Kantian autonomy and the moral self.

Wilson, Eric Entrican

KANT'S ACCOUNT OF AUTONOMY is not designed to solve the traditional problem of free will. It is a response to the problem of heteronomy rather than the problem of determinism. And the former pertains to concerns about the structure of practical reason rather than the scope of nature's causal laws. So his theory of practical reason, rather than his metaphysics, provides the proper context for understanding his account of autonomy.

Unfortunately Kant did not always appreciate this feature of his own views. He says conflicting things about the relation between the practical concept of autonomy and the theoretical concept of "transcendental freedom." He sometimes claims that the former depends on the latter, which he characterizes as "the faculty of beginning a state from itself, the causality of which does not in turn stand under another cause determining it in time in accordance with the law of nature." (1) In other words, he sometimes says that our autonomy depends upon our ability to exempt ourselves from the workings of nature in order to initiate a chain of events that does not itself function as a link to earlier causal states. Yet on other occasions he claims that autonomy and transcendental freedom are two separate issues: "The question about transcendental freedom concerns merely speculative knowledge, which we can set aside as quite indifferent if we are concerned with what is practical, and about which there is already sufficient discussion in the Antinomy of Pure Reason." (2)

Kant himself did not always follow through on this claim, but I propose that we take him at his word here. Further remarks recommend this interpretive approach. Consider, for example, his review of Johann Heinrich Schulz's work, Attempt at An Introduction to a Doctrine of Morals for all Human Beings, which Kant wrote and published in 1783, just before the Groundwork of the Metaphysics of Morals. After summarizing Schulz's views in a friendly and not unsympathetic manner, Kant draws the reader's attention to "the general fatalism which is the most prominent principle in this work and the most powerful one, affecting all morality, [since it] turns all human conduct into a mere puppet show and thereby does away altogether with the concept of obligation." (3) What is interesting is that Kant criticizes Schulz's position not by insisting on the reality of freedom, but, rather, by arguing that the viability and legitimacy of our moral practices do not depend on fatalism or determinism being false. (4) Despite having claimed in the Critique of Pure Reason that autonomy depends on transcendental freedom, and thus a solution to the free will problem, he says here that "the practical concept of freedom has nothing to do with the speculative concept, which is abandoned entirely to metaphysicians." (5) Bearing in mind that Kant regards autonomy as a species of practical freedom, this claim is worth repeating in a way that makes its relevance to his moral theory crystal clear: the concept of autonomy, he tells us, has nothing to do with the concept of transcendental freedom.

This line of thought will surely not impress those who are committed to the importance of the traditional problem of free will. They will insist that the will is either free or not--no matter the perspective or standpoint from which one chooses to examine the issue. And whether the will is free or not depends either on the truth or falsity of determinism or on whether freedom can be made compatible with determinism. In the same review of Schulz Kant indicates a line of response to this objection that he will pursue throughout his practical philosophy:

 The most confirmed fatalist, who is a fatalist as long as he gives
 himself up to mere speculation, must still, as soon as he has to do
 with wisdom and duty, always act as if he were free, and this idea
 also actually produces the deed that accords with it and can alone
 produce it. It is hard to cease altogether to be human. (6)

The idea here seems to be as follows. It is a reality of human life that we weigh our options, reflect on our impulses and desires, evaluate reasons, and make judgments about how we should conduct ourselves. In other words, it is a fact about human life that we face head on, as agents, the world as much as we observe our transactions with it, from a sideways perspective, as prospective knowers (or makers of theoretical judgments). Of course we do not always do this, but the point is that we can and this ability is important to us and central to our lives. When we deliberate about how to act, we cannot help but take ourselves to be capable of exerting influence on the world. We cannot help but regard ourselves as capable of acting according to how we decide and deciding according to how we will. Moreover, we cannot help but think that, when we do this, our options are somewhat open and that the outcomes of our volitions and actions are at least somewhat up to us. This is why Kant claims that "for practical purposes the footpath of freedom is the only one on which it is possible to make use of our reason in our conduct." (7) The fatalist or determinist fails to own up to this. The rejection of freedom is a mark of false consciousness regarding the reality of the practical stance we must all take, as agents, vis-a-vis the world. To deny one's freedom is not to face courageously the hard truths of science or theology. It is to renounce one's agency--and, with that, both one's ability to take responsibility for one's actions and one's "higher vocation" as a person.

If we take this line of thought seriously, then the best way to understand and appreciate Kant's account of autonomy is to isolate it within the practical context to which it rightfully belongs. In that context, the pressing problem is not freedom versus determinism.

Instead, the real issue is autonomy versus heteronomy. What Kant is most concerned about is the prospect that practical reason (or the will) is exclusively heteronomous in nature. In this essay, I intend to show how he appeals to the notion of self-respect in order to argue against this prospect. In Kant's view, our capacity for autonomy depends on our capacity to identify with and feel respect for an ideal, moral version of ourselves. Genuine self-determination, in other words, rests on a form of self-respect.

Allow me to begin with a sketch of Kant's theory of practical reason. Like most philosophers, Kant takes practical reason to be about means and ends. Because his moral theory is normally categorized as "nonteleological" or "deontological" by contemporary taxonomies of ethical thought, this is often overlooked. In fact, it is widely believed that Kant's view of moral reasoning has nothing to do with ends at all. But it is easy to show that this is wrong. In the Groundwork Kant characterizes means, ends, and the relation between them as follows:

 Now, what serves the will as the objective ground of its
 self-determination is an end [der Zweck], and this, if it is given
 by reason alone, must hold equally for all rational beings. What,
 on the other hand contains merely the ground of the possibility of
 an action the effect of which is an end is called a means [das
 Mittel]. (8)

In the Metaphysics of Morals he offers a more straightforward characterization of what he means by "end." There he says that, "an end is an object of the choice (of a rational being), through the representation of which choice is determined to an action to bring this object about." (9) Or more simply: "an end is an object of free choice, the representation of which determines it to action." (10) The word "object" should not mislead here. It is being used in as broad a sense as possible to indicate that at which the will is directed. All Kant means is that the end of a decision or choice is that for the sake of which it is undertaken. It is in this sense that the end of a choice "serves the will as the objective ground of its self-determination."

