EMOTION, REASON, AND ACTION IN KANT

Though Kant never used the word ‘emotion’ in his writings, they are of vital significance to understanding his philosophy. In this talk I offer a reading for the emotion in his work, taking into account its many manifestations in Kant`s work including affect and passion. I explore how, in Kant’s world view, our actions are informed, contextualized and dependent on the tension between emotion and reason. On the one hand, there are positive moral emotions that can and should be cultivated. On the other hand, affects and passions are considered illnesses of the mind, in that they lead to the weakness of the will, in the case of affects, and evil, in the case of passions. Seeing the role of these emotions enriches our understanding of Kant’s moral theory. Key-words: emotion, sympathy, action, affect, passion Recebido em: 06/2018 Aprovado em: 07/2018


Introduction
Th is is a book about practical reason, action, and emotion in Kant In his work, there is a continuum from uncontrollable emotions, like anger, to those which can be cultivated and rationally controlled. Th e voluntariness and involuntariness of emotions as well as their capacity for being cultivated depends on their relation to the passive, reactive, or active self. Second, I argue that Kant's account of emotion includes both physiological aspects and cognitive contents, mainly evaluative beliefs. However, the variety of emotions presents us with diff erent proportions of these two elements. I conclude that Kantian moral theory contributes an outstanding theory of emotions to contemporary debates, one which acknowledges physiological as well as cognitive aspects, without forgetting their diversity.
In Chapter 5, I discuss Kant's theory of aff ects, particularly the possibility of controlling them. I claim that aff ects are not easy to control and some are even uncontrollable through the power of the mind. Th e possibility of controlling aff ects depends upon a mild temperament. Although in some cases Kant admits cultivation of character, the limits of this cultivation will depend on the natural temperament of the agent. I will argue that Kant's theory of aff ects is connected to the seventeenth-century physiology idea of excited states, which make aff ects diffi cult to control merely by the force of the mind.
In Chapter 6, I analyze virtue as a cure for strong aff ects, and refi nement as a propaedeutic to virtue. Kantian virtue is apart from happiness and does not aim at any telos , such as the achievement of a happy life. If virtue does not lead necessarily to happiness or pleasure, what is then virtue? What is the aim of Kantian virtue? Kantian virtue is the fortitude, strength, and selfconstraint to attain full rationality. Th e development of virtue over time will be necessary to control instances of outbursts of feelings that could oppose the accomplishment of moral actions. In Chapter 7, I investigate the aesthetical conditions for morality. I will begin with Critique of Judgment and the thesis of § 59, stating that the beautiful can be considered a symbol of the morally good. Aft er that, I try to sort out the relation between refi nement and morality. I also ask about the presence of the feelings of pleasure and displeasure in morality and analyze moral feeling in Th e Metaphysics of Morals . I then examine the relation between the realm of taste and the realm of virtues in Anthropology from a Pragmatic Point of View .
In Chapter 8, my aim is to explore the relations Kant establishes between women, emotions, and morality, in order to show that the female sex is useful in the moral education of men. Kant has oft en been criticized for holding a very negative vision of women, according to which they are less rational and less morally valuable than men. In this chapter, I will show that, in spite of some pejorative comments, Kant held that women have some characteristics that can be useful to morality. Th is is due to some qualities of the female sex, mainly women's capacity for self-control and the capacity to have moral feelings like sympathy and compassion. Moreover, women demonstrate their mastery of emotions and passions when they are able to use their emotional sensitivity and self-control to master the feelings and passions of men. Since the moral agent presupposes the capacity of mastering his/her inclinations in order to follow the moral law, at least in this particular area, women seem to fi t this role better than men.
In Chapter 9, I argue that the evil of emotions resides in passions and I try to point to a possible cure for this evil through an ethical community. Kant claims that both aff ects and passions are illnesses of the mind, because both hinder the sovereignty of reason. I show that passions are worse than aff ects for the purpose of pure reason. I then relate aff ects and passions to the degrees of propensity to evil in Religion and I analyze the idea of an ethical community as a way to overcome evil, which goes beyond the political and anthropological solutions suggested by Kant. In Conclusion, I off er an idea of Kantian philosophy that does not deny the reality of human emotion, although it is faithful to the claim that the moral Compulsion can be either pathological or practical, the fi rst is the necessitation of an action per stimulus , and the second is the necessitation per motiva .
Human choice cannot be necessitated per stimulus , since it is an arbitrium liberum . Animals are necessitated per stimulus: "So that a dog must eat if he is hungry and has something in front of him; but man, in the same situation, can restrain himself " (LE, 27: 267). To claim that human choice is an arbitrium liberum is to accept that human beings can only be compelled per motiva , not per stimulus. impulses, although they should be independent of determination by sensory impulses. Th is point will come later, when we discuss the possibility of acting without feelings.

Incentives, motives, and the overdetermination of maxims
Th e distinction between Triebfeder and Bewegungsgrund is crucial in contrasting actions according to duty with actions done from duty, because it underscores the distinction between what constitutes moral motive and incentive. Th e diff erence between moral and nonmoral incentives is explored in the philanthropist example, where we fi nd two agents with diff erent incentives for being benevolent. Neither of them has a motive of vanity or selfinterest; nevertheless, the fi rst has a natural inclination to do good for other persons, an inner happiness in making other people happier. Although his action has a moral motive, Kant maintains that "in such a case an action of this kind, however right and however amiable it may be, has still no moral worth" (G, 4: 398). Hence, a moral motive is a necessary, but not suffi cient, condition for a moral action.
When does an action have moral worth? Kant answers with the case of the second philanthropist (G, 4: 398). Unlike the sympathetic philanthropist, the insensible one performs an action with moral worth; consequently, the absence of sympathy seems to make an action morally worthy. In order to avoid the criticism of insensibility, some authors have discussed what makes an action morally worthy. Th ey were obviously trying to avoid the uncomfortable conclusion that moral insensibility is something good.
One of these authors is Henson. Henson 2 tries to answer two questions related to this example of the philanthropist: (1) What does it mean to ascribe moral worth to an act? (2) Under what circumstances are we to say that one acts from duty?
Th ere are two possible answers in Kant Herman agrees with Henson that it is natural to accept that in a morally worthy action nonmoral motives may be present, as long as they are not the reason for the agent to act. However, she points out that it is not obvious how a motive could be present and yet not operative.
In order to understand Kant's idea of moral worth, Herman makes a distinction between motives, incentives, desires, and causes. Kantian motives are not desires or causes in the sense of vector-like forces. Desires are incentives ( Triebfedern ), not motives for acting. Following this line of reasoning, she concludes that the doctrine of moral worth can accept overdetermination with respect to incentives, not motives.
Paul Guyer disagrees with Herman, in that he sees no reason to admit that we can have overdetermination of incentives: "An incentive cannot coexist or cooperate with any other motive as one suffi cient or independent cause might coexist or cooperate with another, but can become a cause of action only by a maxim that makes it into a reason for that action. " 6 Since an incentive is not a cause, it means that there cannot be overdetermination at all, since overdetermination means that you can have two possible causes for the same event. If an incentive is not a cause per se, then the existence of many incentives cannot be seen as overdetermination.
It is very diffi cult, however, to see how an incentive can be present and not be taken as a motive. Perhaps what Kant really wanted to show in the philanthropist example in the Groundwork is that only the complete absence of feelings can assure that a certain act was done from duty. If one is immerged in a sea of feelings, it is a hard task to acknowledge that these feelings haven't played any part in the actual accomplishment of the action. Th e only way to be sure that a feeling did not have any role as a motive is not to have had it at all.
Once love is there, or hatred, or jealousy, or even sympathy, we can never say if the action has a pure motive.
How could we know that we are not acting from friendship, for instance, when we vote in a provost election for a friend of ours? Although we can be persuaded that we are doing so because we really think that this friend has the best platform for the university in mind, it is still the case that perhaps it is not our reason, but our feelings that are inclining us to vote for him/her. Although it is good to have feelings for our friends, the best reason to vote for a provost is not the feeling we have for him, but the fact that we consider him to be the best choice for the university.
I think, however, that Kant is not saying that only the absence of feelings will make the action morally worthy. Aft er a long discussion with commentators, Henson's fi tness report model has proved itself to be the correct interpretation of the Groundwork example. In a dutiful action, other inclinations may be present, provided that respect for duty is present and would have suffi ced by itself for the accomplishment of the dutiful action.
Allen Wood, in the book Kantian Ethics , expresses the Kantian position in a correct way, saying that we have a duty to strive for a pure disposition, so that the motive of duty alone is suffi cient. We have this duty because it is "hazardous" to rely on motives besides duty, because the performance of duty on such motives besides is always only "contingent and precarious" (G 4,: 390; KpV, 5: 73).
We have no duty at all, however, to exclude other motives we might have for doing our duty. (MS, 6: 393) 7 Kantian virtue, in Wood's conception, is not only to have the strength to follow the moral law, but also to strive for the purity of motives that lead to that end.
Th at is not the same as having a duty to be a cold moral person. Other incentives could coexist with the moral one as long as the moral law would have been enough to accomplish the moral action. However, I disagree with Wood when he states that "we are not in the least to blame if we require incentives others than duty, so long as we in fact do it. " 8 I agree that it is better to act according to duty moved by sensible incentives than to act in opposition to the moral law. However, the philanthropist example undoubtedly shows that the genuine moral action is the one which is done only because of duty and not because of any pleasure the agent may have. Th en we could be blamed if we need other incentives besides moral law, since this would not be a pure moral action.

Incorporation thesis and weakness of the will
One problem in any theory that claims that agents act from reasons is weakness of the will. According to the traditional doctrine of the weakness of the will, the agent has a reason to act in a certain way and decide to perform the action A1, but she does A2, because she is driven by a powerful inclination to do it.

Th is apparently contradicts what Kant states in Religion :
Freedom of power of choice has the characteristic, entirely peculiar to it, that it cannot be determined 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); only in this way can an incentive, whatever it may be, coexist with the absolute spontaneity of the power of choice (of freedom). (Rel,(6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24) And some pages later, while analyzing the three degrees of evil, he mentions the weakness of the will ( fragilitas ): I incorporate the good (the law) into the maxim of my power of choice; but this good, which is an irresistible incentive objectively or ideally ( in thesi ), is subjectively ( in hypothesi ) the weaker (in comparison with inclination) whenever the maxim is to be followed. (Rel,6: 29) Th e fi rst quotation was baptized by Henry Allison as the incorporation thesis. If we accept the incorporation thesis, an incentive can never constitute, by itself, a suffi cient reason to act, but only if it is taken into the maxim, or, in another words, if it is taken as a motive. If this is the case, weakness of the will can never occur, since it is exactly the opposite case: we act from an incentive, against the maxim. In weakness of the will, an incentive seems to constitute a suffi cient reason to act. If we accept that weakness of the will can be the case sometimes, then, at least in these situations, the incorporation thesis does not hold.
How can one solve this problem in Kant's philosophy? Should we accept the incorporation thesis and deny the possibility of the weakness of the will? Or should we accept the undeniable fact of weakness of the will and accept at least some exceptions to the incorporation thesis.
In a discussion with Marcia Baron, Allison suggested that weakness of the will is not really weakness, but self-deception. When an agent explains her action based on weakness of the will, she is misleading others and herself about what her motive, and her maxim, really was. 10 Th e contradiction between the incorporation thesis and the weakness of the will can be explained in the following four propositions: (1) S knows the principle P1 (2) S can act according to P1 (3) S freely chooses P1 as a Maxim M1

(4) S acts contrary to M1
Allison's choice is to deny 3; that is, the agent S really chose P1 as a Maxim.
I do not think that Kant really presents a solution to this puzzle in his work, although I will try to see what kind of answer could be taken as a Kantian solution.

