Gibbard on Korsgaard

Review: Morality as Consistency in Living: Korsgaard's Kantian LecturesAuthor(s): Allan GibbardSource: Ethics, Vol. 110, No. 1 (Oct., 1999), pp. 140-164Published by: The University of Chicago PressStable URL: http://www.jstor.org/stable/2672649Accessed: 24/11/2010 23:56

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available athttp://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unlessyou have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and youmay use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained athttp://www.jstor.org/action/showPublisher?publisherCode=ucpress.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printedpage of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

The University of Chicago Press is collaborating with JSTOR to digitize, preserve and extend access to Ethics.

http://www.jstor.org

http://www.jstor.org/action/showPublisher?publisherCode=ucpress

http://www.jstor.org/stable/2672649?origin=JSTOR-pdf

http://www.jstor.org/page/info/about/policies/terms.jsp

http://www.jstor.org/action/showPublisher?publisherCode=ucpress

REVIEW ESSAYS

Morality as Consistency in Living: Korsgaard's Kantian Lectures*

Allan Gibbard

We all confront "the normative question": told, say, not to cheat when I could get rich by cheating, I can ask why; why forgo wealth for honesty? Any answer can prompt a further why-and so I face a regress of whys. How, we must ask, could any answer end this regress? And a specter of unmasking can haunt our thinking: rival versions of the Modern Scien- tific World View tell us what accepting a must consists in and offer psy- chosocial etiologies of this motivation-laden state. The moral skeptic, writes Christine Korsgaard in The Sources of Normativity, "thinks that once we see what is really behind morality, we won't care about it any more"

(p. 13). Kant, though, discovered a way to terminate the regress of whys and

face down the specter of unmasking-so maintains Korsgaard. His ethics of autonomy is "the logical consequence of the theory of normativity shared by Hume, Mill, and Williams" (p. 51), and it is the only ethic "consistent with the metaphysics of the modern world" (p. 5). Kors- gaard, though, is far from a simple mouthpiece of Kant; she reworks Kant in an extended argument, starting from the plight of anyone who reflects on what to do and why. She writes with eloquence, clarity, and humanity, showing Kantian doctrines as flowing from a perplexity we can each accept as our own. The upshot constitutes a modern classic in ethi- cal thought.

A frustrating classic, though, at least for me: Korsgaard's starting points I find correct and illuminating; she presents "the normative problem" as one that lies in wait for anyone who reflects and acts. Moral re-

" A review of Christine M. Korsgaard, The Sources of Normativity, with commentary by G. A. Cohen, Raymond Geuss, Thomas Nagel, and Bernard Williams with a reply by Kors- gaard; edited by Onora O'Neill with an introduction by her (New York: Cambridge Univer- sity Press, 1996). All references in the text are to this work. These stem from Korsgaard's 1992 Tanner Lectures and commentary on them at Clare Hall, Cambridge. Work on this review was supported by a National Foundations for the Humanities Fellowship and an American Council of Learned Societies Fellowship, at the Centre for Philosophy of Natural and Social Science, London School of Economics.

Ethics 110 (October 1999): 140-164 (C 1999 by The University of Chicago. All rights reserved. 0014-1704/2000/11001-0006 $02.00

140

Gibbard Morality as Consistency in Living 141

alism of a metaphysical kind she rightly rejects: "The assumption of a realm of inherently normative entities or objective values is not needed to explain the existence of normative concepts" (p. 47). Normative questions, instead, are practical, questions of "what to think, what to like, what to say, what to do, and what to be"-and why (p. 9). "Moral properties are projections of human dispositions," she writes (p. 91), and though I find this projection metaphor misleading, I very much accept the claim behind it: that moral statements express practical states of mind-policies for living and the like. This is the doctrine I call expressivism. The quest for reasons to act stems, Korsgaard tells us, from "the reflective structure of human consciousness" (p. 104), from our "capacity to distance ourselves" from our mental activities, from our desires and plans, "and call them into question" (p. 93). Hence our why questions and the whole regress of whys, and hence our quest to terminate the regress in an unconditional answer, "one that makes it impossible, un- necessary, or incoherent to ask why again" (p. 33). Practical thinking, she believes, can give rise to practical arguments, showing the commitments that acting carries with it.

These starting points will be controversial, but not with me; I applaud. My topic will be not whether her starting points are right or wrong, but what follows from them. Korsgaard thinks that what follows is an elaborate Kantian package; I think that what follows is something more modest.

Acting at all, for any reason whatsoever, commits you to far-reaching consequences: that is what Korsgaard sets out to demonstrate. Any reflective agent, it turns out, must embrace the moral law and act as a Citi- zen of a Kingdom of Ends. Korsgaard's argument is rich and difficult, and though I'll attempt a reading, I very much can't feel assured that I'm getting it right. There seem, though, to be three chief stages, deriving (i) the Categorical Imperative, then (ii) the moral law, and, as a bit of an afterthought, (iii) duties to others. The Categorical Imperative isn't itself the moral law, and the moral law has content apart from how we must treat others. I'll spend much of this essay struggling with what these self- commands are supposed to be, and how the derivation might run.

I. SUBSTANCE AND LOGIC

Korsgaard regards herself as in one sense a moral realist, but another kind of moral realism she rejects. The "substantive" moral realist thinks, wrongly, that the moral life is "the application of theoretical knowledge to the solution of human problems," knowledge of the right and the good (p. 44). Instead, she insists, thinking about obligation is a matter of thinking what to do and why. Another, procedural form of moral realism, though, is perfectly fine: "The procedural moral realist thinks that there are answers to moral questions because there are correct procedures for arriving at them. But the substantive moral realist thinks that there

142 Ethics October 1999

are correct procedures for answering moral questions because there are moral truths or facts which exist independently of those procedures, and which those procedures track" (p. 36). This last Korsgaard rejects.

Korsgaard, then, distinguishes the two forms of moral realism by two distinct features. First, the "substantive" moral realist appeals to metaphysics, to "theoretical" investigation of a nonnatural realm of normative properties, which parallel the natural properties studied by science. In place of sense perception, he employs a sense-like power of intuition, a power, as it were, to sense the layout of nonnatural entities. The "procedural" moral realist holds that moral thinking is practical, not theoretical; it is thinking what to do. This we can call the contrast between a metaphysical moral realist and an expressivist, who, in Korsgaard's version, contends that moral claims express one's reflective endorsement of policies for living. Moral reflection, says Korsgaard, "is practical and not theoretical: it is reflection about what to do, not reflection about what is to be found in the normative part of the world" (p. 116). The second contrast is the substantive/procedural contrast proper. The substantivalist terminates the regress of whys with substantive principles; the proceduralist, with a procedure.

Substantivalism needn't go with metaphysical moral realism-or if the two must stand or fall together, that would need to be shown. On my own view, metaphysical moral realism has all the defects that Korsgaard charges it has.' But rejecting it doesn't establish proceduralism. Both Korsgaard and the proceduralist recognize that the regress of whys must terminate somehow.2 Korsgaard thinks that the regress ends with a procedure, with "reflective endorsement" somehow tied to "autonomy" and the Categorical Imperative. We need to ask whether procedures claim any advantage, as why-stoppers, over substantive principles-those substantive principles that seem most evident, those policies for living that we find most clearly unproblematic. This is a distinct issue from the one that pits Korsgaard's expressivism against a gratuitous, metaphysical moral realism.3

1. Whether historical intuitionists have been metaphysical moral realists may be a

difficult question. Ronald Dworkin derides metaphysical realism as the "moron" theory and

denies that anyone has ever attempted this explanatory strategy. See his "Objectivity and

Truth: You'd Better Believe It," Philosophy and Public Affairs 25 (1996): 87-139, pp. 103-5.

2. This needn't disallow a form of coherentism: the regress might come to an end

not with one or more principles that are fully self-evident, but with a set of mutually rein-

forcing considerations, each with some degree of self-evidence.