Contrary to the tendency to describe his views as nonteleological, Kant explicitly claims that, "every action ... has an end." (11) In the second Critique he puts the point like this: "it is indeed undeniable that every volition must also have an object and hence a matter." (12) This is in fact a central pillar of Kant's entire practical philosophy, for it is in these terms that Kant characterizes practical reason as such. "Rational nature is distinguished from the rest of nature by this, that it sets itself an end." (13) In the Metaphysics of Morals he goes so far as to say that,

 pure practical reason is a faculty of ends generally, and for it to
 be indifferent to ends, that is, to take no interest in them, would
 therefore be a contradiction, since then it would not determine
 maxims for action either (because every maxim of action contains an
 end) and would not be practical reason. (14)

Because of its centrality to rationality, end-setting is the essential mark of our humanity: "The capacity to set oneself an end--any end whatsoever--is what characterizes humanity (as distinguished from animality)." (15)

In our actual practical reasoning the relationship between means and ends is mediated by incentives (Triebfeder). (16) We select what we take to be the best means toward prescribed ends, but we must also in some sense be "moved" to adopt those ends as our goals. This is the job of an incentive, which Kant characterizes as the "subjective determining ground of the will of a being whose reason does not by its nature necessarily conform with the objective law." (17) Incentives figure into the explanation of intentional actions in that they help account for why we adopt the ends that we do. To cite an agent's incentive is to cite what moves her to treat something as her end. But it is essential to Kant's theory of agency that we are not merely pushed and pulled by our incentives. They move us in a particular sort of way. This theoretical commitment is expressed most clearly in what Henry Allison calls the "Incorporation Thesis," which Kant presents in the first part of Religion Within the Boundaries of Mere Reason: (18)

 Freedom of the power of choice has the characteristic, entirely
 peculiar to it, that it cannot be determined to action through any
 incentive except so far as the human being has incorporated it into
 his maxim (has made it into a universal rule for himself, according
 to which he wills to conduct himself. (19)

There are more and less plausible ways of reading this claim. The least plausible version has Kant committed to the idea that human beings cannot be "determined to act" by incentives unless they incorporate them into their maxims. This is obviously false. Say I hear a scary noise in the woods and am moved by my desire (or "instinct") for self-preservation to run as fast as I can back to camp. This desire can obviously function as an incentive without me having to first incorporate it into a maxim. It is a good thing we do not have to incorporate our incentives into maxims in order to act. We would not survive for long if we did. Surely it must be the case that the vast majority of our actions do not require the incorporation of incentives.

But this leaves room for a more modest and plausible version of Kant's claim. While it must be true that we do not have to incorporate our incentives into our maxims, it also seems true that we can do this. An important range of our behaviors is based on deliberation. That is, many of our actions are the outcome of the process of considering reasons and drawing practical (and sometimes moral) conclusions about how we should conduct ourselves. If we accept this much about human agency, I think we can see that Kant's point is simply that our desires and other sensible impulses do not come prepackaged as reasons. They enjoy this status only by being treated as reasons, that is, only by being taken up and considered to count in favor of or against doing one thing rather than another. As rational beings capable of reflection, we are capable of asking whether a given incentive should move us to pursue a particular end. This is not to deny that sensible impulses of various kinds can also overwhelm us, sometimes to the point where the distance required for reflection is impossible. The point is only that it is in principle always possible for us to respond to a given incentive by asking whether we ought to or should act on it--what Christine Korsgaard calls "the normative question." (20) And it is not until we pose this question that we confer the status of "reason for action" on these incentives. In other words, an incentive becomes for an individual agent a reason for action only when she poses and then answers the normative question. An incentive can certainly force its way in and overwhelm us; this happens frequently, but in such a case it is moving us to act without functioning as a reason. Kant does not have to deny this for his Incorporation Thesis to make sense. In fact, any plausible reading of the Thesis accommodates the point quite easily.

It is important to bear in mind that the Incorporation Thesis belongs to Kant's attempt to work out a general framework for understanding the nature of deliberating about what one should or ought to do and what makes such deliberation possible. It does not express any commitment to an ontological distinction between things that can count as reasons for action and things that cannot. It belongs to his theory of practical reason. And a major concern of that theory is to provide an account of the difference between being moved to action by what one takes, upon reflection, to count in favor of that action and being moved to action "without thought" or simply "by instinct." I think Kant's point here is deceptively simple. Moreover, it seems to conform to some common intuitions about what is involved in considering how to act. Under normal circumstances we take it for granted that there is a genuine difference between deciding what to do and wondering what will befall us. The Incorporation Thesis is designed to account for this difference by articulating a necessary condition for genuine deliberation. According to Kant, we are deliberating only if reflecting on whether a given incentive provides us with a consideration that counts for or against performing some action. And, in his terminology, we do this by incorporating the incentive into one of our maxims. (21)

It is in this context that the problem of heteronomy arises for Kant. Most theorists of practical reason would be willing to grant some version of the Incorporation Thesis. The idea that we are able to achieve some measure of distance from our incentives and ask how they stand with respect to our more general practical principles (our maxims) is not terribly controversial. Our capacity to ask the normative question and weigh reasons for action is not under much suspicion. But how far down does this capacity for reflection and distance go? Do the incentives provided by inclinations and desires function as a limit beyond which practical reason is unable to venture? This is precisely the line of thought that Kant wants to resist. Our incentives explain, in part, why we adopt the ends that we do. If you hold that all incentives come from things such as inclinations and desires, then you must also maintain that the capacity to set ends is limited to operating on these more or less passively experienced incentives. On this widely accepted view, practical reason is essentially a matter of reasoning about means to ends that are themselves simply given. (22) The inclinations and desires that account for the ends, which move us to adopt them as our ends, are not "up to us." If this were true, then I could select from among a range of options afforded me by a range of inclinations and desires. (23) But I could not distance myself from my desires and inclinations as a whole and ask whether some other kind of incentive should move me to set a particular end as that for the sake of which I am considering undertaking some action. Kant thinks that on this view genuine autonomy would be impossible. There is no genuine self-legislation if the self is limited to responding to incentives over which it has no control. To be sure, the capacity to incorporate incentives into maxims and then select appropriate means to desired ends manifests some degree of self-control or freedom, but this is not sufficient for Kantian autonomy. According to Kant it would be sufficient merely for prudential reasoning, which traffics only in hypothetical imperatives.

Kant calls this view the "empiricism of practical reason." (24) His dislike of empiricism is palpable in the Critique of Practical Reason, where he says that empiricism is "dangerous" in that it threatens to "degrade humanity." (25) Why is the view so troubling, and even "dangerous"? Kant worries that it entails what Christine Korsgaard calls "skepticism about practical reason." He is worried in particular about "motivational skepticism." (26) The concern is that theories of practical reason which hold that all incentives spring from inclinations and other sensible forms of desire automatically exclude the possibility of genuinely moral motivation. For Kant, of course, we have a case of moral motivation when, and only when, someone is moved to undertake a particular course of action for the sake of the moral law. It is not enough to act in accordance with the law; one has to take as one's reason for action the fact that the law prescribes the action. When this happens we do what is morally right because it is morally right. We adopt the demand of the law as our end.