Rational agency and irrational actions
In his famous article "Action, Reasons and Causes, " Donald Davidson argues that actions have causes and these are composed of desires and beliefs.
Following Davidson, we are now used to talking about causes of actions.
However, many contemporary authors defend the thesis that actions are not caused by desires and beliefs. In a recent book, Rationality in Action , John Searle claims, against the so-called classical model of action, that rational actions are not caused by beliefs and desires and that, in general, only irrational and nonrational actions are caused by beliefs and desires.
He writes: In the normal case of rational action, we have to presuppose that the antecedent set of beliefs and desires is not causally suffi cient to determine the action. . . . We presuppose that there is a gap between the causes of the action in the form of beliefs and desires and the "eff ect" in the form of the action. Th is gap has a traditional name; it is called "the freedom of the will. " 11 In Kant's works, the diff erence between irrational and rational actions is expressed in the well-known distinction between arbitrium brutum , arbitrium sensitivum , and arbitrium liberum in the Critique of Pure Reason . Human beings have arbitrium sensitivum , meaning they can be aff ected by sensibility, but the suffi cient reason can only be given by reason. Th is is related to the presupposition of practical freedom, as the capacity for independence of being determined by the inclinations.
If a set of beliefs and desires is not enough to determine the action, should one consider that reason is the cause of an action? If so, how can one admit weakness of the will, when an incentive moves the agent to act contrary to what has determined the will? Although some commentators argue that reasons can be the cause of an action, I will try a diff erent solution.
Kant does not use the expression "cause of an action, " with the term "cause" being reserved for the phenomenal world. When discussing motives and incentives, he talks about the ground of determination of the will ( Bestimmungsgrund ), which is a motive ( Bewegungsgrund ). Incentives ( Triebfedern ) are responsible for the subjective determination of the faculty of desire. In the passage of Religion on which the so-called incorporation theory is based, Kant does not state that only motives can be a cause of an action, but that only motives can determine the will.
If I have a strong desire D, but I decide to act for reason R, I will perform the action Ar. In another case, if I decide to perform the action Ar, and nonetheless I act according to my desire and perform the action Ad, I act contrarily to the previous determination of my will. It could be that I changed my mind between the fi rst determination of the will and the action. Th is will not be a case for weakness, but I just changed my mind about which action I wanted to perform, meaning that the determination of will has changed. But there are cases in which I really decided to perform the action Ar, based on any previous maxim, rationally decided and-in the cases of weakness-I do Ad. If the actual action is not the same as the decided action, it falls off the strict model of rational agency, understood as the domain of actions that are maxim based. are principles of will. However, they diff er in the extension of their validity: "Th ey are subjective, or maxims , when the condition is regarded by the subject as valid only for his own will, but are objective, or practical laws , when the condition is recognized as objective, that is, valid for the will of every rational being" (KpV, 5: 19).
According to this defi nition, maxims are principles which apply to the will of a particular subject, and therefore are not a practical law valid for every rational subject. In the Groundwork , Kant off ers the defi nition of maxims at two points: "A maxim is the subjective principle of volition; the objective principle . . . is the practical law" (G, 4: 400n); "a maxim is a subjective principle of acting, and must be distinguished from the objective principle, namely, practical law" (G, 4: 421n); it is "the principle in accordance with which the subject acts ; but the law is the objective principle valid for every rational being, and the principle in accordance with which he ought to act , i.e., an imperative" (G, 4: 421n).
In the fi rst section of Groundwork , we are told that actions are morally judged according to their maxim: An action done from duty has its moral worth not in the purpose which is to be attained by it but in the maxim in accordance with which it is decided upon, and therefore does not depend upon the realization of the object of the action but merely upon the principle of volition in accordance with which the action is done without regard to any object of the faculty of desire. (G, 4: 400) In other words, maxims are principles of action which are judged by the categorical imperative for the ascertainment of their capacity to be considered a law. Th us, the action of the agent who makes a false promise is not judged specifi cally as an action, but instead as a principle that underlies this action.
In this case the principle can be formulated as a maxim to make a promise with the intention of not keeping it . It is not the action, but the maxim that is morally judged.
In the second section of the Groundwork , Kant submits a few maxims to the categorical imperative, expressed in the formula of nature's law: "Act as if the maxim of your action were to become by your will a universal law of nature" (G, AA 4: 421). Again, it is not the specifi c action of a particular suicide which is judged, but the principle "from self-love I make it my principle to shorten my life when its longer duration threatens more troubles than it promises agreeableness" (G, AA 4: 422).
Th e diff erence between the rule of a particular action and a maxim as a subjective practical principle is more easily understood when we examine imperfect duties. An example of an imperfect duty toward others is the duty of benevolence. Th at being benevolent is an imperfect duty shown in the is simply a regularity of action, but not a regularity which is desired as such.
Th is is diff erent from the example given by Kant himself when he says that somebody may have as a maxim not letting any insult go unpunished. In this case it is not a question of a mere regularity verifi ed by an external agent, but the subject himself wants this regularity. Th e maxim, therefore, is a desired regularity, a principle which the subject provides for himself.
Th e Kantian theory of action presupposes that the rational subject acts through self-imposed principles which dictate the kind of action to be realized in certain situations. Such regularity in his conduct is not an empirical or psychological regularity, but one desired as such. However, a rule desired by my own will is not enough to constitute a maxim. We can think of a rule desired by my own will which establishes only the means to a particular end, for example "I should exercise three times a week. " Th is kind of regularity is desired as a prescription which serves as a means to attain a certain particular end, be it health or good physical shape. One could give another example of rules of this kind: "I want to dine every Monday at friends' houses. " 14 Or "I want to look for shelter from the rain so as to not get wet. " Maxims are more than mere regularities of conduct, even if desired by the subject himself . Th ey are connected to life goals, to the comprehension of who   I am and what I want for myself. Th us, "Not making false promises when I'm in trouble" is articulated as a rule of life for me: honesty and truth are more important than benefi ting from every situation.

Maxims and aff ects
Could actions caused by intense aff ects, 15  an act of freedom. Th is means that an act of spontaneity or self-determination is involved even in actions based on dispositions and desires. Th is theory implies that one may, therefore, act through inclination, but that, even so, there was an act of spontaneity by the individual which took this inclination as motive.
Even if the theory of incorporation seems extremely abstract to those who sustain the theory that desires and beliefs are suffi cient reasons for action, it refl ects the common confl ict between reason and sensibility. In order to illustrate this confl ict, suppose, for example, that a married man feels desire for another woman. His desire will not be a suffi cient reason for action. Between the desired and the concupiscent action there will be a free decision on the individual's part, in which other reasons, morals, and discretions will be taken into account. Only then will this process of deliberation, or self-determination, result in a decision regarding the action. Th e theory of incorporation is a little stronger still: it does not simply address a decision regarding a particular action, but the choice of a principle of action. When the individual becomes determined to cheat on his wife due to desire, he is electing a principle, aside from a particular course of action.
Kant, however, mentions weakness of the will, in which we are aware of the maxim but lack suffi cient strength to follow it. Weakness of the will, the fi rst degree of evil, is an example of an action which cannot be submitted to a maxim precisely because it was realized contrary to the maxim consciously assumed. Here we have two possibilities: either we admit that the actions which follow weakness, realized contrary to the maxims previously decided by the agent, such as in the case of strong emotions, are not to be subjected to moral accessibility as they do not have an adequate formal structure, or we admit that there is no such thing as weakness, which is self-deception.
Th e fi rst position brings two consequences: a positive one, as it would solve the alleged contradiction between the incorporation thesis and weakness of will; and a negative one, since it would mean that Kantian morality only has adequate tools for judging actions which have been expressed in a certain formal structure.
Th e second position answers that weakness itself should be seen as something for which we are responsible; in other words, it is self-imposed.
Undoubtedly, this answer is a way of maintaining the internal coherence of the Kantian system. Th is position was sustained by Allison, who held an insightful discussion about this point with M á rcia Baron. 19 In Idealism and Freedom , Allison states that weakness should be seen as something for which we are responsible; in other words, it is self-imposed. If the tendency to subordinate the cause of morality to the causes of self-love is evil, weakness must be understood as the opening to temptation. It is only our tendency for self-deception which leads us to see it as a hard fact. Weakness of the will, as a natural error, presents a problem to Allison. One of Baron's objections to this lies precisely in the incompatibility between the theory of incorporation and this fi rst moment of evil. Allison answers by saying that weakness should not be taken as an error or as hard fact, but identifi ed as an opening to temptation; this opening or susceptibility, on the other hand, is the condition which facilitates something like weakness, which is selfdeceptively seen by the individual as hard fact, part of his nature, which he laments but is not responsible for. Baron insists that, in weakness, there is a real commitment to moral law, which would not exist in the other two degrees: impurity and perversity.
If the interpretation of weakness as self-deception elucidates the relation between the theory of incorporation and weakness through the denial of the latter, it faces, however, two problems: the Kantian texts, which really do address the classical issue of moral weakness, and the experience of the common human being, for whom weakness of the will is a fact. I will now attempt to investigate which would be the Kantian answer to these problems. My hypothesis, to be investigated, is that weakness exists, but must be fought with virtue. Th en virtue is to be defi ned as fortitudo , as the strength to fi ght strong inclinations.
Another query relevant to the relation between maxims and aff ects has to do with the fact that, at times, the agent adopts a maxim and chooses to act against it. Now it is no longer a question of weakness of the will, as weakness is an exception not refl ected in the maxim. Let us consider, for example, that we have as maxim "Not Lying. " Suppose that in a certain situation we conclude that lying would be profi table to us, and we eff ectively lie. How would Kant analyze the situation? He would probably claim that our maxim is not "Not lying, " but instead "lying when it is profi table to us. " Th inking that our maxim was "Not lying" was simply the result of an act of self-deception. Kant goes beyond that, expressing in Religion that we have a prior commitment toward a disposition to adopt maxims of self-love or moral maxims: "Th e disposition, i.e., the ultimate subjective ground of the adoption of maxims, "-in other words, the many maxims particularly adopted by an individual-"can be one only and applies universally to the whole use of freedom" (Rel,6: 25).
Not only does Kant state that actions must be coherent with the chosen maxim, but that the maxims chosen by an individual are not independent from each other in their moral character. To Kant, it would be contradictory to suppose that an individual might be committed to a universal principle of morality which prescribes a certain kind of action and, at the same time, to another maxim which prescribes an action which is incompatible to the maxim of morality, "nor can a human being be morally good in some parts, and at the same time evil in others" (Rel,6: 24).
With the exception of the actions which can fall under the description of weakness of the will, the relation between actions and their maxims is one of a previous moral commitment, a commitment which is either one of not making any exceptions of any kind to the moral law, or of doing so, in case one is inclined toward personal gain in some way.
Although the majority of Kantian moral work presupposes the model of an agent who acts according to her will, it leaves open the possibility of an agent who does not have the suffi cient strength to do what she has decided to do.
Virtue is the only way to make all actions conform to the rationality of the model of rational agency. In analogy with Aristotle, only virtue could heal the akratic. But virtue in Kant, unlike Aristotle, will not be based on habituation and cultivation of good character, but on the eff ort to build a strong will that could fi ght against the force of inclinations.
But are all acratic actions irrational? In a line of argument that goes back to Davidson, acratic action has the following characteristics: 1 It is intentional. Since the agent acts according to a reason and not by compulsion, one can say that the agent has acted intentionally.
2 It is confl icting. Since the agent has a better reason not to have done what she has done, one has a confl ict of reasons.
3 It is irrational. Since the agent has acted for a reason while having a better reason to have done otherwise, then it is incoherent, and, consequently, irrational.
Th ere are, however, many authors who maintain that acratic actions are not irrational. One of the better arguments for this position is the argument of internal reasons. For Bernard Williams, the only authentic reasons for action are founded in a motivational system that includes beliefs, preferences, personal projects, and our emotional states. Contrastingly, external reasons are not necessarily based on our personal motivational system and can sometimes be excessively impersonal and abstract. For supporters of this view, since our best judgment is not founded on our motivational personal system, to act against it, is not irrational.
Th is solution is possible for the non-Kantian philosopher, who admits that our commitment to our rational maxims will be the same as our external reasons. Since sometimes they are not based on our own desires and inclinations, we may end up acting in a diff erent way from how we had decided to act.
For the Kantian point of view, however, the problem is already there.

