satisfactory philosophical defense of practical oomph. That is, a substantive antirationalist in my sense is someone who thinks that our sense of practical oomph must have some other explanation than that we are somehow tapping into a source of practical authority that is metaphysically independent of our motivations or ends.13 Upon coming to the error-theoretic conclusion, an error theorist needn’t think we ought to abolish moral discourse. One might propose some suitable revision of the discourse or even that we only pretend to believe our moral judgments, so as to avoid false belief or utterances.14 Indeed, one might even propose that simply making moral claims one believes to be false is justified by the overwhelming value of doing so. So error theories are all consistent with recommending the continued employment of moral discourse, and several have done so. Nevertheless, some error theorists have argued that in addition to generating untrue or epistemically unwarranted claims, engaging in moral discourse is a normative or practical mistake and that we should abandon or abolish it.15 But what could be normatively problematic about moral discourse? There have been a number of answers given to this question, but most if not all seem to involve the idea that moral discourse promotes ignorance or self-deception about one’s motives, or some form of bad faith. Nietzsche saw in moral ðand religiousÞ discourse a deep drive to selfdeception, especially of one’s own motives. Bernard Williams complained of a lack of transparency in moral discourse and argued that such a lack was an obstacle to normatively important forms of self12. This was Joyce’s view in The Myth of Morality, following Mackie’s Ethics. Joyce has now moved to the position that the semantic presupposition is inchoate. Richard Joyce, “The Error in ‘The Error in the Error Theory,’” Australasian Journal of Philosophy 89 ð2011Þ: 519–34. 13. I can only argue for substantive antirationalism here to the extent that the explanation for practical oomph I offer in the next section is meant to be more plausible than the idea that the sense of oomph is the result of our tapping into metaphysically endindependent reasons ðwhich I take to be implausible on philosophical grounds I don’t discuss hereÞ. However, my committed substantive rationalist readers can nevertheless profit from this article by, among other things, seeing how much plausibly hangs on the issue of substantive rationalism. For though I think my normative case against moral discourse can be made even on substantive rationalist assumptions, I think it is much more powerful and straightforward if we assume substantive antirationalism. 14. See especially Joyce, The Myth of Morality. 15. Hinckfuss, The Moral Society; Garner, Beyond Morality, and “Abolishing Morality”; Greene, “The Terrible Truth”; Marks, Ethics without Morals. This content downloaded from 132.239.1.231 on Fri, 28 Mar 2014 10:55:31 AM All use subject to JSTOR Terms and Conditions Campbell Breakdown of Moral Judgment 453 16 knowledge. R. D. Alexander thought that moral “righteousness provides . . . a rationale for sinking deeper into self-deception about our motives and for justifying acts that could not otherwise be justified.”17 Some Marxists argue that moral judgments are or were to a large extent the result of an ideology meant to promote the interests of the power elite at the expense of the powerless.18 Nietzsche, by contrast, thought that morality was a tool of the weaker to subdue the stronger ðto put his idea very crudelyÞ. It might be objected that some of these critiques, especially the last two, are of the content of folk ðor bourgeois ChristianÞ morality, not of moral judgment itself. And it is quite true that several ðespecially ContinentalÞ philosophers criticize the substance of folk morality as well as its form, none more than Nietzsche. However, Nietzsche’s attack was at least as much directed at the notion of intrinsic ðnonrelationalÞ value that he believed promoted or allowed for ðself-Þdeception about the nature of the motivations that lay behind people’s moral judgments. Also, though Marx and Marxists often critique the substance of morality, if they are right about the function of substantive morality, it would obviously not do for its function to be generally known. Indeed, we should expect any fundamental critique of the content of moral norms to be joined with criticisms of the form of moral judgment, and for these latter criticisms to have something to do with its practical oomph. For if the normative commitments of some “morality” were, for example, systematically detrimental to its adherents, it would plausibly be crucial for their moral judgments to be experienced as somehow legitimately demanding actions quite independent of their desires or interests. Indeed, Hinckfuss’s primary normative critique of “morality” is that it promotes elitism and authoritarianism by means of the “absolutist” or “objective” meaning of moral judgments—terms which are attempts to articulate and conceptualize what we are calling practical oomph. At the end of his book, Hinckfuss explicitly wishes to replace the “guilt, denigration and self-deception” of “the moral society” with “understanding, both of oneself and others.”19 Garner also associates moral concepts with “self-deception, superstition ½and$ duplicity.”20 As we will see, even 16. Bernard Williams, Ethics and the Limits of Philosophy ðLondon: Fontana, 1985Þ. Williams spoke of “ethical discourse” in this connection, but he was clearly talking about moral discourse in my sense. The vast majority of modern ethical discourse, especially in the academy, is moral in my sense. 17. Alexander, The Biology of Moral Systems, 123. 18. See Raymond Geuss, The Idea of a Critical Theory ðCambridge: Cambridge University Press, 1981Þ. 19. Hinckfuss, The Moral Society, 66. 20. Introduction to the revised version of Beyond Morality, accessed from Garner’s website on 11/15/2012. This content downloaded from 132.239.1.231 on Fri, 28 Mar 2014 10:55:31 AM All use subject to JSTOR Terms and Conditions 454 Ethics April 2014 Joyce, who argues in favor of employing moral judgments in spite of his error theory, worries that adherents of the “moral institution” are selfdeluded and that this self-delusion is a direct result of the practical oomph of moral judgments.21 II. MOR AL JUDGMENT, COMMITMENT, AND STRATEGIC IGNOR ANCE My argument that moral judgments function as commitment devices will be abbreviated, as the most common conceptions of the ðprimaryÞ biological and/or cultural functionðsÞ of moral discourse are individual or collective motivation and coordination.22 Moreover, Frank, Blackburn, Gibbard, Richerson, Boyd, and Joyce all hold that moral sentiments and/or concepts serveðdÞ a specifically committing function.23 Further, most or all these theorists think that moral judgment serves a psychological, in addition to an evolutionary, committing function. I think these commitment devices work at the levels of culturally and/or biologically evolved mechanisms and also at the level of individual psychology. But call the general thesis that a central function of moral discourse is to commit its users to ways of thinking, feeling, and/or acting the Commitment Device Thesis, or Commitment. Commitments are fundamentally reductions in flexibility of choice.24 They are essential to bargaining, negotiation, deterrence, and contractual relationships. Interpersonally, they operate by altering others’ expectations of one’s future behavior by rearranging one’s incentives in publicly 21. Joyce ðMyth of Morality, ixÞ cites Camus, Wittgenstein, Russell, Nietzsche, Hume, Mandeville, Hobbes, Antiphon, and the characters Callicles and Thrasymachus as all thinking of the practical oomph of moral discourse as fictional. 22. For variations on this theme, see Robert Axelrod, “An Evolutionary Approach to Norms,” American Political Science Review 80 ð1986Þ: 1095–1111; Alexander, The Biology of Moral Systems; Richard Dawkins, The Selfish Gene ðOxford: Oxford University Press, 1976Þ; Ruse, Taking Darwin Seriously, and Evolutionary Naturalism; Simon Blackburn, “Errors and the Phenomenology of Value,” in Morality and Objectivity: A Tribute to John Mackie, ed. Ted Honderich ðLondon: Routledge & Kegan Paul, 1985Þ, 1–22; Robert Frank, Passions within Reason ðNew York: Norton, 1988Þ; Samuel Scheffler, Human Morality ðOxford: Oxford University Press, 1992Þ; Wright, The Moral Animal; Allan Gibbard, Wise Choices, Apt Feelings ðCambridge, MA: Harvard University Press, 1990Þ, and Thinking How to Live ðCambridge, MA: Harvard University Press, 2003Þ; Joyce, The Myth of Morality, and The Evolution of Morality; Peter J. Richerson and Robert Boyd, Not by Genes Alone ðChicago: University of Chicago Press, 2005Þ. 23. Not all these theorists hold the commitment thesis explicitly; sometimes the thesis is implicit, e.g., Gibbard, Wise Choices, Apt Feelings, 166 and 186Þ; also Blackburn, “Errors,” 11. Also see my description of Richerson and Boyd’s view below. 24. Thomas Schelling, The Strategy of Conflict ðCambridge, MA: Harvard University Press, 1960Þ. This content downloaded from 132.239.1.231 on Fri, 28 Mar 2014 10:55:31 AM All use subject to JSTOR Terms and Conditions Campbell Breakdown of Moral Judgment 455 perceivable ways. But commitments are also very important intrapersonally, and they operate both by rearranging incentives and self-signaling. Some commitments rearrange one’s incentives in straightforward ways, such as when one signs a contract to give money to a despised organization upon being caught using an addictive substance. As we’ll see, others very plausibly take the form of the creation and/or maintenance of personal rules and motivated beliefs by means of which one alters expectations of one’s own future actions. The more original and normatively significant claim presented in this article is that moral judgments are generally threatened by an awareness of their status as commitment devices, as well as by awareness of one’s own ðand often others’Þ motivations in the moral realm. For this reason we should expect successful employers of the discourse to be generally unaware of—and resistant to becoming aware of—their own ðand often others’Þ motivations in making moral judgments, as well as the motivational function of those judgments themselves. Indeed, I think that an important part of theories attempting to ground the authority of morality—as well as features of the discourse itself—are profitably understood as serving the function of deflecting attention from many of our motives and commitments in the moral realm. But call the general thesis that a functionally important component of moral discourse is the deflection of attention from our motivations the Deflectionof-Attention Thesis, or Deflection. Commitment and Deflection may be seen as theoretical elaborations and explanations of the practical oomph of moral discourse. The argument below begins with the importance of intrapersonal, broadly prudential commitments. Our focus will be on commitments in the form of personal rules and principles as this most clearly illustrates the committing value of rules and principles, especially when one lacks an awareness of their function. But moral judgment centrally includes both the moral emotions and concepts which come together in paradigmatic moral judgments.25 I’ll argue that we have good reason to think that these emotions and concepts evolved as commitment devices at one or more explanatory levels. My account will also explain the motivational value of moral principles and the concept of intrinsic value. I’ll try to show that it’s quite plausible to think of the functional heart of our central moral concepts as subjective commitment strategies, which by their nature motivate a lack of awareness of their general functioning as well as of the motivations to which they are in fact related. 