The Roots of Reasons.
--e. e. cummings, "pity this busy monster, manunkind"
Normative reasons for action are considerations in favor of doing something.
When a desire or impulse comes over us, we can coherently ask whether it provides us with any normative reason to act to satisfy it. Some desires we classify as reasons to pursue the objects of desire; others we discount altogether. I think a desire to hear a concert or visit with friends is some reason to go to the concert or pay the visit. I don't think odd whims--the passing urge to stick my finger into a gooey substance, or the vertiginous urge to fling myself from a balcony--are considerations in favor of doing those things at all. We also reject objectionable desires. Should I feel a hateful desire to attack someone or cause someone pain, I would take that as no reason to do so, even in the absence of countervailing reasons. Some regard the urge to loaf as no reason to loaf; others may see it as a reason, but one that is overridden by other reasons on many occasions. Some see the desire for revenge as no reason at all to act vengefully. A Christian will not say, "My longing for vengeance is one reason to strike this person, but on the other hand it's outweighed by the reason I have to turn the other cheek." She will think it no
reason at all, but just an impulse to be resisted.
Why do we systematically regard some desires and impulses as the roots of reasons, and others not? Can these classifications be correct or mistaken? And does our assessment of our own desires and impulses itself play a role in making some desires normative reasons and others not? These are important questions for a theory of normative reasons.
Christine Korsgaard, in The Sources of Normativity,(1) develops a theory of normative reasons for action that takes a provocative position in response to the questions of how we regard our desires and impulses and what connection there may be between our seeing reasons and having them. (Of course, Korsgaard is primarily interested in answering a question about the normativity of morals. "what justifies the claims morality makes on us?" (9-10). But she proposes an account of the sources of all normativity, including and especially the sources of normative reasons.) The theory says we assess our desires and other impulses to act in light of practical conceptions of our identities, and then legislate reasons for ourselves on their basis. So our assessment and endorsement of a particular desire at least in part determines whether that desire provides us with a reason. In legislating for ourselves we create our reasons. And in the process of giving ourselves particular reasons, we also generate reasons to respect our own nature as reflective beings.
This account has considerable appeal. It resonates with those who think that at least some of our impulses provide us with reasons to act, but who doubt that the mere having of an impulse or desire, as such, counts in favor of acting. It elaborates the attractive notion that norms are not imposed from outside the self but are created by it. And it appeals to those seeking an alternative to the neo-Humean picture of agency as nothing more than being pushed about by conflicting psychological forces.
I shall criticize two aspects of this theory. First, I shall question whether it is, as it purports to be, a voluntaristic theory, in which we make an impulse a reason by giving ourselves a law. Secondly, and at greater length, I shall raise some doubts about the argument intended to show that if we are to have any practical reasons, we must value ourselves as reflective beings. Read in one way, the argument fails. Read in another way, it may work, but then it further undermines the theory's voluntarism.
2. The Korsgaardian View of Reasons
Korsgaard draws our attention to the fact that we human beings have a consciousness that is by nature reflective. We do not merely have impulses and desires that incline us to act, and perceptions that incline us to believe, but in addition we have the capacity to examine those impulses or perceptions and decide whether to act on them or believe on their basis. In reflection we step back from our inclinations to consider whether they are reasons for action. Then we endorse or reject them. Indeed, Korsgaard claims, our reflective consciousness requires us to do so. Only when we act on impulses that we reflectively endorse do we act for reasons, or even act at all. To be pushed hither and yon by impulses within us is not to act. Human agency comes into being with the "I" that can ask "Shall I act on this impulse or that one?" and then can decide.
Why do I endorse or reject any given desire? On Korsgaard's account, endorsement or rejection is mediated by a conception of my practical identity, a description of myself under which I find my life worth living and my actions worth undertaking (101). Apparently, in directing my reflective scrutiny upon my impulse, I consider whether it is the sort of impulse someone would act on who had my practical identity of daughter, professional artist, Chicana, or political conservative, to take some examples at random. Perhaps I also consider such descriptions as musical, kind, athletic, or witty, if I value myself under those, although Korsgaard does not give such examples. If acting as this impulse directs me would be consistent with my being such a person, then I will see reason to act this way. If acting this way is forbidden by one or another of my practical identities, then not only do I have reason not to act this way, but I am obligated not to.(2) To act in a way that undermines my practical conception of my identity is in effect to destroy my sense that my life is worth living and my deeds worth doing. This may not happen every time I so act, or in an instant, but it will happen eventually. Thus my practical identities determine my reasons (101, 129).
From the practical perspective, I must act under the idea of freedom, as Kant puts it. This freedom is manifested in the reflective structure of my consciousness: I can step back from my impulses and consider whether to act on them, and in so doing I am not under their control (or at least, so I conceive myself). But my free will cannot choose at random; it must be governed by some law, although of course not a causal law that renders my choice the effect of some impulse. Kant's solution, echoed approvingly by Korsgaard, is that my free will is governed by its own law, and furthermore this is a law it gives itself. The thinking self makes a law that commands the acting self (104). Because I can distance myself from my impulse and decide whether to be governed by it, I am required to be a law to myself: to identify myself with some law or principle that will govern my choices. The law may have any content whatever so long as it has the form of a law, which for Korsgaard means that it must be universal over some (possibly limited) domain. The particular domain will be set by the relevant practical conception of the agent's identity. So the law of a free will, for Korsgaard, is roughly this: Act only on that maxim that you will to be a law for all [Phi]'s, where [Phi] stands for the relevant practical identity, for example, brother, friend of Jim's, citizen of Spain. When I endorse an impulse to act, I make the maxim of so acting a law for myself as a [Phi]. My thinking self, because of the reflective structure of my consciousness, has (legislative) authority and not merely power over my acting self. This is the source of normativity. My practical conceptions of my identity show me what it would be a good idea for me to do (107), but do not yet make those considerations normative reasons. "Of course we discover that the maxim is fit to be a law; but the maxim isn't a law until we will it, and in that sense create the resulting value" (112).