Trying to solve the puzzle
Aft er the discussion between Allison and Baron, some other Kantian philosophers tried to solve the puzzle from a Kantian point of view: if we cannot act without taking incentives in the maxim, or taking incentives as motives, how is the weakness of the will possible in Kantian terms?
Iain Morrison tries to solve this problem. He addresses the problem by explaining the apparent contradiction in weakness "typically, weakness of the will is understood as the phenomenon whereby an agent is somehow overcome by a desire, upon which she knows she should not act. But such an occurrence does not seem possible when all desires must be incorporated into maxims before they can be acted upon. " 20 Th e contradiction lies in the fact that weakness implies the presence of unmaximed or unprincipled action. He explains that to be committed to a maxim is to regard the actions as justifi ed or good, in some sense, and that maxims seem to make actions good. Morrisson proposes the possibility: Kant can solve this problem by claiming that having a maxim (i.e., committing to a course of action) is not exactly identical to regarding an action as justifi ed or good (in some sense). Th at is to say, he can reconcile the ubiquity of principled action and the existence of weakness of the will by modifying the notion of principled (i.e., maximed), such there is principled (i.e., maximed), and yet weak, action. 21 In order to show that a weak action can also be a principled action, Morrison makes a diff erence between happiness-related maxims and pleasure-related maxims. An agent may fail to follow a happiness-related maxim and follow a pleasure-related maxim instead. Morrison takes the example of an agent who desires to eat a lot of chocolates but also desires to lose weight. She will have to select which desire is more important and will form an interest or a maxim based on that. Suppose that losing weight is more important to this agent, she In the Groundwork , the answer seems to be also positive. Not only can one, but one should act without any moral feeling. Sympathy 3 for other people's fortune, as a feeling that leads to benefi cence, is analyzed in the well-known example of the Groundwork. When explaining the diff erence between acting from duty and according to duty, Kant presents the example of two philanthropists, distinguishing the one who possess a strong pleasure in spreading joy to his fellow human beings from the one who helps other people out of duty: Suppose, then, that the mind of this philanthropist were overclouded by his own grief, which extinguished all sympathy with the fate of the others, and that while he still had the means to benefi t others in distress their troubles did not move him because he had enough to do with his own; and suppose that now, when no longer incited to it by any inclination, he nevertheless tears himself out of this deadly insensibility and does the action without any inclination, simply from duty. (G, 4: 398) Kant also asks if we should not consider that his action would have a higher worth if nature had put little sympathy in his heart, but the answer is negative: "By all means! It is just then that the worth of character comes out, which is moral and incomparably the highest, namely, that he is benefi cent not from inclination but from duty" (G, 4: 399).
We can clearly distinguish in the example of the two philanthropists an action Reason , the necessity to support morality in a nonmaterial practical principle leads, obviously, to the refusal to ascribe the role of moral incentives to feelings such as love, benevolence, and aff ection.
Th e aim of the Critique of Practical Reason is to prove at least the possibility of practical reason; that is, that reason is capable of driving us to act morally, in spite of the good or bad feelings we have. To prove that pure reason can be practical is to prove that it can, alone, determine the will. We would fail to prove it if the will were always dependent on empirical conditions. If the will were always based on feelings or passions, this would mean that the pure reason cannot be practical and that the causality of freedom is impossible. Th e Groundwork , as well as the Critique of Practical Reason , has the aim of obtaining, respectively, the categorical imperative and the moral law, in an attempt to prove that reason can determine the will, without the help of empirical incentives. In this context, Kant refuses to ascribe the role of an incentive to benevolent feelings, since these would be empirical and contingent, not being able to be taken as a ground for the determination of the will. In these texts, Kant states clearly that a feeling is a subjective incentive, being inappropriate for the establishment of morality and its foundation upon reason.
In respect is a feeling caused by moral law and is not the incentive to act morally, but it is only an eff ect of the acceptance of moral law.
Frierson, in the book Kant's Empirical Psychology , calls attention to the debate between "intellectualists" and "aff ectionists": Intellectualists (Allison,Guyer ,6 Reath) claim that morally good action is motivated solely by cognition or consciousness of moral law, with a feeling (of respect) generally seen as an eff ect of moral motivation rather than its cause.
Aff ectionists (McCarthy, Singleton, Hererra, Morrisson, Nauckhoff ) argue that the feeling is the immediate cause of moral motivation, the means by which an otherwise inert cognition of the moral law can give rise to an action. 7 According to Frierson, the intellectualist-aff ectionist debate presents textual evidence for both sides. However, I maintain that the stronger evidence is on the side of the intellectualists. In the Groundwork , Kant claims that "immediate determination of the will by means of the law and consciousness of this is called respect, so that this is regarded as the eff ect of the law on the subject and not as the cause of the law" (G, 4: 402). Also, in chapter III of the Critique of Practical Reason , Kant claims the following: If the determination of the will takes place conformably with the moral law but only by means of a feeling, of whatever kind, that has to be presupposed in order for the law to become a suffi cient determining ground of the will, so that the action is not done for the sake of the law , then the action will contain legality indeed, but not morality . ( He also shows that the treatment of respect and moral feeling in Kant's treatment of respect in the third Critique is consistent with the characterization of these emotions in the Groundwork and the second Critique . In the Critique of Judgment (KU, 5: 289), Kant claims that when an a priori principle determines the will, there is pleasure associated with that, the moral feeling, that is the consequence of that determination. Th is is not in contradiction with what is stated in the Groundwork , where respect is said to be "a feeling self-wrought by means of a rational concept" (4: 401n).

Sympathy in Th e Metaphysics of Morals
Sympathy, an aff ect that does not have any intrinsic moral value in the Groundwork , comes out in the Doctrine of Virtue as a feeling of pleasure and displeasure that should be used to promote benevolence, being itself an incentive for moral actions: Sympathetic joy and sadness ( sympathia moralis ) are sensible feelings of pleasure and displeasure (which are therefore to be called "aesthetic") at another s state of joy or pain (shared feelings, sympathetic feeling). Nature has already implanted in human beings receptivity to these feelings. But to use this as a means to promoting active and rational benevolence is still a particular, though only a conditional, duty. (MS, 6: 456) In this quotation, Kant explicitly admits the possibility of using the feeling of sympathy as an incentive, a way to activate benevolent actions. More than that, to use sensible feelings is a conditional duty called humanity. It seems that  To answer this question, it will be necessary to correctly understand the conception of a metaphysics of morals, as that doctrine that contains in it principles of application of the universal law to the "particular nature of the human beings, which is only known by experience" (MS, 6: 217). Th e other side of the metaphysics of morals is a moral anthropology, which gives the conditions of the acceptance or rejection of the moral law by human beings. Kant claims that "a metaphysics of morals cannot be based upon anthropology, " "but can still be applied to this" (MS, 6: 217). In the Groundwork , Kant clearly distinguishes between a twofold metaphysics: a metaphysics of nature and a metaphysics of morals (G, 4: 388). Both parts of metaphysics belong to pure philosophy and refer to a priori principles.
Twelve years later, however, the idea of a metaphysics of morals includes in itself an empirical knowledge on the nature of human beings, without which it would not be possible to determine a concrete system of duties for the human beings. Allen Wood correctly analyzes this displacement in the conception of a metaphysics of morals that occurs between 1785 and 1797, regarding the separation between the empirical and pure part of the ethics. 10 According to Wood, when Kant alters the content of a metaphysics of morals in order to encompass the empirical nature of human beings, he is not abandoning or modifying his basic thesis, that the basic principle of morality is totally a priori . He is only restricting its previous thesis that a metaphysics of morals is only related to the ideas and principles of a possible pure will. In other words, Kant does not consider anymore that a metaphysics of morals is composed only of a set of pure moral principles, but it is a system of duties that result when the pure moral principle is applied to the empirical nature of the man.
Th e application of the pure moral principle to the empirical nature of the man gives us a system of virtues, defi ned as ends that are also duties. Kant enumerates two ends that should be considered as duties: self-perfection and other people's happiness. Th ese two ends lead to two diff erent kinds of duties: the duties of man related to him, and duties related to others, among which we fi nd the duty to love, which consists in promoting the happiness of others.
However, this virtuous love is not a love related to the pleasure experienced in the presence of other person, but it is a principle to do benevolent actions: In this context, however, love is not to be understood as feeling , that is, as a pleasure in the perfection of others, love is not to be understood as delight Th e duty to love must be understood as a principle of benevolence, which consists not in wanting the good of others without practically contributing to this, but in a practical benevolence, or benefi cence, which consists in considering the good of others as end in itself. Th e benevolence principle will produce, in turn, duties of benefi cence (to help the ones in need to fi nd happiness) and of recognition (to honor a person due to a favor that was received) and of aff ection ( Teilnehmung ). Kant accepts that to participate in the pain or joy of others is, without a doubt, a feeling, apparently falling again in a material determination for morality. Th e introduction of this feeling of sympathy must be, however, interpreted, not as a ground of determination for the action, but as a natural feeling that we must used in order to accomplish benevolent actions. It will be our duty, therefore, to cultivate in us those sympathetic feelings, although the moral law should not be based on that, but on the pure reason.
In the Doctrine of Virtue , Kant presents a more complex moral theory on the role of the feelings related to moral actions. Even though sympathy can be an incentive to the accomplishment of a moral action (or an incitement to practical love), this does not mean that all sharing of feelings is positive. We can see it in the division of humanity in humanitas practica , "the capacity and the will to share in others' feelings" and humanitas aesthetica , "the receptivity, given by nature itself, to the feel of joy and sadness in common with others" (TL, 6: 456). Th e fi rst is desirable, but not the second, because the fi rst is free and depends on the will, while second is spread among people "as the susceptibility to heat or the contagious diseases" (TL, 6: 457).
Th e reason for praising humanitas practica and disapproving of humanitas aesthetica is that compassion, when not followed by a practical action, is a way to increase the evil in the world. If a friend is suff ering and I can do nothing to diminish his pain, I do not have a duty to be sympathetic to his feelings, because this would only make me increase the suff ering and troubles of the world.
Kant without a doubt recognizes that feelings of sympathy may play the role of a moral incentive, when the representation of duty by itself is not enough, "for this is still one of the impulses that nature has implanted in us to do what the representation of duty alone might not accomplish" (TL, 6: 457). If the representation of the law will not be enough to bring about the action, it is a duty to promote our natural good feelings to add a natural incentive to a rational moral one. Going, therefore, beyond the spirit of the Groundwork , Kant admits that sympathy, duly cultivated to answer to the correct situations, can be the incentive of a correct action. In this case, this duty must be understood on two levels: fi rst, one to carry out moral actions; second, a derived duty to use natural feelings when consideration about the moral correction of the action is not enough to start the action.
Th e role that Kant attributes to sympathy is, therefore, of a provisory moral feeling, which can assist in the accomplishment of good actions, when the feeling of respect for the moral law is not yet developed enough. As Nancy Sherman notes, this is a morality faute de mieux, that is, a kind of provisory morality: it is a morality of an inferior kind, an immature morality that fi nally will be substituted in the progress of the individual. Nancy Sherman, however, admits that feelings such as sympathy, compassion, and love possess a perceptive moral role in Kant, that is, that "we still need the pathological emotions to decide where and when these ends (of the moral law and its spheres of justice and the virtue) are appropriate. " 11 Sherman seems to be correct and faithful to the texts when she examines the provisory role of feelings such as compassion, love, and aff ection, since Kant really admits a function for these in the accomplishment of moral actions, when mere respect for the law is not strong enough to trigger the action. Th e perceptive role, however, is more doubtful, since the idea that emotions are blind seems to remain a constant in Kant's work, without variations from the Groundwork to the Doctrine of Virtue . Th e critique of sympathy as a possible incentive for a moral action was based, in the case of the philanthropist, not on the contempt for sympathy in itself, but on the idea that sympathy alone could not show us which course of action is the moral one. A good example given in contemporary literature is supplied by Barbara Herman: we hear somebody crying out for aid to load something heavy, we help this person, and later we come to know that a sculpture at an art museum was stolen by a thief. In this example, one ended up helping a thief to carry out his wrong act. And this was done out of sympathy.
In the Doctrine of Virtue , sympathy can play the role of a moral incentive, if it is trained and controlled by the will, which will also inform us when this feeling must be activated. Th is is the reason why humanity is divided into free and non-free humanity. Free humanity ( humanitas practica ) is the capacity and the will to use the feeling of sympathy to promote the happiness of others, which includes a procedure to decide in which cases I must set in motion this feeling. A stoic who decides that he will not set in motion his feelings of So, in order to make sympathy eff ective and turn it into benefi cence, one should go to hospitals and other places, in order to see other people's suff ering; it is a duty, says Kant "not to avoid the places where the poor who lack the most basic necessities are to be found but rather to seek them out, and not to shun sickrooms or debtors' prisons and so forth in order to avoid sharing painful feelings one may not be able to resist"(TL, 6: 457). Th is habitus does not aim at developing compassionate personalities, but at training our feelings of compassion and sympathy so that they can be used as a means to accomplish good actions. However, the feelings of love, sympathy, and compassion are, in themselves, morally blind, depending on moral principles to be set in motion in the correct situation.