25. Though I think of emotions as occurrent mental states and ðmoralÞ sentiments as dispositions to have ðmoralÞ emotions, I use the terms here roughly interchangeably since the distinction isn’t relevant to my argument. This content downloaded from 132.239.1.231 on Fri, 28 Mar 2014 10:55:31 AM All use subject to JSTOR Terms and Conditions 456 Ethics April 2014 Commitments Help Avert Deep Tendency toward Diachronic Inconsistency People ðand animalsÞ standardly display patterns of preference such that they temporarily prefer smaller, sooner ðSSÞ rewards to larger, later ðLLÞ rewards when the SS rewards are close enough in time. That is, there is a deep tendency for people to change many of their preferences as a function of time, then change them again in retrospect. Here I’ll call any such apparent discounting of the future hyperbolic discounting.26 Attempts at case-by-case optimization of one’s preferences are highly susceptible to the negative effects of hyperbolic discounting, which centrally involve diachronic inconsistency and therefore frustration of long-term goals. George Ainslie has done much to illuminate both the nature and importance of the strategies we ðoften unwittinglyÞ use to secure our longer-term interests against competing shorter-term interests.27 For example, hyperbolic discounting suggests a rationale for why choosing according to principles has motivational benefits. If your current decision is treated as a test case, that is, as providing a prediction about what you are likely to do in similar situations in the future, then the value of those future decisions becomes added to your present decision. There is experimental evidence, in animals and humans, that when rewards are added in this way, hyperbolic curves begin to resemble exponential ones in that there is no preference switching. As illustration, imagine thinking of whether to smoke this cigarette as the question whether to smoke for the rest of your life. If you can conceptualize your decision this way, then even at the moment when the cigarette becomes most attractive and available, your preference might well be not to smoke this cigarette, since even at that moment you probably prefer never smoking to always smoking. However, if you consider each cigarette in isolation, it will always be possible to both have this particular cigarette and resolve to quit smoking for good—after this cigarette. The ubiquitous availability of that logic is one of the things that makes quitting so hard, and the strategic heart of making decisions according to rules or principles is precisely to remove the psychological availability of that logic.28 26. It’s called hyperbolic discounting because the curves drawn to represent such preferences have a highly bowed shape, unlike exponential curves, which cannot represent changes in preference. I will use the phrases “hyperbolic discounting” and “diachronic inconsistency of preference” interchangeably. In doing do, I mean to bypass disagreements about whether such discounting is “quasi-hyperbolic” or not, as this distinction isn’t relevant here. 27. See especially George Ainslie, Picoeconomics ðCambridge: Cambridge University Press, 1992Þ, and Breakdown of Will ðCambridge: Cambridge University Press, 2001Þ. 28. For an Ainslie-inspired discussion of how to use principles to perform committing functions by having specific actions symbolize classes of actions, see chap. 1 of Robert Nozick, The Nature of Rationality ðPrinceton, NJ: Princeton University Press, 1993Þ. This content downloaded from 132.239.1.231 on Fri, 28 Mar 2014 10:55:31 AM All use subject to JSTOR Terms and Conditions Campbell Breakdown of Moral Judgment 457 It’s important to see that this “test-case” strategy is rarely if ever conscious; indeed, in many contexts these commitment devices can be threatened by an awareness of their motivational function. If we feel that some rule ðor belief Þ is ultimately in our service, then we are more likely to feel we ought to violate it if it seems rational to do so—and our powers of rationalization are often very good at leading us to judge some action best that both in foresight and hindsight we judge to be a ðsometimes very badÞ mistake. One of the ways that Ainslie thinks these commitment devices avoid detection as such is by manifesting as beliefs about being watched or helped by gods ðor saintsÞ, as well as beliefs in objects or states of affairs having intrinsic value.29 On his model of the will, such beliefs are best explained due to their serving motivational, committing functions. By contrast to commitments one might hold for self-consciously prudential or more broadly instrumental reasons, these commitments-as-beliefs can add a layer of motivational stability by deflecting attention from that instrumental function. Such deflection of attention can be very motivationally important in the context of strong contrary inclinations—the kind that moral injunctions are centrally concerned to govern—and so we are more likely to find such deflection of attention in these contexts. Evolution of Moral Sentiments as Commitment Strategies At the beginning of this section I cited a large number of theorists who think that moral discourse evolved as motivational, or even specifically commitment strategies. In this subsection, I’ll argue that anyone who thinks that moral sentiments are adaptations should think that they evolved specifically as commitment strategies geared to combat hyperbolic discounting, and that anyone who believes that people are hyperbolic discounters should also believe that the moral sentiments are commitment strategies geared to combat hyperbolic discounting. I’ll point out that this holds for the moral sentiments connected with moral condemnation and punishment as well as those connected with altruism and moral conscience. This is mostly to prepare the ground for my argument in the next subsection that moral concepts are likely culturally evolved commitment strategies that rely on a deflection of attention from our motivations. Robert Frank explains why anyone arguing that moral sentiments are adaptations at all should think of them as evolutionary commitments strategies.30 His argument for the evolution of moral sentiments is grounded in the claim that it must have been crucial to be able to cred29. Ainslie, Breakdown of Will, 107–12. By “intrinsic value” I mean nonrelational, not final or ultimate value. 30. Frank, Passions within Reason. This content downloaded from 132.239.1.231 on Fri, 28 Mar 2014 10:55:31 AM All use subject to JSTOR Terms and Conditions 458 Ethics April 2014 ibly signal to others in one’s social environment that one is committed to act in certain ways. To be able to reliably communicate that one has certain commitments is a large social advantage because if one’s potential allies, partners, mates, or enemies have no strong reason for thinking that their sacrifices will be repaid, that they and their children won’t be abandoned when convenient, or that their thefts or other assaults won’t be aggressively deterred or avenged, then one will be at a severe social disadvantage relative to others who can reliably signal these various commitments. One important way to effectively signal to others that one has the relevant sort of commitments is to have an appropriate reputation. However, precisely because we would expect reputations to be obviously important, one might think that basic prudence would have effectively motivated people to avoid cheating ðbroadly construedÞ, even when the likelihood of getting caught was fairly low. A general capacity for prudence would then undermine any need to develop novel sentiments to help secure one’s reputation. But this reasoning only has force if one is unaware that future rewards are heavily discounted relative to the present. Agents who reason in a purely prudential or instrumental way can be expected to regularly get caught cheating, lying, and/or stealing, be uncooperative or unhelpful when doing so would be onerous, and fail to avenge or deter thefts or other violations when doing so would be dangerous. And such people would be quite likely to develop corresponding reputations as a result of the ubiquitous human proclivity for gossip. So, if you think the socio-moral sentiments are adaptations geared toward motivational functions, as most theorists do, you should specifically think they are evolutionary commitment strategies geared toward combatting the downsides of hyperbolic discounting. Now, it is also the case that if you start out believing that people have a deep tendency toward diachronic inconsistency of preference, then you should believe that the moral sentiments are likely commitment strategies geared to combating hyperbolic discounting. For we can predict that hyperbolic discounters will fare poorly in social environments for just the sorts of reasons explained above ðand there are more to comeÞ. We also know that there are two broad strategies that are most effective in mitigating the downsides of hyperbolic discounting—rearranging incentive structures and reducing the salience or availability of the SS rewards. Knowing that people are hyperbolic discounters living in large social groups allows us to predict that they will likely have developed some version of at least one of these two broad strategies. And just as Frank’s model rightly emphasizes, the socio-moral sentiments seem to represent a clear case of rearrangement of incentive structure. Their motivational power can manifest at the very contemplation of the relevant transgressions, allowing them to compete with very highly salient and This content downloaded from 132.239.1.231 on Fri, 28 Mar 2014 10:55:31 AM All use subject to JSTOR Terms and Conditions Campbell Breakdown of Moral Judgment 459 powerful motivations such as those connected to food, sex, avoidance of bodily harm, ðperception of Þ status or power, and so forth. One of the central weaknesses of Frank’s model is that it leaves some central moral emotions and practices totally unexplained. Frank’s argument, like most evolutionary accounts of the development of the moral emotions, focuses on biological evolution and on the sentiments associated with moral conscience or altruism.31 Guilt and the emotions associated with benevolence or sympathy take center stage in explanations of the development of moral