So to sum up, on Korsgaard's view I take an impulse to be a reason for me to act, and I make it my reason to act, when, on reflection, I endorse it. First I identify with it, which is, at least in part, finding it to be connected in the right way with one or more of my practical identities. In endorsing the consideration as a reason, however, I also make a law for myself to act in this way. The normativity I thus impart to my impulse is imposed by my will, in that I give myself this maxim as a law. I have the authority to give myself a law, in part because I have power over myself, including power to punish myself with painful emotions should I disobey. Desires, then, are not automatically normative for me. I make some of them normative by means of my practical identification and self-legislation. The others are not reasons.
Korsgaard does not stop here. While most of the practical conceptions we have of our identity are contingent and vary from one person to the next, there is one description under which we will all value ourselves, of necessity, if we value anything. This is as reflective beings. Anyone who values anything at all, on her view, will consequently value his or her own reflective nature (121-25). Since we all are reflective beings, this is a true conception of our practical identity.(3) Thus, there is one consideration we must treat as a reason, one we should be mistaken to omit. And it is a foundational reason, in two ways. First, our other practical conceptions of our identity must be brought in line with it when there is any incompatibility. Second, it is the condition of all our other practical identities, and so all our other reasons. The normativity of the reasons we have to philosophize, to relax with our friends, and even to eat and sleep depends upon the normativity of our status as reflective beings.(4)
3. Making Reasons or Finding Them?
The idea that we identify with some descriptions of ourselves, and with the desires and other impulses that fit those descriptions, in such a way that these things impart meaning to our lives, is a powerful one. The loss of a valued identity is a grave loss indeed, one to be feared and avoided. And if I say that I acted in a certain way because this is who (or what) I am, that sounds like a reason. However, a question arises about this.
Korsgaard's basic account of reasons for acting is both psychological and voluntaristic. An impulse becomes a reason when we identify with it on reflection and will its maxim in a universal form. On this account, when we reflect, do we find reasons that are already there, or do we make them by our will? According to Korsgaard, someone's practical identity (as a Mormon or a judge, for example) is enough to show that something would be a good idea to do, but not enough to make it normative. For that we need autonomy--self-command. I propose that a genuine practical identity of the kind Korsgaard describes is at times such a life-shaping psychological feature that it is quite enough by itself to give us reasons to act. Self-legislation is unnecessary.
I do not mean to deny that we ever give ourselves commands. "Get out of bed!" I say to myself on some mornings. Sometimes my acting self even obeys. Nor do I deny that when we give ourselves commands they are universal in form. I will grant this for purposes of the argument. The content of my morning injunction to myself may well have to be "Get out of bed and prepare your lecture, as anyone with teaching responsibilities must do when the time comes." Perhaps every self-command is tacitly universal in this way. Nor do I deny that reasons are inherently general in some important way. What I do not see is how commanding myself, or even willing, makes a reason of a consideration rooted in my practical identity that was not a reason before. If I have reason to get up and prepare my lecture, I have that reason even if I will nothing, even if I sleep till noon without a moment's reflection.
Conceptions of our practical identity are typically contingent, and often come about without any choice on our part, as Korsgaard notes. I happen to be the daughter of certain parents, the mother of a particular child, a certain person's neighbor, even a fan of Gilbert and Sullivan, all through no choice of my own. These things are not and never have been up to me in the way that some practical identities are (being married, for example). And many of these things could have been otherwise. I could have produced a different child, had a different neighbor, hated light opera. At times, practical identities are very deep, however, even those that are not chosen and those that could have been otherwise. It is not a deep part of who I am that I am Terry's neighbor, but it is a deep part of who I am that I am the mother of this particular child. It is very important to me that I be a good mother to this child; and should I fail or cease to be one, I might well experience this as a partial disintegration of my self. This fact about me determines, as Korsgaard says, what I will take to be a reason: among other things, I will take as a reason my impulse to act in a way that, by my lights, is required for me to be a good mother to my child. If I think that to be a good mother I must teach my child not to steal, I will take the desire to do that as a reason. Now, on Korsgaard's account a certain procedure is necessary as well if I am actually to have a reason to teach my child not to steal: I must reflect on the impulse and identify with it in such a way that I give myself a law. Now, in general it is a good idea to reflect on my impulses. But it is hard to see how, in endorsing the consideration and commanding myself to act on it, I make it a reason. The reason was there all along. If teaching my child not to steal is necessary if I am to be a good mother as I conceive of it, and I care so deeply about being a good mother that to cease to be one would threaten my very identity, then I have a reason to do it before I give myself a law to that effect. The reason is rooted in my practical self-conception, which was present, and in part made me what I was and am, before I reflected, willed, or legislated. I do something by endorsing the impulse: perhaps I make my action intentional, or thoughtful, or I take control over it. Perhaps reflective endorsement is necessary for acting for this reason. But one can, of course, have a reason without acting for that reason. And I do not make it my reason by reflective endorsement.