The motivational skeptic doubts that, in the absence of some inclination-based incentive, there could ever be such a reason to fulfill the demands of the moral law. He or she doubts, in other words, that purely rational considerations could provide genuine reasons for action. This follows from a combination of the assumption that all incentives stem from common motivational sources such as inclination and what Korsgaard calls the "internalism requirement" on practical reason. (27) According to this requirement, R is a reason for performing some action only if R implies that a rational individual would be actually motivated, at least to some extent, to perform the action in question. It is easy to see how practical reasons springing from sources such as inclinations and desires meet this requirement. But it is more difficult to see how purely rational considerations, which find their source in deliberation rather than in antecedently given inclinations, could as well. The empiricist exploits this difficulty by arguing that purely rational requirements cannot provide reasons for action precisely because they cannot meet the internalism requirement. And they cannot do this because they cannot provide incentives, that is, "subjective grounds" for the determination of the will. Purely rational considerations cannot be reasons, in other words, because they cannot ever motivate anyone to act. Although Hutcheson's moral sense theory is most directly in Kant's sights at this point, an extreme version of this view is crystallized in Hume's famous claim that, "reason is, and ought only to be the slave of the passions, and can never pretend to any other office than to serve and obey them." (28)

This is the context in which the problem of autonomy emerges for Kant. He is less worried about determinism than he is about a view of practical reason that would deny the possibility of being motivated to act by purely rational considerations. Such a view would deny that pure practical reason could, of itself, determine the will by denying that incentives could spring from a source other than inclination or other sensible forms of desire. According to Kant, such a view of practical reason would condemn us to heteronomy. If we were entirely dependent upon incentives given prior to rational deliberation by our inclinations, then the will would be able to determine itself "only by means of an incentive that the anticipated effect of the action has upon the will," that is, an incentive based on a momentary desire or inclination. (29) If this were the case, then autonomy would be impossible. Instead, we would have "always only heteronomy of the will." (30) And by condemning us to heteronomy, motivational skepticism would undermine the basis of morality altogether, since "autonomy of the will is the sole principle of all moral laws and of duties in keeping with them." (31)

Kant's account of autonomy is his response to this problem. To appreciate further his understanding of the difference between autonomy and heteronomy in this context, we can take our cue from the words themselves. If I can be forgiven for pointing out the obvious; the prefix of the word "heteronomy" comes from the Greek word for "other" (hetero), and the prefix of the word "autonomy" comes from the Greek for "self" (auto). Bearing this in mind, it is hardly surprising that the language Kant uses to characterize heteronomous volition and action reflect the etymology of the word itself. If the will were restricted to heteronomy, he argues, then "the will would not give itself the law but a foreign impulse [ein fremder Antrieb] would give the law to it by means of the subject's nature, which is attuned to be receptive to it." (32) This notion of a "foreign impulse" is indeed quite common to the terms in which he discusses inclinations and other sources of sensuous incentives. He repeatedly and consistently distinguishes between volition that is determined by respect for the moral law and volition that stems from "alien influences" or "alien causes." (33) Thus the obvious contrast: autonomy is self-determination, while heteronomy is other-determination.

The point is familiar to any reader of Kant. There is a deeper and less obvious conceptual point here, however. Namely, it is impossible to clarify the difference between autonomy and heteronomy without clarifying the distinction between self and other. We cannot make sense of the concept of autonomy, in other words, without establishing what it is that properly belongs to the self. As with so many other pivotal issues in modern philosophy, the distinction between autonomy and heteronomy thus rests on a conceptual distinction between what is internal and what is external to the self. This explains why Kant sometimes characterizes autonomy as "inner freedom," in which the will is determined by something that springs from within, as opposed to heteronomy, in which the will owes its end to something that comes from without, something "alien." (34)

This is by no means a simple issue. There are of course many different ways to characterize the difference between what is internal and what is external to the self, or what properly belongs to it and what does not, since these concepts have many different senses, depending on context and usage. There is, for example, a literal sense in which each and every one of my sneezes and itches belongs to me. If you notice me sneeze and then say, "bless you," I understand that you are addressing me and that you are ascribing the sneeze to me. I acknowledge and endorse that ascription when I say "thanks" in response. But it is just as obvious that if a historian were to someday write my biography, he or she would not include such episodes. Nor would anyone think that my story would be incomplete if they were missing. Yet one should resist the temptation to infer from this sort of thing that physical episodes never belong to the person in the sense that they are not "internal" to her self. Changes in personality structure wrought by brain injuries, for example, make this perfectly clear. Similarly, objects and events located outside one's epidermis are external or foreign in an obvious literal sense. Yet in other respects this is not always the case. There is a perfectly meaningful and uncontroversial sense in which some things that happen to my wife or my best friend also happen to me, even partially constitute who I am. Analogously, the rules, standards, and ideals that characterize the various institutions and practices in which we participate are outside each of us in one sense, yet we sometimes meaningfully describe them as "ours" in another respect. In cases where we do not, we sometimes say that we are "alienated" or "estranged" from them, and such language indicates that something is amiss. It often indicates the recognition that one can expand the boundaries of one's self by belonging to something that exists outside of it. Properly belonging to something external can alter the criteria of what is to count as internal. (35)

Of course, this barely scratches the surface. The point is that it is difficult to demarcate the self. Perhaps it is impossible to do so in a fixed way. The self expands and contracts, as it internalizes and disavows, accepts and rejects--sometimes consciously, sometimes unconsciously. As a result we draw the line between inside and outside in a variety of ways for a variety of purposes. (36) What we need to know now, however, is how Kant draws this line. What version of the distinction between self and other underwrites his conception of the difference between autonomy and heteronomy? Kant does not offer a direct answer to this question. I am not sure that he always looked at the issue this way. But, assuming I am right about the need to clarify the distinction between self and other in order to clarify the difference between autonomy and heteronomy, any attempt to understand his conception of autonomy calls for a reconstruction of the demarcation project that underlies it.

Historically, philosophers have often distinguished between what is internal and what is external to the self on the basis of substance dualism. This doctrine provides a quick and easy criterion for demarcating the boundaries of the self in a fixed and a priori manner. Simply put, according to this cartesian strategy, that which belongs to the body is outside the self, while that which belongs to the mind belongs inside it. (37) Whatever one thinks of the cartesian strategy, it is clear that Kant would not accept it. As we know from the Critique of Pure Reason, in particular the Refutation of Idealism and the Paralogisms, he regards substance dualism as untenable. He does not quite reject it as false; instead, he undermines the philosophical enterprise from which it springs. In its place he puts a healthy agnosticism regarding the issue. We can find this view expressed in his practical philosophy as well. In the Metaphysics of Morals, for example, he argues that, "neither experience nor inferences of reason give us adequate grounds for deciding whether the human being has a soul (in the sense of a substance dwelling in him, distinct from the body and capable of thinking independently of it, that is a spiritual substance), or whether life may not well be, instead, a property of matter." (38)

How, then, does Kant circumscribe the boundaries of the self for the purposes of developing the concept of autonomy? Many passages in his work suggest that the self/other or inner/outer distinction lines up evenly with the noumenal/phenomenal distinction. For example, when he describes the human being as a being endowed with "inner freedom" he characterizes each of us as a "homo noumenon." (39) Much of what he says suggests that the Kantian strategy of demarcation is of a piece with the "two-worlds" metaphysics that sometimes seems to underwrite his moral theory as a whole. In the second Critique, for example, he says that, "causality through freedom must always be sought outside the sensible world in the intelligible world." (40) This points to an "intelligible substratum in us." (41) The spatial imagery here is quite complex: that which is "outside" the sensible world is actually "in" us, qua intelligible characters or noumenal selves. Because we are rational beings (vernunftige Wesen), the internal contents of the sphere of our inner selves are outside the empirical world within which we live as empirical or animal beings.