conscience. But without significant further assumptions, these explanations all leave out any explanation of the development of emotions associated with moral condemnation and punishment. By contrast, beginning with an attempt to explain moral condemnation and punishment is potentially more fruitful, since in social contexts where punishers exist, we can view the development of moral conscience as, at least in part, a strategy to avoid condemnation and punishment.32 At the very least, the focus on moral conscience has neglected the plausibly central role of moral condemnation and punishment. Call norms about punishing those who defect from norms metanorms.33 Metanorms are an important mechanism for establishing and maintaining first-order norms because even in an environment where temptations to defect are low ðperhaps due to emotions such as guilt and shameÞ, if nobody is sufficiently motivated to punish the small amount of defectors, a norm can still collapse.34 Therefore third-party punishment— punishment of norm-transgressors where the punisher is not the victim or family member of the victim of the transgression—was plausibly crucial to the development of large-scale human societies.35 Third-party punishment can be restricted to social exclusion, or it can extend to what 31. Alexander, The Biology of Moral Systems; Franz de Waal, Good Natured ðCambridge, MA: Harvard University Press, 1996Þ; Matt Ridley, The Origins of Virtue ðNew York: Viking, 1997Þ; Wright, The Moral Animal; Marc Hauser, Moral Minds ðNew York: Harper Collins, 2006Þ; Joyce, The Evolution of Morality; Jonathan Haidt, “The New Synthesis in Moral Psychology,” Science 316 ð2007Þ: 998–1002; Joshua Greene, “The Secret Joke of Kant’s Soul,” in Moral Psychology, vol. 3, The Neuroscience of Morality: Emotion, Brain Disorders, and Development, ed. Walter Sinnott-Armstrong ðCambridge, MA: MIT Press, 2008Þ, 35–80; Gerd Gigerenzer, “Moral Satisficing: Rethinking Moral Behavior as Bounded Rationality,” Topics in Cognitive Science 2 ð2010Þ: 528–54. 32. Peter DeScioli and Robert Kurzban, “Mysteries of Morality,” Cognition 112 ð2009Þ: 281–99. 33. Axelrod, “An Evolutionary Approach to Norms.” 34. Ibid., 1104. 35. Ibid.; Herbert Gintis, “Strong Reciprocity and Human Sociality,” Journal of Theoretical Biology 206 ð2000Þ: 169–79; Samuel Bowles and Herbert Gintis, “The Evolution of Strong Reciprocity: Cooperation in Heterogeneous Populations,” Theoretical Population Biology 65 ð2004Þ: 17–28. This content downloaded from 132.239.1.231 on Fri, 28 Mar 2014 10:55:31 AM All use subject to JSTOR Terms and Conditions 460 Ethics April 2014 Robert Trivers calls “moralistic punishment,”36 which I’ll understand to include being actively attacked and killed by one’s “erstwhile compatriots.”37 Such severe punishments are central to why moralistic punishment can provide an additionally effective means of stabilizing large-scale cooperation over and above “mere” exclusion ðwhich can be tantamout to death in some social circumstancesÞ. For even if moralistic punishers are relatively rare, the severity of the penalties they are disposed to mete out can yet make for sufficiently effective inducements to avoid punishable activities.38 At the time of Axelrod’s original writing, costly third-party punishment was only theoretical, in the sense that there had not been experiments done to test for its existence. But now it is clear that we do engage in third-party punishment both in the laboratory39 and in the field,40 even where doing so is clearly costly to the punisher, and that doing so significantly raises overall cooperation levels. There is also good evidence that this general willingness to undertake costly punishment is culturally universal, though there is significant variability between cultures with respect to how severely and at what cost third parties are willing to punish what degree of uncooperativeness.41 Robert Boyd and Peter Richerson have developed a model that is an improvement over Frank’s primarily in recognizing the evolutionary importance of the sentiments associated with moral condemnation and punishment, as well as the likely importance of cultural evolution.42 On their view, cooperative groups in conflict with one another created an arms race in which cultural variants ðincluding socio-moral sentiments 36. Robert Trivers, “The Evolution of Reciprocal Altruism,” Quarterly Review of Biology 46 ð1971Þ: 35–57. 37. Richerson and Boyd, Not by Genes Alone, 200. 38. Ibid. 39. Ernst Fehr and Simon Gachter, “Altruistic Punishment in Humans,” Nature 417 ð2002Þ: 137– 40; Elinor Ostrom, James Walker, and Roy Gardner, “Covenants With and Without a Sword: Self-Governance Is Possible,” American Political Science Review 86 ð1992Þ: 404 –17; Toshio Yamagishi, “The Provision of a Sanctioning System as a Public Good,” Journal of Personality and Social Psycholog y 51 ð1986Þ: 110 –16; Ernst Fehr and Urs Fischbacher, “Third-Party Punishment and Social Norms,” Evolution and Human Behavior 25 ð2004Þ: 63–87. 40. Elliott Sober and David Wilson, Unto Others ðCambridge, MA: Harvard University Press, 1999Þ; Abigail Barr, “Social Dilemmas and Shame-Based Sanctions: Experimental Results from Rural Zimbabwe,” The Centre for the Study of African Economies Working Paper Series ðOxford, 2001Þ; Michael Price, “Punitive Sentiment among the Shuar and in Industrialized Societies: Cross-Cultural Similarities,” Evolution and Human Behavior 26 ð2005Þ: 279–87. 41. Joe Henrich et al., “Costly Punishment across Human Societies,” Science 312 ð2006Þ: 1767–70. 42. Robert Boyd, Culture and the Evolutionary Process ðChicago: University of Chicago Press, 1985Þ; Richerson and Boyd, Not by Genes Alone. This content downloaded from 132.239.1.231 on Fri, 28 Mar 2014 10:55:31 AM All use subject to JSTOR Terms and Conditions Campbell Breakdown of Moral Judgment 461 as well as different kinds of ½meta-$ norms and punishment strategiesÞ that could generate ever-greater in-group cooperation would tend to proliferate.43 The environments created by such groups selectively favored those with sentiments geared to life in such groups, including psychological structures prepared to internalize the norms of one’s groupðsÞ. Those without emotions such as shame, guilt, and perhaps ðmoralÞ anger were on average more likely to transgress against social ðmeta-Þ norms and be subject to a variety of punishments as a result, reducing their fitness.44 Although I find this general account compelling, my point here is not to sign on to any particular account of the evolution of moral sentiments. I have three goals in briefly summarizing the above research. First, I want to push back against the common tendency to focus on moral conscience to the exclusion of moral condemnation in discussions of the value of the commitment devices associated with moral discourse. I think this common practice is both a cause and effect of the almost universally unquestioned idea that “morality” is a Good Thing. Second, I want to point out that Richerson and Boyd’s model is every bit as reliant on hyperbolic discounting as Frank’s, and that they recognize this.45 Therefore, this quite different model of the evolution of the moral sentiments—including those associated with moral condemnation and punishment—must also think of them as evolutionary commitment strategies. Third, I want to introduce the idea of cultural variants and point out that cultural evolution might well have had a large role to play in the development of our moral judgments—if not of the sentiments themselves, then of our moral concepts. Any account of moral judgments that sees them as commitment devices plausibly needs a commitment-based story about the nature of moral concepts. For even if we had some entirely satisfactory evolutionary account of moral sentiments, it would fail to provide us with an account of the evolution of moral judgment. We would have only an explanation for our suite of motivationally loaded perceptions, our desires and aversions in reponse to, or perhaps as part of, perceptions of happenings in the social world. But moral judgments seem to be something more than desires and aversions. Were they not, we would not 43. Cultural variants are roughly analogous to genes. They can range in complexity from individual phonological rules to entire grammars, from entire religious or moral systems to the specific morality of contraception ðRicherson and Boyd, Not by Genes Alone, 90–91Þ. 44. Summary from Richerson and Boyd, Not by Genes Alone, 214–15. 45. Ibid. Though Richerson and Boyd recognize the need to appeal to hyperbolic discounting to make their account plausible, they do not cite Ainslie or Frank, the latter of whom had already developed this rationale for why novel sentiments are required, and in much more detail. This content downloaded from 132.239.1.231 on Fri, 28 Mar 2014 10:55:31 AM All use subject to JSTOR Terms and Conditions 462 Ethics April 2014 have the puzzle that opens this article—the seeming lack of contingency on motive that moral judgments possess. For it is a commonplace that the sense of acting from a motive of fulfilling a moral obligation is not experienced as simply doing or avoiding what one most strongly wants to do or avoid at the time, which is what one should expect from an account that restricts itself to the evolution of moral sentiments.46 Moral Concepts as Commitment Devices First, a word about the viability of cultural group selection in general. Evolution needn’t work only on genetic material. Wherever there are heritable variations in traits that differentially affect reproduction rates, there will be selection.47 In order for group selection to work, there must be sufficient similiarity within groups and sufficient variability among groups. While it is very difficult to achieve these conditions in genetically based groups, it is not in cultural groups.48 Cultural evolutionary theory offers a way of studying cultural change that applies Darwin’s idea of inherited biological features as “small consequences of one general law,” which he summarized as “multiply, vary, let the strongest live and weakest die.” There is now general agreement among many researchers that we can take talk of cultural evolution literally as describing consequences of the same general law, though operating in the realm of culture rather than biology.49 Once we take seriously the possibility that our moral concepts are products of cultural evolution, we shall see that the same argument I ran in the last section with respect to the evolution of moral sentiments holds with respect to the evolution of moral concepts. That is, recognition of our deep tendency for preference reversal and akratic rational46. Joyce, The Evolution of Morality, makes a similar critique of Frank’s claim in Passions within Reason that the latter had given an account of the evolution of moral conscience. 47. Richard Lewontin, “The Units of Selection,” Annual Review of Ecology and Systematics 1 ð1970Þ: 1–18. 48. Indeed, punishment strategies play an important role in maintaining intragroup similarity. 