3. Many other issues too can be labeled ones of moral "realism" vs. "antirealism,"

but I must leave these aside. "Realism" might be the claim that moral claims are true or

false. If truth is minimal, so that 'It's true that torture is wrong' just means that torture is

wrong, an expressivist can agree (see Paul Horwich, Truth [Cambridge, Mass.: Blackwell,

1990]). "Realism" might also be the claim that moral truths, or some of them, hold inde-

pendently of us. Again, an expresssivist can agree. Suppose, for instance, the "moral realist"

proclaims, "It is a fact, independent of us, that kicking dogs for fun is wrong." This, the


What makes a principle of action substantive, as opposed to procedural, formal, or logical? Thinking what to do is disciplined thinking, subject to formal requirements of its own, a logic of its own. This claim that a logic governs action is characteristically Kantian, and to me it seems utterly right. There is, for instance, something incoherent about the following set of thoughts:

i) Let me spend the whole morning at the beach, ii) Let me spend the whole morning writing, iii) I can't write at the beach.

If I accept all three of these, I am then under pressure to revise what I accept-the same kind of pressure as I would be under with inconsistent beliefs in prosaic fact.

Korsgaard and I agree on this, but we disagree sharply about the powers of logical requirements to settle what to do. Korsgaard labels herself a proceduralist, but she may be something much stronger. Is the right procedure, after all, just self-evident? What if I just don't see that it is? Korsgaard says I must: "If you acknowledge the existence of any practical reasons, then you must value your humanity as an end in itself" (p. 125), and this commits me to the full package of enlightenment morality (p. 123). "The price of denying that humanity is of value is com- plete practical normative skepticism" (p. 163). This price is exacted, it seems, with the full force of logic: anyone who reasons what to do and why must acknowledge moral claims, on pain of contradiction in his maxims for action. This we might call a moral logicism: just as mathemati- cal logicists hold that mathematics is part of logic, so Korsgaard appears to hold that morals are a part of the very logic of what to do.4

expressivist can say, just means "Kicking dogs for fun is wrong, and would be even if we regarded such fun as innocent." Saying this expresses one's reflective endorsement of opposing kicking dogs for fun-even kicking dogs in the counterfactual circumstance that no human reflectively endorses opposing it. Importantly, though, on this view, the seeming metaphysics doesn't explain the morals; it just restates moral stances and policies, perhaps in a misleading way. Nagel takes this view in his commentary (Sources, p. 205); see also Simon Blackburn, "Errors and the Phenomenology of Value," in Morality and Objectivity: A Tribute to J L. Mackie, ed. Ted Honderich (London: Routledge & Kegan Paul, 1985), pp. 1-22, p. 11, whose example this is; Allan Gibbard, Wise Choices, Apt Feelings: A Theory of NormativeJudgmeent (Cambridge, Mass.: Harvard University Press, 1990), pp. 165-66; and Dworkin, p. 109. Alternatively, a "moral realist" might be someone who thinks that funda- mental moral truth extends beyond our powers to discover it. But some who regard themselves as "normative realists" reject this possibility; see Thomas Nagel, The Viewfrorn Nowhere

(New York: Oxford University Press, 1986), p. 139. 4. More precisely, she appears to think it a theorem of logic that any nonskeptical

normative claim whatsoever (when conjoined with the nonnormative facts) entails the whole of morality. One might conceivably be a moral logicist in this sense and yet be a

metaphysical moral realist, one who thinks that moral claims are "theoretical" claims; Locke might be an example. If so, one would think that the logic in question is the ordinary logic of the "theoretical." Korsgaard, as I say, is an expressivist; she thinks that moral claims


Perhaps, though, interpreting Korsgaard in this way is a mistake. Kant himself might appear sometimes to rest arguments on a necessity that isn't a pure matter of logic; a rational being, for instance, "necessar- ily wills that all his faculties should be developed." Korsgaard concludes such things as this: "Not every form of practical identity is contingent or relative after all: moral identity is necessary" (p. 121). Perhaps this in- vokes some form of nonlogical necessity. But if so, that form of necessity would have to be explained: What kind of inescapability is Korsgaard claiming for morality? Much of her argument reads as if morality is sheer consistency in one's maxims. That is a strong claim, and it would be a marvelous boost to morality if she or anyone else could establish it. I'll be reading Korsgaard as making this strong claim, and I'll be asking whether it's a claim that we need-and whether anyone could make good on it.5

II. SUBSTANTIVE MORAL GROUNDINGS

On my own view, logic does indeed yield requirements of consistency, not only in belief but in plan-but there are many possible ways of being consistent. Most of them would be crazy, and some are downright vicious. Consistency has real bite, to be sure: often we'll find ourselves inconsistent, and then we're in for rethinking. But consistency, I would insist, is no guarantee of truth: we can coherently disagree with someone we recognize as fully consistent.6 Special creationists, for instance, are mistaken in their rejection of Darwinian evolution as accounting for the origins of

are "practical." Hence if she is a moral logicist, then the logic that, according to her, entails morality is practical logic, the logic of what to do. I'll use the term 'moral logicism' to mean this doctrine. (G. A. Cohen too, in his commentary, may read Korsgaard as claiming logical status for her arguments; he summarizes her crucial arguments on p. 185.)

5. Does moral logicism, according to Korsgaard, perhaps apply only to perfect duties and not to imperfect ones? Previous readings of EKant by Korsgaard fit reading her as a moral logicist for all duties. The crucial distinction for Kant is between duties ofjustice and duties of virtue. Duties of virtue stem from the "contradiction in the will" test (Christine Korsgaard, Creating the Kingdom of Ends [New York: Cambridge University Press, 1996], p. 83), and they involve "the free adoption of some end which pure practical reason di- rects," namely humanity (ibid., p. 19). They are deduced from a feature of human beings, that we are purposive, that we act only for ends. On Korsgaard's reading of Kant, I conclude, you are inconsistent if (i) you act, (ii) you are by your nature, you know, a being who only acts for ends, and (iii) you do not have humanity as your end. The inconsistency is one of pure practical reason-or as I would put it, to be this way is to violate practical logic, the logic of acting. This appears to fit Korsgaard's own aspirations in these later lectures: For duties of virtue as well as duties of justice, I read her as purporting to establish a theorem in practical logic. The papers from Creating the Kingdom of Ends cited in this note date from 1985-89. See especially pp. 15-22, 82-84, and 176-79.

6. Korsgaard too thinks there are many ways of being consistent; she thinks that what to do depends partly on one's inclinations (see ibid., p. 20). As I understand her, though, if your inclinations are worked into logically consistent ends which you endorse, there can be nothing to criticize, and the rest of us can't coherently disagree with your choice of ends.


our species. Often, their views are inconsistent or rooted in ignorance of the evidence, and in that case, contemplating more evidence and rethinking in light of inconsistencies could lead a special creationist to correct his views. Logic and evidence, nevertheless, won't guarantee truth: there is no logical guarantee that one couldn't be fully consistent and fully aware of the evidence, and still accept special creation. If one did, then no matter how consistent one's views, one would be wrong. Or this, at least, is the view of any Darwinian like me, and it is a coherent view: we can recognize that the rare special creationist might be consistent, and still disagree.

A like view is available for thinking what to do. Imagine I'm a binge alcoholic: I get blind drunk every Saturday night, and often I end up in jail, or wreck my car, or suffer other calamities. Now in all likelihood when I take that first drink, I'm wrong on the facts or inconsistent in my policies for living. I'm kidding myself. And so heeding the facts and the requirements of practical consistency might lead me to change my mind on whether to drink; if I developed a formally coherent policy for living, then in light of what always happens once I begin to drink, this coherent policy might well include never drinking. Still, though, some possible, formally coherent policies for living do include weekend binges with all their consequences. Such a policy is terrible in its upshot, given my proclivities, but its defect isn't formal inconsistency. The bite of logic is sharp; the decision theoretic tradition, indeed, argues that formal coherence requires that one's policy for living take the form of maximizing an index of expected value.7 What constitutes "value," though, I would insist, isn't settled by requirements of practical consistency alone.8 Caligula, imagine, aims solely to maximize the suffering of others. That is a hor- rendous life policy, but it needn't be formally inconsistent. We decent people might recognize such a policy as consistent, but still disagree with it; we are coherent to do so.