In this respect I think Korsgaard cannot really be a voluntarist. The normativity of my reason, its status as a reason, does not stem from my will. My will does not make my reason a reason, because the reason was already there. To say we give ourselves a law is at best a figure of speech, like saying we make broccoli for dinner. "What are you doing in the kitchen?" I ask, and you answer, "Making broccoli." But what are you making it out of? Well, broccoli. The broccoli is already there. You are doing something with it, but to say you are making it is just a figure of speech.
These remarks in no way count against Korsgaard's contention that we are a law to ourselves. My identity may be a law to me: it determines what I have reason to do and even what I must do on pain of disintegration. Nor do they count against the thesis that the self punishes itself with painful emotions for failing to act on its reasons. If I see my life as worth living in part under the description "responsible teacher," when I fail to behave as a responsible teacher would, say, by sleeping late rather than preparing my lecture, my very identity will sanction me with guilt and remorse. But none of this makes the law-giving or the penalty infliction voluntary.
Now, perhaps Korsgaard has a different process in mind here in the activity of reflective endorsement. Perhaps until we endorse an impulse by way of a practical conception of our identity, we have no practical identity. This version of the view would say: at every juncture where I am confronted with an impulse, I must elect whether to take on, or to sustain, the practical identity associated with it. And until I do, that practical identity is not fixed for me, or perhaps is not really mine. To make the impulse a reason, I must choose or reaffirm the practical identity that sanctions action on it. This would indeed make autonomy crucial to the creation of reasons.
However, this is psychologically unrealistic. On some few occasions, a present impulse may offer me the option to slip out of a practical identity; but surely that option is not always available. In many instances, a conception of my practical identity is not a garment waiting to be slipped on or off. As we have seen, some such identities run deep. We can try to repudiate them, but it may not work. People who contemplate divorce sometimes find, to their surprise, that the thought of a life in which they are not so-and-so's spouse is unacceptable, not a live option. People who consider changing their citizenship sometimes find that while they wish to become citizens of the United States, for example, they cannot bring themselves to renounce their original citizenship as U. S. law requires. These identities do not need to be reaffirmed each time we reflect on a new impulse. They have taken root in us and impose their own requirements. When we try to disavow them, they can be as hard to eradicate as weeds. (Even a formal act of divorce or renunciation of citizenship can fail to stop a person from seeing herself as John's wife, or as Dutch, in the relevant way.) A conception of myself may become and remain one that makes my life worth living and my actions worth undertaking quite apart from any volition of mine. I may not give my heart to a person or a country, but rather discover that he, or it, has it.(5)
This de facto quality of some of our reasons is exactly as it should be. Indeed, to regard all my practical identities as unfixed would be a psychological pathology. Only a disturbed person would find his attachment to his beloved parents, his spouse and children, his life's work, and his moral and political causes all open to revision or revocation at any moment, all conditioned by his endorsement every day of the impulses they generate. This interpretation of the voluntarist picture gives us a parody of the reflective life.
Now, Korsgaard's proposal may be somewhat different. One's practical identities, and/or the putative reasons they give rise to, while they have some stability, nonetheless can be undermined by further reflection performed in light of other values. For example, I love someone and am moved to forgive him for treating me badly, but I recognize that the love is unhealthy. Or, I am attached to my profession, but should I be, when it so threatens my family life? Perhaps the Korsgaardian view is that while we do have some settled practical identifies prior to reflection and endorsement, these unexamined identifies do not yet generate full-blown reasons, but only reason-candidates, or the seeds of reasons, which in the press of events we inattentively take for full-grown ones. On reflection they may or may not come to fruition.
This view has two serious drawbacks, however. First, it commits us to the implausible claim that even though I value myself under the descriptions "lover of N" and "psychologically healthy person," until I reflect upon and endorse their resulting impulses, I have no reason either to forgive N and reconcile with him, or not to. Second, on this view the notion of a practical conception of one's identity plays no role in the theory. For now practical identifies are of the same status as impulses and desires: mere reason-proposals, candidates for our endorsement. The philosophical appeal of practical identities is that they distinguish the impulses that do not give reasons from those that do, by tying reasons to what gives our lives meaning--to what, practically speaking, we are, rather than simply what we want. If practical identities have no normative status in their own right, then they do not perform that function.(6)
A third view that might be Korsgaard's is that before I give myself a law, the impulse now confronting me is not yet a reason for me to act, even though it is compatible with my practical identity, because at that point there is no I, no acting self, whose reason it can be. The process of giving myself a law is not one of taking on a new practical identity; my practical identities may well be in place already, and they determine which impulses are candidates for my reasons. Giving myself a law is, rather, an act of creating an active self. By making an active self for which this impulse can be a reason, I make the impulse a reason. An agent is something that can act through time, so to make a reason I must make myself an agent whose existence extends forward into possible future times by giving myself a universal law, a principle that pertains to all relevantly similar situations and not just to the one at hand.(7)
This picture, however, is metaphysically very mysterious, and perhaps incoherent, for two reasons. First, how can I give a law to a self that does not yet exist? Indeed, how can making a law for someone bring that someone into existence? Second, on this picture each of us is not one acting self but a whole network of overlapping acting selves, perhaps one for each type of action we choose to perform. This is a steep metaphysical price to pay to hold on to voluntarism.