49. Kate Distin, Cultural Evolution ðNew York: Cambridge University Press, 2011Þ, 4. In support of this claim she cites Marion Blute, “The Role of Memes in Cultural Evolution: Memes If Necessary, but Not Necessarily Memes” ðOpen Semiotics Resource Centre, Virtual Symposium, Imitation, Memory and Cultural Change: Probing the Meme Hypothesis, May 2007Þ. Terrence W. Deacon, “Editorial: Memes as Signs,” Semiotic Review of Books 10 ð1999Þ: 1–3; Daniel Dennett, “From Typo to Thinko: When Evolution Graduated to Semantic Norms,” in Evolution and Culture, ed. Stephen C. Levinson and Pierre Jaisson ðCambridge, MA: MIT Press, 2006Þ, 133–45; Joseph Henrich, Robert Boyd, and Peter J. Richerson, “Five Misunderstandings about Cultural Evolution,” Human Nature 19 ð2008Þ: 119–37; Paul Marsden, “Memetics and Social Contagion: Two Sides of the Same Coin?” Journal of Memetics— Evolutionary Models of Information Transmission 2 ð1998Þ, http://cfpm.org/jom-emit/1998 /vol2/marsden _ p.html; Alex Mesoudi, “Biological and Cultural Evolution,” Biological Theory 2 ð2007Þ: 119–23. This content downloaded from 132.239.1.231 on Fri, 28 Mar 2014 10:55:31 AM All use subject to JSTOR Terms and Conditions Campbell Breakdown of Moral Judgment 463 ization generates a prediction that we will find socio-moral concepts that reflect one or both of our two general strategies—rearrangement of incentive structure and/or reduction of the salience or availability of specious ðor dangerousÞ rewards. Again, knowing that people are hyperbolic discounters living in large social groups allows us to predict that they will likely have developed some version of at least one of these two broad strategies. And since it is plausible that we evolved in the context of an arms race toward ever-greater in-group cooperation, we should expect to find ourselves heavily equipped to behave in ways that were ðand may still beÞ geared toward the survival and growth of large groups of unrelated members. Further, for the last couple hundred thousand years or so, our ancestors both have been hyperbolic discounters and had linguistic resources to employ in the service of “rationalizing” in favor of ancient and extremely powerful motivations.50 We should therefore expect successful groups to have developed cultural variants which employ these very conceptual/linguistic resources to aid in the commitment of their members to the social norms. Further, in the context of third-party and moralistic punishers, we should expect individuals to be ðbiologicallyÞ designed to readily take up and employ concepts that help them combat the dangers of their hyperbolically discounting motivational psychology. We have seen that case-by-case reasoning opens up a dangerous space for rationalization. Therefore Commitment strongly predicts that in moral discourse we will find concepts, and generally ways of talking and thinking, bearing the marks of an attempt to reduce the scope for rationalization. Since these reductions of scope are fundamentally subjective—that is, the concepts of moral discourse do not by themselves alter one’s material incentives as contracts do—we should also expect typical users of the discourse to ðstronglyÞ resist awareness of their motivational function, and more generally to ðstronglyÞ resist the notion that the value or legitimacy of their moral judgments must be understood in relation to their own contingent motivations. I think these predictions are very clearly and robustly supported. Further, it is mysterious why moral discourse would have these features unless something like Commitment and Deflection were correct. Let us see why moral concepts look just as we would expect them to if they were fundamentally in the commitment business. Moral concepts bear the marks of both broad sorts of commitment strategy. First, moral language straightforwardly suggests a style of thinking in which options are being closed off, that is, their availability is being subjectively removed. “I can’t keep the money; it’s wrong” is a likely thing to say when 50. I place “rationalizing” in quotes because I don’t mean to imply that those motivations are never genuine sources of reasons. This content downloaded from 132.239.1.231 on Fri, 28 Mar 2014 10:55:31 AM All use subject to JSTOR Terms and Conditions 464 Ethics April 2014 making a self-consciously moral judgment. The idea that one must ðnotÞ do something, or that one has a moral duty ðnotÞ to, or that something is impermissible, morally required, or even unthinkable, clearly suggests not a weighing of options but rather taking some of them off the table. This is just the sort of thing we should expect to see if moral concepts were in the business of reducing the saliency or apparent availability of dangerous options. Second, we can remind ourselves of the central importance of principles in moral thinking. When we say that we are against something on ðmoralÞ principle, this does not generally indicate that one option is simply better than another option, but that the other is out of bounds. According to Ainslie’s framework, this is precisely the function of acting on principle—one categorizes actions into classes and has a rule about how to act with respect to any action in the class. This precludes case-by-case evaluation of the actions and, thereby, the threat of weakening of the will by means of rationalization of particular cases. The central role of moral principles in moral discourse is, again, much what we should expect if Commitment were true. Moral principles appear to employ both broad commitment strategies. Not only do they seem to reduce the apparent availability of options, they also seem to employ the same kind of rearrangement of incentive structure we saw in Ainslie’s work. There we saw that bundling decisions into classes alters the incentive structure associated with decisions about whether to, for example, smoke in a particular instance. Bundling achieves this rearrangement by having individual instances symbolize or represent entire classes of actions. So, just as this sort of strategy can help conceptualize whether to smoke this cigarette as the question whether one will be a smoker, moral concepts help conceptualize whether to steal, cheat, lie, and so on in particular cases as questions about whether one will be a thief, cheater, or liar. And such a negative self-conception comes at a very high cost, both in terms of guilt and shame for most people, but also in terms of social currency. For if one thinks of oneself in these terms, it is more likely that one will act in accordance with this self-conception and more likely that others will come to share that conception. The above remarks are admittedly brief and rough. But bear in mind that we should have a high prior probability that our moral concepts act as commitment devices, based simply on the belief that we are creatures with a tendency for preference reversal evolved to operate in social environments where punishment for norm-transgression is common. As we saw, it is quite plausible that the ability to perceive norm violations in a motivationally laden way ði.e., via emotions such as shame, guilt, and angerÞ is an evolved capacity geared toward avoiding norm violations, including perhaps those norms calling for the punishment of norm violators. The avoidance of norm violations must be at least fairly This content downloaded from 132.239.1.231 on Fri, 28 Mar 2014 10:55:31 AM All use subject to JSTOR Terms and Conditions Campbell Breakdown of Moral Judgment 465 reliably achievable, despite the demands of those norms often running directly contrary to other very powerful aspects of our motivational psychology, including and especially those connected to food, sex, power, and avoidance of bodily harm. So, where these contrary motivations are present, there is ðthe potential forÞ motivational instability and norm violation. Therefore, especially in the context of within-group punishment for norm violation and intergroup competition, we would expect significant evolutionary pressures for individuals and groups to achieve stability in the direction of the motivations geared to norm adherence. And now when we look at our moral concepts, they appear to be just the sorts of concepts we would have predicted to find, insofar as they bear the clear marks of both of the independently understood strategies for overcoming hyperbolic discounting. Now I want to emphasize that the commitment strategies at the heart of moral concepts will ðat leastÞ often be threatened if one is aware of them. Since the removal or limiting of options is subjective,51 once one is aware of the strategy, many unthinkable things may become thinkable, and once thinkable can be difficult to unthink in important cases. This is especially true in the context of a society becoming aware of the moralization of their norms as a kind of commitment strategy. Therefore, we should also expect to find that those individuals or groups that ðmostÞ effectively employ these strategies have ðthe mostÞ effective methods of remaining relevantly unaware. Another way of putting the point is that awareness of the relevant sort will be likely selected against at both individual and group levels. Now, by what means would moral concepts regulate awareness in ways geared toward maintaining stability of the motivations associated with norm adherence? It is well known that direction of attention can be pivotal in determining which competing motivation controls behavior, so one of the primary ways of not succumbing to temptation is to direct attention away from the tempting thought or object.52 However, the 51. By this I mean that these strategies don’t involve changes to material incentives. Removing or limiting the option to smoke by flushing cigarettes down the toilet is not a subjective commitment strategy, while thinking or saying “smoking is simply not an option” is. As I’ve indicated above, this kind of thought can have the effect of rearranging incentives as well, since if one were to smoke subsequent to such an assertion, it would lessen the effectiveness of future such assertions. However, unlike flushing cigarettes, this rearrangement of incentives operates in a way that is fundamentally dependent on our interpretations of events. If we subsequently interpret ourselves as having rightly changed our minds about whether it was a ðgoodÞ option to smoke, the value of such thoughts or assertions could be undiminished. 