The view I'm adopting is Kantian in one respect: against Hume, it recognizes a role for reasoning and logic in deciding how to live, a role that has genuine bite and that isn't confined to thinking about the facts. In another way, though, the spirit of this view is Humean. Hume is sus- picious of any philosophy that claims to force us to strong, substantive conclusions by sheer force of logic alone. In the realm of belief, what we

7. Peter Hammond presents perhaps the most refined argument to this effect, though most readers will find his presentation daunting; see his "Consequentialist Foun- dations for Expected Utility," Theory and Decision 25 (1988): 25-78. Frank Ramsey, "Truth and Probability" (in his Foundations of Mathematics and Other Logical Essays [London: Rout- ledge & Kegan Paul, 1931]), and Leonard Savage, The Foundations of Statistics (New York: Wiley, 1954), are the classical arguments to this conclusion.

8. Or again, more precisely, what I'm denying is this: that what constitutes values is settled, logically, by the requirements of practical consistency plus the facts of what aims various people reflectively endorse.


accept and what we reject depends not only on heeding logic, but on our proclivities to find some things more credible than others. If these proclivities go wrong, we come to the wrong beliefs, no matter how consistent we might be. Likewise, our judgments of how to live depend notjust on a logic of what to do, but on our proclivities to "see" some eventuali- ties as goals to pursue, and others as dangers to shun. If our proclivities go wrong, then we go wrong in our life policies, be they ever so consistent. Illogic isn't the only way to be wrong.

What besides practical logic could do the job of getting our life policies right? Substantive principles, one reply might be-and this reply is somewhat in the spirit of Clarke, Sidgwick, and other intuitionists. Take, for instance, what I'll call the weak pleasure principle: "The fact that I would enjoy doing something weighs in favor of doing it." This is a practical principle: Accepting it amounts to a determination, whenever one would enjoy doing something, to weigh this fact in favor of doing it. On what basis can I accept the principle? I find the question puzzling; what further basis do I need? True, nothing in the sheer logic of what to do dictates the principle; the principle is substantive. True, I might come to reject this principle, if I find it clear that there are exceptions. I could then look for deeper substantive principles that discriminate which pleasures to seek-but I would then still be a substantivalist. And still, surely, indifference to all pleasure would be perverse. As Korsgaard says, "If you think reasons and values are unreal, go and make a choice, and you will change your mind" (p. 125). One real value you'll find is the value of pleasure.

I find in Korsgaard three objections to grounding moral reasoning on substance. One is that doing so requires a metaphysical moral realism; to this I have already spoken. Second, calling the principle self- evident, she charges, is no more than an expression of confidence- which can't help anyone who doesn't share this confidence. Realism, says Korsgaard, "refuses to answer the normative question. It is a way of saying that it cannot be done" (p. 39). This complaint targets moral substantivalism itself, not just metaphysical moral realism. What, then, Korsgaard's worry seems to be, if you don't confidently favor your own pleasure? To this, though, the answer is clear: If you don't, true enough, philosophical argument can't make you.9 What you need is perhaps not philosophy but therapy. Or perhaps you do need philosophy, but only in a negative way: you need to be disabused of some bogus philosophical line that has led you to reject a goal you'd otherwise embrace with confidence.

9. This is a theme in the commentaries: Cohen distinguishes forcing a morally alien- ated person into morality by sheer force of logic-which, he thinks, can't be done by Kors- gaard or anyone else-from justifying a moral commitment one has by appeal to one's substantive commitments. Moral argument, says Nagel, "is a matter of being faced with the alternatives, and having to decide which is more credible" (Sources, p. 208).


Third, though, a more serious worry: pleasure is easy but morality is hard. For morality you may have to die! Moral skepticism needn't be practical skepticism wholesale, a blanket doubt that anything at all is worth pursuing. It can focus on stringent moral demands as such, skep- tical that these can be founded on substantive grounds we'll find unproblematic in the face of hard choices. This third challenge is fair enough, and I can offer no quick fix. One job of moral philosophers is to propose candidate grounding principles for morality and display their virtues. Korsgaard herself works to explain the stringency of morality: Moral vio- lations threaten loss of one's "practical identity" as a rational being. And because, as Wittgenstein indicates, reasons are public, because of our social natures, my practical identity must be responsive to reasons of others. Korsgaard is eloquent in explaining what a practical identity is and why it matters. Now the considerations she offers, at this point in her argument, are available just as much to the substantivalist as to the proceduralist. It's a good question what can ground the stringency of moral demands; this is the prime question of normative moral theory. Kors- gaard herself, though, offers some possible kinds of answers.

Korsgaard herself would think that the weak pleasure principle badly misfires as an example of a self-evident substantive principle. First, not all pleasures are of value. Second, she can explain why pleasures do have value when they do, proceeding without appeal to substantive bases. So Korsgaard maintains, and I return later to her treatment of pleasure and its value. Before that, though, I examine the arguments she presents as central to Kantian ethics.

III. REFLECTIVE SUBJECTIVISM

Like Kant, Korsgaard moves first to the Categorical Imperative -though what's at stake at this stage in the argument I find puzzling. On a modest reading, a substantivalist can gladly accept the Categorical Imperative; it just tells us to be consistent in our rationales for action. That's fine, the substantivalist can say, but remember too, consistency won't on its own carry us to morality. Korsgaard, though, seems at times to be committed to a stronger position, one a substantivalist would reject; I'll label it a logically constrained reflective subjectivism. This doctrine, as I'll explain, does tell us what to do, and it might seem to follow from Korsgaard's starting points-but to think this would be a mistake. And a good thing too, I claim, for some things this doctrine could tell us to do are vicious.

Begin, then, with the reflective agent, who has impulses that he can accept or reject. On what basis can he do this? "According to Kant, as each impulse to action presents itself to us, we should subject it to the test of reflection, to see whether it really is a reason to act. Since a reason is supposed to be intrinsically normative, we test a motive to see whether it is a reason by seeing whether we should allow it to be a law to us" (p. 89). So far, the substantivalist can applaud. (Her commentators in


the volume have their qualms, but these I'll skip over, because my own worries don't come at this point.) A substantivalist who believes in practical logic-that there is such a thing as consistency and inconsistency in thoughts of what to do-can read "law" here just as a consistent rationale for doing what one settles on doing. What, then, should the agent treat as a law? In asking this, the substantivalist will say, the buck must stop with substantive groundings, axioms, for example, such as that suffering is worth preventing. Korsgaard, though, insists on something firmer, something that would have to satisfy an agent who ques- tioned even such things as this.

We see whether we should allow a motive to be a law to us, Korsgaard says, "by asking whether the maxim of acting on it can be willed as a law" (p. 89). The free will faces a problem: it "must have a law, but because the will is free, it must be its own law. And nothing determines what that law must be. All that it has to be is a law" (p. 98). Thus, "if the action and the purpose are related to one another so that the maxim can be willed as a law, then the maxim is good" (p. 108).

On one strained reading of these words, all this should still be acceptable, even to a substantivalist: The only purely logical constraint on action, we might agree, is that it fit a consistent policy for living that one might adopt. But Korsgaard seems in these passages to be saying something far stronger: that if an agent adopts a maxim and the maxim passes this logical test, it is then satisfactory. Not every conceivable logically consistent maxim is good, to be sure, for some will have no appeal to an agent, and others will have their appeal but lose out to other maxims in the agent's reflective endorsement. But still, a maxim will be satisfactory, on this reading, if it satisfies both of two conditions: (i) it somehow encapsulates an impulse to action that the agent actually has and reflectively endorses, and (ii) it can be willed as a law.

