Moral cognitivism and motivation

Philosophical Review

Volumn 108, Issue 2, 1999, Pages 161-218

  Svavarsdóttir, Sigrún ^a  

^a NONE

Author keywords

[No Author keywords available]

Indexed keywords

EID: 57049143820     PISSN: 00318108     EISSN: 15581470     Source Type: Journal    
DOI: 10.2307/2998300     Document Type: Article

References (117)

1. Here I have in mind Dostoyevsky's fictional character Lisa (the little she-devil) in
2. For useful general discussions of these positions as well as surrounding issues, see William K. Frankena's classic article "Obligation and Motivation in Recent Moral Philosophy" in Essays in Moral Philosophy, ed. A. I. Melden (Seattle: University of Washington Press, 1958), 40-81;
3. Stephen L. Darwall, Impartial Reason (Ithaca: Cornell University Press, 1983), chap. 5;
4. David O. Brink, Moral Realism and the Foundations of Ethics (Cambridge: Cambridge University Press, 1989), 37-80;
5. Michael Smith, The Moral Problem (Oxford: Basil Blackwell, 1994). In Darwall's terminology the thesis under discussion is judgment internalism about moral judgements in Brink's terminolog it is appraiser internalism about motives.
6. Michael Stocker, "Desiring the Bad: An Essay in Moral Psychology" Journal of Philosophy 76 (1979): 738-53.
7. Smith, The Moral Problem, sect. 3.1. Besides introducing the rationality condition, Smith's statement of the practicality requirement differs from my statement of motivational internalism in two ways. First, Smith does not use any modal notion in stating the content of the practicality requirement. However, he claims that the requirement has the status of a conceptual platitude. But presumably if it is a conceptual truth that a certain connection holds between moral judgment and motivation, then that connection holds in every conceptually possible world; that is, the connection holds of conceptual necessity. Second, Smith's practicality requirement specifies that the connection in question holds only between moral judgments about the behavioral options of the agent making the judgment and motivation. I do not restrict motivational internalism in this way because it seems artificially to weaken its putative support for noncognitivism (see next section). But such a restriction would not affect my arguments. In his development of his practicality requirement, Smith is influenced by Christine Korsgaard's work on practical reason.
8. See her "Skepticism about Practical Reason," Journal of Philosophy 83 (1986): 5-25.
9. Maybe these two issues cannot be pried apart. Maybe one cannot plausibly argue that the internalist constraint is or reflects a requirement of rationality unless one gives an account of moral judgments that reveals why that would be the case. Smith at least seems to think that these two issues go together; he gives an account of the content of moral judgments that, he argues, makes sense of why the practicality requirement is a conceptual truth (see note 61, below). Still, I counsel caution in formulating the qualification in terms of conditions of practical rationality unless, of course, a more informative formulation of it is forthcoming. Needless to say if conditions of rationality are just conditions under which the agent's conduct is interpretable as an intentional action, then I have no qualms about incorporating it into the internalist thesis. However, such an addendum is unnecessary, since it adds nothing of substance to the thesis.
10. Notice that as I understand motivational internalism, it makes a claim about the connection between motivation and moral judgments, rather than evaluative or normative judgments in general. Some would advocate an internalist constraint on all evaluative or normative judgments. Obviously, I am contesting both that moral judgments are necessarily connected to motivation and that the distinguishing mark of evaluative or normative judgments in general is their connection to motivation. But I would like to leave it open (though I remain doubtful) whether there is a subcategory of evaluative or normative judgments for which the internalist constraint holds. Now, it is of course notoriously difficult to say what distinguishes moral judgments from other evaluative or normative judgments. If pressed, I would gesture toward such factors as their "aspiring" to some sort of impartiality and being vaguely "concerned with" human well-being. But here I will have to rely on an intuitive understanding of moral judgments' being distinct from prudential, aesthetic, and other categories of evaluation. Thanks are due to John Richardson for pressing me to make this clarification.
11. James Dreier offers a weaker internalist thesis, modest internalismn, as a constraint of adequacy on accounts of moral language.
12. See his "Internalism and Speaker Relativism," Ethics 101 (1990): 6-26.
13. Modest internalism adds to motivational internalism the condition that the moral judge be "in normal context." This proposal need not be taken seriously if "normal contexts" are specified merely as "those circumstances under which a person is motivated by his moral judgments." Dreier recognizes this, but admits that he does not know how to specify rigorously the independent conception of normality that he needs. In the absence of such specification, the condition is too obscure for modest internalism to play an important role in deciding between competing accounts of moral thought and language. For further remarks on Dreier's view, see note 30, below.