The first of these difficulties may be surmountable. One cannot create a human being by giving it a law; this is indeed both impossible and incoherent. But the active self is not an animal, not a product of nature, on this view. Perhaps one can create a new entity of some kind by giving a law to that (not yet existent) entity, in the manner in which people create a nation by adopting a constitution, or a legislature creates a new government agency by writing and enacting its charter. The second puzzle, however, is, I think, insurmountable. If in giving myself a law to act on a particular impulse I charter a new active self, then I recreate my active self each time I have (that is, create) a reason to act, and I am a series of distinct active selves rather than a unified agent. But surely the active life of a single human being is not composed of a series of agents. For one thing, if I really were a series of many selves or agents, each corresponding to one of the reasons I have, then I would not have internal conflicts of reasons. The different reasons would be the reasons of distinct selves or agents. For another, my normative reasons could not extend into my future as reasons are supposed to do, since a new "I" would replace the one for which they were reasons.
Thus we should not adopt the view that in giving oneself a law one creates an active self that did not exist before.
Having ruled out various ways to support the thesis, we must conclude that if we have practical conceptions of our identity of the sorts Korsgaard envisions, we do not need to will a law in order to have reasons.
We could see Korsgaard's account of reflectivity and self-legislation as providing an attractive analysis of volition. We step back from our impulse, and consider whether we identify with it in virtue of our practical identifies. If we do, then we issue a universal command to ourselves to act on the impulse on this ground. This is what it is to act voluntarily, perhaps. Or perhaps it is what it is to act voluntarily for a reason. But this is an analysis of volition, not of the creation of reasons. It does not show that the reasons are thus brought into existence. The impulse either is or is not compatible with the agent's practical identities. The reasons are already there. We can choose to act for those reasons or not, but we do not choose whether or not they shall be reasons.(8)
4. The Foundational Reason: Valuing Our Reflective Nature
Korsgaard claims that in valuing anything, we must value ourselves as reflective beings. To have any practical identities at all, and so any reasons, our identity as beings with a consciousness that demands reflection must be normative for us; we must see it as a source of reasons. I now turn to her argument for this dramatic claim.
At first sight the argument appears to have the following structure. I think there is more to it than this, and I indicate the gap to be filled by skipping a number in listing its steps. (Other refinements will emerge later.)
1. My reflective consciousness gives me a need for reasons to act and to live.
2. In order to have reasons, I need some practical conceptions of my identity.
3. Therefore, I have reason to have and be governed by some such conceptions of my identity.
4. This reason does not spring from the contingent practical identities I have. Rather, I have it in virtue of having reflective consciousness.
6. Therefore, in order to have any practical identities, I must value my reflective consciousness. (120-23)(9)
The conclusion is a non sequitur. I could have a reason to have some practical identities--to value myself under some descriptions or other--and I could do so, without valuing the feature of myself that gave me reason to do so. The gap is obvious in the following analogous argument:
1. My illness gives me a need for medicine.
2. In order to have medicine, I need a doctor's prescription.
3. Therefore, I have reason to obtain a doctor's prescription.
4. This reason does not spring from any doctor's prescriptions I have. Rather, I have it in virtue of having the illness.
6. Therefore, in order to obtain a doctor's prescription, I must value my illness.
The gap, however, can be filled (and the analogy defeated) in at least two plausible ways suggested by other remarks Korsgaard makes. I will give two reconstructions of the argument, using completely different versions of step 5 to fill the gap. I call them the Reflective Persistence Version of the argument and the Implication Version. I take them up in turn.
4.1 The Reflective Persistence Version
The first four steps are as before. Step 5 is my first guess as to how step 6 is supposed to follow from what went before (see especially 123). Step 6 is given an appropriate interpretation.
5. In order to sustain any practical identities--in order to continue to preserve them in the face of further reflection--I must confirm or re-endorse my original endorsement of them. This requires me to endorse my very faculty of endorsement: my reflectivity.
6. Therefore in order to have any practical identities on reflection--in order to see my life as worth living and my actions as worth undertaking under any description whatever, and so to have any reasons for acting, on reflection--I must value my reflective nature.
On this account, it does not follow logically from the fact that I have some practical identity or other that I value my own reflective nature. Rather, in order to sustain any practical identity I in fact possess in the face of further reflective challenge, I must come to value my reflectivity.
If this paraphrase is right, then we must imagine the argument to work in the following way.
Imagine I am a musician about to give a performance. If I reflect, I will ask myself, "Why give the performance? Is my impulse to do it a reason?" Suppose my truthful answer is, "Yes, because performing is what a musician does, and I care about being a musician." My reflective consciousness continues to press. I go on to ask, "Why be a musician?" Suppose my actual answer is, "In spite of its pitfalls, it's a good thing for me to be." So I see reason to give the performance. Does this require that I endorse my reflective faculty, which just found this reason? Well, I could reflect further. I could ask, "Why be what seems good to me?" According to the Reflective Persistence argument, my answer then has to be: "I trust myself to find what is best; I endorse my evaluation." And that is an endorsement of my own reflectivity.(10)
That is one way the process may work. Or perhaps the process is supposed to end in this way instead. I ask myself, "Why should I care about being a musician?" And my answer is: "I have to care about something if I am to act at all, because my consciousness is reflective." For once I pose a reflective question to myself of the form "Why should I do what I do?" I cannot simply go on playing music as before, without an answer. I will need to find something I endorse, some reason. I reflect in turn on this need and ask, "Why do what my reflective consciousness requires?" As Korsgaard at one point puts it (123), "Does it really matter ... whether II] find some ways of identifying [myself] and stand by them?" If I am to continue to act intentionally, and to have any practical reasons, my answer will be, "Yes, it does matter, because I value the reflective nature that requires me to find reasons and stand by them."
I shall argue that it is not true that if my reason for acting is to persist on reflection, my answer to the reflective challenge must be either "I endorse my evaluation" or "I value the consciousness that requires me to have reasons." There are other answers that one may in fact have, ones that enable a practical identity to stand up to reflective scrutiny without need for self-endorsement.