According to this logically constrained reflective subjectivism, reflectively deciding to do something makes it the thing to do, so long as your policies for action are logically consistent. Is this Korsgaard's position? I'm not entirely sure; some passages in these lectures seem carefully crafted to avoid forcing such a reading.'0 The passages I have cited, though, do seem to commit her to this view, and we'll need to keep asking whether any plausible reading avoids it. Reading her this way does give sense to her rejection of substantivalism, and to her claim that the source of normativity lies in the agent's will. I'll be finding later, more- over, that her main argument seems indeed to require this reading. With

10. At one point she may be rejecting it: "Certainly I am not saying that reflective endorsement-the mere fact of reflective endorsement-is enough to make an action right" (ibid., p. 161). This still leaves it open, though, that all that's further required is full practical consistency. Korsgaard, then, could say this and still be a logically constrained reflective subjectivist.


some hesitation, then, I'll mostly read Korsgaard as indeed just this: a logically constrained reflective subjectivist.

Is this a bad thing? Much will depend on the power of the logic of acting to constrain the will to what's reasonable and morally decent; Korsgaard thinks these powers far greater than do I. Many maxims, we substantivalists think, can be willed as law which it would be a very bad idea to will- even if one in fact does. The binge alcoholic chooses badly, even if his policies for living have perfect logical form, and likewise with Caligula. If we can't derive something like the moral law as a theorem of practical logic, then logically constrained reflective subjectivism will de- clare maxims good that are repugnant. And in that case, we'll reject the view-unless we think it somehow forced on us, which nothing we've seen so far should lead us to think.

Reflective subjectivism goes beyond another set of claims that strike me as entirely acceptable. Reflective endorsement is, in a sense, inescapable; if you think about what to do and come to a conclusion, you have thereby reflectively endorsed your conclusion. And, we might say, you will be acting autonomously, if autonomy is what you exercise if you reflect fully on what to do. To reflect on what to do and come to a conclusion is an exercise of autonomy, and terminates in a reflective endorsement-to all this, who could object? And if I reflectively endorse, for instance, my impulse to have fun by swimming, I thereby regard its object (fun) as of value.

Reflective subjectivism, though, involves something far stronger: that fun, it follows, is truly of value. Reflective endorsement of an impulse makes its object good. Korsgaard calls her view an "ethics of autonomy" and says, "The source of the normativity of moral claims must be found in the agent's own will" (p. 19). "Autonomy is the source of obligation, and in particular of our ability to obligate ourselves" (p. 91). What do these words mean? An expressivist can agree that the will is the source of thinking oneself obligated; that to think oneself obligated is to reflectively accept a demand or requirement. It's quite another claim that the will is the source of validly being obligated. Is this last what Kors- gaard means by an "ethics of autonomy"?

Korsgaard offers this further gloss on her position: "Were it not for our desires and inclinations ... we would not find their objects good.... Kant saw that we take things to be important because they are important to us-and he concluded that we must therefore take ourselves to be important" (p. 122). Think how this might work with an example: say, the value of biodiversity. The passage begins unobjectionably: "Were it not for our desires and inclinations. .. we would not find their objects good." Thinking biodiversity good consists in reflectively desiring it; this is Korsgaard's expressivism, which I accept. The next part I find multiply ambiguous: "We take things to be important because they are important to us." This could just be read as the empty claim that because we regard


things as important, we take them to be important. If, though, it is to lead us to think that we ourselves are important as the source of all im- portance, it must mean this: as an agent, I believe, "What makes biodiversity important is the fact that I reflectively desire it." For this, gener- alized, is the claim that what makes things important is that they are "important to us," in the sense that we consistently and reflectively desire them.

But if this is meant as a theorem in the logic of acting, it is fallacious. From the premise,

We wouldn't think biodiversity important unless we reflectively desired it,

we can't derive the conclusion,

Biodiversity wouldn't be important unless we reflectively desired it.

The premise comes from Korsgaard's version of expressivism: that thinking biodiversity important consists in desiring it and reflectively endorsing this desire. This is a conceptual claim, a metaethical claim, a claim about the concept of being important. The second, though, seems to be a crucial claim of Korsgaard's; it encapsulates the position I'm calling "reflective subjectivism." This is a claim not about the concept of being important, but about what is important or isn't. Expressivists since Ayer and Hare have striven to impress on us that expressivism does not entail subjectivism. That humanity is the source of all valuing is fine, but does valuing make for value? An argument like this conflates the two."

Even, though, if Korsgaard is a reflective subjectivist, and even if she needs to be in order to derive the value of humanity, she may not be depending on this fallacy. The quick gloss on Kant that I have quoted is more or less an afterthought, and it follows a far more elaborate argument for the claim that we must value our humanity. I display this fallacious derivation of subjectivism from expressivism not because Kors- gaard clearly embraces it-it's unclear whether she does when she's not speaking quickly-but to stress the contrast between expressivism and reflective subjectivism.

IV. PRACTICAL IDENTITY

Whatever the Categorical Imperative turns out to be, the moral law is something further. It tells us to join in a Kingdom of Ends, a scheme that

11. As Nagel says in his commentary: "We do not make these things true by taking some kind of leap" (ibid., p. 208). Hare, whom I read as an expressivist (see my "Hare's Analysis of 'Ought' and Its Implications," in Hare and Critics: Essays in Moral Th1.inking, ed. D. Seanor and N. Fotion [Oxford: Oxford University Press, 1988]), has been at pains to

distinguish his position from any form of subjectivism; see, e.g., R. M. Hare, Freedom1I and Reason (Oxford: Oxford University Press, 1963), p. 199. The distinction is central to my own

views; see W'ise Choices, pp. 153-54.


accommodates each person's reasons with the reasons of others. As for just what these rules of accommodation are, Korsgaard is noncommittal; that's a question for elsewhere. Her core argument for the moral law deliberately leaves another loose end too, tied down in the following lecture: In her core argument, Korsgaard undertakes to show that if you act at all, you must value your own humanity. But a common misconception must later be discredited to show that if you value your own humanity, you must value the humanity of others. What work the core argument itself is doing will be one of my puzzles with these lectures.

The Categorical Imperative safely established, Korsgaard explains what still needs to be shown:

Any law is universal, but the argument Ijust gave doesn't settle the question of the domain over which the law of free will must range. And there are various possibilities here. If the law is the law of acting on the desire of the moment, then the agent will treat each desire as a reason, and her conduct will be that of a wanton. If the law ranges over the agent's whole life, then the agent will be some sort of egoist. It is only if the law ranges over every rational being that the resulting law will be the moral law. (P. 99)

What, then, is the "domain" of a law? The term might be read in either of two ways: First, the domain might be the set of situations the law applies to, its domain of application. Does the law command just me at this moment, in my actual situation? Does it command any reflective agent, at any time, in any possible situation? Or is its domain of application something in between? Consider the law, "Get yours and to hell with anyone else!" Thrasymachus accepts this with the widest possible domain of application: "No matter who you are, when, and in what possible situation, get yours and give no intrinsic heed to anyone else! " This won't take him far toward the Kingdom of Ends. In the passage above, though, Korsgaard seems to have something else in mind, what we might call the law's domain of concern. Does the law tell me to act for the sake of myself just now, or of all sentient beings at any time, or what? The wanton acts on the narrowest kind of law; the good utilitarian on the widest. Kors- gaard's examples favor this latter reading of "domain" as domain of concern, but we'll have to ask which reading fits her argument. (And another worry: if the "domain" of a law is its domain of concern, and if Korsgaard proves that one domain of my laws must be humanity in general, then we're done. We don't have the further worry that I might only need to value my own humanity. We won't then need the separate, Witt- gensteinian discrediting of private reasons that comes later.)