14. Richard V. Miller suggests that it is a condition on the ascription of moral judgments that the agent be motivated by her cognizance of moral considerations when it has no cost to herself. See his Moral Differences: Truth,Justice and Conscience in a World of Conflict (Princeton: Princeton University Press, 1992), 95. I am inclined to read him as accepting motivational internalism as formulated in the text: moral judgments necessarily import some motivation, however weak and easily overridden it may be in situations that involve some cost to the agent (see, however, note 32, below). But Miller can also be read as accepting a weaker version of motivational internalism: moral judgments need only motivate in circumstances when the moral judge perceives no cost to himself. This weaker version also falls within the target of my argument in subsequent sections.
15. Notice that I have formulated cognitivism so that it does not involve any ontic commitment to moral facts or properties. It is a view about the nature of moral thought and the semantics of moral language. Prima facie, a cognitivist could be either a realist or an error theorist about moral thought and discourse. The cognitivist camp includes thinkers as diverse as Richard N. Boyd, "How to be a Moral Realist," in Essays on Moral Realism, ed. Geoffrey Sayre-McCord (Ithaca: Cornell University Press, 1988), 181- 228;
16. Brink, Moral Realism;
17. John L. Mackie, Ethics: Inventing Right and Wrong (New York: Penguin, 1977);
18. John McDowell, "Projection and Truth in Ethics," The Lindley Lecture (University of Kansas: Lawrence, 1988);
19. G. E. Moore, Principia Ethica (Cambridge: Cambridge University Press, 1903);
20. Peter Railton, "Moral Realism," Philosophical Review 95 (1986): 163-207;
21. Nicholas L. Sturgeon, "Moral Explanations," in Morality, Reason and Truth ed. David Copp and David Zimmerman (Totowa: Rowman and Allenheld, 1985), 49-78;
22. David Wiggins, "A Sensible Subjectivism?" in his Needs, Values, Truth: Essays in the Philosophy of Value (Oxford: Blackwell, 1987), 185- 214;
23. Smith, The Moral Problem;
24. David Copp, "Moral Obligation and Moral Motivation," in New Essays on Metaethics, ed. J. Couture and K. Nielson (canadianjournal of Philosophy, supp. vol., 1996);
25. Dreier, "Speaker Relativism." The last three thinkers, unlike the rest, accept reductivism about moral concepts: they believe that an adequate analysis can be given of moral concepts. However, all three use concepts from other parts of normative discourse in the analysans. Reductive (or analytic) cognitivism should not be equated with reductivism about moral properties or facts, which is, for example, accepted hy Railton.
26. Moral judgment has become a term of art in the metaethical literature. It is used to refer to the mental and speech acts central to moral evaluation, whatever their nature may be. Since I am concentrating on moral motivation in this paper, I will be mostly using it to refer to the relevant mental acts. It seems more accurate to think ofjudgments as mental acts rather than mental states, although they are, of course, the onsets, expressions, or activations of mental states. Perhaps they should be equated with occurrent mental states. Nevertheless, I distinguish between moral judgments and the corresponding mental states in this paper, and talk about the former as manifesting the latter.
27. I have been careful to formulate the noncognitivist position so that it encompasses not only the emotivist view of A.J. Ayer (see Language, Truth, and Logic (London: Gollancz, 1946), chap. 6)
28. Charles L. Stevenson (see "The Emotive Meaning of Ethical Terms," Mind 46 (1937): 14 -31),
29. R. M. Hare's prescriptivism (see The Language of Morals (Oxford: Oxford University Press, Clarendon Press, 1952)
30. but also Simon Blackburn's quasi-realism (see Spreading the Word (Oxford: Oxford University Press, 1984), chap. 6,
31. and Essays in Quasi-Realism (Oxford: Oxford University Press, 1993)
32. and Allan Gibbard's norm-expressivism (see Wise Choices, Apt Feelings: A Theory of Normative Judgement (Cambridge: Harvard University Press, 1990)).