The interpretation I want to propose rests on reading such language figuratively and taking Kant at his word when he says that "the concept of a world of understanding is ... only a standpoint." (42) I must admit that the texts by themselves do not decide unequivocally in favor of this interpretation. There are undeniably good textual reasons for taking the talk of "worlds" literally. But I think the exegetical virtues of this interpretation are cancelled out by its philosophical vices. For this reason, I ask the reader to indulge what is generally called the "two-standpoints" interpretation as he or she examines the remarks I shall marshal in support of my proposal. (43)

First consider the following passage from the Groundwork:

 So it is that the human being claims for himself a will which lets
 nothing be put to his account that belongs merely to his desires
 and inclinations, and on the contrary thinks as possible by means
 of it--indeed as necessary--actions that can be done only by
 disregarding all desires and sensible incitements [sinnliche
 Anreizungen]. The causality of such actions lies in him as
 intelligence and in the laws of effects and actions in accordance
 with principles of an intelligible world, of which he knows nothing
 more than that in it reason alone gives the law, and ... it is
 there, as intelligence only, that he is his proper self. (44)

An implicit demarcation strategy is clearly at work here. Kant draws the line between self and other sharply and decisively. On the side of the "proper self" (das eigentlich Selbst) (45) we have intelligence, reason, and will. And on other side we have desires, inclinations, and all "sensible incitements." This could reasonably be read as just another version of the cartesian strategy; one might perfectly well expect an eighteenth-century philosopher to group inclinations, desires, and other impulses under "body" or "extension" and then group will, intelligence, and reason under "mind" or "thought." But we should resist this interpretation. We know that Kant rejects substance dualism. Moreover, this passage should be read in the context of others that make clear this is not what he has in mind.

Further down the same page, Kant continues by claiming that:

 What inclinations and impulses ... incite him to cannot infringe
 upon the laws of his volition as intelligence; indeed, he does not
 hold himself accountable for the former or ascribe them to his
 proper self, that is, to his will, though he does ascribe to it
 [his will] the indulgence he would show them if he allowed them to
 influence his maxims to the detriment of the rational laws of his
 will. (46)

Here we can see that the distinction between what is internal and external to the proper self is not made metaphysically. That is, it is not the result of classifying inclinations and desires, on the one hand, and will and reason, on the other, in terms of their respective essential attributes (for example, "extension" and "thought"). Instead, the line is drawn from the first-person perspective by means of acts of identification and disavowal. (47) It is we as deliberating, judging, acting individuals--as agents--who draw the line. We draw the boundaries around our proper selves by disavowing the incentives provided by our inclinations and desires, by refusing to endorse them as sufficient reasons for action or guides to conduct--that is, by not "ascribing them" to our "proper selves." To be sure, my desires and inclinations provide me with incentives; they have motivational force and they can provide me with reasons (assuming I "incorporate" them into my maxims). But I am capable of questioning their directives, and I am capable of denying that they belong to my proper self. When I disavow them this way, by denying that they are part of me, I constitute, at least partially, my proper self by identifying instead with the directives supplied by my intelligence, my reason, and my will.

This act of identification is normative in the sense that it is at bottom evaluative. In identifying with the reasons supplied by my intelligence and will and disavowing those supplied by my inclinations I express my commitment to what I take to be the greater value of the former. I thereby show that I place greater stock in what my intelligence provides reason for than I do in my desires and inclinations. That Kant subscribes to such a normative demarcation strategy is evident in passages such as the following, which couches the issues in explicitly evaluative terms. Reflecting on what happens when an agent deliberates about action from the moral perspective, Kant writes:

 With a will free from impulses of sensibility he transfers himself
 in thought into an order of things altogether different from that
 of his desires in the field of sensibility ... ; [from this] he can
 expect only a greater inner worth of his person. This better person
 [Diese bessere Person], however, he believes himself to be when he
 transfers himself to the standpoint of a member of the world of
 understanding, as the idea of freedom, that is, of independence
 from determining causes of the world of sense, constrains him
 involuntarily to do; and from this
 standpoint he is conscious of a good will that, by his own
 acknowledgements, constitutes the law for his evil will as a member
 of the world of sense--a law whose authority he is cognizant even
 while he transgresses it. The moral ought is then his own necessary
 will as a member of an intelligible world, and is thought by him as
 ought only insofar as he regards himself at the same time as a
 member of the world of sense. (48)

This long dense remark raises many questions that cannot be answered here. For our purposes, what is most interesting is the way in which it makes what I am calling the normative strategy seem so central to so many fundamental issues in Kant's moral theory, particularly to the question of autonomy. The Verstandeswelt ("intelligible world") is "only a standpoint" and I, as a deliberating and judging agent, adopt that standpoint when I take myself for this "better person" that I find articulated in the idea of a will that is motivated by the demands of the moral law rather than by incentives that stem from desires and inclinations. In doing this, I identify myself with the demands of morality; I take them to be expressions of my own volition. More precisely, I take them to be expressions of the ideal to which I aspire and about which I care deeply. (49) That is why the imperative of morality (das moralische Sollen) is not foreign to my will, something imposed upon me from outside.

According to this view, the distinction between what is internal and what is external to the self is not metaphysical. It is normative in that it reflects a set of commitments concerning the respective values of the various sources of practical reasons. Reasons that stem from the pure will (or pure practical reason) are considered more valuable or important than those that stem from the incentives provided by desire and inclination. They express our better selves. Kant's term for this ideal self is "moral personality." (50) To be a person in a moral sense is to possess several characteristics and capacities: a person is a "lawgiving being." (51) In fact, as a "subject of pure practical reason," a person is "the supreme lawgiver." (52) A person, unlike a mere thing, is also an "end in itself." (53) And a moral person is a being that is accountable or responsible (zurechnungsfahig) for its actions. (54) According to Kant, when we think of an individual human being in terms of his moral personality, we think of his "invisible self," (55) his "freedom and independence from the mechanism of the whole of nature," (56) and his "freedom [as] a rational being under moral laws." (57)

While Kant defines "humanity" as "the capacity to set oneself an end--any end whatsoever," (58) he thinks of moral personality as "the idea of humanity considered wholly intellectually." (59) So we can say that the idea of a moral person is the idea of an individual who sets ends and reasons about how to achieve them on the basis of "formal" rather than "material" principles, someone who decides and acts on the basis of what practical reason recognizes as right, rather than on the basis of sensuous inclinations and idiosyncratic interests. This is the idea of who each of us would be if we had the strength and goodness of will (virtue) to regard the demands of morality as a sufficient and overriding guide to conduct. In short, this better person we believe ourselves to be when we "transfer ourselves in thought" to the standpoint of the intelligible world is a conception of ourselves as autonomous.