The argument Korsgaard seems to present as her central one draws on practical identity, "a description under which you find your life to be worth living and your actions to be worth undertaking" (p. 101). One of your practical identities, she argues, must be as a "human" in general, a


reflective chooser. From this she derives a requirement to value your humanity and obey the moral law. Her argument divides into roughly these three steps; she argues that:

i) Acting at all requires some practical identity or other; ii) Any practical identity whatsoever entails the practical identity

of a human, as a reflective agent; iii) A human practical identity entails valuing one's humanity, in a

sense that commits one to morality.

Korsgaard in this argument, I should remind you, doesn't distinguish valuing your own humanity from valuing humanity in general; later she argues that she needn't make this distinction.

First, then, acting requires a practical identity: "Unless you are committed to some conception of your practical identity, you will lose your grip on yourself as having any reason to do one thing rather than another-and with it, your grip on yourself as having any reasons to live and act at all" (p. 121). The term 'practical identity' suggests a focus on oneself, tying one's sense of self-worth to standards of conduct. That the sheer logic of acting requires anything like this might seem dubious. A connoisseur may act from pride in his connoisseurship, from his self- image as a connoisseur, but children seek out pictures from straight lik- ing, and an adult might visit the Art Institute in the same spirit. Not all motivations focus on maintaining a view of oneself as valuable, and a reflective agent might conceivably have no self-focused, esteem-driven motives at all. Without these motives he wouldn't, perhaps, be a normal member of our species-but with Korsgaard, as I read her, we're examining the sheer logic of agency. Korsgaard also gives the need for practical identity a more Kantian tone: "The reflective structure of human consciousness requires that you identify yourself with some law or principle which will govern your choices. It requires you to be a law to yourself" (p. 104). I have agreed that when the reflective agent stops to look at a picture, he's consistent only if his vague, implicit rationale fits some general, consistent rationale that he doesn't reject. Must he, though, "identify himself" with this law? He does, to be sure, in a very thin sense: he allows that that's the law for him to follow. Must he identify himself with a law in any stronger sense than this?

Perhaps, though, this thin sense is all that Korsgaard needs. Her rhetoric suggests to me a strong sense of identity, tied to one's sense of one's own intrinsic worth-and this, I agree, is crucial to us of our species. Her official definition of "practical identity," though, exacts only the thin reading: a "description under which you find . .. your actions to be worth undertaking." This seems fine: if you act, that commits you logically to the claim that under some description, your act is worth undertaking. Perhaps you even need some implicit view of what some such description is, and reflection can call that view into question. And


this commits you logically to a description of yourself: you are a being such that for you, acts of a certain kind are worth undertaking. The child eager to look at a picture book may little reflect, but logically she is committed to the claim that she is a being such that looking at pictures, in these circumstances, is worth doing. In this thin sense, she's clearly committed to "a conception of who she is which is normative for her" (p. 123).

We should ask, as we proceed, whether this thin reading gets Kors- gaard what she wants, or whether she needs the stronger reading, the reading tied to a sense of one's worth. Either way, if she can proceed to derive enlightenment morality, that will be important. The reading tied to self-esteem won't apply to every conceivable reflective agent, but clearly it applies to you and me, probably as men, or at least as men of a modern ethos.'2 If we ourselves are so inescapably committed to morality-if there's no consistent pride to be had without pride in one's moral de- cency- that is result enough to shake the world. A thin reading would ac- complish even more; it would meet the full ambitions of moral logicism, giving us a derivation that applies to any conceivable reflective agent.

V SELF-WORTH

On, then, to step 2: that any practical identity whatsoever entails identi- fying with one's humanity, that it entails a practical identity as a reflective chooser. One reason for conforming to any particular practical identity is that, as established in step 1, you need some practical identity or other. But this, writes Korsgaard, "is not a reason that springs from one of those particular practical identities. It is a reason that springs from your humanity itself, from your identity simply as a human being, a reflective animal who needs reasons to act and to live. And so it is a reason you have only if you treat your humanity as a practical, normative, form of identity, that is, if you value yourself as a human being" (p. 121).

One aspect of Korsgaard's argument will be controversial, but not with me. It is transcendental: it takes something we can't act without accepting, derives a consequence, and then embraces the consequence. "I show you that rational action is possible only if human beings find their own humanity to be valuable. But rational action is possible, and we are the human beings in question. Therefore we find ourselves to be valu-

12. See, e.g., Claude Steele, "Race and the Schooling of Black Americans," Atlantic 269 (1992): 68-78. I of course don't intend "men" to be confined to males; I need a term for our own, terrestrial species. Esteem-tied identity isn't, in Korsgaard's sense, a "human" universal, I'm saying; it isn't something that any conceivable reflective agent would have. Whether it is universal among men is an important question, closely linked to the question of which aspects of morality, as we know it in the wake of the worldwide diffusion of Euro- pean ways of thought, are universal among men. The capacity for esteem-tied identity must clearly be universal among men; whether it is realized in all cultures is an important question for anthropology.


able. Therefore, of course, we are valuable" (p. 124). Even, though, if the argument goes through as intended, its conclusion doesn't follow logically from its premises-that's the worry.'3 Mightn't it be that al- though merely acting at all commits us to thinking that humanity has value, in fact it doesn't have value? Korsgaard, though, say I, has every right to rely on such arguments. Suppose she is right, and in settling whether to act, I've settled whether to believe humanity valuable. I'll then act and voice the conviction to which acting commits me: Humanity is valuable. What other conceivable access can I have, after all, to the question of whether humanity is valuable, but to reflect on what to do? The value of humanity or its lack isn't a feature of nonnatural space, glimpsed by intuition. Thinking humanity valuable, if Korsgaard is right, is an inseparable part of thinking what to do and why. Whether to think humanity valuable is just the question, whether to value humanity.

Must I, then, identify myself practically as a human, as a reflective chooser in general? Proceed first with the strong reading of a practical identity as something I need, that I must act from a conception of myself that I tie to a sense of my intrinsic worth. I tie my self-esteem, say, to a conception of myself as a brave, unflinching Achaean warrior; I value being that way. Acting as befits an Achaean is one particular way of being a reflective chooser-as living up to any particular, contingent identity entails being a reflective chooser. Korsgaard's argument seems to be, then, that in valuing myself as a brave Achaean, I'm committed to valuing myself as a reflective chooser. And so the more general form of the argument seems to be this; in a Kantian spirit, I'll put it in terms of imperatives you address yourself. The premise, "Value being a brave Achaean!" has the form, "Value X!" Next we have "X entails Y. " being a brave Achaean entails being human. The conclusion is "Value Y!" The validity of Korsgaard's particular argument depends, it seems, on whether this general form fits and is valid.

Why, though, couldn't I think of reflective choice as a burden, only mitigated by some admirable way that people like me handle it? As an ascetic, to take a parallel, I might value being someone who resists the cravings of the flesh. I can't be someone who does that without having cravings of the flesh; am I then committed to valuing having these cravings? Amn I committed to valuing this, though I share this feature with the sensualist I despise? Return, then, to me as an Achaean: I value being someone who deals with my reflective choices valorously; couldn't I still disvalue the sheer state of being a reflective chooser?

The question is logically tricky. Perhaps the form of argument I've displayed is valid, but much depends on the precise form of what I value. Suppose it's right that if I value something, I must value anything it entails. What I value as an ascetic, though, could be this: resisting any

13. Cohen worries inconclusively about this transcendental step (Sources, p. 186).


cravings I might have. This doesn't entail having any cravings at all, and so it doesn't commit me to valuing having cravings. If, on the other hand, what I value is resisting cravings in large numbers, then this, the argument-form tells us, commits me to valuing having those cravings at least as an indispensable part of a valuable whole. Perhaps I must value my cravings as opportunities to resist-though that's backhanded valuing. Transfer this logic chopping, then, to Korsgaard's crucial argument: Whether my particular practical identity as an Achaean warrior commits me to a practical identity as human will depend, it seems, on the exact form of the particular identity. I could value making whatever reflective choices I might make in a way that befits the brave Achaean. This doesn't entail making any reflective choices at all, and so doesn't commit me to valuing my "humanity." There is, then, a form my practical identity could take that wouldn't commit me to valuing my humanity. Is this the form the warrior's identity is likely to take? The question seems ridiculous: the warrior on the field just assumes that he makes reflective choices, like anyone else, and against this background, he values making them valorously. In this, does he value making reflective choices as such, something even the helot or tradesman does? It doesn't seem that his thoughts commit him to this. Whether to put his values in a form that carries this entailment is a question that hasn't arisen for him.