33. All but Ayer invoke the internalist constraint in favor of their view, but it is most explicitly done by Stevenson ("The Emotive Meaning," 13)
34. cf. his magnetism requirement-and Blackburn (Spreading the Word, 188). However, in a 1991 article ("Just Causes," reprinted in Essays in Quasi-Realism, 198-209),
35. Blackburn modifies his position in such a way that one may wonder whether he remains a motivational internalist. When responding to an objection by Nicholas Sturgeon (see note 41, below), Blackburn seems to grant that Socrates and Thrasymachus (in Plato's Republic I) are both making judgments about justice, although only Socrates is motivationally affected by them. But Blackburn has not really given up on motivational internalism; for he holds that Satan can want evil to be his good only so long as he attaches some negative emotion to what he deems evil: "Milton's Satan can be represented as wanting to make evil his good because of his actual self-disgust, indeed his suffering, at being forced (by the need to have something over which to reign) to make such a choice. ⋯ But if his strategy is successful, and he succeeds in driving out not only remorse but all the other elements that enable him to see his own plight as desperate, then indeed the interpretation would start to waver. He would no longer be seeing evil as his good, but merely doing evil and seeing it as good" (201)
36. And Blackburn is quite ambivalent about whether Thrasymachus should, indeed, be interpreted as making judgments about justice: "Thrasymachus himself faces this peril [of being best reinterpreted]: if he sticks to the view that justice is what is in the ruler's interest, perhaps he is not best seen as discussing justice at all. Add the quirk that he is contemptuous of whatever it is that he is thinking of, and the difficulty increases" (199).
37. John McDowell is probably the most prominent cognitivist and internalist who reconciles these two iews in this manner. See his "Are Moral Requirements Hypothetical Imperatives?" Proceedings of the Aristotelian Society supp. vol. 52 (1978): 13-29,
38. and his "Virtue and Reason," Monist 62 (1979): 331-50.
39. See also M. Platts, Ways of Meaning (London: Routledge and Kegan Paul, 1979).
40. And for an argument against the Humean view of motivation, see T. Nagel, The Possibility of Altruism (Oxford: Oxford University Press, Clarendon Press, 1970).
41. David Lewis, who is a critic of the view, invented the convenient term 'besire' to refer to this unique kind of mental attitude that has both an assertoric and a motivational force. See his "Desire as Belief," Mind 97 (1988): 323-32.
42. For a simple illustration, take a subjective naturalist who claims that the belief that a is good amounts to the belief that a meets one's approval. The subjectivist may attempt to demonstrate that it is metaphysically or even conceptually impossible to have a belief with such content without actually approving of a, which is a type of desire within a Humean framework. Relying on a different analysis of the content of moral beliefs, Michael Smith attempts a reconcilation between cognitivism, the Humean view of motivation, and the practicality requirement (see note 61, below).
43. Miller (Moral D4fferences) attempts a reconciliation along entirely different lines (see note 32, below).
44. I am, moreover, inclined toward nonreductive or nonanalytic cognitivism (see note 9). Notice that the unanalyzability of moral concepts does not preclude the possibility that their semantic values are properties for which we also have (or could develop) other, nonmoral, concepts. I prefer to remain noncommittal on this point.
45. Brink, Moral Realism, 49.
46. Railton, "Moral Realism," 202.
47. My view on moral motivation is closest to W. D. Ross's and H. A. Prichard's, but this does not mean that I accept their metaphysical and epistemological views. In his 1928 article "Duty and Interest" (reprinted in Readings in Ethical Theory, ed. W. Sellars and J. Hospers (New York: Appleton Century Crofts, 1952)),
48. Prichard gives up his earlier view of moral motivation and identifies the desire to do what is right as the source of moral motivation. Ross subscribes to the same view in his The Right and the Good (New York: Oxford University Press, 1930).
49. I identify the desire as a conative.attitude taken towards an object under some moral mode of presentation; it need not involve the concept of rightness. I also depart from Prichard and Ross in that I do not take this desire to be universal or somehow inherent in human nature.
50. See Copp, "Moral Obligation." On Copp's own view, moral motivation has its source in an intention (or a policy) to conform, and support conformity, to certain standards. The relation between this intention and the agent's moral judgments is that the judgments imply that the standards in question are justified and call for (or prohibit) the action under evaluation. His account thus presupposes a partial analysis of the content of moral judgments; mine doesn't.
51. This distinction is closely related to Stephen Darwall's distinction between judgment and existence internalism about moral judgments (Imparhal Reason, 54).
52. (Brink also makes a similar distinction; in his terminology, this is the distinction between appraiser and agent internalism about motives (Moral Realism, 40).)
53. However, my distinction is drawn at the level of thought and concepts, whereas Darwall's is drawn at the level of reference. Application internalism pertains to the application conditions of moral judgments or concepts to an individual, whereas existence internalism pertains to the conditions for a moral obligation or value existing or being instantiated. Prima facie, my distinction leaves it open to noncognitivists and other antirealists to accept application internalism. For an example of application and existence internalism,
54. see Gilbert Harman's "Moral Relativism Defended," Philosophical Review 85 (1975): 3-22.