Desire, aff ect, and passion: Th e anthropologic modalities of love
In Th e love-aff ect resembles the feeling of falling in love with someone, denoting a romantic, uncontrollable love, whose manifestation can make the person blind to the defects of the objects of desire: "Whoever loves can keep his vision intact; but the person who is in love is inevitably blind to the mistakes of the beloved object, although the latter will usually regain his vision a week aft er the wedding" (Ant, 7: 253). Th e emotion of this passionate person is an aff ect, in Kantian terms. Th e term passion is reserved for more deliberative attitudes, being able to coexist with a cunning dissimulation, since this, as shown in the Love, in the form of aff ect or passion, even in its most violent manifestation, is not as harmful as the passions of ambition, vanity, and greed. However, it is not as helpful to morality as the feeling of sympathy, since love implies a feeling between dissimilar people. Or, Kant writes in one of the Refl exionen grouped in the Nachlass on anthropology: "We need more to be honored than to be loved, but we also need something to love with who we are not in rivalry. Th en we love birds, dogs or a young, fi ckle and darling person" (R 1471, Nachlass 15: 649).
Apparently, this claim denounces a prejudice of the time regarding women.
However, in another Refl exion , Kant affi rms that "men and women possess a reciprocal superiority one in relation to the other" (R 1100, Nachlass 15: 490). Despite the fact that this superiority of each one is relative to diff erent aspects, the reciprocal inequality is what stimulates and promotes love as aff ect or passion. Th e fact that these feelings need a reciprocal moral inequality indicates that their place is strange to morality, which consists of considering the other as equal and promoting her happiness.
Th e fi gures of love assume diff erent positions in Kant's philosophy-some have moral value, others do not. Love as benevolence can be considered a practical principle: to do good and to help people, from which the love for others can also be awakened. Th is was clear in the analysis of the Kant's texts, where it is said that it was not necessary to love and, due to this, to do good to human beings, but to act morally, and through this habit, to promote feelings for human beings. Th e feeling of sympathy can also be used by the agent to stimulate moral actions in which the respect for the moral law was not strong enough as an incentive. Th is is not in opposition to what is explained in the Groundwork , in which the moral value of an action resides in the respect for the law. To use the feeling of sympathy is only a provisional morality that, empirically, can and must use these feelings of pleasure and displeasure for other people's luck to encourage good actions, until our respect for the law is suffi ciently strong to be a possible incentive.
Relative to aff ects and passions, even if both were criticized as illnesses of the reason, the negative eff ect of the love-aff ect would be less dangerous than the persistence and inversion of principles in the love-passion. However, since the passion of love ceases when its physical desire is satisfi ed, it does not have the persistence of other cultural passions. But such feelings of love are not useful to morality, since the love-aff ect or love-passion is awakened from an idea of inequality alien to morality.
Finally, it is important to emphasize that the analysis of feelings, inclinations, and passions in the Doctrine of Virtue and Anthropology does not contradict the spirit of the Groundwork , since the action with true moral value is still the one whose incentive is respect for the law, which does not hinder us in using our sensible feelings, such as sympathy, for the purposes of reason.

What should we take for granted?
In his book Kant and the historical turn , Karl Ameriks argues that Kantians should prove moral law can move the agent to act, because action is not a matter of mere judgment. If one has an impulse to do something, this could not be a mere thought. 12 If we take for granted that only feeling could move us to action, Kantians should explain what kind of feeling moves the agent, or how can we act without feelings. As Ameriks himself stressed in an earlier work, "Since the ground of duty is defi ned independently of all our natural inclination, it seems that Kantian morality leaves the very motivation of moral activity unexplainable. " 13 However, if we do not take for granted that one should act from feelings, then we do not have to prove anything.
In fact, Kant does not take for granted that we need feelings in order to accomplish moral actions. On the contrary, tender feelings make the heart weaker and not stronger. Since virtue for Kant is before anything else strength, teaching tender feelings will build a weak character, which cannot meet the exigencies of morality: In our times, when one hopes to have more infl uence on the mind through melting, tender feelings or high-fl own, puff ed-up pretensions, which make the heart languid instead of strengthening it, than by a dry and earnest representation of duty, which is more suited to human imperfection and to progress in goodness, it is more necessary than ever to direct attention to this method. (KpV, 5: 157) Th e method mentioned here is the moral education of young men. In order to really develop moral character in children, it is not useful to tell stories about magnanimous and noble actions. Is it worthwhile to call attention to the holiness of duty alone? In pedagogical terms, this is more useful, because feelings do not develop character since they calm down and the organism tends to go back to its natural vital motion.
All feelings, especially those that are to produce unusual exertions, must accomplish their eff ect at the moment they are at their height and before they calm down; otherwise they accomplish nothing because the heart naturally returns to its natural moderate vital motion and accordingly falls back into the langor that was proper to it before, since something was applied that indeed stimulated it, but nothing that strengthened it. (KpV,

5: 158)
A possible objection from a sentimentalist would refer, not to the duration of the incentive, but to its force. Even if feelings cannot last for a long time, they give us a more intense incentive to the moral action. Kant would not disagree with that, he would even give us an example of someone who tries in extreme danger to save people from a shipwreck, fi nally losing their own life in the attempt. In this case, there is more "subjective moving force as an incentive if the action is represented as a noble and magnanimous one than if it is represented merely as a duty in relation to the earnest moral law" (KpV, 5: 158). However, the incentive presented in the pure law of duty is the most elevated of all.
Not only it is possible to act from the motive of duty alone, but it is also desirable. And the conscience of moral law should be a suffi cient motive for us to act from it.

What is wrong in acting morally out of emotions?
People usually make a portrait of Kant's philosophy and the relation between action and emotion as if Kant has said that we know what to do, but sometimes, because of the weakness of the will, we can fail to accomplish doing the right thing. But is this always true? First, is it a good picture of human nature? Does Kant really say that?
Here there are two diff erent situations. We can act out of emotions that are commonly taken as anti-moral emotions, such as rage, envy, ambition, jealousy, or greed. First, we have the possibility of someone taken by a strong emotion who cannot fi gure out what is the right thing to do. People taken by the strong emotion of rage not only have the tendency to fi ght someone, but also think that revenge is the right thing to do.
But we can also have emotions that lead us to moral actions, such as sympathy. Although Kant acknowledges that sympathy can be part of a morality faute de mieux, the benevolent action done out of sympathy does not have intrinsic moral value.
Kant has been criticized by many authors who support the view that acting from duty is repugnant. Th e anti-Kantian literature has illustrated this assumption with a well-known example: the example of someone who visits her friend in the hospital out of duty.
Th is example was fi rst formulated by Stocker and discussed, among others, in Baron's book Kantian Ethics Almost without Apology. 14 Suppose that someone goes to the hospital to visit a friend and, when asked by her friend why she is doing that, she answers that she is visiting her friend out of duty. Th is example is supposed to be a critique of Kantian system, in which it is supposed that to be cold or insensible is a virtue.
But is it really so? In fact, Kant is not concerned with what is more comfortable or warm in a psychological way, he is asking what is the right thing to do and what should be a right motive. In saying that the right motive is duty, Kant is not condemning warm feelings, but is only saying that we should visit our friends in the hospital even if we do not have any inclination to go to the hospital. In fact, few people really like to go to hospitals. If going to the hospital should be dependent upon a feeling of friendship, compared to the bad feeling of entering a hospital, perhaps in most cases the feeling of sympathy for a friend will not be enough to counterbalance the feeling to avoid being in a hospital.
One of the main criticisms against the supposed coldness of Kantian morality has come from feminist philosophy. According to this critique, some ingredients that are important to female identity, such as emotion, love, empathy, and cooperation, are not in consideration in Kantian ethics.
Sally Sedwick expresses this criticism when she says that because moral agency in Kant's view is a function of acting from reason rather than from feeling, it is said to refl ect features more of a male than of a female identity. 15 She, however, supports Kant against the critique of a misunderstanding of human psychology. What Kant is saying is not that we are-or should be-cold people without any feeling or that in our meaningful relations feelings should not play an important part. What he is claiming is that empirical motives do not have moral weight.