This concept of moral personality works together with Kant's notion of "respect" to solve the problem of heteronomy. Recall our earlier discussion of the internalism requirement on reasons for action. According to this idea, R can play the role of a practical reason for a rational agent, only if R could, to some degree at least, motivate that agent to act in light of it. Reasons for action, in other words, can count as genuine reasons only if they include some motivational component. Without this, R cannot function as a reason for the agent. Typically, the reasons that characterize heteronomous volition easily meet this requirement. They concern ends that are set by incentives that stem from familiar motivational sources such as desire and inclination. It is no mystery, for example, how my wanting a drink can furnish me with a reason to order a beer. Nor is it hard to see how my hoping for a comfortable retirement gives me a reason to invest wisely now. This picture does not assume that desires or inclinations cause the decision or action in question. On the contrary, it assumes that the Incorporation Thesis is true. We are "affected" but not "determined" by our sensible impulses; (60) we can step back from them and ask whether they provide considerations in favor of or against some course of action. But if we are only heteronomous then this is all we can do. I doubt that would make us slaves to our passions, but it certainly would limit practical reason to a narrower scope than we might like. According to Kant it would rule out the possibility of categorical imperatives and prevent us from fulfilling our "highest vocation." This is the sense in which empiricism threatens to "degrade" our humanity.

The difficult question for Kant's view is whether the reasons that characterize autonomous volition can also meet the internalism requirement. The question, in other words, is whether there really can be reasons of this sort. Can a consideration of pure practical reason really be a reason for someone like you or me? If not, then Kantian autonomy is impossible. Recall that, according to Kant's official definition, the autonomy of the will is "the will's property of being a law to itself." (61) Less technically, an individual is autonomous insofar as he or she is capable of self-legislation or self-determination. A person exercises this capacity by acting or deciding on the basis of reasons that stem from purely rational considerations--"formal" rather than "material" principles--regarding how one ought to behave. The motivational skeptic doubts that such considerations could possibly have a motivational component. The only thing that could motivate anyone, even a rational agent, is a feeling of some sort. Our cognitive grasp of a general principle, in other words, will not actually move us in any particular direction unless it is contains or is accompanied by some sort of affective component. What is doubted here is the idea that we could ever actually be moved to do anything once we have subtracted from our deliberations all inclination and desire. If the motivational skeptic is right about this, and the internalism requirement is true, then, even if we have the capacity to grasp the implications of the moral law for our behavior, that grasp could not actually provide us with genuine reasons for action.

This is where the notion of respect comes in. Kant agrees with the empiricist of practical reason--the motivational skeptic--that motivation requires an incentive. And he thinks this requires a feeling of some sort. We must feel something in order to be moved to act. But he rejects the claim that reason cannot play this role. According to Kant, a "feeling" (Gefuhl) is a modification of "sensibility" (Sinnlichkeit). It is "the effect of a representation [Vorstellung] ... upon a subject." (62) So when one has a representation of something--say, a sensation of red--that representation may produce in the perceiver a subjectively felt response. The sensation of red may, for example, trigger some degree of irritation or anger. Kant also says that the representation that produces the feeling may be either "sensible or intellectual." (63) Thus, his claim is that it is not just our traffic with empirical objects and our sensuous representations that produce feelings in us. On the contrary, even intellectual episodes can do this. Respect is an example of this sort of feeling, in that it is produced by a purely intellectual or rational grasp of the demands of morality. It is difficult to characterize this feeling precisely. Kant says that feeling in general is "the capacity for having pleasure or displeasure in a representation," (64) but respect is not a clear case of either. It seems to straddle the border between the two; rather, it seems to begin as a feeling of displeasure and then give way to a sort of pleasure, one that we "can't get enough of." (65)

For our purposes, the key claim is that it is the feeling of respect "which connects a ground for determining choice to this action subjectively with the representation of the law." (66) Respect thus functions as an incentive (Triebfeder) to adopt the moral law as one's maxim. (67) Respect, in fact, is the moral incentive. Kant's thought here is that as incentive the feeling of respect is capable of motivating an individual "subjectively" to subordinate the principle upon which she acts (her maxim) to the demands of the moral law. Respect, in other words, is the affective component that is capable of making the moral law a reason for me to act in a particular way. (68) This is not the place to defend the doctrine of respect in full, but if Kant is right about this, then this explains how purely rational considerations--namely, those that result from drawing conclusions about the implications of the moral law for one's conduct--can meet the internalism requirement and thus function as genuine reasons for action. For, if the doctrine of respect is plausible as a piece of moral anthropology, then it explains how "formal" considerations can include a motivational component.

The concepts of respect and moral personality go hand in hand. The ideal self is the object of respect. This is obscured by Kant's talk of "respect for the moral law," an expression which has settled into one of the stock phrases of Kantian moral philosophy. The problem is not that this is wrong, but that it is potentially misleading. Talk of "respect for the law" tends to encourage the impression that respect has to do with something abstract, impersonal, and transcendent. And this cannot be right. While the moral law is certainly not subjective in the sense of being private or idiosyncratic, it comes from within. Let us not forget one of the most oft-quoted and moving passages in all of Kant's work: "Two things fill the mind with ever new and increasing admiration and reverence, the more often and more steadily one reflects on them: the starry heavens above me and the moral law within me." (69) The rhetorical symmetry of "uber mir" and "in mir" makes it clear where the law resides. This is why Kant explicitly says in the Groundwork that "our own will ... is the proper object of respect." (70) Yet, it is not simply the will as it is found--characterized by volition on the basis of prudential or instrumental reasons, and adopted in response to ends set by inclination-based incentives. Rather, it is "our own will insofar as it would act only under the condition of a possible giving of universal law through its maxims." (71) In other words, it is the will considered as moral personality, our better self, which Kant describes as "this respect-inspiring idea of personality" (diese Achtung erweckende Idee der Personlichkeit). (72) This idea of an ideal version of the self is capable of "setting before our eyes the sublimity of our nature (in its vocation) while at the same time showing us the lack of accord of our conduct with respect to it and thus striking down self-conceit." (73) There is an important sense, then, in which respect for the law is really self-respect, and self-respect has both a negative and a positive component; it motivates by providing both a sense of shame (a feeling that "humbles" or "humiliates") (74) and an object of aspiration.