55. To be more exact, Hare maintains that the prescriptivity of a moral judgment consists in its entailing an imperative or command addressed to the person to whom the judgment applies; but it is distinctive of commands that an assent to a command addressed to oneself is sincere only if one abides by it. Notice that Hare maintains that there is a necessary connection between a moral judgment and action-or, translated into the language of motives, between moral judgment and overriding motivation. This is stronger than motivational internalism as stated above. Hare, as far as I know, is alone in holding this stronger thesis. See his The Language of Morals, § 2.2 and § 11.2; and his Moral Thinking: Its Levels, Method and Point (Oxford: Oxford University Press, Clarendon Press, 1981), § 1.6.
56. Gibbard, Wise Choices, 101.
57. Ayer (Language, Truth, and Logic),
58. Stevenson ("The Emotive Meaning"),
59. and Blackburn (Spreading the Word) maintain that moral judgments are expressive of emotive states: certain motivationally charged sentiments.
60. Hare (Moral Thinking takes them to express preferences of a unique type.
61. While Gibbard (Wise Choices) contends that they express the acceptance of norms for certain sentiments, which is a state of a motivational system he refers to as a normative control system.
62. Admittedly it is not clear that Gibbard's view has this upshot. On his ie I am expressing a positive attitude towards my friend's feeling guilty about treating his partner as he does and towards others' feeling resentful about his treating the partner as he does. Gibbard thinks that I can have this second-order attitude even if I do not feel any resentment against my friend's conduct. So maybe he would contend that this second-order attitude need not be manifested in the sort of inclinations mentioned in the text. Still, Gibbard must rely on motivational internalism rather than application internalism for the second reason given in the text.
63. Harman ("Moral Relativism") indeed arrives at application internalism in such a way. But a Kantian would presumably resist the implication that the applicability of a moral judgment to an individual is conditioned by his motivational states or capacities. Kantians have attempted to get around this by maintaining that a moral judgment correctly applies to an agent only if he would be motivated to act in a suitable way insofar as he were rational. In that way a triadic connection is established between the correct applicability of a moral judgment to an individual, the individual's having certain reasons for actions, and the individual's having certain motivations under conditions of rationality.
64. See Korsgaard, "Skepticism"; and Smith, The Moral Problem.
65. For exampterms of a necessary connection between an agent's having moral beliefs and "moral considerations ... provid[ing] her with reasons for choice in costfree situations," but he obviously takes this trivially to imply a necessary connection between moral belief and motivation in costfree situations.
66. It seems to me that Nagel, in The Possibility of Altruism, is best read as advocating motivational internalism on the grounds of the second rationalist position discussed in this paragraph; of course, this is not to deny that he also accepts the rationalist thesis discussed in the last paragraph.
67. Brink's objection against motivational internalism consists essentially in bringing up the case of the amoralist (Moral Realism, 46-48)
68. "someone who recognizes certain considerations as moral considerations and yet remains unmoved by them and sees no reason to act on them" (27).
69. This is essentially Smith's response to Brink's amoralism objection (Smith, The Moral Problem, 68-71).
70. Smith goes on to shoulder his part of the burden of argument by giving an argument against externalism. I address that argument in section 6 of this paper.
71. Dreier acknowledges the episternic possibility externalists advocate and, thus, rejects the internalist position that I am investigating in this paper. However, as explained in note 7, he accepts a weaker form of motivational internalism, one that includes the condition that the moral judge must be motivated in the normal context. His putative evidence for the claim that we operate with some sort of normality condition is that the most serious counterexamples to motivational internalism could not be construed as externalists interpret them except against "the background of central internalist cases" conceived of as providing the normal situation ("Speaker Relativism," 13).
72. Dreier would think it significant that my description implies that the cynic in question used to take moral judgments more seriously. He seems to think that our understanding of the cynic as making genuine moral judgments in spite of his lack of motivation depends on seeing his state of mind against the background of a more normal one he used to have and that most other people in his community have (cf. his point against Stocker, at 12).
73. Dreier might be on to something important here, but it is far from clear that it is best captured by a normality condition on motivational internalism.
74. Thanks are due to Ruth Chang and anonymous referees for the Philosophical Review for pressing this objection. And thanks are due to Carsten Hansen for a helpful discussion of it.
75. Miller (Moral Differences) would contest this claim. He provides a neo Davidsonian argument for motivational internalism: he derives it from a general condition on the ascription of belief rooted in norms of rationality to which an individual must to a great extent conform for his behavior to be interpretable as intentional action. Miller does not offer the thesis as a conceptual truth about moral judgments in the sense that it would fall out of an analysis of either the concept of a moral judgment or the concepts employed in moral judgments. But then it is doubiful that it can be invoked as a condition of adequacy on accounts of moral judgments that prima facie favors noncognitivism. Indeed, Miller can be seen as engaged in the project of explaining how a cognitivist can accept motivational in ternalism, while denying that moral judgments are ways of conceiving of the world that depend on or affect the motivational states of the thinker. I am not inclined to follow his lead, since I am convinced that we can readily interpret an individual as making sincere and competent moral judgments even if he fails to be motivated by them when he perceives it to be cost-free. At the end of section 5, I respond to Miller's argument for the claim that this is impossible.