52. Michael Posner and Mary Rothbart, “Attention, Self-Regulation and Consciousness,” Philosophical Transactions of the Royal Society B: Biological Sciences 353 ð1998Þ: 1915–27. This content downloaded from 132.239.1.231 on Fri, 28 Mar 2014 10:55:31 AM All use subject to JSTOR Terms and Conditions 466 Ethics April 2014 conscious direction of attention away from temptations itself appears to deplete a limited resource necessary for willpower.53 Therefore we should expect moral concepts to be designed to deflect attention away from contrary motivations without requiring—indeed by precluding— that this process be deliberate or conscious. In my view, this is the primary function of ðparadigmaticallyÞ moral judgments. They play a committing role by deflecting attention away from our motivations and onto a realm which is apparently independent of those motivations. Divine commands have traditionally played ðand for billions of people still playÞ this role, since what God commands is paradigmatically independent of our motivations. The central moral concepts ðwhich largely arose from religious onesÞ that do this kind of work for most of my readers are those of moral obligations, requirements, duties, rights, ðimÞpermissibility, blameworthiness, responsibility, desert, and the like.54 It is these concepts, and others suitably related to them, which help provide the sense of practical oomph for which naturalistic accounts of moral obligation struggle ðand in my view failÞ to provide an account, and yet which any acceptable naturalistic account must do in order to retain the paradigmatically moral character of these concepts. On my hypothesis, they retain this kind of felt normative force due in large part to their function of deflecting attention ðand therefore investigationÞ from the very motivations that they are in the business of supporting. So practical oomph is just what we should expect if Commitment and Deflection are correct, for these theses together maintain that the central moral concepts are essentially commitment devices that require for their operation a lack of awareness of the relational nature of ðspecifically moralÞ value judgments. And when I say that we should expect practical oomph, I mean we should expect only a vague sense of “mustness,” one that is likely opaque to analysis. For Deflection suggests that the fundamental business of moral discourse is not to commit people to any particular presuppositionðsÞ, but rather to the avoidance of thoughts about their own motivations and commitments. Therefore Deflection predicts that Philippa Foot was right to doubt that there is some “fugitive thought” behind morality’s peculiar normative force. Rather, she supposed that the sense of “must” or “have to” we associate with morality 53. For an excellent general survey of the relevant literature, see Mark Muraven and Roy Baumeister, “Self-Regulation and Depletion of Limited Resources: Does Self-Control Resemble a Muscle?” Psychological Bulletin 126 ð2000Þ: 247–59. 54. The “first-order” concepts seem roughly geared toward nonviolation of first-order norms, and the second four “second-order” concepts seem geared toward nonviolation of metanorms, mostly to do with punishment. This content downloaded from 132.239.1.231 on Fri, 28 Mar 2014 10:55:31 AM All use subject to JSTOR Terms and Conditions Campbell Breakdown of Moral Judgment 467 55 reflects our feelings about morality. We can feel as if we have to do something, she said, without having any belief that we are under some form of compulsion or other.56 Finally, it’s worth being clear that Commitment and Deflection could be true even if the concepts and sentiments didn’t evolve at all. Recall that on Ainslie’s view, the relevant sense of mustness develops at the individual psychological level, as largely unwitting strategic responses to the problems of hyperbolic discounting. However, this is not my view. I discussed cultural evolution partly because I want to guard against the impression that I think moral discourse has a single function. In focusing exclusively on moral judgments as commitment devices, it might seem that I have left out the kind of social functionðsÞ of morality pointed to by the social contract tradition. For example, Hume held that the rules and norms connected to justice were both explained and justified by their being useful—even necessary—for society in general to function well. This strikes me as quite plausible. So let me be very clear that I do not think moral judgments have the sole function of acting as commitment devices. Like Hume and some of the researchers listed above, I think moral discourse has a fundamentally social, coordinating function as well. I put so much emphasis on Commitment and Deflection for two primary reasons. First, I have used them to explain the peculiarly moral character of moral discourse, specifically the sense of end-independent authority that seems to provide especially powerful forms of motivation and has been the target of the most prominent forms of moral skepticism. Commitment and Deflection in no way imply that the only function of moral discourse is to act as commitment devices. Rather, they imply that these other functions could be accomplished by a quite different discourse, for example, a version of “straight talk” conducted squarely in terms of our ðcollectiveÞ long-term preferences, goals, values, concerns, commitments, their priority rankings, and so on. Commitment and Deflection are therefore perfectly consistent with there being other functions of moral discourse, but these theses are particularly well suited to explain why moral discourse is so different from any such straight talk, and also why moral discourse has been criticized as inherently self-deceptive, hypocritical, manipulative, or otherwise involving some kind of bad faith. However, and this is the second reason I have focused on them, I think Commitment and Deflection are plausibly prerequisites for at least some of the fundamentally social, coordinating functions of moral discourse. I have argued that moral concepts plausibly contributed to cre55. Philippa Foot, “Morality as a System of Hypothetical Imperatives,” Philosophical Review 81 ð1972Þ: 305–16. 56. Ibid., 312. This content downloaded from 132.239.1.231 on Fri, 28 Mar 2014 10:55:31 AM All use subject to JSTOR Terms and Conditions 468 Ethics April 2014 ating and maintaining large groups. That general kind of cultural evolutionary story is well suited to explain why people would not tend to justify the requirements of justice by means of the kind of general, instrumental arguments Hume offered. Christine Korsgaard criticizes Hume because he “forgot that knowing that our hatred of injustice was based on general rules would have a destabilizing effect on the obligation always to be just.”57 I don’t know whether Hume forgot this or not, but I am confident that it’s true; this awareness does tend toward destabilization, which is precisely why Commitment and Deflection predict that most people would resist thinking of the badness of ðcausing, allowingÞ injustice in general, instrumental terms. So I do not see the committing function of moral discourse as only a matter of preventing individual backsliding, or as operating only in parallel to any fundamentally social functionðsÞ of moral discourse, but rather as what is plausibly common to and ðloosely speakingÞ required for either of these functions to develop. This concludes the argument for my solution to the first of the questions we started with, though there will be further considerations in favor of Deflection in the next section. We have made sense of the seeming lack of contingency on motive that moral judgments possess, even if they are in fact so dependent. My primary goal now will be to suggest an answer to the second question, namely, the value of moral discourse. Of course I’ve already argued that its value can be understood in terms of Commitment and Deflection. But the next section will explore downsides to these commitment strategies that are both potentially severe and almost entirely invisible until one learns how to look. III. A PRETEND SOLUTION Joyce’s normative moral fictionalism is at heart the project of keeping dangerous thoughts at bay.58 He judges all moral claims to be false since they presuppose nonexistent end-independent reasons but argues that we should nevertheless continue employing the discourse ðveryÞ much as normal—that we should pretend to believe our own moral claims and judgments. The point of pretending to believe the judgments is to retain the motivational benefits of moral discourse without believing falsehoods. He thinks that we can think the same thoughts as standard users of the discourse, such as “Stealing is immoral!” or “That’s just ðmorallyÞ wrong!” without believing such thoughts. For a moral fictionalist, thoughts like 57. Christine Korsgaard, The Sources of Normativity ðCambridge: Cambridge Universty Press, 1996Þ, 103 n. 12, my emphasis. 58. Joyce, The Myth of Morality. This content downloaded from 132.239.1.231 on Fri, 28 Mar 2014 10:55:31 AM All use subject to JSTOR Terms and Conditions Campbell Breakdown of Moral Judgment 469 these would continue to be in the business of reducing or preventing weak-willed rationalization, now unblemished by false belief. One thing I find particularly valuable about Joyce’s proposal is that it helps us see that the question whether to be a moral fictionalist amounts to the question whether to systematically deflect attention from the values and goals our moral judgments promote, as well as from the general fact that they are such commitment strategies. In this section, I’ll show how Joyce’s fictionalism is an excellent though unwitting argument for Deflection and then examine a side of the practical coin that Joyce misses. That is, I’ll show how moral judgment—on our shared conception of it—can give rise to severe but typically invisible pathologies. The explanation of the pathologies will also explain why they remain nearly invisible, especially among moralists. According to Joyce, thinking “stealing is just wrong” serves to silence thoughts to the effect that one ought to steal in certain cases because one is unlikely to get caught, or that this is in some other important respect an exceptional circumstance. On Joyce’s fictionalist proposal, the difference between pretending to believe moral judgments and actually believing them is that if one is pretending, one will be disposed to dissent from the judgment in one’s most “critical context,” such as a philosophy seminar.59 I agree that Joyce succeeds in distinguishing pretense from belief in general, using the example of pretending that Sherlock Holmes really existed in order to have more fun on vacation in London. However, the distinction collapses in contexts of “deep immersion”—precisely the kind of immersion that Joyce is insightful and forthright enough to admit will be required for a moral fictionalist.60 Unlike someone passing a jolly day in London pretending to be touring areas where the “real” Sherlock Holmes lived and sleuthed, a moral fictionalist does not have the luxury of casually stepping out of her fiction if someone—especially herself!—asks her if she really believes what she’s saying. Unlike Holmes fictionalism, moral fictionalism is specifically designed to prevent weak-willed rationalization in contexts of significant temptations to do something “immoral.” Therefore it is only worth going to the effort to overcome the other myriad obstacles to being a moral fictionalist if one can reliably count on not stepping out of the fiction when strong temptations arise. But it is in just such circumstances that one is likely to remind oneself that one does not really believe what one is saying. There is simply no way to make the fictionalism a remotely viable strategy unless one is willing and able to ðsomehow!