This account of the relation between moral personality and the moral feeling of self-respect constitutes Kant's solution to the problem of heteronomy. When you think of yourself in terms of your moral personality you think of yourself as actively autonomous or self-determining. Kant's claim is that thinking of yourself in this way can motivate you to be autonomous. Self-respect provides the incentive to adopt the moral law as the basis of one's volition. It thus provides the incentive to act on the basis of purely "formal" or rational considerations, to abstract from sensibility in order to treat the demands of morality as sufficient reasons for action. In this manner, self-respect makes possible the exercise of autonomy, wherein the will or pure practical reason is a law unto itself. This means that our capacity to exercise autonomy rests on our ability to think of ourselves as autonomous--to imagine ourselves in terms of our moral personality, and then identify with and be moved by the idea of that better version of ourselves. You cannot actually be autonomous unless you can first think of yourself that way. For only by doing this can you awaken the feeling of respect for the moral law that can provide you with the motivation to obey that law.

On this view, the capacity for autonomy does not rest on transcendental freedom or freedom in the "cosmological sense." (75) It does not require the ability to exempt oneself from the natural order and then function as an uncaused cause. Instead, it rests on the capacity to identify with one's "better person" and distance oneself from or even disavow the dictates of one's desires and inclinations. One might reasonably object that this view rests on an a priori and unjustified disparagement of feeling and sentiment. Kant thinks that "all inclination and every sensible impulse is based on feeling," (76) and his attitude toward feeling has earned him a bad reputation among philosophers who believe that he has drastically underestimated the moral importance of sentiment, passions, desires, and "affect" in general. It is tempting to think that the normative strategy sketched above reflects an unwarranted disparagement of a valuable and necessary part of moral and practical life. It is easy to find passages that seem to encourage this reading. In the Groundwork, for instance, Kant claims that, "the inclinations themselves ... are so far from having an absolute worth ... that it must instead be the universal wish of every rational being to be altogether free of them." (77) In a similar vein, he describes them in the Critique of Practical Reason as "blind and servile," arguing, in a passage that seems to confirm his critics' worst suspicions, that the feeling of compassion and tender sympathy "is itself burdensome to right-thinking persons" who "wish to be freed from them [that is, inclinations altogether] and subject to lawgiving reason alone." (78)

But Kant's view is not so straightforward. He does not, in fact, simply label inclination in general as "bad"--in contrast with pure practical reason, which is "good." (79) His concern is not with inclination as such. It is with our attitude towards our inclinations and desires. That is, he is interested in how we treat the latter in our practical reasoning, the roles we assign them during the course of deliberating about how we ought to behave. "Considered in themselves," he argues, "natural inclinations are good, that is, not reprehensible, and to want to extirpate them would not only be futile but harmful and blameworthy as well." (80) Indeed, inclinations would seem to belong to "the predisposition to animality in the human being" which Kant regards as essential to the "original predisposition to good in human nature." (81) The question is how we, as deliberating and judging individuals, incorporate them into our maxims. The question, in other words, is what we do with them. In fact, for Kant, the difference between good and evil turns precisely on this question. (82)

I realize that this response goes only so far. It is undeniably true that Kant values the will (practical reason) more than inclination and desire, or sensibility in general. He also argues that all rational agents do as well. We cannot help but care more about our reason than our inclinations as sources of practical reasons. Only the former connects us to our "highest vocation" as human beings, and only it can function as the source of moral obligation, and, therefore, of duty. These issues go right to the heart of his entire practical philosophy. His devaluation of sensibility in favor of reason stems from his conviction that moral requirements can only be expressed in the form of categorical imperatives and from the idea that the autonomy of the will is the basis of morality in general. This is certainly not the place to defend either of these deeply Kantian commitments. I can only hope that what I have said in response to the above objection may give some critics pause.

According to Thomas Nagel, we want to know whether or not we have the capacity for autonomy because "we want to be able to will the sources of our actions down to the very bottom." (83) My account of Kantian autonomy does not scratch this particular itch. I have argued that Kant does not rest the capacity for autonomy on transcendental freedom. Instead, it rests on the capacity for identification--more specifically, on the ability to identify with and be motivated by the thought of one's self as a moral person. However, this does not address the question of how we come to have this idea of moral personality in the first place and why we might find ourselves disposed to identify with and be motivated by it--to feel respect for it. In response I can only admit that I do not have the answer to this question. Despite his sense that self-love usually wins out in the end, Kant seems confident that the presence of this idea of ourselves as autonomous moral agents and the capacity to be moved by it are universal. I find it impossible to share that confidence. Yet I also think they are more common than a widespread cynicism regarding practical reason would have us believe. And I think that any theory of practical reason that cannot accommodate these concepts is hopelessly impoverished. At any rate, their origins must be at least in part psychological and sociological. What is needed is a story about how the idea of moral personality and the disposition to feel respect for it can result from the internalization of standards provided early on by parents and other sources of authority. (84) It is true that this sort of answer still would not give us what Nagel says we want. But I doubt any account of autonomy could do that. Kant himself seemed to regard the ultimate ground of freedom as inscrutable, and I tend to think he is right. (85) But that should not prevent us from digging as deep as we can. (86)

Loyola College in Maryland

(1) Immanuel Kant, Critique of Pure Reason, translated and edited by Paul Guyer and Allen Wood (New York: Cambridge University Press, 1998), A533/ B561, hereafter KrV. Except for the KrV, which is cited according to the standard "A/B" pagination, all references to Kant are to the volume and page of Kant's Gesammelte Schriften, ed. Koniglichpreusssischen Akademie der Wissenschaften (Berlin and Leipzig: Walter de Gruyter, 1902--).

(2) Ibid. A803/B831-A804/B832.

(3) Immanuel Kant, "Review of Schulz's Attempt at An Introduction to a Doctrine of Morals for all Human Beings regardless of different religions," in Immanuel Kant, Practical Philosophy, translated and edited by Mary Gregor (New York: Cambridge University, 1996), 8:13, hereafter Schulz.

(4) Kant does not, to my knowledge, distinguish between fatalism and determinism. He seems to take the threats to freedom posed by the one as equivalent to those posed by the other.

(5) Schultz, 8:13.

(6) Ibid. Unless otherwise noted, all emphases belong to the quoted author.

(7) Immanuel Kant, Groundwork of the Metaphysics of Morals, translated and edited by Mary Gregor (New York: Cambridge University, 1997), 4:455-6, hereafter GMS.

(8) Ibid. 4:427.

(9) Immanuel Kant, The Metaphysics of Morals, translated and edited by Mary Gregor (New York: Cambridge University Press, 1996), 6:381, hereafter MdS.

(10) Ibid. 6:384-5.

(11) Ibid. 6:385.

(12) Immanuel Kant, Critique of Practical Reason, translated and edited by Mary Gregor (New York: Cambridge University Press, 1997), 5:34, hereafter KpV.

(13) GMS 4:437.

(14) MdS 6:395.