76. These arc genuinely distinct candidates. The best way to appreciate this is to consider the theoretical possibility that moral judgments employ concepts like good, right, and just, but share these concepts with nonmoral judgments; what distinguishes them from judgments that use the same conceptual resources is their attitudinal force, which is in part motivational. On this view, motivational internalism falls out of the concept of a moral judgment rather than out of the concepts employed in moral judgments. Noncognitivists would also prefer the former candidate, since they do not think moral judgments employ any distinct concepts (assuming concepts are representational devices). It is worth noting in this context that Brink interprets motivational internalists (or "appraiser internalism about motives," as he refers to the position) as claiming that the internalist thesis holds "in virtue of the concept of morality" (Brink, Moral Realism, 40).
77. Undoubtedly, some philosophers have the hunch that motivational internalism provides a genuine insight into the nature of moral judgments. And proceeding on that hunch might prove fruitful. I have my doubts. Be that as it may, my main point is that it is only a hunch and that the internalist thesis cannot play a crucial role in the context of justification, namely, in evaluating competing accounts of moral judgments.
78. Obviously, the commitment skeptic is not wondering whether he morally should take into consideration the moral value of the alternatives. Traditionally, the commitment skeptic has been understood as asking whether he rationally should commit to morality; in other words, he is implicitly asking about the relation between morality and rationality.
79. Both Nicholas Sturgeon and David Brink have drawn the above distinction between two types of moral skepticism. Sturgeon uses the traditional term 'amoralism' to refer to skepticism about moral commitment, while Brink uses the term 'amoralist skepticism'. I do not use the traditional term 'amoralist' because it has been used both to refer to commitment skeptics and (as Brink does) to people who are indifferent to their moral judgments-cynics, in my terminology. But I want to emphasize that commitment skepticism and moral cynicism are two distinct phenomena, as I explain in the text. Brink draws attention to commitment skepticism in order to convince internalists to take his amoralism objection seriously {Moral Realism, 46-48). Sturgeon, on the other hand, uses commitment skepticism to make a point similar to the one I have just made. Sturgeon observes that noncognitivists are committed to dismissing amoralism as impossible, and notes that their tactic can be seen as an instance of a discredited anti-skeptical strategy. (See his "What Difference Does it Make Whether Moral Realism is True?" Southern Journal of Philosophy 24, supp. (1986): 121.)
80. Peter Railton has correctly pointed out to me that an explanation is needed of why it is reasonable to operate with this assumption. This should, I believe, be done in the context of explaining how the conceptual resources employed in moral thinking are more intimately bound up with the regulation of conduct than are those involved in conveying information, say, about climatic conditions.
81. Interestingly, David McNaughton rests his case for motivational internalism on the claim that an adequate moral theory must account for the connection between moral commitment and action. See his Moral Vision (Oxford: Basil Blackwell, 1988), 20-22. Thanks are due to an anonymous referee for the Philosophical Review for pointing this out to me.
82. Thanks are due to Paul Boghossian for raising this case.
83. The Language of Morals (New York: Oxford University Press, 1952), especially, 124-25 and 164-65. Hare's exact words are: "Thus it is possible to say 'You ought to go and call on the So-and-sos' meaning by it no value-judgement at all, but simply the descriptive judgement that such an action is required in order to conform to a standard which people in general, or a certain kind of people not specified but well understood, accept" (164).
84. I have come to realize that this point is a close relative to Sturgeon's objection to noncognitivism in "What Difference Does it Make Whether Moral Realism is True?" There he points out that it is implausible to deny that Socrates and Thrasymachus (in Plato's Republic I) agree on various things concerning justice; for example, "that the actions of the aspiring despot, in 'appropriating the possessions of the citizens' while enslaving the owners of these possessions as well (344b-c), are unjust" (120). But noncognitivists seem committed to doing so; for while Socrates disapproves of that action as well as any other action he calls unjust, Thrasymachus admires it and other actions he calls unjust. Noncognitivists will have to claim that Socrates is expressing his disapproval, while Thrasymachus is "using the terms 'just' and 'unjust' (that is: the corresponding Greek terms) in some secondary 'descriptive' sense, and hence that his argument with Socrates is only apparent" (121). In other words, Sturgeon's objection is that noncognitivists will implausibly have to treat the agreement, as well as the disagreement, between Socrates and Thrasymachus as a case of massive miscommunication.