Only motivations motivate?
Paul Guyer off ers a challenging version of Kant's philosophy. 16   Even if we can agree with Paton regarding the necessity of distinguishing these three levels, Kant was not always clear on this. Th e identity or diff erence of psychology and anthropology, the place of anthropology in the system, and even the late conception of a pragmatic anthropology show us that the distinction between these levels is not so precise in Kant. Our purpose now will be to examine some of the moments of this distinction.
Th e provisory refuge of empirical psychology in metaphysics Kant   We have three faculties according to the three things that belong to the self: representations, desires (appetites), and pleasure and displeasure. Th e self is therefore composed of three faculties (cognitive, desire and pleasure, and Th e division of philosophy into a pure and an empirical part is described as well in Mrongovius's notes from 1785, 7 the same year as the Groundwork 's publication. However, as Allen Wood has pointed out, 8  Even using examples in which a few elements relating to human nature can be verifi ed, it is possible to affi rm that the procedure of obtaining the categorical imperative is achieved without a substantial contribution of these elements.
If the attainment of that which should be done is obtained without empirical elements, nowhere does Kant state that moral philosophy does not include an empirical part. On the contrary, moral philosophy is composed of metaphysics of morals and a practical anthropology.

Metaphysics of M orals and principles of application
If, in the 1770s, empirical psychology was able to fi nd a place in metaphysics, even if temporarily, as with the publication of Groundwork (1785) Comments on race and sex occupy the second part, called characteristic. in Doctrine of Virtues and in Doctrine of Right . All these texts discuss, not exhaustively, that which seems to be the object of a moral metaphysics: a practical anthropology, in other words, the nature of the rational sensitive being.

Th e impure part of ethics
Kantian moral theory has a pure part and another part which, by contrast, we may call impure. 9 We are able to verify this union of two parts in the oft en uses the term "anthropology" to refer to pragmatic anthropology and "transcendental" for the conditions of possibility of experience; however, he explains his use of the term: transcendental anthropology "provides a useful term to contrast Kant's approach to human being in his a priori philosophical works with empirical and pragmatic approaches elsewhere. " 11 Frierson understands the three critiques as parts of transcendental anthropology: the fi rst critique would give us a "transcendental anthropology of cognition, " the second critique will be the "transcendental anthropology of volition, " and the Critique of Judgment will supply us with "transcendental anthropology of feeling. " 12 He considers that the answer to the question: "What is the human being?" is given in his transcendental philosophy, "where he develops his metaphysical account of humans' free and fi nite natures and lays out norms that should govern cognition, feelings, and volitions, " and in his empirical anthropology, "where he provides detailed descriptions of how human beings actually think, feel, and choose. " 13  contains merely the necessary moral law of free will in general" (KrV, A 55).
On this level of total abstraction, no information regarding the peculiar nature of the human being or of another rational being is given. However, not even the Groundwork itself would fi t into a pure ethics in this more strict sense, since this text discusses subjective limitations and obstacles, as well as the way in which moral law should be received as an imperative, which is not valid for every rational being.
Th e second level, present in the Groundwork , would be named morality for fi nite rational beings . In this case, none of the enunciated principles depend on specifi c information regarding human culture and nature, even if the categorical imperative is valid for rational fi nite subjects, who are conscious of the moral principle but whose inclinations oppose it. We then have a third level, represented by Th e Metaphysics of Morals , whose objective is to determine moral duties for human beings as such. Determining duties, as human duties, is only possible when we know the constitution of human beings (MS, 6: 217), which requires minimal empirical information about human nature.
Which empirical information would be required to determine human duties?
In order to apply the moral law to human beings, we should have general knowledge regarding human nature, such as the instincts, tendencies, abilities, and faculties of such beings. Th e project of determining specifi c duties to human beings is still a part of metaphysics, since empirical knowledge is not incorporated in the system (MS, 6: 205).
If the determination of the specifi c duties of human beings is an object of metaphysics, however, the specifi c study of the human peculiarities which assist or hinder the exercise of morality will be the object of a practical or moral anthropology, as the text establishes at various moments (MS, 6: 217).
What is the specifi c locus of moral anthropology? In order to answer that, we should fi rst answer the following questions: What are the passions and tendencies which hinder or assist adherence to moral principles? How should these principles be taught to human beings? How can political, cultural, and religious institutions be organized so that they can realize moral objectives?
Are there specifi c aspects of modern time which assist in the establishment and development of morality?
Th e Anthropology , especially in its fi rst part, answers the fi rst question.
Th e pedagogical texts, along with the texts on religion and history, appear to be the right place to answer the other questions of moral anthropology. On the other hand, the second part of the Anthropology presents a more specifi c description of subgroups within the human species, which implies a more detailed empirical knowledge than what is necessary in a moral anthropology.

Impure ethics and sensibility
In order to analyze the relation between reason and emotion, three central Th e most important of these presuppositions is moral feeling, defi ned as "the susceptibility to feel pleasure and displeasure merely from being aware that our actions are consistent with or contrary to the law of duty" (TL, 6: 399).
Th is feeling can be pathological or moral; in the fi rst case, it seems to precede the representation of law. In the second, it is posterior to the law, and is an eff ect of a concept regarding the faculty of feeling pleasure or displeasure.
Since it refers to a natural predisposition of the mind to be aff ected by the concept of duty, we are in the realm of practical anthropology and no longer of metaphysica pura . Th is natural predisposition is a fact about human nature: "No human being is entirely without moral feeling, for were he completely lacking in receptivity to it he would be morally dead" (MS, 6: 399).

Emotions in the map of the self
First of all, Kant does not use the term "emotion. " 9 What we pre-analytically call emotion refers to at least three diff erent kinds of phenomena: aff ects, 10 moral feelings, and passions. Th ese inclinations 11  between the meaning of aff ects and passion, a diff erence that is also illustrated in outstanding literary works, such as Dangerous Liaisons. 15 While aff ects are outbursts of feelings, which cannot coexist even with a prudential rationality, passions show their cunning. Th e same diff erence can be seen when one compares the aff ect of hate with the passion of anger: while the latter is fi ckle, the former is a permanent disposition, which inclines the agent to plot a cold blood vengeance.
In order to improve our understanding of Kant's model or models for emotions, it is helpful to locate them in relation to a Kantian sketch of the self. 16  action. We cannot decide whether we will feel pain or not, while we can decide whether we will feel a moral pleasure, because this requires just that we act according to the moral law.
Th e inferior faculty of pleasure is responsible for purely sensible phenomena such as pain, hunger, and thirst. Th e superior faculty of pleasure relates to the reactive and active self. Th e sensation of displeasure that we feel when we know that our actions are wrong is associated with the active self, since it is a feeling connected with a concept: the concept of duty. Between the pure passive and the pure active part of the self, abide the reactive feelings, which still belong to the superior faculty of feeling, and refer to aff ects and sensibility.
Kant, who was a well-known hypochondriac, illustrates in his Lectures on Anthropology the diff erence between passive, reactive, and active feelings with a case of a man who suff ers from gout: We can clearly see, in this quotation, the diff erence between three levels of displeasure. Th e fi rst one is purely physical, beyond control, even indirectly.
Since it is purely sensible, it is independent of any cognitive content; the subject feels a certain wound as pain. Th is is an example of the inferior faculty of pleasure and displeasure, where the feeling is given through sensation alone. Th us, the pain model will not even explain the emotions Kant ascribes to animals.
Humans have another level of emotions, the one connected with moral judgments. Th is third level, the displeasure of spirit, the pure active part of the self, is completely absent in animals, since it depends on reason that awakes self-reproach. Th e moral conscience would create in the agent the feeling of displeasure because she knows that she has not acted well. Self-reproach is, then, an illustration of an intellectual displeasure, caused just by a concept.
As we have seen, in the classifi cation of the faculty of pleasure and displeasure (Ant, 7: 230), we have the division between sensuous and intellectual pleasure.
While the former can be produced by sensation or imagination, the latter is conceived by concepts or by ideas. In the case of self-reproach, displeasure is produced by the idea of moral law. Feelings from the animal (or passive) part of the self are involuntary, and cannot have their sensation controlled. Th is is the case with pain, thirst, hunger, and so on. In the other extreme, we have the pleasure and displeasure of the active self, such as moral feelings and the feeling of respect, which are controllable through our actions, since they are outcomes of good or evil actions. In between lies aff ect, a phenomenon of the reactive part of the self, which is connected with the imagination.
When we mention reactive feelings, such as aff ects, we should consider also that some strong aff ects have involuntary outcomes, which involve strong physiological arousal. Kant cites two highly illustrative cases in this connection, related to anger and fright. About anger, he mentions the situation of a man who enters one's room in anger in order to say harsh words: If a person comes to your room in anger in order to say harsh words in great wrath, politely ask him to be seated, and, if you succeed in this, his scolding will already be milder because the comfort of sitting is a relaxation which does not really conform to the menacing gesticulations and screaming while one is standing. (Ant, 7: 252) In this case, one should make the angry man sit down, because the comfort of sitting is a relaxation, which does not conform to a great wrath. As for fright, he mentions the disagreeable digestive eff ects that can aff ect soldiers before a battle (Ant, 7: 257). Kant's account of the incontrollable and even highly undesirable outcomes of these emotions agrees with contemporary fi ndings in physiology, according to which emotions may cause eff ects related to the autonomic neural system. 19 Th ese physiological components of emotions are responsible for the diffi culty to controlling them, 20 since they have an inertial component, which is not easy to handle.

Between the propositional attitude school and the feeling theory
For Kant, emotions do have evaluative components, but, with the exception of passion, they are also feelings, which have physiological features. Th is is why Kant recommends that we should soothe the movements of a mad man, making him sit down and relax. Th e soothing of movements will help to calm the aff ect of anger itself.
We can say that Kant is between the propositional attitude school and the feeling theory. 21 According to Griffi ths, 22  or directed to something in the world beyond themselves. We are not only mad, as we feel pain, but we are mad with someone or at something, we are not in love without an intentional object, we are in love with someone. As intentional states, they have a propositional content, and also bring evaluation and cognitive elements, as feelings they present physiological arousal. In this way, he can overcome the dichotomy of thought and feeling. Kant will not deny that an emotion has mental content. However, it is not only a desire for, or a judgment about something. Physiological processes also accompany it.
He would agree with the propositional attitude school, in that there is no sense in attributing sadness without the idea that something valuable was lost, or anger without the idea of something off ensive done by someone.
We are angry with a specifi c person or at a determinate situation; anger consequently has an intentional object. However, he seems to acknowledge that for some strong emotions, such as anger, a physiological arousal seems to be a necessary condition for emotion. In the case of anger, the man has to sit in order to calm himself, because anger is accompanied by a strong physiological arousal, which cannot be instantly overcome by the sole reasoning about the emotion.
Will this physiological arousal also accompany emotions related to the active part of the self, such as moral feelings? Th e answer is yes. Although moral feelings are ultimately caused by an idea, the idea of moral correction of an action, they are still feelings of pleasure and displeasure. Although here the physiological arousal is not as strong as in anger, the idea of feelings brings with it pleasure and displeasure, which is impossible without any physiological component. assessment about the unfairness of a situation; however, since it is an aff ect, it will make deliberation more diffi cult and will be a hindrance to moral action.
Sympathy, although sharing with aff ects some features, because it is still a feeling of the reactive self, can be shaped by the will in order to help morality.
Moral feeling, as a feeling of the active self, is directly connected with morality and with the moral evaluation of a situation.
Passions have a special relation with morality: they are, with no exception, evil. Th ey are not, like aff ects, only agitations of the soul, which hinder refl ection and moral deliberation. Moreover, some aff ects can give us moral information about the situation, although they are immoderate and diffi cult to control. Passions, however, are quite the opposite: they are immoral desires for something. Th irst for vengeance and power, greed and ambition present a perversion of moral goodness, since they are related to an evil deliberation of a mind that is not disturbed by aff ections.