Facing the question "Why be a musician?" I might have a number of different, truthful answers. First, I might say, "Because music is sublime," or, "Playing music is the Lord's work." Now I could try to reflect further from there. But notice that the questions "Why do the Lord's work?" and "Does what is sublime really matter?" are odd ones. These descriptions are what Anscombe calls desirability characterizations.(11) They have settled the matter. We have come to the end of the chain of reasons. No further answer seems to be either available or needed. What is sublime really matters because it is sublime. We should do the Lord's work because it is the Lord's work. And even if, somehow, these questions still press, the answer will not be an endorsement of my own reflectivity. I could, of course, ask if the music I play really is sublime, or if I am mistaken about that. But if I am right that playing music is the Lord's work, or is sublime, my answer to the question why I should do that work will not be, "Because my own reflective nature is of value." The value is in the music, not in me.(12)
This possibility draws attention to an important way in which we see value in things. There are occasions on which we see reasons to act where our concern for or interest in something is just what we take to be the reason to pursue it. But there are other occasions in which it is not our interest in something that presents itself as a reason, but the thing's importance. Of course, it is important in our view;, but it is the first-person perspective that counts here. Reflection, as Korsgaard reminds us, occurs in the first-person perspective. What an outsider would describe as valued by me, I perceive as valuable. From the first-person perspective, however, some things are valuable because they satisfy me, and others are important in their own right. It is only from the third-person perspective that the distinction between these two ways of valuing disappears. Take the musician who chooses playing Mozart over playing saleable pop songs because playing Mozart is a purer and nobler endeavor. I can say of him, "His reason is that he cares more about Mozart than about having fun or making money." But he won't regard his own choice this way. From his point of view, he does it because playing Mozart is nobler and purer, not because of anything about himself.(13)
Second, in answer to "Why be a musician?" I might make the response, "Playing music brings people joy." "Why give people joy? Is this really important?" the voice of reflection interrogates me further. The answer might have to be, "Because people are worth it; their pleasure counts for something." In a sense this makes the sequence of endorsements end in humanity, but not my own humanity in particular, and not my or anyone's reflective capacities. The foundational value is the sensibilities of people, and their pleasure. This is another way my identity as a musician might persist upon reflection: it might be grounded in my estimation of other people.
Third, in answer to the question, "Why be a musician?" I might say, "I know, it's only rock `n' roll, but I like it."(14) It isn't sublime or the Lord's work; it isn't of transcendent value. It may bring pleasure to some, but it irritates others. Anyway I am not trying to improve the human condition. I do it because I like it. On further reflection I ask myself, "Why should I do what I like? Does what I like really matter?" And here I think the answer will be, "It does, because I'm worth it." Only a person overwhelmed by feelings of her own unworthiness will not think she has reason (at least some reason) to do what she likes. This does locate the roots of my reason firmly in myself, for once. So reflection on my reasons can lead me to endorse myself, although it does not have to.
Since reflection on my reasons need not lead me to endorse myself, as we saw in the first and second alternative responses to the reflective challenge, this version of the argument does not establish that we must value our own reflective nature if any of our reasons are to persist through further reflection.
Furthermore, consider just what this answer comes to where it is appropriate. "What I like matters, because I am worth it" is different from "I endorse my evaluation," which was an endorsement of my reflective and endorsing faculty. To continue to value anything at all after reflective questioning, I must trust my evaluating capacities; to cease to trust those would undermine my evaluations of everything else.(15) But to think I have reason to do what I like because I am worth pleasing is not to endorse my faculty of reflection and endorsement. It is not equivalent to the judgment that my evaluations are reliable. To see this, note that I could judge my evaluative capacities to be highly reliable, and use those capacities to assess myself as contemptible and not worth pleasing in the least. My judgments that I am, or am not, worth pleasing are just evaluations that I make. The unfavorable one does not itself raise skeptical doubts about the reliability of my evaluative faculty, and the favorable one does not allay them. So even supposing that the successful resolution of skeptical reflection must lead us to trust our evaluations, there is no guarantee that a successful outcome of such reflection will lead us to value ourselves as worth pleasing, or as in other ways the roots of reasons to act.
Nor is "I am worth pleasing" a judgement that my reflective faculty, which demands that I have reasons, is valuable, and its demands worth satisfying. The rock musician thinks, "I am worth pleasing, so the fact that I like rock and roll is a reason to play it." This is not thinking, "My reflective nature is valuable and its requirements are worth satisfying; thus, its requirement that I have some practical identity or other is a reason to play rock and roll." The positive evaluation of my reflective consciousness does not support being a rock musician as distinct from anything else. It does not even support doing what I like as distinct from what I don't like.
Korsgaard might reply: Regardless of what reflection-resistant reason the musician finds for playing music, he, like every other person, nonetheless has to admit that his reflective nature does impose a requirement on him to have some practical identity or other. Setting aside the specific issue of whether to be a musician, he can reflect separately about the requirement imposed by his reflective nature: "Why should I conform to it? Does it really matter that I have some practical identity, and so some reasons, rather than none?" At this juncture, according to Korsgaard, he must find some reasons, or he will no longer be able to act, not even to play music. Thus, if he does not become a nihilistic suicide, he will find some reasons. And once he does, he must endorse the part of him that requires him to have reasons: he must answer, "Yes, it does matter that I have some practical identity or other, because it matters that I satisfy this requirement of my nature." Thus, her view could be, not that reflection on his actual, particular reasons will of necessity lead him to endorse his reflective nature (for we have seen that it need not), but that should he reflect instead on the demands of his reflective nature itself, this process will force him to endorse that nature on pain of nihilism.