85. In The Moral Problem, Smith takes a somewhat similar track. When responding to Brink's objection to internalism, Smith claims diat amoralists (cynics, in my terminology) "do not really make moral judgements⋯. The point is not that amoralists really make judgements of some other kind: about what other people judge to be right and wrong, for example. The point is rather that the very best we can say about amoralists is that they try to make moral judgements but fail" (68). Of course, there is a difference between Smith's proposals and the one in the text: Smith's suggestion is that cynics make honest attempts, but fail at making moral judgments, whereas my suggestion, on behalf of internalists, is that cynics are play-acting or parodying moral judgments. But the proposals are similar in that both reject Hare's original suggestion that the cynics are making judgments, albeit not moral ones. Now, Smith's proposal would not fare any better than the one I have suggested. As Smith himself acknowledges, the issue of whether his or Brink's description of the cynic's (amoralist's) employment of moral language is the right one boils down to the issue of whether "being suitably motivated under the appropriate conditions is ⋯ a condition of mastery of moral terms" (70). I agree. My main point in this paper is that the latter issue can only be setded by defending an account of moral thought and language on grounds that are neutral as far as the debate between motivational internalists and externalists goes.
86. Miller's moral nihilist is this type of cynic {Moral Differences, 84). I am in agreement with Miller that his moral nihilist cannot but use moral language in inverted commas. However, given his acceptance of motivational internalism, Miller is committed to claiming diat no cynic can make a genuine moral judgment. I disagree with him on this point. Momentarily, I will address Miller's argument to the effect that it is incorrect to ascribe moral belief to any cynic.
87. These characters are much like a scientist who for illustrative purposes gives an explanation of some phenomena in terms of a scientific theory she does not accept. The scientist is not strictly speaking describing how the phenomena in question would be explained by the rejected theory, but rather framing an explanation within that theory. However, she is not offering it as an explanation of the phenomena; for giving an explanation in terms of a theory that one disowns hardly amounts to offering an explanation of the phenomena. It can at most be an act of illustration or deception.
88. Obviously, I am rejecting global and other strong versions of semantic holism. If the reader finds it implausible that noncognitivists, error-theorists, and realists about moral discourse are making the same sort of judgments within that discourse so long as they continue to participate earnesdy in it, I urge him or her to consider whether the same is to be said about belief-ascriptions made by philosophers who have different views about belief-ascriptions.
89. Moral Differences, 95.
90. Moral Differences, 95 Ibid.
91. Moral Differences, 95 Ibid.
92. Thanks are due to a referee for the Philosophical Review for helping me to formulate this point.
93. The first version is given in §3.5 of his The Moral Problem, while the second version appears in his "The Argument for Internalism: Reply to Miller," Analysis 56 (1996): 175-83.
94. Parenthetical references in the text to "MP" and "RM" are to these two works, respectively. Smith accepts a thesis closely related to motivational internalism (as I have characterized that position), which he calls "the practicality requirement on moral judgment" (see note 4 for a discussion of the subtle difference between these two positions). Smith's argument does not hang on whether externalism be understood as contrasting with motivational internalism or with the practicality requirement. A third statement of the argument-a hybrid of the two versions that does not avoid the problems with either one-appears in "In Defense of The Moral Problem: A Reply to Brink, Copp, and Sayre-McCord," Ethics 108 (1997-98): 84-119.
95. Smith has declared in print that he regrets his choice of terminology when giving the anti-externalist argument in his book. He claims he did not really mean "the term 'good' to pick out those who are good in the more substantive sense of having the motivations that one true morality tells them that they should have" ("Reply to Miller," 177). In this context, Smith refers to objections he has received from David Copp, objections that depend on the above reading of "good people" (see Copp, "Moral Obligation"). I suspect that Smith will also dismiss my response to his argument on similar grounds. However, I am confident that, as the quotes from his book demonstrate, his mishap is not merely an unfortunate choice of terminology. I, uierefore, stand by my response. For different criticisms of Smith's argument see Alexander Miller's "An Objection to Smith's Argument for Internalism," Analysis 56 (1996): 169-74;
96. David Brink, "Moral Motivation," Ethics 108 (1997-98): 4-32;
97. David Copp, "Belief, Reason, and Motivation: Michael Smith's The Moral Problem," Ethics 108 (1997-98): 33-54.
98. Smith would add "in rational people." Since by this qualification he intends to exclude, among others, the weak-willed, he specifies that the person would have to be strong-willed as well as good in order for the connection to hold reliably.