Th e role of imagination
One of the problems regarding the propositional attitude theory is that it Imagination can intensify the feeling of sadness, as the example of animals that are sad, but never miserable, shows, because they don't have the intensity of imagination human beings have.
In the text "On the Philosophers' Medicine of the Body, " Kant ascribes to imagination an even stronger force: the power of imagination can heal or produce a disease. On the one hand, the confi dence sick people put in their doctors helps them through the imagination that they will be healed. On the other hand, imagination could spread diseases like epilepsy, since it is able to alter bodily movements. 27  to him as a nicer one, and these images of a good and nice future will arouse in him the feeling of relief. Moral feelings, like the pleasure or displeasure in the rightness or wrongness of our conduct, can be more related to judgment than to imagination. If the sick man were to reproach himself for his situation in the former case, now that he remembers he has health insurance, he could feel pleasure and self-satisfaction in being prudent and wise.

Can we be held responsible for our emotions and actions?
Can we be held responsible for our emotions? It is true that some emotions of the reactive self, such as joy, sadness, and longing, are intimately connected with imagination and this faculty does have the power to modulate aff ects.
However, the power of imagination cannot deny the evidence for my belief in a bad or good future. Although we can modulate some aff ects through imagination, we cannot decide not to have them or to have them at ease.
Imagination of a future state of pleasure or displeasure should be connected with a possible state of aff airs. If the sick man could not work in the future because of his gout, he cannot prevent the feeling of sadness for the bad future he imagines. Or, he can even trigger for some moments the aff ect of joy when he remembers his past moments, but as soon as he focuses on his future situation, this joy will be quickly turned into sadness.
If we are not responsible for our emotions, can we be responsible for our actions? Emotions are said to be an obstacle to moral deliberation, and Kant uses phrases such as water that breaks a dam in order to show the irrational force of these feelings. However, he also draws on a strong picture of freedom, which does not allow for pathological compulsion. 28 We have strong emotions, diffi cult to control, but we have means to tame them in order to act the way we want. Virtue is one of these means. We are not responsible for emotions, although we can be held responsible for actions, since the strength of emotions cannot be equated to compulsion, such as the compulsion for drugs. Recently, some philosophers have tried to establish this kind of parallel between strong emotions and addiction. 29 Kant would not accept this picture, because as strong as emotions can be, and as much of a problem for morality they can portray, the very idea of practical reason presupposes that agents can decide how to act. In fact, the Kantian picture here is more likely to be accepted by moral common sense, since a strong emotion can never be taken as a total excuse for a bad action. Agents can mention strong anger to explain their violent acts, yet not to forgive them. People are held responsible for wrong actions due to strong emotions, because it is presupposed that they could have acted otherwise.

What can Kant really teach us about emotions?
In order to explain emotions, philosophers have tried, with few exceptions, 30 to build a single model for something that they supposed was a single kind of mental event. But perhaps the key to the understanding of emotions is to grasp their complexity and the diff erent weight of rational and irrational components each one has. Although Kant did not write one specifi c book on emotion, I think he was very attentive to this variety of emotional events, particularly in his anthropological works.

In this brief reconstruction of what could be a Kantian theory of emotion
and its relation to the passive, reactive, and active self, I have showed that this kind of feeling presents diff erent relations with our passive and active self. Kant's theory is more likely to be accepted by our common moral sense, because it explains emotions as people actually feel them. It is a very sophisticated theory that confi rms the complexity of its object and does not allow for unrefi ned explanations. And perhaps he shows his cunning in not using the word emotion, but instead, other terms such as "aff ects, " "sympathy, " "moral feelings, " and "passions, " in order to show that what we pre-analytically call emotions refer to a wide variety of states, which call for diff erent philosophical categories. Th is insight agrees with some contemporary philosophers who have recently challenged the idea that emotions form a unique class of events. Although in some cases Kant allows for cultivation of character, the limits of this cultivation will depend on the natural temperament of the agent. Kant presents some indirect-or even medical-ways to deal with the strong aff ects that are not under our control, since apathy is essential for Kantian virtue.
I conclude that Kantian philosophy not only is a matter of a priori principles, but is also dependent on empirical anthropology. Th e task of Kantian ethics is to attain freedom of strong aff ects, in order to better attain virtue.

Virtue and the controlling of inclinations
Kant is categorical about the relation between virtue and the controlling of inclinations: Since virtue is based on inner freedom it contains a positive command to a human being, namely to bring all his capacities and inclinations under his reason's control and so to rule over himself. 2 Virtue presupposes apathy, in the sense of absence of aff ects. 3 Kant brings around the stoic ideal of tranquilitas as a necessary condition for virtue: "Th e true strength of virtue is a tranquil mind . " 4 In the Anthropology and the Doctrine of Virtue , apathy is taken in the sense of freedom from aff ects. 5 In these texts, Kant maintains that we must strive toward a state in which aff ects are absent. Kant praises the stoic aim of apathy, as a desired state of self-control and self-possession when emotions are suppressed: Th e principle of apathy, that is, that the prudent man must at no time be in a state of aff ect, not even in that of sympathy with the woes of his best friend, is an entirely correct and sublime moral precept of the stoic school, because aff ect makes one (more or less) blind. 6 Although Kant undoubtedly sides with the stoic moral aim, he points out that to control aff ects is a diffi cult task. I will show that his strategies for controlling aff ects rely on a special physiology of emotions, in which they are related to certain bodily movements that cannot be directly controlled by reason. Th is is so because these movements, once they begin, depend upon physical causation that acts on organs and fl uids. Such a standpoint can be found, for instance, in Descartes's Passions of the Soul. 7 According to Descartes, an essential ingredient of passions is the movement of fl uids called spirits, contained in the cavities of the brain. In the passion of fear, for instance, the spirits go from the brain to the nerves that move the legs and allow us to run. Th e soul cannot have full control over its passions because they are accompanied by disturbances, which take place in the heart and through the blood and spirits: "Until the disturbance ceases, they remain present to our mind in the same way as the objects of the senses are present to it while they are acting upon our sense organs. " 8 Th e soul can overcome the lesser passions, but not the violent ones, and it can also prevent itself from feeling a slight pain; however, it cannot stop the pain caused by a fi re that burns the hand. We can only overcome strong passions aft er the disturbance of blood and spirits has gone away. While the disturbance is still agitating the blood and spirits, the only thing the will can do is to inhibit the movements of which the disturbance disposes the body: in anger, it can stop the hand to strike a blow, in fear, it can stop the legs. Th e way to control aff ects proposed here is the relaxation of the body. If the movements are lessened, so are the feelings. For stronger emotions, when even the control of bodily movements is useless, we should wait until they go away, since our rational mastery is impotent.

Physiology in the seventeenth and eighteenth centuries
Although there is no evidence that Kant read Th e Passions of the Soul , my hypothesis is that Descartes and Kant share a common background whose roots lie in the physiology of emotions. I will also argue that Kant is discoverable from the annexed marks: Th e languor of the heart and arteries is discernible in the pulse; as is also that of the extreme vessels on the surface, from the paleness, the dryness of the skin, the shrinking of tumours, the drying up of the ulcers, and the manifest absence of sthenic diathesis, to produce any symptoms like these. 22 Th erapy consists in giving sedatives for sthenic diseases and stimulants for asthenic ones. 23  However, he should also acknowledge that some mental phenomena are beyond his control: In many diseases of the mind, when imagination turns savage and the patient's head resounds with great, unheard of things, or he is cast into the depths of depression and tormented by empty terrors, the mind has been dethroned and bleeding the patient is likely to produce better results than reasoning with him. 26 When the mind is assaulted by strong aff ects like a profound sorrow or a strong fright, the possibility of controlling these emotions by the discipline of the mind should give room to physiological intervention. Th e same applies to strong aff ects like anger. Th at is the reason why Kant proposes to calm the angry man by making him sit and relax and does not propose a direct control of this aff ect by the will. 27 Th is example from Anthropology has an interesting parallel with Passions of the Soul , where Descartes argues that we can easily overcome the lesser passions, but not the violent ones, except aft er the disturbance of the blood and spirits has died down. Th e most the will can do while this disturbance is at its full strength is to inhibit many of the movements to which it disposes the body. 28 In the example of the angry man, Kant seems to accept the inertial component of aff ects. Th is example shows us something about the functioning of a strong aff ect in general: while we are taken by these strong feelings, we cannot do anything but wait for them to go away. Evidently, what Kant says about anger cannot be generalized to all aff ects. As we have seen in the previous chapter, he presents a continuum of emotions, one that goes from the most strong and irrational aff ects, like anger, to those that are less passive and likely to be modulated by reason. Th is is also the case with moral feelings in general, including sympathy, moral courage, and enthusiasm for the moral law (Ant, 7: 254). Another example of an emotion created by reason is the astonishment felt by men while contemplating the wisdom in Nature. Th is is an emotion that can only be aroused by nature: "It is a kind of sacred awe at seeing the abyss of the supersensible opening before one's feet" (Ant, 7: 261).
As a general term, "emotions" refers to a wide variety of states, which Kant calls aff ects, sympathy, moral feelings, and passions. In general, they all have intentional objects, which is to say, they are about something. Besides, they are also connected with physiological arousals. Even among aff ects there are some whose physiological arousals are stronger than others. Th is is the case of anger if compared with shyness. If it is possible to control shyness by habituation, Kant is not that optimist regarding aff ects whose physiological arousals are very strong. Th is is the case with anger, fear in battle, and so on.

Cultivation of emotions
In the last decades, many commentators have pointed out that Kant makes One of the most common strategies for controlling the emotions is found in Aristotle's theory of virtue. Aristotle claims that to be virtuous is to have the appropriate emotion befi tting the situation, that is, to have these emotions "at the right time, about the right things, towards the right people, for the right end and in the right way. " 32  having the emotion of anger as a paradoxical task: Socrates was in doubt whether it would not be good to be angry sometimes, but to have emotion so much under control that one could coldbloodedly deliberate whether or not one ought to be angry appears to be something paradoxical. 33 We can surely deliberate about what to do, in spite of our anger, but the occurrence of this aff ect escapes the rational control of the will. With respect to such a feeling we are passive, and it is beyond our power to control it. Sympathetic joy and sadness ( sympathia moralis ) are sensible feelings of pleasure or displeasure (which are therefore to be called "aesthetic") at another's state of joy or pain (shared feeling, sympathetic feeling). Nature has already implanted in human beings receptivity to those feelings. But to use this as a means to promoting active and rational benevolence is still a particular, though only a conditional, duty. 34 In relation to sympathy, Kant  whenever the maxim is to be followed. " 43 Were inclinations and aff ects under complete control, there would not be a fi rst degree of evil-that is, the weakness of the will. In Kant's view, both passions and aff ects are diseases of the will. Th e only diff erence between them is that the former is a persistent perversion of reason while the latter is ephemeral. Moreover, "yet virtue here is worth so much only because it costs so much, not because it brings any profi t" (KpV, 5: 156). Virtue shows its worth, because of its pureness, and deserves approval and admiration because moral actions are performed without any pretension to happiness or even magnanimity. 1 In this chapter, I analyze the idea of virtue in Kant and how it relates to the controlling of aff ects and passions. I begin by showing the relation between virtue and happiness and then I explore virtue as strength.