Move on, then, to step 3. "It follows from this argument that human beings are valuable. Enlightenment morality is true" (p. 123). Suppose the warrior does value his humanity, in the technical sense that he values being a reflective chooser as such. Being human, I've been saying, doesn't by itself commit him to this; he can be a reflective chooser whose pride and the like are all invested elsewhere. But reflective choice seems a plausible point on which to focus some of one's sense of self-worth; who would want not to be a reflective chooser? If pride in being a reflective chooser as such entails pride in being morally decent, that in itself will be an important theorem.

Korsgaard thinks that it does. "To value yourself just as a human being is to have moral identity, as the Enlightenment understood it" (p. 121). As I mentioned, though, a crucial aspect of morality she leaves aside for later treatment: that if you must value your own humanity, you must value everyone's. The Golden Rule isn't what's now been established. This aside, then, what does it mean to "treat your humanity as a practical, normative, form of identity" (p. 121)? The question isn't what these words would suggest in ordinary rhetoric, but what has been proved by Korsgaard's argument: universal concern aside, what must any person do on pain of inconsistency in plan of life? Korsgaard's later argument that reasons must be shared does have clear import for morality, if it works. My puzzle is whether this later argument does all of her work, or whether the conclusion that I must value my own humanity has some force of its own in bringing me to the moral law.


Here I'm genuinely puzzled. If valuing my humanity is taking pride in being a reflective chooser, how does that constrain what I do, what I reflectively choose? Perhaps I must nurture my powers of reflective agency. There's still the worry that I might take pride just in being a reflective agent right now, not in the future-and there's not much I can do to nurture my reflective agency at this very moment except to choose reflectively, in any way whatsoever. But combine this requirement with Korsgaard's later arguments that my concern can't be confined to myself right now, and the requirement to nurture reflective agency will have some teeth. Still, if that's all we must do as human beings, then enlightenment morality is far too narrow: we'll need to oppose pain and seek fulfillment and enjoyment only when they affect our powers of reflective agency.

Korsgaard herself, of course, finds far more in "valuing humanity," so that anyone who values humanity will indeed oppose unbearable pain and favor innocent pleasures and fulfillments. Valuing humanity as Kors- gaard pictures it, then, is surely an excellent thing-but if to fail to do all that's in this package is to have inconsistent maxims, we need a systematic characterization of the package, and a proof of how consistency in acting requires all that's in it. One systematic view would be what I have called logically constrained reflective subjectivism. But whether or not Korsgaard embraces this position, it isn't a view that follows from the reading I'm now examining. I value being a worthy Achaean in my choices, but I needn't value the way I'd be if I chose in some other way. What, I ask, if when battle called, I reflectively chose to turn the other cheek? I don't now think that my doing so would make pacifism in any way good. My humanity I would retain if, anachronistically, I choose as a Quaker tradesman, but in what manner must I, the brave Achaean, value this?

VI. ACTIONS WORTH UNDERTAKING

I have been examining Korsgaard's argument on a strong reading of "practical identity" as tying standards of conduct to a sense of my worth. But Korsgaard's initial definition strictly read, I noted, offered a thinner version: your practical identity is "a description under which you find your life to be worth living and your actions to be worth undertaking" (p. 101). Acting with consistency does involve finding your acts to be worth undertaking; that seems unobjectionable. And your feelings toward your life and your actions must fit with this, we might say, or they will have a content that is inconsistent with your plans of action. Not every philosopher would allow this, but it is a rich and promising line of thought that may well be right. From these materials, could we get a derivation of the moral law in some crucial aspect?

The target of the derivation is to conclude that, so long as I find my actions worth undertaking, then I must cherish my "humanity" as one


of my practical identities. A practical identity as a human will now involve finding acts of a certain kind to be worth undertaking. What kind? The ordinary suggestions of the term "humanity" are clear in some ways and contestable in others, but again, we are seeking a systematic characterization that plugs into a logical derivation.

Again, one systematic candidate is logically constrained reflective subjectivism. As an Achaean warrior, I find unflinching battle worth undertaking. I come to realize that, at base, it is worth undertaking because it is what I most deeply and reflectively want, because my thirst for battle proceeds from a practical identity. Realizing this, I note that the Quaker tradesman's goals of prosperity within the constraints of peaceful, honest dealing spring likewise from his practical identity. I haven't yet concluded that I must value his humanity; that comes later. But perhaps I come to realize that what I reflectively value is of value precisely because I value it- or because I value it in a way that proceeds from a deep conception of myself. Consistency then forces me to think, at least, that if my deep conception of myself were far different, if it were that of a Quaker tradesman, then my prosperity under constraints of peaceful, honest dealing would be of value. It's a further step, perhaps, to say that his prosperity is actually of value, but I have reached an important and surprising conclusion, one that moves me part way to the Kingdom of Ends. I know now that were I such an honest man of peace, my prosper- ing in peace and honest dealing would be of value. "Valuing my humanity" now means valuing things on the grounds that I, as a human, choose them.

This is a form of subjectivism, confined, perhaps, to consistent, reflective desires that proceed from a deep conception of oneself. This view has its attractions; it may be the best view to take of what, at base, renders anything of value. I'm not at all sure; my previous worries are still worries: If an argument that comes later doesn't work, if we can't show that concern for like goals of others is logically mandatory, the view will allow knaves and fanatics whose identities are deep enough. The view raises questions of coherence: can I see having fun, say, as valuable and reflectively endorse that view, and still think that, really, fun is valuable only because of my reflective endorsement- or because it stems from a deep, normative view of myself? Our question now, though, isn't whether subjectivism is acceptable in this form, but whether it is logically mandatory. Does finding your actions worth undertaking commit you to such a view, on pain of sheer practical inconsistency?

I think it is clear that it doesn't. This form of subjectivism no more follows from expressivism than does any other form. The Achaean warrior tells himself, "Battle unflinchingly, with a deep sense that it's worth doing!" He may even hold his sense of worth responsible to this injunc- tion: "View myself as worthy if I battle unflinchingly, and unworthy if I don't." Consistency doesn't require him to generalize this to "Do what-


ever I view reflectively as worth doing!" The same goes for "Do whatever I view reflectively as worth doing, so long as doing so proceeds from standards of worthiness that I reflectively endorse!" He can consistently reject these injunctions as they apply, say, to the hypothetical case of having the ethos of a Quaker tradesman.

We have canvassed, in this section and the last one, two readings of "valuing your humanity": roughly, pride in being "human" and reflective subjectivism. The first is some positive attitude, pride or something akin, in being a reflective chooser. The second is favoring the goals one reflectively adopts because or on the grounds that one reflectively adopts them. On the pride reading, none of the three steps of the argument has strictly logical force. (i) Thinking you have reasons doesn't by itself commit you to thinking you have reasons that focus on your qualities as a chooser. (ii) Pride in using certain standards of choice needn't commit you to pride in being a reflective chooser in general, and (iii) pride in being a reflective chooser doesn't commit you to anything approaching the moral law. On the reflective subjectivist track, the steps go through, apart from the derivation of reflective subjectivism itself: (i) thinking you have reasons- reflectively endorsing certain considerations -doesn't

commit you to thinking that this reflective endorsement itself constitutes a reason.