99. Copp ("Moral Obligation") does not grant Smith this. He argues that the connection between moral judgment and motivation would not be reliable in a good, but weak-wiled, person, and concludes that it is strength of will rather than goodness of character that is responsible for the reliable connection. He claims that an externalist must and can explain the reliable connection in terms of the nature of the strong-willed person. Now, Smith would not object to the claim that the reliable connection is absent in a good, but weak-wiled, person; his practicality requirement (as well as motivational internalism, in my formulation) has a built-in excepdon for the weak-willed person, be he good or evil. One might, therefore, take Copp's conclusion that it is strength of will rather than goodness of character that is responsible for the reliable connecdon as a concession to Smith's version of motivational internalism; for Smith, strength of will is a part of pracdcal rationality. But if I understand Copp correctly, he rejects the idea diat strength of will is a necessary part of practical rationality; it is a distinct characteristic that only some practically radonal moral judges have. Thus, Copp's conclusion should not be interpreted as a concession to Smith's internalism. Copp's point is really that the credibility of Smith's initial premise (that "a change in motivation follows reliably in the wake of a change in moral judgement⋯ in the good and strong-willed person") depends on a different reading of 'strong-willed' than Smith is ready to give. I am not taking this line, since I am willing to grant Smith's initial premise, given his reading of 'strong-willed'. Nodce that this issue between Copp and Smith illustrates what sort of can of worms is opened by the introduction of die practical radonality condidon on the internalist thesis. It could, for example, be conceded to Smith that strength of will is a necessary ingredient in practical rationality, but then argued that Copp is right that strengdi of will is a disdncdve characterisdc that is responsible for the reliable connection between moral judgment and motivation and diat, moreover, is not necessary for having the capacity to make moral judgments. Whoever takes that posidon accepts Smith's practicality requirement, but rejects his idea that it reveals an important fact about the preconditions for die mastery of moral concepts.
100. See Brink's and Railton's suggestions reviewed in secdon 1.
101. The conceptual content of this conative state in each individual must, I submit, reflect the moral concepts he employs. For someone may be a good person, even if he holds a somewhat mistaken moral view. Say the correct moral view is not duty-based. Still, someone who holds a duty-based morality may be a good person, so long as he displays the virtues diat are crucial to being a good person (say, kindness, compassion, and honesty) and cares about doing what, by his lights, is of moral value. However, there might be limits to how mistaken the moral view could be. Is it, for example, possible to be a good person, yet hold a moral view that has strong racist or sexist implications? This is a tricky question, as we can come to appreciate if we reflect on individuals from an entirely different era. I tend to think that in evaluating persons who hold reprehensible moral views, we take into account whether they are in a good epistemic position for discovering the mistakes in their moral views. A person who is epistemically in a position to recognize the serious problems with his moral view, but doesn't, does not count as good, even if he is morally conscientious, kind, compassionate, and honest.
102. Copp ("Moral Obligation") makes essentially the same point with his example of Dana.
103. Indeed, Smith makes the stronger claim that externalists are committed to giving an account of the motivational system of the good person that amounts to elevating this moral vice into the one and only moral virtue. This stronger claim relies in part on the mistaken assumption that externalists are committed to the view that the desire to be moral is the only defining characteristic of the good person. Again, I do not see why externalists need to be committed to this.
104. In "Duty and Interest," Prichard actually argues that given the presence of the desire to be moral, it makes no sense to raise the question 'Why be moral?' His argument depends on interpreting the question as a request for an incentive to act morally. But that is a mistaken premise. The question may also serve to initiate critical reflections on our commitment to morality. As such it can be intelligibly raised by people who already have the desire to be moral.
105. Smith certainly agrees with me on this point. We both reject what he and Philip Pettit have coined "the foreground view of desires." See P. Pettit and M. Smith, "Backgrounding Desire," Philosophical Review 99 (1990): 565-92.
106. It is very tempting to use the notions of reason and justification in this context. I am resisting the temptation, though, since that move raises a lot of subtle questions that I want to avoid at this point. I believe an agent can see himself as doing something because it is β without thinking that its being β carries much justificatory force or gives him anything like a reason for action. Philosophers have been far too ready to equate justifying one's actions to oneself or others with explaining one's actions to oneself or others.
107. According to Smith, all internalists (or defenders of the practicality requirement) depict the connection between moral judgment and motivation as follows: "Thus, if an agent judges it right to φ in C, and if she has not derived this judgement from some more fundamental judgement about what it is right to do in C, then, absent weakness of will and the like, defenders of the practicality requirement can insist that she will be motivated non-derivatively to φ in C. This is because on the rationalist alternative, a non-derivative desire to φ in C is what her judgement that it is right to φ in C causes in her, or because, on the expressivist alternative, the judgement that it is right to φ in C is itself just the expression of such a non-derivative desire" (The Moral Problem, 73). Smith accepts the rationalist alternative. He then argues that this connection holds at least in rational agents because the content of the moral judgment that it is right for S to φ in C is (skipping one qualification that need not detain us here) equivalent to the proposition that S would desire that he φ in C, if S were fully rational. It would be a sign of irrationality to fail to want de re to do what by one's own lights one would want to do if one were rational; that is, part of rationality is to be disposed to form the desires one (correctly or incorrectly) believes that one would have if one were rational.