I do not see how this last point follows, however. If he should reflect on the demand for reasons that his reflectivity makes upon him, and he does not subsequently cease to see any reason to live, he will find some practical identity and stand by it. But how does finding a particular identity and endorsing it commit him to endorsing the reflectivity that put him into this bind in the first place? The earlier gap in the reasoning reappears here. Why couldn't he think of his reflective nature as a weakness, imposing a need for identification rather like an addict's need for her drug or a diabetic's need for insulin? He can make the best of a bad situation by at least selecting an identity he likes, as the caffeine addict chooses a tasty coffee drink rather than vending-machine coffee to feed her habit. He can make a virtue of a necessity. But I do not see that he must value his status as one who has this need, rather than regard the need as a weakness to cope with as best he can.(16)
There is an additional problem. We have seen that this reflective process is, in some cases, utterly independent of reflection on the particular reasons any person has. Our reflectivity demands of us that we identify ourselves in some way or other. But if I ask myself, "Is it really important, after all, to be a musician?" I am not asking, "Is it really important to be anything at all?" In reflecting on my contingent practical identities, the task before me is not to find some way of identifying myself and stand by that. It is to figure out whether the way in which I already identify myself is a good one by which I should continue to stand. If I figure out that my current practical identity is worth having, it may well have its own support. If it does, it does not need a support of the form "I've got to have something, and this is something, so it will do." If you buy a painting to hang on your wall because it is a magnificent painting, you need not have done so because you had an empty space on your wall that needed to be filled. That is quite a different reason, and one that might be absent in your particular case. (You could instead take something down to make space.)(17)
I have argued that the Reflective Persistence version of Korsgaard's argument fails to establish its conclusion. Now let us look back briefly to see what that version of the argument would establish even were it to succeed. The thesis it defends is that in order to sustain our practical identity when we examine it reflectively, we have to endorse our reflectivity. Thus, if we reflect, and we reject or refuse to endorse our reflectivity (we find we cannot trust our endorsements, or do not think we are "worth it," or do not endorse satisfying the demand of our reflective nature for reasons), this will undermine all our reasons. On this understanding, whether we continue to value X depends crucially on whether we endorse X in response to the onslaught of reflection, something that, when the time comes, we do. If we do not actually subject our contingent practical identities to reflective challenge, then we need not actually endorse our reflective nature. In that case we have some reasons without having come to endorse our reflectivity. True, we have not carried reflection to its limit, so our reflective nature may some day drive us to go on. In the mean time, though, we do not yet value our reflective nature, since we have not yet endorsed that nature. So it is not yet normative for us. Its status as a source of reasons awaits our act of endorsement. Perhaps this makes it depend, in a way, upon our will.
On this view, however, the valuing of the valuers is not inescapable, as Korsgaard says it is. We can avoid it by failing to ask ourselves "Why be a musician?" and other such questions. We can continue merrily playing music, and when such probing thoughts cross our minds we can dismiss them. We can avoid entering into the requisite stretch of reflection. Without it, we need not value our reflectivity.
However, Korsgaard claims we cannot but take our reflective identity as normative for us. Thus, I think we have to read the claim in a different way. We have to interpret it as saying that if we value anything at all, we already value our reflectivity, whether or not we have performed any process of reflection or act of endorsement. Our valuing of our reflective selves must be implied by our valuing of anything else. For this the Reflective Persistence Version of the argument is not adequate; we need the Implication Version.
4.2 The Implication Version
On this version, the first four steps of the argument are the same as before. Instead of 5 I insert 5'.
1. My reflective consciousness gives me a need to have reasons.
2. In order to have any reasons, I must have some conception (s) or other of my practical identity to which I conform.(18)
3. Therefore, I have reason to possess and be governed by some practical conception (s) of my identity.
4. My reason to have and be governed by these springs from my reflective nature, not from any contingent practical identity I happen to have.
5'. All reasons are rooted in a practical identity of the agent whose reasons they are.
6. Therefore, if I am to have any practical identifies, my reflective nature must itself be a practical identity of mine. Since it is the source of a reason that everyone has who has any reasons, it must be something I value, a source of reasons for me.
Step 5' adds that every reason is rooted in some practical conception of one's identity. We then see the relevance of step 4, which says that our reason to have practical identities is not rooted in any other practical identity, but only in our reflective nature. Thus, the conclusion follows: since my reflectivity itself is a source of reasons, it must be a practical identity of mine. The illness and prescription argument is not a logical paraphrase of this argument either. For the reason I have to obtain a prescription is not in this sense rooted in the illness itself, or alone; presumably it is rooted in something else about myself that I do value, and which the illness threatens. Thus, the Implication Version requires that we also reinterpret step 4. It says not only that my reason to have practical identities springs from my reflective nature rather than my contingent practical identifies, but that this reason originates entirely in my reflectivity and in no other (more basic) fact about me.