Þ render oneself unaware of one’s belief that reasons are related to one’s goals or 59. Ibid., 92. 60. Ibid., 218–19. This content downloaded from 132.239.1.231 on Fri, 28 Mar 2014 10:55:31 AM All use subject to JSTOR Terms and Conditions 470 Ethics April 2014 values—or of one’s underlying motivations for becoming a fictionalist in the first place. And one needs to remain unaware of these things precisely in contexts in which one is strongly tempted to act “immorally.” I think that Joyce’s normative fictionalism is psychologically infeasible for this and other reasons, but that is not my point here. Rather, my point is that it is crucial to a workable moral fictionalism that one achieve a stable lack of awareness of one’s own underlying motivations and commitments in the “moral” realm. This is not something that Joyce denies; he is clear-sighted enough to see how deeply one will have to be “immersed” in the fiction for it to work well enough to be worth the considerable difficulties involved. Crucially, it is the whole point of Joyce’s moral fictionalism to attempt to do better ðmotivationally speakingÞ than conscious attempts to act in accordance with commitments one understands in instrumental or prudential terms.61 Joyce wishes to convince us that we can maintain this lack of awareness when it counts without falling into self-deception. However, he relies on Kent Bach’s analysis of self-deception as evidence that fictionalist thoughts can motivate while falling short of belief.62 According to Bach, a primary technique of self-deception is “jamming.” Jamming is having the thought that not-p as a means of avoiding the thought that p. If the thought that p happens to occur to a person, he can have the thought that not-p. If the unwanted thought of p arises, out of a motivation that not-p, “he focuses his attention on what it would be like if not-p . . . ½and may$ perhaps even go so far as to ½act$ as if not-p were the case. . . . Jamming is particularly effective in self-deception about one’s feelings or motives.”63 My main point here is not that we should see moral fictionalism as essentially self-deceptive. What I am pointing out is that on Joyce’s own model of fictionalist thinking, moralistic thoughts are essentially in the business of “jamming” the awareness of or inquiry into one’s own motives where moral matters are concerned. This is further evidence that cultivating a lack of awareness of one’s motives is at the heart of Joyce’s fictionalism. And of course I think that Joyce is on to something deep about moral discourse, since Joyce’s model of ideal fictionalist thinking is essentially a restatement of Deflection. Therefore, to the extent that Joyce is ðveryÞ successful in capturing what is important for the 61. In fact, it must perform even better than someone who understands the importance of commitments quite well and has taken less philosophically and psychologically tortuous steps than those involved in becoming a fictionalist. 62. Kent Bach, “An Analysis of Self-Deception,” Philosophy and Phenomenological Research 41 ð1981Þ: 351–70. 63. Ibid., 361–62. This content downloaded from 132.239.1.231 on Fri, 28 Mar 2014 10:55:31 AM All use subject to JSTOR Terms and Conditions Campbell Breakdown of Moral Judgment 471 practical success of normative moral fictionalism, Deflection is ðstronglyÞ supported. It is worth being very clear that Joyce had no intention whatever of supporting a thesis like Deflection. Indeed, if Deflection is correct, then the distinction between a true believer and a successful moral fictionalist is one without any important difference, something he was very keen to deny. Rather, what Joyce has helped to reveal is that belief in moral judgments and success in being a moral fictionalist are both centrally understood and explained in terms of a general and systematic lack of awareness of one’s underlying motivations. Joyce acknowledges that most people might be “too deeply embedded in the ‘morality cult’” to see that moral requirements have no more intrinsic rational authority than wacky cult rules.64 On the same page, he acknowledges that our motivation to resist the conclusion that someone could ever perfectly reasonably scoff “Morality, Schmorality!” might lie in our “being immersed in a particular normative framework.” Later, returning to the same theme, he suggests that the reason we are uncomfortable granting that immorality could be practically rational might be “due simply to the fact that we are adherents of the moral institution. . . . Yet this in itself seems troubling, for it seems to suggest a degree of self-delusion in the moral adherent’s relationship to moral precepts.65 In other words, Joyce thinks that being deeply immersed in the normative framework of morality can lead to, or even amounts to, self-delusion. The most natural response to the above line of argument is to say that if moral discourse is so valuable, why be so fussy about not deluding ourselves? But is moral discourse valuable on the whole? Joyce’s arguments for this claim rightly address the value of subjective commitment strategies but fail to address any of the downsides of these strategies, especially where those strategies rely on a sustained lack of awareness of their status as commitments strategies. Recall that Joyce’s central argument for normative fictionalism relies on the claim that we will have less scope for weak-willed rationalization if we can employ “conversationstoppers” such as “Stealing is just wrong!” Unfortunately, Joyce did not avail himself of George Ainslie’s work, which provides an empirically grounded theoretical framework capable of explaining and strongly supporting Joyce’s intuition and so would have made Joyce’s arguments as to the motivational function of moral discourse much stronger.66 How64. Joyce, The Evolution of Morality, 63. 65. Ibid., 204. 66. In The Myth of Morality, Joyce cites George Ainslie, “Impulsiveness and Impulse Control,” Psychological Bulletin 82 ð1975Þ: 463–96, but only in support of the general inef- This content downloaded from 132.239.1.231 on Fri, 28 Mar 2014 10:55:31 AM All use subject to JSTOR Terms and Conditions 472 Ethics April 2014 ever, it would also have provided good reasons for thinking that this motivational function comes with grave and unrecognized risks. Ainslie’s model of the will predicts and explains several significant downsides of willpower, pathologies that are rarely recognized as connected to willpower at all. I think we can see these downsides in the moral realm as well, though they are also rarely recognized as connected to moral thinking as such. Though there are multiple side effects to willpower, I will focus on those involving the motivated misperception of our own motives and the overshadowing of ultimate ði.e., finalÞ ends or goods. The motivated failure to perceive violations of one’s own moral rules is one of the greatest potential drawbacks of moralistic commitment strategies, but it goes entirely unrecognized by Joyce. And Joyce is very far from alone in this, as it is rarely if ever recognized as a drawback of moral thinking as such. Ainslie’s work gives us a way to recognize it. On his account of the will, great amounts of willpower are often organized on the basis of personal rules. Perceived violations of a rule threaten the ability of that rule to organize motivation in the future. However, if you catch yourself ignoring a violation, you damage your credibility with yourself, so there can be very strong motivation not to catch yourself, even in the intrapersonal realm. In the context of moralized forms of guilt and anger, there can be especially powerful motivation for the misperception of our own motives, actions, and values. In the intrapersonal realm, motivations to avoid the perception of violations can lead to failure to save money or lose weight despite “strict” budgets and diets. In the interpersonal or political realm it can lead to systematic deception despite a strict policy of never lying, or systematic usage and valuing of aggression and terrorism despite a lack of awareness of this fact, and indeed strong beliefs ðand perhaps other valuesÞ which condemn it in the harshest terms. Making matters worse, long-term failure to notice moral violations is much more likely than long-term failure to notice violations of prudential rules. If one continues to lose money or not to lose weight, it will be hard not to notice, and there will be consequent pressure to perceive that one is likely violating one’s own rules. But very often, there aren’t any such unavoidably apparent consequences of violating one’s moral rules or values, or even of having a value structure that is quite different than one supposes it is. And since the perception of moral transgressions is generally at least as aversive as that of prudential transgressions, the motivations to avoid the former perceptions will tend to be ðmuchÞ more efficacious. fectiveness of prudential reasoning in securing long-term goals. In Joyce’s The Evolution of Morality, he cites Ainslie’s Picoeconomics to the same end, but without any acknowledgment of the downsides of willpower Ainslie discusses there and in Breakdown of Will. This content downloaded from 132.239.1.231 on Fri, 28 Mar 2014 10:55:31 AM All use subject to JSTOR Terms and Conditions Campbell Breakdown of Moral Judgment 473 Understanding this point allows us to see the double-edged nature of the committing strategies associated with the moral sentiments, as well as moral rules, principles, and concepts. By hypothesis, it makes evolutionary sense for moral sentiments like guilt to develop because they can compete in a hyperbolically discounted marketplace with the rewards associated with transgressions. To compete, the aversiveness must be sufficiently strong, as well as reliably present or predictable before, at the same time, or soon after the relevant transgression. To feel guilty about something one must perceive that one has transgressed ðnot in those terms of courseÞ. But there are always two ways to avoid such a perception. One is not to transgress and the other is not to perceive one’s action as a transgression. So the very motivational efficacy of guilt generates powerful motivations to systematically misperceive one’s own actions and motives. When we see that these motivations ðsynergistically?