Is there a more promising reading of Korsgaard's argument from practical identity? I hope so, but I haven't found it. Clearly in her treatment of "practical identity," Korsgaard brings to our attention important phenomena to explore, and things she says about it are intriguing and suggestive. Motivations that stem from a sense of one's worth and focus on what one does are powerful, I agree.'4 If you have these motivations and reflectively endorse them, you thereby regard the question "Who would I be then?" as specially deep and weighty. My worry is with Korsgaard's further claim: that a reflective chooser, taken purely as a reflective chooser, must so view matters. I'm open to the transcendental form of her argument: acting does, I agree, carry logical commitments- but I don't see how to argue that these by themselves carry us far toward morality. I'm open to arguments that we ourselves must choose morally, as normal members of our species or of modern culture, on pain of inconsistency in our rationales for action. I'm open to Socratic forms of argument that show this: resentment, for instance, which Korsgaard in- vokes, may well commit one to important aspects of morality. It may be inconsistent, for instance, for egoists to resent. Korsgaard, though, hasn't explicitly shown us how to construct such a more modest, Socratic argu-

14. Geuss worries that these motivations might be egoistic, and Korsgaard replies

(ibid., pp. 242-51). They are, to be sure, self-focused, but they aren't, I would think, what we usually have in mind when we talk of self-interest, and it isn't a clear intuitive datum that

concerns of integrity aren't properly moral concerns.


meant from practical identity, and it would require ingenuity at the very least, to do so. Her own ambitions, in any case, seem to reach farther.

VII. PLEASURE AND PAIN

The logic of acting doesn't entail "valuing your humanity," not in any sense that includes the lofty connotations of the phrase. It entails no requirements to be decent or even sensible. The promise, then, of a more solidly based alternative to substantivalism is vain. In particular, the weak pleasure principle, or something in its vicinity, is a principle it would be crazy or perverse to reject. Hence if no plausible account of the value of pleasure is a theorem of practical logic, then the sheer logic of acting isn't enough, by itself, to guide us in how to live. Practical logicism fails-or so I have been maintaining.

Korsgaard, though, offers her own account of what's wrong with pain and what's good about pleasure, when pain indeed is bad or pleasure is good. "If you don't value your animal nature, you can value nothing" (p. 152), and valuing your animal nature includes a due regard for pleasures and pains. This, I take it, is a compressed, suggestive statement of a theorem of practical logic; Korsgaard I read as a logicist with regard to why pains are bad and pleasures are good, when they are.

If her treatment of pleasure and pain works, then practical logicism will be a far more credible view than I have allowed. By practical logicism, I mean the view that all demands of reason on how to live stem from the facts and practical logic alone. (The relevant facts crucially include the psychological facts of what our impulses are and which ones we reflectively endorse.) When a way of life would be crazy, perverse, unduly self- denying, or unduly self-centered, then undertaking to live that way is incoherent; that is what the practical logicist claims. Such a claim is in- credible if an extreme asceticism, a blanket scorn for all innocent pleasure, would be logically coherent in a plan for life. But Korsgaard sketches an argument that such a plan wouldn't be coherent. If her argument works, then some of the substantivalist's strongest examples turn out after all to be friendly to Korsgaard's position.

Though it is fairly brief, Korsgaard's treatment of pain and pleasure is intriguing and subtle, insightful and even profound. I can't do justice to it, but I'll summarize as best I can her view and my grounds for doubt.

Pleasure, argues Korsgaard, is "the perception of a reason," just as is pain; pain, she says, is "your perception that you have a reason to change your condition" (p. 148). I have reason not to bang my thumb with a hammer: it injures my thumb, and this to some degree threatens my survival. The pain in my thumb is a perception of this reason; I feel the badness of the injury.

Pain isn't always bad, though if I'm not misfeeling things, then that which pains me is bad. To grieve a loved one, for instance, may be in no respect intrinsically bad, so long as the grief isn't beyond bearing; I grieve


veridically in that the loss that I grieve is itself bad. (Thus the weak pleasure principle needs revision at the very least; pleasure, say, in a snuff film is in no way good.) What, then, of misperceptions of goods or bads, of pain in a phantom limb or pleasure in a Diet Coke? Does it follow that when pain is misfelt it isn't bad, or that the contrived mistasting of a drink as sugary is a pleasure of no value however small?

Pain itself, Korsgaard answers, can be bad apart from the badness of what pains us, and enjoyment itself can be good. When, then, is enjoyment and not just its object intrinsically good? Hedonists, of course, think always. An alternative view is that enjoyment is the perception (or misperception) of a reason that is independent of the perception-and this at first appears to be what Korsgaard is saying. But both views are wrong, she tells us.

Here I lose the thread, and I look forward to seeing a fuller working out of Korsgaard's treatment. The deep point, she tells us, is this: "It is a pain to be in pain" (p. 154). We can feel averse to being in a state of felt aversion: When I'm horrified at a photo of concentration camp victims, I'm pained at their plight, but I'm also pained at my state in being pained. "It is bad to look at a picture of something bad" (p. 155). The reason that pain, then, is nearly always bad is that "the creatures who suffer from it object to it" (p. 154). But how does this position work? We have already learned that paining me doesn't make a thing bad. Rather, my pain, if veridical, is correctly perceiving that it's bad. So even if I'm averse to feeling pained by the photo, this second-order aversion is veridical only if my seeing the photo and feeling pained by it was, for some reason, itself bad. Not all being pained is bad, as shown by the case of grief. What, then, makes the difference between being pained in a way that's good and a way that's bad?

We can't rely on a substantive sense of plausibility to answer this; our answer must stem from the logic of what to seek or shun. If grief is good but horror bad, why is that? Pleasure and pain, Korsgaard tells us, are "expressive of the value that an animal places on itself" (p. 152), but we need to understand when this valuing is something to which an animal (or a human) is committed by the logic of what to do.

The reasons, Korsgaard seems to say, come down to survival. "A living thing is a thing for which the preservation of its identity is imperative" (p. 152). No one will doubt that survival greatly matters, but can that be a matter of logic, the logic of what to do? To value anything you must of course exist-but only for the moment, and injuring your thumb doesn't threaten that. The pain might mean, in a way, that you are "not yourself," but the same might go for a drop of Dionysian ecstasy. There's a difference, perhaps, but we need to learn more. And anyway, even to wish you were dead you must be alive, but that doesn't make such a wish logically incoherent. The value of survival, then, seems substantive, and not a matter of sheer logic.


"A living thing is so designed as to maintain and reproduce itself," and Korsgaard at this point mentions natural selection (p. 149). But this can't logically ground the value of pleasure. In the first place, natural "design" is for survival only as needed for reproduction; the salmon is "designed" to reproduce but not to survive after that. But the crucial, familiar point is this: our "design" may causally explain what we end up doing, but what to do is a different question; its answer doesn't follow logically, all by itself, from how natural selection "designed" us. The lion is "designed," when he takes over a pride, to kill the cubs his predecessor sired, and this motivation promotes his genetic reproduction; the lioness is sooner fertile and turns sooner to nurturing his own cub. Human males, alas, may under certain circumstances have analogous impulses, with a similar evolutionary shaping of emotional mechanisms.15 In a strictly positive, Darwinian sense, our "end" is reproduction, but that doesn't settle whether to reproduce.

None of this precludes another kind of argument that Korsgaard might offer: she might convince us Socratically that, in an extended sense, it is survival that matters, and that this value stands behind large parts of the demands of reason on how to live-including the demands of morality. She might show that survival, including survival as an animal, somehow accounts for how it's reasonable to live and how it isn't, for what's plausible in this realm. Working out her intriguing view of the value of pleasure and pain would be an important part of such a demonstration. I'd be amazed if such an argument worked, but I haven't ruled it out. My point is that if this succeeded, it would be a substantivalist success; it would trade on our sense of plausibility in matters of how to live. It wouldn't show that, as a theorem of practical logic, "If you don't value your animal nature, you can value nothing" (p. 152).