Virtue is not happiness
In the Anthropology from a Pragmatic Point of View , while discussing the faculty of desire and its relation to the feelings of joy and sorrow, Kant recommends moderation of feelings. His Stoic advice, however, is accompanied by a particularly awkward vision: an intemperate joy is worse than an extreme sorrow: Exuberant joy (untempered by any concern for grief) and Kantian defi nition is the commonsensical one: " Happiness is the state of a rational being in the world in the whole of whose existence everything goes according to his wish and will , and rests, therefore, on the harmony of nature with his whole end as well as with the essential determining ground of his will" (KpV, 5: 124). Th e accomplishment of happiness will come from the satisfaction of our desires.
Schneewind points out that the problem of how to attain happiness is the problem of how to satisfy our desires, which is not within our power to do: "Since happiness for Kant comes from the satisfaction of desires, the range of components out of which we can choose to fl esh out our conception of happiness is not up to us. " 4 He also shows in the same article that the conception of happiness as being beyond our control accords to the modern lack of confi dence in the natural world. In that pessimistic "lack of confi dence" he would disagree with a major metaphysical Stoic position: Th e metaphysics of Stoicism is profoundly important for its ethics. Regardless of the extent to which any particular moral principle is derived from the metaphysics, Stoic metaphysics grounds at least the a priori assurance that when we act from reason as far as we can, everything of concern to us will be well. Kant simply takes for granted an anti-Leibnizian, anti-Stoic acceptance of the indiff erence of the natural world to human concerns. 5 Th e accomplishment of the demands of moral law will not necessarily bring us any happiness, for the connection between morality and happiness is only contingent. According to Kant, for a fi nite being, there is no correspondence between happiness and morality, because such a being cannot be a cause of nature: Consequently, there is not the least ground in the moral law for a necessary connection between the morality and the proportionate happiness of a being belonging to the world as part of it and hence dependent upon it, who for that reason cannot by his will be a cause of this nature and, as far as his happiness is concerned, cannot by his own powers make it harmonize thoroughly with his practical principles. (5: 125) A fi nite being that belongs to the world and is dependent on it cannot be a cause of nature, nor can it harmonize its happiness with morality. If pure reason needs the latter connection for the pursuit of the highest good, a being which is the cause of nature and also connected to morality should be postulated. Th e second postulate of pure practical reason, the existence of God, is advanced as a result of the need for a connection which fi nite beings would not be able to produce: the attachment of happiness to morality.
In the defi nition given in the KpV, external goods are certainly essential parts of happiness for Kant. Happiness is attained when events in the world conform to our wishes. As a result, happiness cannot be the aim of a moral life, because happiness is dependent on contingent goods, which makes it a fragile and unstable base on which to ground morality. Th at happiness is contingent, however, is not its sole diffi culty. Kant claims the following: 1 Happiness cannot be universally defi ned 2 Even if happiness could be universally defi ned, it will lead to disagreement and not to harmony

Happiness cannot have a universal defi nition
In the Groundwork , Kant distinguishes imperatives of skill from imperatives of prudence: the former command an action as necessary to accomplish an end; for the latter, this end is happiness. Although happiness is undoubtedly the end of all rational beings, unfortunately for the eudaemonist philosopher, it is impossible to give a determinate concept of happiness: "But it is a misfortune that the concept of happiness is such an indeterminate concept that, although every human being wishes to attain this, he can still never say determinately and consistently with himself what he really wishes and wills" (G, 4: 419).
Th e claim that one is not able to produce one's own happiness seems more acceptable than the claim that we do not know what can count as happiness.
It is reasonable to think that even if we do not have the power to produce Happiness cannot give us a universal law, whether the object of pleasure for multiple parties is the same or diff erent. Happiness proves to be an inadequate basis for the moral life. Its relation to the empirical and particular object of pleasure and joy, the incapacity of fi nite beings to promote their own desired ends, and the confl ict that would still result if we overcame these diffi culties seem to banish forever the term happiness from the moral domain.
Nevertheless, Kant redefi nes happiness as an object of duty in the second part of Th e Metaphysics of Morals , the Doctrine of Virtue. In this book, Kant introduces the concept of virtue as an end that is also a duty. Th ere are two ends that are also duties: one's own perfection and the happiness of others.
Th at happiness of others should be prescribed as an end that is also a duty is not, at fi rst glance, clear. Wouldn't it be more reasonable to assign some moral end as a duty? Shouldn't we expect something like "promote others' moral lives" as a fi nal virtuous aim? Kant remains consistent with his earlier stated view in which he refuses to defi ne happiness in terms of a virtuous life and to reduce natural happiness to moral happiness.
Were we talking about persons whose desires were identical to moral actions, we would have perfect moral beings. We have, however, sensible rational beings for whom happiness follows from the correspondence between events and their will, as expressed in the Critique of Practical Reason . Th eir will could be anything. When we talk about promoting someone's happiness, it is her task to decide what her happiness consists of. Of course, I could refuse to satisfy desires that are not permitted or contrary to the moral law.
When it comes to my promoting happiness as an end that is also a duty, this must therefore be the happiness of other human beings, whose (permitted) end I thus make my own end as well . It is for them to decide what they count as belonging to their happiness, but it's open to me to refuse them many things they think that will make them happy but I do not, as long as they his own examples. One such maxim is to love one's neighbor; another is to love one's parents. According to Kant, to be a wide duty is not to permit exceptions to these maxims. Rather the agent can limit one maxim by another. If an action performed from the maxim "to love of one's neighbor or friend, " for instance, to house a jobless, homeless, and noisy friend in your parents' home, would damage the well-being of one's parents, then we are permitted not to carry out this action, because one maxim (the love of one's parents) would limit the other (the love of one's friends).
If someone failed to fulfi ll the duties of love, this would not be considered a vice, but only a lack of virtue.
Failure to fulfi ll mere duties of love is lack of virtue ( peccatum ). But failure to fulfi ll the duty arising from the respect owed to every human being as such is a vice ( vitium ). For no one is wronged if duties of love are neglected; but failure in the duty of respect infringes upon one's lawful claim. (MS,

6: 465)
Th is quote suggests that the priority of the right over the good should probably be the correct interpretation of Kant's philosophy, even if some recent authors have claimed to leave deontology behind. 7 In the Doctrine of Virtue , Kant goes further than the mere formal negative view that is sometimes attributed to him. Kant did provide us with a theory of virtue, although his theory does not provide us with an exhaustive list of virtues, as the Aristotelian-type theory does. He certainly recognizes that there is moral merit in doing more than the negative commands of the categorical imperative, that being benevolent and benefi cent is certainly better than not being so. Nevertheless, there is room to decide how our own values will accord with the demands of virtue and to follow these demands. If a person is benefi cent in her acts, she is undoubtedly following the demands of virtue. However, if she decides not to be so, she is not doing something wrong because nobody is wronged by her actions.
Th e duty to promote others' happiness, as a wide duty, is subordinate to the negative demands of the moral law and to the narrow obligations of the duties of right. Th is entails two things: First, the latter should be satisfi ed prior to the former; one should not lie to promote the happiness of a friend, for instance, because the duty to promote others' happiness is subordinate to the narrow obligation not to lie. Second, once the narrow obligations are fulfi lled, we have space to choose what to do to fulfi ll the demands of virtue. If we decided to do nothing, at least in a particular situation, to promote others' happiness, it would mean a lack of moral worth, but not culpability.
Th e duty of love, understood as practical love, shows that blaming a mere formal theory is wrong, since Kant shows the directions of the virtuous life.
However, it does not lead to the abandonment of deontology, because the good (the wide duty) will always be subordinate to the right (the strict duties).

Virtue and pleasure
In We should consider that benevolence (as the promotion of others' happiness) is obtained as a generalization of the maxim of being helped in case of need: Th e reason that it is a duty to be benefi cent is this: since our self-love cannot be separated from our need to be loved (helped in case of need) by others as well, we therefore make ourselves an end for others; and the only way this maxim can be binding is through its qualifi cation as a universal law, hence through our will to make others our ends as well. Th e happiness of others is therefore an end that is also a duty. (MS, 6: 393) Th e need to be helped, in order to be a universal obligation, should be transformed into a general rule of benefi cence. Moreover, to choose the happiness of others as an end is a way to indicate that we should do more than we are morally bound to do, in order to be fully virtuous: If someone does more in the way of duty than he can be constrained by law Th e center of Kantian theory will comprise the moral law and the narrow duties. Th is is nothing but morality in a narrow sense of right and wrong. In this central nucleus, the end of promoting others' happiness plays no role. Th e happiness of others operates in a wider realm of morality. 9 But the wide duties of virtue are subordinate to the duties of right, showing that we cannot forget the essential deontological character of the Kantian ethics. If the Doctrine of Virtue shows how to go beyond the formalism of the moral law, it does not go so far as to transgress the limits of deontology.
Kant has been criticized 10 for being incapable of providing an ethics with moral content. Th e Doctrine of Virtue is an eff ort to provide this content; however, it is nothing more than general outlines that will guide us to promote others happiness or our moral perfection. Virtues like courage, prudence, justice, and wisdom give place to virtues related to these two ends. We have mostly a list of vices rather than a list of virtues. Lewdness, excessive eating or drinking, lying, avarice, and servility are vices opposed to our virtuous purpose to search for our natural and moral perfection. Arrogance, defamation, and ridicule are vices opposed to the respect of others. In any case, the Kantian theory of virtue comprises mostly interdictions rather than prescriptions.
Th e indetermination of what is a good life is related to the changing of the defi nition of happiness. Happiness is not defi ned by virtue, but in terms of pleasure and joy, which makes it variable according to personal preferences.