(15) Ibid. 6:392. It is a mistake to think that the focus on ends drops out in Kant's moral theory. He, in fact, defines ethics as "the system of the ends of pure practical reason" (Ibid. 6:381). For important criticisms of the deontological reading of Kant's moral theory, see Barbara Herman, The Practice of Moral Judgment (Cambridge, Mass.: Harvard University Press, 1993) and Allen Wood, Kant's Ethical Thought (New York: Cambridge University Press, 1999). For an insightful discussion of the importance of ends to the Categorical Imperative, see Jens Timmerman, "Value Without Regress: Kant's 'Formula of Humanity' Revisited," European Journal of Philosophy 14, no. 1 (2006): 69-93.

(16) At one point in the Groundwork Kant distinguishes between incentives and motives: "The subjective ground of desire is an incentive [die Triebfeder]; the objective ground of volition is a motive [der Bewegungsgrund]; hence the distinction between subjective ends, which rest on incentives, and objective ends, which depend on motives, which hold for every rational being" (GMS 4:427). Roughly, it seems that a motive is an incentive that has been endorsed in some sense. Nothing in my interpretation turns on this distinction, so it can be safely ignored for the purposes of the present essay.

(17) KpV 5:72.

(18) Henry Allison, Kant's Theory of Freedom (New York: Cambridge University Press, 1990). Allison convincingly argues that the view captured in this Thesis is not unique to the Religion book or to Kant's later moral theory. On the contrary, it expresses a line of thought about agency already present in Groundwork of the Metaphysics of Morals.

(19) Immanuel Kant, Religion within the Boundaries of Mere Reason and Other Writings, trans, and ed. by Allen Wood and George di Giovanni (New York: Cambridge University Press, 1998), 6:23-4, hereafter Religion.

(20) Christine Korsgaard, Sources of Normativity (New York: Cambridge University Press, 1996), 7-21.

(21) On Allison's reading, the Incorporation Thesis assumes that we are transcendentally free. Though it is not without textual support, I do not think this interpretation is necessary. The Thesis assumes the reality of "negative" practical freedom, that is, the idea that our volition is "affected" but not "determined" by sensibility. But this by itself does not require the "total independence" from nature allegedly provided by transcendental freedom. All it requires is that we have the capacity to distance ourselves from our inclinations in order to consider how we should respond to them. On such a view we are certainly connected to nature, just not in a merely mechanical way.

(22) As Gilbert Harman presents the view, "practical reasoning is always means-ends reasoning ...; you can come to want something as a means to something else you already want. But ... you cannot reason yourself into having something new as an ultimate end, since you always reason from your ends to things that are means to those ends." The Nature of Morality (New York: Oxford University Press, 1977), 31.

(23) Actually, pushed to extremes, views that limit practical reason to instrumental reasoning about means seem to rule out even this. Consider David Hume's famous remark: "'Tis not contrary to reason to prefer the destruction of the whole world to the scratching of my finger. 'Tis not contrary to reason for me to chuse my total ruin, to prevent the least uneasiness of an Indian or person wholly unknown to me. 'Tis as little contrary to reason to prefer even my own acknowleg'd lesser good to my greater, and have a more ardent affection for the former than the latter." A Treatise of Human Nature (New York: Oxford University Press), 267, hereafter Treatise. The claim seems to be that if there is no stronger passion to counteract the preference for world destruction, then practical reason, being limited to the selection of means, cannot offer any considerations that would count against adopting such an end.

(24) Kp V 5: 70.

(25) Ibid. 5:71.

(26) Christine Korsgaard, "Skepticism About Practical Reason," in Creating the Kingdom of Ends (New York: Cambridge University Press, 1996), 311-34. My discussion of this issue draws directly from Korsgaard's influential essay.

(27) Korsgaard, "Skepticism About Practical Reasoning," 317.

(28) Hume, Treatise, 266.

(29) GMS 4:444.

(30) Ibid. 4:444.

(31) KpV 5:33. I do not mean to suggest that Kant views motivational skepticism, or the empiricism of practical reason, as the only threat to autonomy. On the contrary, he argues that traditional rationalist concepts of morality and practical reason also endanger the autonomy of the person. Empiricism is only one version of the problem. For a concise and illuminating discussion of this issue, see Allen Wood's "The Supreme Principle of Morality," in The Cambridge Companion to Kant and Modern Philosophy, edited by Paul Guyer (New York: Cambridge University Press, 2006), 342-80.

(32) GMS 4:444.

(33) GMS 4:446, 448.

(34) MdS 6:418-420.

(35) For an interesting account of how demarcation relates to social practices, see Meir Dan-Cohen, "Socializing Harry," in Harry Frankfurt, Taking Ourselves Seriously, Getting it Right (Stanford: Stanford University Press, 2006).

(36) Commentators often fail to appreciate the prevalence of this issue in Kant's work as a whole. His discussion of the self, and the problem of how to determine its structure, is by no means limited to his theoretical treatment of cognition in the Critique of Pure Reason. Throughout his corpus, Kant seems concerned to explore the ways in which the structure of the self is determined by a complex interplay of the active and the passive, the internal and external, the individual and the social. One important place where this can be seen is his treatment of the sensus communis in the Critique of Judgment. Rudolf Makkreel argues convincingly that the idea of a sensus communis has implications not just for aesthetic judgment, but also for reflective judgment as a whole. According to Makkreel, this idea, which underwrites the capacity for reflective judgment, allows us to "abstract from the private empirical aspects of our subjective representations in order to generate what might be called a communal or intersubjective perspective." Makkreel, Imagination and Interpretation in Kant (Chicago: University of Chicago Press, 1990), 158. Also see Stanley Cavell, "The Aesthetic Problems of Modern Philosophy," in his Must We Mean What We Say? (New York: Cambridge University Press, 2002). Kant also treats this issue in his discussion of conscience in the Metaphysics of Morals. See MdS 6:437-40, where he seems to argue that the force of conscience is explained by our having internalized some sort of external authority figure, such as God.

(37) I write "cartesian" instead of "Cartesian" in order to avoid attributing this view to Descartes himself, who held a more complex view of the self than he is often given credit for.

(38) MdS 6:419.

(39) Ibid. 6:418.

(40) KpV 5:105, my emphasis.

(41) Ibid. 5:99, my emphasis.

(42) GMS 4:458.