VIII. PRIVATE AND PUBLIC REASONS

In her fourth and final lecture, Korsgaard rules out the autonomous knave. "In the last lecture . . . I took it for granted ... that valuing one's humanity amounts to having what I called 'moral identity'." But her derivation of morality has gone through, she allows, "only if we can conclude that valuing humanity in your own person somehow ... involves valuing it in that of others" (p. 132). To philosophers steeped in current projects of deriving morality and fending off egoism, this may seem the chief gap she needed to span in the first place. "Consistency can force me to grant that your humanity is normative for you just as mine is normative for me," Korsgaard has a critic saying. "But it does not force me to share in your reasons, or make your humanity normative for me. It could still be

15. See Martin Daly and Margot Wilson, Thle Trutth about Cinderella: A Darwinian View of Parental Love (London: Weidenfeld & Nicolson, 1998), on child abuse. It is important to state their findings carefully, since as they stress, most stepparents are loving and nurturing.


true that I have my reasons and you have yours, and indeed that they leave us eternally at odds" (p. 134). In Korsgaard's view, though, the critic will think this only because he harbors a misconception, a distorted view of the nature of reasons. Cast off the misconception, she may be saying, and we'll see that the argument for moral identity worked all along, that it took for granted only what all must grant.

Hobbesians, neo-Kantians, and others assume that reasons are "es- sentially private." This, Korsgaard thinks, would rule out any satisfactory derivation of morality whatsoever, her own included. But as Wittgen- stein's private language argument allows us to see, reasons are "inherently sharable"; they "are public in their very essence" (p. 135). "What both enables us and forces us to share our reasons is, in a deep sense, our social nature" (p. 135). Acknowledge "that our reasons were never more than incidentally private in the first place" (p. 136), and the specter of egoism dissolves in the light.

No one, in my experience, knows quite what Wittgenstein's private language argument is, but everyone knows that it works: in some sense, we all know, there can be no private language.16 When I work out my reasons, I can do so only in a language I can share. Grant all this; has egoism vanished?

I'm not sure; people differ in their accounts of what the private language argument shows. A frequent view of the matter, however, is this: when I think of such things as visions and visceral sensations, I can do so only in a language in which I could, in principle, talk with others. I can't name a sensation unless I could teach you that name and its meaning. Applied to reasons, this will mean that when I think through my reasons to do one thing and avoid another, I must do so in a language that I could at least teach you to share.

Suppose, then, I think of my reasons in my native English, a shared, public language that you too understand. Does this mean that we "share our reasons" (p. 135)? In one sense, it certainly does: if you think you have reason to enslave me if you can, you can tell me so, if you choose, and if I think I have reason to try to prevent you, then I can tell you that. What morality needs, though, is reasons we share in a different sense: roughly at least, that you have genuine reason to try to enslave me only if I have some reason to submit. (What morality requires more precisely is perhaps, as Korsgaard suggests, an ideal of a "Kingdom of Ends": that our reasons all feed into a scheme of accommodation that we all have reason, all told, to join.) From our ability to talk together about your

16. Korsgaard herself offers a form of the argument that doesn't seem to work: Mean- ings and reasons are normative, one must be able to get things right or wrong, and norms require both a legislator and a citizen. "It takes two to make a reason." But as she goes on to say, "Here the two are the two elements of reflective consciousness, the thinking self and the acting self" (Sources, p. 138). It seems, then, that within a single human skin, the two of me can do it all. If reasons stretch from me to you, some further argument is needed.


reasons and about mine, can we derive reasons we share in this strong sense?

A bereaved Kwakiutl chief sets off for mass slaughter in the village across the water; "Shall I mourn or shall they?" 17 We find his loss no reason to slaughter the innocent, but his talk of reasons is public enough. His victims understand, though they see every reason to resist-no kingdom of ends in this sad tale! I myself (perhaps unlike the victims) think the chief wrong about what he has reason to do, but I would need some demonstration that he or his victims botch the sheer logic of acting. Why must reasons they could claim to each other to have all be reasons they must think they have in common? "I have my reasons and you have yours" is a statement in public language, however unsavory. "Nagel char- acterized the egoist as a practical solipsist and of course he is right," Korsgaard pronounces (p. 143). But he isn't right if this means that an egoist can't explain himself if he likes-or if he is, that needs further argument.

There may well be something to be gleaned, philosophically, from the pressures we face to share our reasoning. Korsgaard here identifies an important topic. Notoriously, an egoist often will deny his own theory. If we all were consistent egoists, then, would we all lose the mental resources to think egoism through? Perhaps, but I don't know how to show this, or how to disprove it. And what would follow if this were so; is there a transcendental refutation of egoism to be had along these lines? Again, I don't know. We do have public languages with words for reasons, though the Kingdom of Ends on Earth is nowhere at hand; does this all by itself entail that we are inconsistent in our thoughts of what to do? How would an argument that showed this go?

You can obligate me, Korsgaard argues, because you can "get under my skin" and "intrude on my reflections" (p. 136). The egoist retorts that, as Korsgaard puts it, "I am merely describing a deep psychological fact-that human beings are very susceptible to one another's pressures.... But nothing I have said so far shows that we have to treat the demands of others as reasons" (p. 141). Bullies, as Williams observes, don't give me reason to obey, even if I can't help myself (p. 217). In re- sponse, Korsgaard depicts how we can reason jointly on what we're to do-say, on arranging an appointment. "Why shouldn't language force us to reason practically together, in just the same way that it forces us to think together?" (p. 142). Again, though, we need proof that it does. We can reason together, and often we do, but does anything force us that precludes egoism? 18

Korsgaard does suggest how deeply difficult it would be to be a con-

17. My memory of an incident recounted in Ruth Benedict, Patterns of Culture (Bos- ton: Houghton Mifflin, 1934). Similar practices were known in Africa.

18. See Geuss's commentary (Sources, pp. 197-98).


sistent egoist. For one thing, resentment has to go. To this consideration two responses are possible: like Hume at times, we can think that the reasonable view of matters is psychologically impossible to maintain. Or we can exploit such facts as that the self-proclaimed egoist won't forswear resentment, and use them Socratically to reason him out of egoism: since the hurtful, selfish acts of others are worth resenting, he comes to see, egoism isn't a view to hold.

This last, however, is the approach of a logically constrained substantivalist, not of a moral logicist. The moral logicist thinks that requirements of consistency alone, in one's plans for how to live and act, entail the moral law, that moral requirements are inescapable in this very strong sense. One of Korsgaard's initial complaints with "substantive moral realism" was that it preaches to the confident-and to be sure, some forms of it do. Moral substantivalists, though, have other resources besides an initial, unshaken confidence in the whole of morality. (Think, e.g., of Sidgwick's long treatise.) Substantivalists can look for deep, under- lying principles that are plausible, and aim to show that no coherent alternative to morality remains plausible on examination. Korsgaard's initial complaints seemed directed at this kind of "long, hard slog" substantivalist too; even this deep substantivalist won't demonstrate a kind of inescapability you'd recognize if you lacked the right sense of credi- bility. By the end of her argument, though, Korsgaard's own road to enlightenment morality is paved with the same brick. To my eye, this isn't a defect: moral logicism doesn't work, and so any philosophical treatment of moral demands must make appeals, in the end, to our sense of plausibility. Korsgaard may manage this better than the substantivalists she attacks-we'd have to see-but this is the game that she too must play. It isn't the sheer logic of what to do that precludes egoism, though it is, perhaps, the implausibility of some of the things an egoist is committed to accepting.

Korsgaard's lectures are full of insights that could fit well into a moral substantivalist's way of thinking, insights I have no further space to explore. My complaints have been with moral logicism: her form of moral proceduralism, Korsgaard seems convinced, answers the normative question with a force that no moral substantivalist could match. In- stead, I have argued, proceduralisms and substantivalisms alike must rest on more than the sheer logic of what to do. Accepting any coherent view of the sources of normativity requires finding some things credible and others not. No surprise, perhaps-but even in the humanized form that Korsgaard gives him, Kant thinks otherwise.

Gibbard on Korsgaard

Documents