Þ combine with the motivations to misperceive one’s motivations that I described above in connection with the employment of moral concepts and principles, we have the foundation for an explanation of the staggeringly deep, widespread, and intransigent hypocrisy and self-delusion characteristic of so much moral discourse. The problem of misperception of motives and actions is great, but it is possible to overcome it. Unfortunately, doing so can lead to another pathology of willpower—and of paradigmatically moral cognition. In order to avoid the interpretation that she has lapsed, that is, violated her moral principles, a moralist ðincluding a “successful” moral fictionalistÞ may be prone to overcautiousness. At the extreme, she may resemble Kant, who, as a consequence of his ðinÞfamous view that the will should operate according to an absolutely inviolate moral law, proscribed lying in any situation whatever. Reliance on moral principles “can make large categories of differential reward hinge on decisions of little intrinsic importance.”67 This feature underpins the fact that not only is it the case that lapses reduce one’s ability to follow rules, but nonlapses reduce one’s ability not to. Therefore not only can we feel that much is at stake in decisions of intrinsically little importance, but we can be led to do what is ðradicallyÞ against our long-term preferences and values in decisions of great intrinsic importance. Pace Kant, sometimes lying is a very good idea. Even if this pathology doesn’t reach such an extreme state, the point is that the commitment to absolutist moral principles has the potential for serious pathology. Ainslie cites Paul Ricoeur’s observation ðalongside those of several philosophers, authors, and psychotherapists recognizing essentially the same problemÞ that the freedom of our wills is undermined not solely by sinful temptations but by “moral law, through 67. Ainslie, Breakdown of Will, 148. This content downloaded from 132.239.1.231 on Fri, 28 Mar 2014 10:55:31 AM All use subject to JSTOR Terms and Conditions 474 Ethics April 2014 the ‘juridization of action’ by which ‘a scrupulous person encloses himself in an inextricable labyrinth of commandments.’”68 He also notes that existentialism was in significant part born of Kierkegaard’s decrying the attack on the “vitality of experience” entailed by Kant’s and Hegel’s insistence that all actions be taken in accordance with universal rational principles.69 Another proto-existentialist and critic of morality, Nietzsche ðinÞfamously perceived that “convictions are prisons.”70 If Joyce and I are right, then moral convictions are prisons by design. Their main point is to limit or prevent ða sense of Þ freedom with respect to actions falling under certain descriptions or categorizations. Joyce rightly sees the motivational benefits to be had by placing ourselves—in the normal case, being placed—in such prisons, but it’s odd that he doesn’t see much in the way of downsides to being in prison. Of course he could argue that the benefits of being in prisons of our own creation outweigh the costs, but what’s strange is that he doesn’t. He doesn’t seem to perceive the need to do so, assuming that if we moralize what is “already useful” that the long-term benefits of being in the right sort of prisons are likely to outweigh the costs.71 But even if we assume that we roughly have good or right values, moralizing them can be deeply problematic. Even such prohibitions as seemingly “useful” as that against killing innocent humans without their consent can lead to potentially very undesirable policies, such as opposition to abortion under any circumstances. And when this moral prohibition runs up against the injunction against violating a woman’s right to do whatever she wants to with her body ðanother very “useful”sounding policyÞ, the stage is set for intense and saddening confusion on all sides, not to mention angry, moralistic divisiveness. One might object that the better principle is against killing innocent persons without their consent. I would entirely agree, but this principle runs into very serious problems too, most obviously in contexts of war and other largescale hostilities. Now my point is not that we can’t create principles or policies to guide us here. Just the reverse; I think we will certainly need them in at least very many contexts. But thinking of moral principles as being true or false rather than better or worse at promoting our ðdeepest, highestÞ values is both a symptom and cause of the lack of awareness of the value of such principles. So my point isn’t that we should do without principles or policies generally, but that the moralization of principles serves to block an understanding that these principles are— 68. Ibid., 144, citing Paul Ricoeur, “Guilt, Ethics, and Religion,” in Moral Evil under Challenge, ed. J. Meta ðNew York: Herder & Herder, 1971Þ, 11. 69. Ainslie, Breakdown of Will, 144. 70. Friedrich Nietzsche, The Antichrist, sec. 54, trans. R. J. Hollingdale ðNew York: Penguin Books, 1984Þ. 71. Joyce, The Myth of Morality, 181. This content downloaded from 132.239.1.231 on Fri, 28 Mar 2014 10:55:31 AM All use subject to JSTOR Terms and Conditions Campbell Breakdown of Moral Judgment 475 at their most valuable and least pathological—in the business of protecting and promoting things that we care about. And once we gain such an understanding, we will be in a position to craft better policies and principles. The critique above assumes that we moralize roughly the right or best values. But it is very far from clear that this is so! This was Ian Hinckfuss’s criticism of morality.72 Hinckfuss noted that there have been innumerable atrocities not only carried out by people who firmly believed themselves in the moral right, but often done in the name of their moral values. As we’ve seen, Joyce responds that this just shows that it is important to moralize the “already useful actions,” that is, it is important to fictionalize the right moral values. Unfortunately, what the right values are is not something to be gotten squared away in an afternoon or fortnight prior to setting about moralizing them or rendering one’s existing moralizations fictional. When we recognize that what would make some values “the right moral values” goes far beyond what is instrumentally useful to us, this is even clearer. What the right values are is something we are always—if we are lucky—in the process of discovering and/or creating, both as individuals and societies. Those values are such that there will almost always be appropriate exceptions to any policy. But let me emphasize that my critique of moral discourse is not dependent on the idea that serious users of moral discourse must fail to see this, or that they generally follow rules “for their own sake,” treating them fetishistically. Rather, it holds that peculiarly moral beliefs and judgments, including but not restricted to belief in moral principles and intrinsic value, play a motivational role that has the tendency to hide the nature of our fundamental values from us by deflecting attention from what we care about for its own sake onto rules or principles or ostensibly intrinsic values. At their best, these tools tend to keep us from acting in ways that are contrary to our deepest or highest values. At their worst, they generate seriously pathological behavior and prevent us from gaining selfknowledge, knowledge of what it is that we do care about and thereby what other, derivative things we should care about. And again, it is for this reason that an attempt to save moral discourse by holding that all we need to do is adhere to good or correct moral principles misses the point that the moralization of these principles can serve to prevent an awareness and understanding of the values we ought to be attempting to promote by means of such principles. Again, the normative problem with moral discourse does not lie in Commitment alone, but with Commitment combined with Deflection. 72. Hinckfuss, The Moral Society. This content downloaded from 132.239.1.231 on Fri, 28 Mar 2014 10:55:31 AM All use subject to JSTOR Terms and Conditions 476 Ethics April 2014 In conclusion, I find it very unlikely to be coincidental that—aside from a handful of philosophers and psychotherapists beginning in the nineteenth century—both willpower and morality have almost universally been thought of as unmixed blessings. Joyce has done a fine job in analyzing the benefits of one in terms of the other, but in failing to examine the other side of the coin, his case is much weaker than it might seem, for the kinds of costs I’ve maintained that moral discourse has are just the sorts of costs one should expect if the discourse has roughly the motivational benefits Joyce and I think it does. And notice that the costs I have described are roughly limited to those associated with moral conscience. But much moral cognition and discourse is concerned with condemnation or punishment, and we very plausibly have commitment devices geared toward motivating punishment of transgressors that can become very deeply pathological. One of the great benefits of recognizing that our moral sentiments and concepts were plausibly developed in part due to their role in maintaining large group sizes ðvia effective punishment strategiesÞ is that we can ask about the extent to which those concepts promote what we care most about now. This is a very important question to ask insofar as moralistic commitment strategies, perhaps especially the concepts associated with condemnation and punishment, are designed not to promote our goals or interests, but rather those of inclusive fitness, or even perhaps group size/strength as such. This gives us yet more reason to attempt to gain an awareness of our concepts and sentiments as commitment strategies, as well as the nature of the specific motivations or values those strategies are supporting. For to the extent that we remain unaware of the nature of the commitments promoted by our moral judgments, we run the serious risk of being committed in ways that are deeply pathological from the point of view of our deepest or highest values. IV. CONCLUSION I have sketched a solution to the puzzle of how a naturalist can explain the apparent motivation-independence of moral judgment and used that solution to address the question of the value of moral discourse. I have explained why moral discourse has significant motivational benefits, but also the potential for serious pathologies of roughly the sort that myriad critics of moral discourse have attempted to pin down. I conclude by pointing out some advantages of critiquing moral discourse without relying on any traditional error theory. My critique of moral discourse has nowhere relied on any traditional error theoretic commitments. On the contrary, my critical approach is consistent with a wide range of metaethical views, which I think has very important advantages over contemporary versions of moral This content downloaded from 132.239.1.231 on Fri, 28 Mar 2014 10:55:31 AM All use subject to JSTOR Terms and Conditions Campbell Breakdown of Moral Judgment 477 abolitionism, all of which rely on moral error theories. One such advantage is that my view is insulated from the seemingly intractable semantic disputes bedeveling traditional error theories. For example, many of those who reject Joyce’s moral error theory are, like him, substantive antirationalists. Unlike Joyce, they are also conceptual antirationalists. That is, they think that moral concepts do not presuppose the end- or motivation-transcendent reasons Joyce thinks they do,73 and so no such error theory is warranted. Unlike traditional error theories, the critique presented here takes no stand on these semantic disputes.74 For example, suppose Stephen Finlay is right to say that moral claims are true just in case they promote the implicit ends embedded in moral judgments—but that people only have reason to promote morality insofar as they care about moral ends.75 Or perhaps Jesse Prinz is correct in thinking that one’s moral judgments are true just in case one has the relevant sentiments of ðdisÞapproval that constitute the judgments.76 I doubt that we should accept either of these semantic views, but nothing in my critique depends on their being false. That is, if either of these or indeed any form of conceptual antirationalism is true, then very many moral claims might be true, but that will not affect my critique. My critique is centrally grounded in the pathologies associated with Commitment combined with Deflection, and these theses entail no semantic view.77 However, these theses do predict that most practically serious employers of moral discourse would be reluctant to accept either of these semantic views; they predict that there would seem to be something deeply wrong about them, even if they end up getting the semantics right. So even if some version of conceptual antirationalism is correct, that by itself will provide no reason at all for engaging in moral discourse generally. Whether to do so is a broadly normative question, and no semantic knife can cut any of the normative ice relevant to Commitment and Deflection that I have discussed here.78 However, these semantic views—as well as noncognitivist semantics—do need to be fought off by 73. Joyce, The Myth of Morality, and The Evolution of Morality. 