108. Interestingly, it is far from clear that other internalist accounts will portray the intentional perspective of the good person as Smith does; they might side with me. Expressivists, according to Smith, maintain that the moral judgment is but an expression of a nonderivative desire to φ- hence, the reliable connection between making a positive moral evaluation of φ and being nonderivatively motivated to φ (see the quote in the previous note). Since, on this account, the good person has a nonderivative concern for doing φ, one might think that she "sees" herself as doing φ simply because it is φ. However, in the public language this person expresses the desire to § with "φ has moral value," and maybe expressivists will maintain that her intentional perspective is also best expressed by these words. A third internalist alternative (which Smith does not mention in this section of his book) is to maintain that moral judgments manifest besires. On this account as on my externalist account, the intentional perspective of the good person is such that she "sees" herself as φ-ing because it is morally valuable. For that is the conception of her action that engages her motivationally. It is just that a besire rather dian a pure conative state is postulated to explain this motivational engagement.
109. "Persons, Character and Morality," in his Moral Luck (Cambridge: Cambridge University Press, 1981), 18.
110. Copp ("Moral Obligation") makes a similar point during his discussion of Dana.
111. Throughout this section I have taken Smith's distinction between derivative and nonderivative desires as intuitively clear and not in need of discussion. Nevertheless, I think that it stands in need of clarification, as this paragraph reveals. One can understand 'a nonderivative desire' as designating a desire that is not acquired as a result of another desire and the belief that this latter desire would be or could only be fulfilled if the object of the former desire were procured. Let's call such desires "acquisition-nonderivative." In this sense, the desire to φ is derivative on my account. But one can also understand 'a nonderivative desire' as designating a desire that at a given point in time is self-standing, in that its role in the psychology of its possessor at that time is not limited to motivating him to do something he believes is needed in order to satisfy another desire. The desire to φ can become nonderivative in this sense in the good person, even if it is acquisition-derivative. For Smith's argument to go through the desire to φ must be acquisition- nonderivative in the good person. But I see no merit in that claim, although I agree that the psychology of the good person has to be such that it is possible for the desire to become nonderivative in the sense of self-standing. See also my discussion below of what Smith means by an "instrumental" desire in his recent reformulation of the anti-externalist argument.
112. "The Argument for Internalism: Reply to Miller" (see note 51). In "In Defense of The Moral Problem," Smith appears to take back this disclaimer. There he claims that the problem with the externalist account of moral motivation is that it does not "square ⋯ with our commonsense idea of moral virtue" (112). As in The Moral Problem, his argument depends on the mistaken assumption that those who explain moral motivation-motivation by moral judgments-as having its source in the desire to be moral must think that this desire is the source of all the motivations characteristic of the virtuous, or at least that this desire takes center stage in the motivational system of the virtuous.
113. Hovvever, notice that it is not a far cry from a more interesting internalist thesis, namely, motivational internalism explicitly restricted to morally committed individuals-a thesis that we may dub MCI and that I accepted in section 5. Morally committed individuals are moralists, although the class of moralists, as defined, is broader than that of the morally committed. Amongst moralists there may be morally uncommitted individuals who for some perverse reason are, as consistently as any of us, motivated to act as they judge morally valuable or required. The domain of the internalist thesis I am willing to accept is, therefore, even more restricted than the domain of WMI. But unlike WMI, the conceptual truth of MCI is not secured by way of a stipulative definition. Its source, I submit, is in our ordinary concept of commitment.
114. I would, indeed, advocate a conception of desire according to which a desire to φ not only grounds the motivational disposition to do φ under certain conditions, but also grounds the cognitive disposition of noticing things that bear on whether φ obtains and how φ can be brought about, as well as emotional dispositions such as being pleased upon doing φ and frustrated when one cannot do φ.
115. And even so, considerations of systematicity and uniformity in our theoretical apparatus, used to make sense of each other, might dictate that we maintain the distinction between belief and desire even in such cases.
116. See McDowell's "Are Moral Requirements Hypothetical Imperatives?" for criticism of the Humean theory of motivation, which charges that it amounts to a mere dogma built on some suspect aprioristic or metaphysical intuition.
117. I would also think that variations in some of their cognitive and emotional dispositions can be explained in a similar way. See note 68 for my ronrention of desire.