(43) Thanks to the work of commentators such as Lewis White Beck, Henry Allison, Christine Korsgaard, and Thomas E. Hill, Jr., it has become quite common, particularly in English-speaking circles, to interpret Kant's talk of two "worlds" in terms of the concept of a standpoint or practical perspective. My reading owes a significant debt to this tradition of commentary. To my knowledge, Allison's Kant's Theory of Freedom represents the most carefully and thoroughly worked-out version of the approach. I cannot do justice to his influential work here, but I should note that my version of the two-standpoints interpretation differs from his in at least one very important respect. Allison proposes this reading in order to offer a nonmetaphysical interpretation of the notion of transcendental freedom. He thinks autonomy rests on the capacity for such freedom, but that making sense of latter does not require taking literally the notions of "intelligible world" and "noumenal self." In contrast to Allison, I adopt the two-standpoints reading in order to detach the question of autonomy from that of transcendental freedom. The reason why is that I do not see the point in retaining the latter notion once we turn away from the metaphysics that seems to require it. My thinking on Allison's account owes much to Karl Ameriks's critique in "Kant and Hegel on Freedom: Two New Interpretations," reprinted in his Interpreting Kant's Critiques (New York: Oxford University Press, 2003). I should add, however, that Ameriks and I draw very different conclusions about the limitations of Allison's approach.

(44) GMS 4:457, my emphasis.

(45) It is tempting to translate "das eigentliche Selbst" as "the authentic self," but this would import issues into the discussion that are best left aside for now. There is no doubt, however, that the concepts of autonomy and authenticity have much in common, and that Kant's work provides a fruitful opportunity to explore the ways in which they overlap. The road from Rousseau to Nietzsche certainly passes through this juncture in Kant's moral philosophy.

(46) GMS 4:457-8, my emphasis.

(47) My thinking on the issues at stake here owes a good deal to Harry Frankfurt's work on the concept of "identification," developed in a number of essays reprinted in The Importance of What We Care About (New York: Cambridge University, 1988) and Necessity, Volition, and Love (New York: Cambridge University Press, 1999).

(48) GMS 4:454-5, my emphasis.

(49) I take myself here to be developing a point made by Korsgaard. See The Sources of Normativity, 238.

(50) The moral concept of personality should not be confused with the psychological concept. Kant defines the latter as "the ability to be conscious of one's identity in different conditions of one's existence" (MdS 6:223). As he puts in the Critique of Pure Reason, "what is conscious of the numerical identity of its self in different times, is to that extent a person" (A361). This pertains to the sense of personhood that is so central to traditional questions regarding the substantiality of the soul, personal identity, and so on--questions pursued by Descartes, Locke, Hume, and Leibniz, among others.

(51) GMS 4:438.

(52) KpV 5:75 and GMS 4:434-5.

(53) GMS 4:428.

(54) MdS 6:223.

(55) KpV 5:162.

(56) Ibid. 5:87.

(57) MdS 6:223 and KpV 5:87.

(58) MdS 6:392 and GMS 4:437. For further discussion of the concept of "humanity," see Thomas E. Hill, Jr., "Humanity as an End in Itself," in his Dignity and Practical Reason in Kant's Moral Theory (Ithaca: Cornell University Press, 1992), 38-57. Also see chapter four of Allen Wood, Kant's Ethical Thought.

(59) Religion 6:27-8.

(60) MdS 6:213 and KrV A534/B562.

(61) GMS 4:446.

(62) MdS 6:212n.

(63) Ibid.

(64) Ibid. 6:211.

(65) KpV 5:78.

(66) MdS 6:218.

(67) KpV 5:76.

(68) Kant, it must be admitted, would not be comfortable with the appearance of the word "affective" in this context. He defines "affects" (Affekte) as feelings that precede rather than follow deliberation (Uberlegung). Far from motivating moral action, affects impede it. See MdS 6:408-409 and Anthropology from a Pragmatic Point of View, trans. and ed. by Robert B. Lowden (New York: Cambridge University Press, 2006), [section]75.

(69) KpV 5:161.

(70) GMS 4:440.

(71) Ibid. 4:440.

(72) KpV 5:87.

(73) Ibid. 5:87.

(74) Ibid. 5:74.

(75) KrV A533/B61.

(76) KpV 5:72.

(77) GMS 4:428.

(78) KpV 5:118. This well known remark makes it tempting to read Kant as a modern Stoic. Compare Seneca, when he writes of "affection" that "the enemy, I say, must be stopped at the very frontier; when he has invaded and rushed on the city gates, there is no 'limit' which his captives can make him accept." See "On Anger," p. 26 in Seneca: Moral and Political Essays, translated and edited John M. Cooper and J.F. Procope (New York: Cambridge University Press, 1995). But as we shall presently see, this comparison is, while apt in important respects, misleading in others.

(79) For a careful and thorough treatment of this issue, see Marcia Baron, Kantian Ethics Almost Without Apology (Ithaca: Cornell University Press, 1995), 194-226. She draws extensively on the Religion book and the Anthropology, in addition to his mainstream works in moral theory, in order to present a picture of Kantian moral psychology nuanced enough to do justice to the role of feeling in our practical lives. For a discussion of how moral feelings relate to aesthetic feelings, see Rudolf Makkreel, "Sublimity, Genius, and the Explication of Aesthetic Ideas," in Kants Asthetik/Kant's Aesthetics/ L'esthetique de Kant, edited by Herman Parret (Berlin: Walter de Gruyter, 1998), 615-29. In Imagination and Interpretation in Kant, Makkreel argues convincingly that feeling, for example, the notion of Lebensgefuhl, is much more central to the entire critical project than is normally thought. An important lesson of Makkreel's work is that Kant's attitude both toward respect and toward the inclinations should not be surprising at all. Once we put his remarks about inclinations in their proper context, which includes the Critique of Judgment and the essays on history and orientation, we can appreciate the fact that Kant rarely disparages feeling as such. On the contrary, it is essential to his project as a whole. It is true that he refuses to base morality on feeling, but that does not mean that he ever intends to eliminate it. The centrality of respect makes that quite clear, as does his discussion of the "original predisposition to good" in the Religion 6:26-8. For an interesting discussion of Kant's attitude toward sentiment and the "Age of Sensibility," see Ernst Cassirer, Kant's Life and Thought, trans. James Haden (New Haven: Yale University Press, 1981), 268-70.

(80) Religion 6:58.

(81) Ibid. 6:26-7.

(82) This marks an important difference between the Kantian and the Stoic views. Compare, again, to Seneca: "The affections are no less evil as subordinates than they are as commanders." See "On Anger," 27. This is not, of course, to deny affinities between Kant and the Stoics.

(83) Nagel, The View from Nowhere (New York: Oxford University Press, 1986), 135.

(84) For a fascinating attempt at this, see J. David Velleman, Self to Self: Selected Essays (New York: Cambridge University Press, 2006), especially chapters 4, 5, and 6.

(85) Nagel himself shares this pessimism as well. He differs from Kant, though, in that Kant seems, at least sometimes, confident in his conviction that the theoretical issue really does not matter for the purposes of practical philosophy. Nagel thinks the practical and theoretical perspectives are more continuous with each other than Kant does. Our tendency to slide from the former into the latter is natural and unavoidable.

(86) I would like to thank Paul Abela, Rudolf Makkreel, and Andrew Norris for their insightful comments on this essay.

Correspondence to: Loyola College in Maryland, 4501 North Charles Street, Baltimore, MD 21210-2699.