74. Again, the critique does rely on substantive antirationalism, so all or nearly all the normative action rests with the issue of substantive, not conceptual, rationalism. This is a central reason why normative critiques of moral discourse should sidestep error theories as such. 75. Stephen Finlay, “Oughts and Ends,” Philosophical Studies 143 ð2008Þ: 315–40. 76. Jesse Prinz, The Emotional Construction of Morals ðNew York: Oxford University Press, 2007Þ. 77. If, on the other hand, substantive rationalism were true, that would affect my critique. I think there would still be an important critique to be made, but it would be different than the one I offer here. 78. Consider an analogy. Suppose that religious discourse is serving a committing function that relies on a deflection of attention from our motivations and that it has serious practical downsides, even worse than those of moral discourse. Further, suppose that belief This content downloaded from 132.239.1.231 on Fri, 28 Mar 2014 10:55:31 AM All use subject to JSTOR Terms and Conditions 478 Ethics April 2014 traditional moral error theorists. This has proven a very difficult task, to say the least. My approach is also unthreatened by the modern sophisticated expressivisms of Simon Blackburn and Allan Gibbard. Both theorists now hold that the distinctive and important features of their respective views are the sorts of explanations they give of normative language, not their former noncognitivist semantic claims that moral judgments are not in the market for truth. In fact, they both accept that attitudeindependent moral claims can be true, and explain that acceptance in terms internal to the functionðsÞ of moral discourse.79 That is, they do not give a philosophical defense of practical oomph by appeal to the nature of moral reasons or facts, but rather an explanation of the motivational roles that attitude-independent reason-claims play in our lives. And on the sort of minimalist conception of truth that they favor, there is no way to sincerely accept the reason-claims while denying their truth. Indeed, minimalism about truth must be rejected by any traditional error theory, though my critique can take it or leave it. Minimalist or deflationist accounts of truth hold that to say that the sentence “Snow is white” is true is just to say that snow is white.80 Likewise, the theoretical temperature does not change between claiming “It is true that slavery is bad” and claiming that slavery is bad. But if to accept that a moral judgment is true is just to accept that very moral judgment, there will be no way to deny that moral judgments are true while nevertheless acin a supernatural creator is central to the committing and deflecting function and that the existence of a supernatural creator is implausible. Based on these ðand perhaps otherÞ considerations, suppose I were to recommend a departure from religious discourse in serious ethical thought. Now suppose that someone sought to challenge my recommendation by successfully showing that divine command judgments can be true despite there being no supernatural creator. This could be by means of a causal theory of reference, or something Finlayesque according to which God’s commands are true just in case they promote the ends “embedded in” or that otherwise generate divine command judgments. I submit that the reaction of the employers of religious discourse should, on reflection, be “Who cares?” I think employers of moral discourse should have the same response to arguments for conceptual antirationalism. What matters most is whether the discourse can and should continue in light of an understanding of its function and overall value, not whether we can get many of the claims made within it to come out true. Of course if moral or religious claims are all false that might be an additional reason not to engage in them. 79. See Gibbard’s Thinking How to Live, 62–63; and Simon Blackburn’s Essays in QuasiRealism ðNew York: Oxford University Press, 1993Þ, 3–6, and Ruling Passions ðNew York: Oxford University Press, 1996Þ, 317–19. 80. W. V. O. Quine, Philosophy of Logic ðNew York: Prentice Hall, 1970Þ, 12. Deflationism is consistent with thinking that there is some important function served by the truth predicate, e.g., the promotion of argumentation and thereby consensus. See Huw Price, Facts and the Function of Truth ðOxford: Blackwell, 1989Þ, and “Truth as Convenient Friction,” Journal of Philosophy 100 ð2003Þ: 167–90. This content downloaded from 132.239.1.231 on Fri, 28 Mar 2014 10:55:31 AM All use subject to JSTOR Terms and Conditions Campbell Breakdown of Moral Judgment 479 cepting them. That is, there will be no way to separate one’s moral error theory from the question whether to accept moral judgments, as normative fictionalism must do. The question whether to consider any moral judgments true will turn on the question whether to make or accept moral judgments in general. And to decide that question, we will have to ask about the value of employing moral discourse generally, as I have done here. Of course all the various deflationist/minimalist views of truth might be wrong. However, I am not aware of any traditional error theorist even acknowledging, much less defending a nondeflationist account of truth. Fortunately, my critique of moral discourse is entirely consistent with but does not require deflationism. A second and related advantage of my approach is that although I believe we cannot locate moral normativity in the natural world, that belief forms no part of the basis of my critique of moral discourse. As I said, sophisticated expressivists do not think normativity is to be located in the natural world, but neither do they think moral judgments involve systematic “reification error.” Their kind of naturalism does not seek to locate normativity in the natural world but, rather, seeks to provide naturalistic use-explanations of problematic philosophical concepts, the moral as well as the modal.81 Now, though expressivists do not tend to do this, it would be open to them to offer criticism of these concepts, grounded in an appreciation of the roles those concepts play in our lives. That is what I have attempted here, though my critique in no way asserts or denies that moral judgments involve any systematic reification error. For anyone wishing to offer a normative or practical critique of moral discourse, the value of being able to sidestep these semantic and metaphysical debates is hard to overstate. A third advantage of my view over other versions of moral abolitionism is that I can—indeed must—recognize the value of moral discourse. All contemporary moral abolitionists regard moral discourse as almost uniformly destructive, leading to authoritarianism, elitism, exacerbation of conflict, self-deception, dishonesty, and so on, without much or anything to recommend it. This is of course the antipode to Joyce’s view, which regards it as almost uniformly beneficent. By contrast, I think it is awfully complicated and difficult. This aspect of my view leads to a final point of departure from contemporary abolitionists, which is that I allow that many, even perhaps most people or groups might be currently 81. For defenses of such a pragmatic, or “subject” naturalism against the orthodox, or metaphysical naturalism that sees morals and modals as either locatable in the natural world, or else bunk, see especially Huw Price, Naturalism without Mirrors ðNew York: Oxford University Press, 2010Þ; and David Macarthur, “Pragmatism, Metaphysical Quietism, and the Problem of Normativity,” Philosophical Topics 36 ð2008Þ: 193–209, 193. This content downloaded from 132.239.1.231 on Fri, 28 Mar 2014 10:55:31 AM All use subject to JSTOR Terms and Conditions 480 Ethics April 2014 best advised to continue employing moral discourse. I do not pretend to know how the balance of considerations will work out for each person or group.82 However, those who think of our reasons as ultimately connected to our contingent values and concerns should be especially attracted to the potential rewards of moving beyond moral discourse. For that discourse is not conducted in terms of what we care about or value.83 Rather, it is conducted in the language of rights, duties, obligations, requirements, ðimÞpermissibility, and the like. Whether one is bound by various duties and such is not thought to depend on one’s contingent values, and therefore such discussion not only does not encourage, but positively discourages, investigation into what it is that we actually care about— how much, in what ways, and with what priority rankings. When we add to this the powerful and systematic tendency for moral discourse to motivate us to misperceive our motivations, we have a potent prescription for avoiding knowing ourselves as well as we might. It is of course beyond the scope of this article to discuss what an alternative to moral discourse would look like, but at its heart would be an attempt at a sustained and honest inquiry into our ðdeepest, highestÞ values and commitments as a central aspect of investigating what values and commitments we ought to have and, more generally, how we ought to feel, act, and live. 82. One important factor will be that accepting ðsomething likeÞ substantive antirationalism, Commitment and Deflection makes the motivational benefits of moral discourse harder to maintain. Another is the priority ranking of values connected with selfknowledge. 83. Of course, moral discourse can take these things into account in some contexts, but they are never the fundamental considerations. This content downloaded from 132.239.1.231 on Fri, 28 Mar 2014 10:55:31 AM All use subject to JSTOR Terms and Conditions