First we should note one thing about the Implication Version of the argument. To begin, it says that because I am reflective, I need practical identities; therefore, I have reason to have practical identifies. Is this because I necessarily have reason to have what I need? The assumption that I do seems to lie behind the transition from steps I and 2 to step 3. If needs give us reasons, however, then we do not have to give ourselves laws in order to have reasons. Also note what is assumed in step 5' itself: every reason is rooted in a practical conception of my identity. This assumes an answer to a question Korsgaard takes up later in the book: whether our animal nature gives us reasons independently of our conceptions of ourselves. Step 5' assumes it does not; only our values do. One might naturally think, however, that some of the things I need as an animal, such as food and water, I therefore have reason to try to get, a reason that need not depend upon any conception under which I find my life worth living and deeds worth doing. Even someone whose life's purpose has come to naught and consequently falls into the deepest dejection and existential despair needs food and water every day. It would be strange to claim that he did not. But if he does, it seems we are committed, by the assumption that needs give reasons, to say that he has reason to nourish and hydrate himself, regardless of whether he has any practical identities at all. So the assumption that needs give reasons comes into sharp tension with step 5'.
Now for a more general assessment of the argument. On the Implication Version of the argument, we interpret the conclusion to say that if we see reason to do anything, then we are logically committed to valuing ourselves as reflective beings, whether we reflect upon our reason or not. On this reading valuing does not depend upon an act of endorsing on reflection. For people see reasons to act all the time without having reflected on those reasons or on anything that has led them to conceive themselves as reflective beings. No one asks herself, "Does what I like really matter?" while still in kindergarten. But on this view everyone who sees any reasons to act at all thereby values herself as a reflective being. Therefore, on this view, valuing is not identical with post-reflective endorsement and does not require it.
This is the primary flaw in the Implication Version of the argument, for it flies directly in the face of the Korsgaardian claim, discussed earlier, that we make our reasons and do not already have them. I argued above that often it is simply a fact that I find my actions worth undertaking under a particular description of myself, and so I have reasons to act, whether, on reflection, I command myself to act on those considerations or not. Now suppose that, as the Implication Version of the argument entails, we cannot but take our reflective identity as normative for us if we have any practical identities at all. Then our foundational reason is also already given to us, already there, if we value anything at all. And we do value something, whether or not we reflect, provided we act for what we take to be reasons. Thus, if I find that I love someone, for example, or that I like rock and roll, and because of that I find some of my deeds worth doing, then it follows as a matter of logic that I value my reflectivity as well. That is what the claim seems to come to that "[t]he value of [reflectivity] is implicit in every human choice" (122). My choice is guided by what I regard as reasons; and regarding those as reasons commits me to valuing my reflective nature, whether or not I have endorsed that nature or made it a law to myself. My reason to respect my own reflectivity is thus a de facto reason par excellence. The law of my reflective nature is not a product of my will after all. It is a law for me simply because, in some way or other, life has taken hold of me and made me care about it. My reasons are not made but born.
State University of New York, Albany
I am grateful to Nomy Arpaly, Michael Bratman, Stephen Darwall, John Kekes, Christine Korsgaard, Timothy Schroeder, Carol Voeller, and members of the audience at the Central APA, May 1998, author-meets-critics session in Chicago, for comments on earlier versions of this paper, and especially to Charlotte Brown for inviting me to speak on this topic.
(1) Cambridge: Cambridge University Press, 1996. Page numbers in parentheses refer to this volume.
(2) Since for Korsgaard obligations can conflict, there is no logical problem here about incompatible obligations arising from one's various practical identifies, though of course there will be practical problems.
(3) As I indicate in note 8, below, it is not clear why this matters. Conceptions that are in fact false of me can nonetheless be sources of reasons, so why is a true conception privileged?
(4) On Korsgaard's view this practical conception of ourselves as reflective beings commits us to something further as well--to valuing the reflective nature of other people, and so to the moral law--and it constrains us not to value ourselves in ways incompatible with that (for example, as assassins). Although this is where the theory ultimately goes, we will stop earlier and just examine nonmoral reasons.
(5) In addition, this picture leaves Korsgaard without a solution to the problem faced by the seventeenth-century voluntarists: why should the ruler command one thing rather than another? My thinking self is not committed to any particular conception of my practical identity; it must always choose anew. If it does not have any particular identity, then there is no reason for it to select one such identity rather than another. (Reasons are supposed to arise from what I am, and there is nothing that I am, in this sense.) Hence it can identify itself with any conception whatsoever on each occasion. It can throw off an old identity and take on a new entirely at random. Its choices will be arbitrary. But how can the arbitrary, unwarranted selection of a new practical identity create genuine reasons? It is hard to see how such a process could yield normativity at all. If the identity I choose could just as well have been another, and I may well choose a different one tomorrow, how does my choice give me a reason to conform to this one now?
(6) We may seem to need something more than practical identities because I could have an impulse rooted in one conception of my practical identity that is at odds with another, more important identity. My impulse to forgive my beloved could be incompatible with the psychological health I value. Of course, the two impulses might yield two conflicting reasons; in that case we need nothing more than the practical identities to generate reasons. But couldn't it turn out that under the circumstances I don't have reason to forgive my loved one after all, even though my urge to do so is rooted in one of my practical identities, because my self-identification as a psychologically healthy individual forbids my giving any weight to it? So we bring in the will, perhaps, to settle the conflict. But in fact the possibility of such conflict does not show that we need the will to settle what reasons we have. I might discover this conflict in reflection, but it is not the activity of reflective rejection of the impulse to forgive that stops its being a reason. My further and more valued practical identity (my commitment to my own psychological health), which may well be just as involuntary as the first, rules it out antecedently.
A further possible situation is one in which there is no antecedent fact about which I value more, or to which I give precedence, and in reflection I simply decide somehow in favor of one or the other. This decision may well create new reasons; see, for example, Michael Bratman, Intentions, Plans, and Practical Reason (Cambridge: Harvard University Press, 1987), chap. 2. But this possibility does not show that no reasons exist prior to reflective endorsement. It does not even show that in deciding in favor of one I render the other a non-reason.
(7) The idea for this reading comes in part from C. Korsgaard, "Personal Identity and the Unity of Agency: A Kantian Response to Parfit," in Korsgaard, Creating the Kingdom of Ends (Cambridge: Cambridge University Press, 1996).
(8) To put my point in terms of one of Korsgaard's distinctions, in this objection I am accusing Korsgaard of being to some extent (what she calls) a substantive normative realist, in spite of her official rejection of that position in favor of a more modest type of procedural normative realism (35-77). According to the substantive normative realist there are correct and incorrect answers to normative questions such as "what should I do?" because there are normative truths or facts that exist independently of any procedures we might carry out to answer such questions. Procedural normative realism does not require that there be such entities; it requires only that there be some correct or best procedure for answering normative questions. Indeed, it says that there are answers to normative questions because there are correct procedures for answering them. Korsgaard seems to favor a version of procedural realism in which all such answers are created by the very procedure of answering the questions and no answers exist independently of it, so the correct procedure is always one of making rather than finding. I have argued that aspects of her picture of reasons make it impossible to sustain the view that all reasons are generated by a certain procedure that we enact. Instead, some of our inclinations are already reasons, on her view, and so in a way reasons are "out there"--are substantively real.
What, I imagine, is supposed to keep substantive realism at bay is that on the Korsgaardian picture, if I have a practical conception of my identity that persists on reflection, so long at it is compatible with my reflective and moral identity, I cannot be mistaken in having it, no matter how idiotic or otherwise objectionable it may be. It does not matter if the description under which I value myself is false. I can value myself as an alchemist or channeler of the spirits of the dead, and in seeing reasons to act to carry out my alchemy or channeling, I am not mistaken about what reasons I have. If I value myself under some rather pathetic description, that too is a source of reasons. Suppose my ruling passion, which I endorse on each occasion, is the desire for Hollywood fame: to be recognized everywhere I go, asked for autographs, fawned over in restaurants. Then I actually have reasons to pursue this, within the limits of morality. I am not mistaken about what reasons I have. Most writers who think there are normative truths or facts to be discovered, by contrast, would think me mistaken about these alleged reasons.
But actually, the difference between Korsgaard, who will not classify these as mistakes, and the more robust substantive realist who will, is a matter only of degree. Korsgaard already countenances some reasons that are not in fact generated by reflection and endorsement. Thus, we do not have pure procedural normative realism without a trace of the substantive kind. To some extent the procedure we use to answer normative questions is not one of creating their answers but rather one of finding them (albeit in our psyches). So the camel's nose of substantive realism is already under the tent-flap.
(9) I take it this is her conclusion because she says, "You must value your own humanity if you are to value anything at all" (123). She understands humanity to be "your identity as ... a reflective animal who needs reason to act and to live" (121), thus as the possession of an essentially reflective type of consciousness (as explained at 92-93).
(10) This point is plausible. Were I to reject or refuse to endorse my own reflective and valuing faculty, that would undermine my continued endorsement of all the things I value. This is analogous to Hume's skeptical argument to show that reflection on the faculty of reason undermines belief.
(11) G. E. M. Anscombe, Intention, 2d ed. (Ithaca: Cornell University Press, 1976), 70-76.
(12) David Copp, in "Korsgaard on Normativity, Identity, and the Ground of Obligation," forthcoming in Rationality, Realism, Revision, ed. Julian Nida-Rumelin (Berlin: Walter de Gruyter, 1999), makes a related but somewhat different point. Copp understands Korsgaard as saying that the only valuing that underwrites our reasons is valuing things about ourselves; valuing such things as a pristine environment without connection to self does not provide reasons. I read Korsgaard more narrowly, as predicting that should reflection press us to question the things we value (which might include a pristine environment), we will find that we need to value an aspect of ourselves in order to sustain our valuing of the other things. This is the point I question in the example.
(13) It is possible that his attitude can be analyzed properly as some complex arrangement of first-order preference for Mozart plus second-order preference that he prefer Mozart to pop music, or these plus some desire that Mozart continue to be played (perhaps even by him) even were he to stop liking it better. But note that this is a thoroughly third-personal description of the situation, not one that he himself would give. Cf. Allan Gibbard, Wise Choices, Apt Feelings (Cambridge: Harvard University Press, 1990), chap. 8.
(14) "It's Only Rock `n' Roll (But I Like It)," The Rolling Stones, EMD/ Virgin, 1974.
(15) This is reminiscent of the reflexivity Korsgaard sees in the early Hume. It is also, interestingly, reminiscent of Allan Gibbard's discussion of self-trust in Wise Choices, Apt Feelings, 176-79.
(16) David Copp makes a similar point, "Korsgaard on Normativity," n. 10.
(17) That is, I deny that our reasons persist through reflection only because we find, on reflection, that we need to have reasons. If we reflect on our reasons themselves, we need not find any such thing; rather, we may find that they persist because they are grounded in practical identities that are valuable in themselves. If we reflect separately on our own reflectivity, we may indeed find we need to have reasons; but that does not show we have or should have the particular reasons we do.
(18) It may seem that I only need to have some such conception, not necessarily to conform to it; but if I have it and I am never guided by it in action, I will not continue to have it, so I need to conform to it as well.
|Printer friendly Cite/link Email Feedback|
|Publication:||The Philosophical Review|
|Date:||Jan 1, 2000|
|Previous Article:||Reflection, Planning, and Temporally Extended Agency.|
|Next Article:||ARISTOTLE: POLITICS, BOOKS I AND II.|