Virtue as fortitude against the inclinations
Kantian virtue does not aim at any telos as the achievement of happiness, even if it is only the happiness of the Stoic tranquilitas . If virtue does not lead necessarily to happiness or pleasure, what is then virtue? What is the aim of Kantian virtue? Th e Kantian answer will be: virtue is self-constraint in order to attain full rationality. For holy beings there will be no doctrine of virtue, because they are never tempted to violate duty. Virtue is autocracy of practical reason, and entails "consciousness of the capacity to master one's inclinations when they rebel against the law" (TL, 6: 383).
Virtue is perfection in choice, where one can freely determine oneself. For inner freedom two things are required: "Being one's own master in a given case ( animus sui compos ), and ruling oneself ( imperium in semetipsum ), that is, subduing one's aff ect and governing one's passions" (TL, 6: 407). Th e fi rst task is easier than the second. Since aff ects are precipitate or rash, they can be taken as something childish or weak. Aff ects do not lead to real vice, but only to weakness of the will, when man cannot control himself in order to accomplish what his will has determined to do. As we saw in Chapter 2, weakness is to act contrarily to what has been determined by the will. If it is a sign of rational agency to act according to the motives which determined the will, weakness can lead to irrational action, actions which the agent had decided not to perform.
In the Doctrine of Virtue , Kant proposes a cure to heal weakness of the will-this momentary inability to act according to reason-and this therapy is attained through the duty of apathy: Since virtue is based on inner freedom it contains a positive command to a human being namely to bring all his capacities and inclinations under his reason's control and so to rule over himself, which goes beyond forbidding him to let himself be governed by his feelings and inclinations (the duty of apathy ); for unless reason holds the reins of government in its own hands, his feelings and inclinations play the master over him. (TL, 6: 408) Kantian apathy should not be taken as indiff erence to the objects of choice.
Th at is the reason why Kant calls it moral apathy: Th is misunderstanding can be prevented by giving the name " moral apathy " to that absence of aff ects which is to be distinguished from indiff erence because in the case of moral apathy feelings arise from sensible impressions lose their infl uence on moral feeling only because respect for the law is more powerful than all such feelings together. (TL, 6. 409)

Virtue as a process, but not an Aristotelian habit
Kant points out that virtue is a process, that occurs over time through which man reaches the highest stage in human morality; however, it is an ideal and unattainable: Virtue is always in progress and yet always starts from the beginning . It is always in progress because, considered objectively, while yet in constant approximation to it is a duty. Th at it always starts from the beginning has a subjective basis in human nature, which is aff ected by inclinations because of which virtue can never settle down in peace and quiet with its maxims adopted once and for all but, if it is not rising, is unavoidably sinking. (TL,

6: 409)
Here it seems that the Kantian virtue is nothing more than an Aristotelian habit, a second nature attained by practice. However, Kant denies that virtue is habituation in the Aristotelian sense. Were virtue a habit, man would lose the power to freely choose maxims for his conduct: For moral maxims, unlike technical ones, cannot be based on habit (since this belongs to the natural constitution of the will's determination): on the contrary, if the practice of virtue were to become a habit the subject would suff er loss to that freedom in adopting his maxims which distinguishes an action one from duty. (TL, 6: 409) Virtue is a process in time, in which the agent fortifi es his will in order to do what he has decided to accomplish. But how can virtue, as fortitude, be a cure for inclinations? Since weakness is a momentary lack of control, it is not diffi cult to see how virtue could be a cure for this. Virtue increases the force of the will and helps to prevent aff ects leading the agent to irrational actions.

Th e beautiful and the good: Refi nement as a propaedeutic to morality
In this chapter, I would like to explore the connection between refi nement and morality, both in its closest relation between the judgment of the beautiful and moral judgments, and in the presence of the feeling of pleasure and displeasure in morality. I will begin with Critique of Judgment and the thesis of § 59, which claims that the beautiful can be considered a symbol of the morally good. I will move on to the examination of moral emotion in Th e Metaphysics of Morals (1797) and, lastly, I will examine the relation between the realms of taste and of virtue in Anthropology from a Pragmatic Point of View (1798). I will try to show, ultimately, that there is a Kantian consideration of aesthetic aspects in morality, which contrasts with those formalities presented in the Groundwork and in the Critique of Practical Reason . I will ask whether there is a possible change in opinion between these two books, or if it is merely a diff erent form of presentation in the various works. If we may present real examples of empirical concepts and if, moreover, we can present schemes of the categories of understanding, the same does not occur with the ideas of reason, which need a symbol in order to make their content indirectly sensitive. Francesca Menegoni calls attention to the diff erence between symbol, example, and schemes: "We cannot take a symbol as a simple example that shows the necessary intuition to prove the reality of an empirical concept, nor can we take it as a schema, whose reference to the concepts of understanding is direct. " 5 Proving the objective reality of the ideas of reason is an impossible task, since there is no intuition which corresponds to them; however, there is a possibility to exhibit these ideas, even if indirectly. Th e scheme is a direct exhibition; the symbol is an indirect exhibition; while the fi rst operates demonstratively, the second operates through an analogy.

Th e Beautiful as a symbol of the good
Allison explains what the indirect exhibition of a concept comprises: According to this account, what is directly presented ( darstellt ) in a case of indirect exhibition or "symbolic hypotyposis " is not the idea to be symbolized but some other (schematizable) concept. Th e representation of the object, which is the sensible realization of this latter concept, then functions as the symbolic exhibition of the initial (unschematizable) idea just in case judgment's refl ection on it is formally analogous to the form of refl ection on the original idea. 6 Reasoning through analogy constitutes a theoretical proof to which one may resort when there is no need for rigorous reasoning. However, even if it is not a rigorous reasoning, the level of proof is superior to mere hypothesis or reasonable opinion. 7 In reasoning through analogy, the faculty of judging, by means of a universal law and a particular principle, accomplishes two distinct operations: fi rst, it applies a concept to the object of a sensitive institution; second, it applies the rule of refl ection, under this fi rst relation, to a diverse object, in respect to which the fi rst accomplishes only the function of a symbol.
Allison stresses that as Kant indicates, this procedure involves a double function of judgment: one is quasi-determinative and the other is refl ective.
Th us, a living organism is a symbol of constitutional monarchy and a mill is a symbol of absolute monarchy. What is the relation between the idea of absolute monarchy and a mill? It is a mere analogical relation: in both cases, we think of the same kind of process, of objects (grains or people) being submitted to powers which are external to themselves (the mill or monarchic power). Th e structure of refl ection, when we consider the way the mill operates, is analogous to the structure of refl ection when we think about monarchy; however, there is no relation between the content of the symbol and its object.
Th e determined symbol, in this case, the mill, is not even the only possibility.

According to Paul Guyer,
Th is fact suggests that the connection between a symbol and its referent will be looser than that between examples or schemata and their respective referents. Nothing but a dog can serve as an example of the concept dog, and given the nature of your sensible intuition, nothing but an objectively valid temporal succession can serve as a schema for the pure concept of ground and consequence. But anything which allows one to relate ideas in the same way as does the handmill-some other mechanical device, perhaps, or other form of human relation-could serve equally well as a symbol of despotism. 8 In claiming that the beautiful is the symbol of good does not mean that beauty has a moral content or can serve as a scheme for morality, nor does the thesis imply a sensible intuition of morality. Th e analogy between aesthetics and morals is due not to a similarity of content (be it a moral content of beauty or sensible intuition of morality), but only to the common elements of both judgments, a similarity in the laws of refl ection. Th e mere thesis that the beautiful is the symbol of the morally good, in itself, would not leave us a viable passage between one and the other from the point of view of their content, since "being a symbol of " simply means an analogy in the refl ection of both types of judgment. However, at the end of § 59, Kant affi rms that taste can be used as assistance for the passage between the sensitive and the moral: "Taste as it were, makes possible the transition from sensible charm to habitual moral interest without too violent a leap. " 9 Kant goes further than that and attributes to taste the turning of moral ideas  11 it seems that the possibility of the former being empirically aff ected is decisive for the eff ective realization of moral action, while respect seems to be a feeling which aff ects the spirit ( mens , Geist ) more than the mind. In order for the action to take place, it is necessary that the mind be aff ected, which is done through moral feeling. Gisela Munzel accurately analyses the importance of aesthetical capacities in the eff ective practice of morality. 12 Beyond this, in moral feeling there is a regard for pleasure thus far nonexistent. As Guyer 13 highlights, the feeling of respect in Groundwork is a feeling of displeasure proportional to the coercion of the law, or, at most, as a feeling of self-approval, which radically diff ers from the feeling of pleasure in the Th ird Critique and the Doctrine of Virtue . A result of the determination of will which opts for the moral action, such a feeling seems to be one of reward.
Th at morality can give us a sort of satisfaction, and, moreover, that were we to lack this capacity-the Doctrine of Virtue tell us-we would be morally dead, seems to invite us to question the role of pleasure in Kantian morality. We should ask if moral feeling is a motive for a moral action. As we have seen, some commentators, such as Morrison, think respect is the incentive to moral action. My position, as explained earlier, is that respect is a consequence of the determination of the will by the moral law, not a motive itself. I think the same works for moral feeling: it's not a motive or incentive, but a consequence of the moral determination of the will. Th is satisfaction of taste derives from an agreement between the feelings of pleasure among agents whose origin is reason, according to a general law. Th e choice of this satisfaction is in accordance with the form of the principle of duty, due to its accordance with a general law. Th us, the exercise of aesthetic taste is a preparation for morality. Th is idea of the Anthropology follows the spirit of the Critique of Judgment , but with some interesting additions.

Aesthetic and moral pleasures
One of the curiosities in the Anthropology is the presentation of good manners as a transition between taste and morality. Th e rules of etiquette, of good hosting, prepare one for virtue. A good host manifests his aesthetic taste when choosing food and beverages, not only according to his/her personal taste but having in mind his guests' tastes. In the composition of these various tastes, the procedure of constructing a common taste which satisfi es all those present has a universal validity (Ant, 7: 242). Something similar takes place in how the host conducts the conversation, in the choice of topic, as in the number of guests, which according to the Chesterfi eld's rules should not be fewer than the number of graces (3) or more than the number of muses (10). Such a rule is not arbitrary, but seeks to construct an ideal group for communication, which would partake not only of the pleasure of table but of good ideas. Th e procedures of sociability according to the rules of good manners belong to the realm of taste, but prepare us for morality, in the sense that they construct a community which shares a common discourse:  All machines designed to accomplish with little power as much as those with great power, must be designed with art. Consequently, one can assume beforehand that Nature's foresight has put more art into the design of female than the male. (Ant,7: 303) Indisputably, Kant thinks men are superior to women in physical power, which is the reason why he claims that under uncivilized conditions, superiority is on the man's side only. Yet in civilization the (so-called) weakness of the feminine sex calls for deeper inquiry. She has the power to control men's inclinations and she brings men to moral behavior.
Kant shows indeed his old-fashioned sexism when he claims that one of the ends of nature regarding women is the preservation of species. However, he also attributes another end to women: the improvement of society and its refi nement. She is endowed with fi ner sensations, since nature "made this sex the ruler of men through modesty and eloquence in speech and expression" (Ant, 7: 306). She demands gentle and polite treatment from men; by doing so, they are brought to moral behavior, which is not morality itself, but is a preparation and introduction to morality.
My aim is to explore these relations Kant establishes between women, emotions, and morality, in order to show that the female sex is useful to moral education and has some moral interesting features.

Beautiful morality
Much has been said about Kant's prejudice against woman. One of the most striking examples is in Th e Observations on the Feeling of the Beautiful and Sublime , when, aft er stating that women could be successful in science, Kant advises them against it, because they will lose their power over men: Laborious learning or painful grubbing, even if a woman could get very far with them, destroy the merits that are proper to her sex, and on account of their rarity may well make her into an object of cold admiration, but at the same time they will weaken the charms by means of which she exercises her great power over the other sex. (Observations,2: 229) Men are sublime, woman are beautiful. Surprising as it is for our contemporary feminist moral sensibility, I will show that the attribution of the beautiful to women has some important consequences.
According to Kant, refi

Women's social virtues
In spite of some awful comments about women, Kant accepts that they have many virtues. Although some of these virtues appear to relate to a submissive character, they are well worth analyzing. One of these virtues is patience, although one could object that this is a very old-fashioned and submissive virtue, it is still proof of self-control.
Feminine virtue or vice is very diff erent from masculine virtue or vice, not only in kind but also in motive. She is expected to be patient; he must be tolerant. She is sensitive; he is perceptive. Th e man's business is to earn, the woman's is to save. (Ant, 7: 307) Th rough patience, women can control men, even for their own purposes. Because men like domestic peace, they prefer to submit to women in the domestic sphere, since they do not want to be hindered in their own aff airs.
"Th e woman wants to dominate, the man wants to be dominated" (Ant, 7: 306): although Kant sometimes attributes the capacity to dominate to some submissive and traditional capacities such as the ability to please (Ant, 7: 306), he acknowledges that women can better master some abilities required in society.
By controlling the rude manners of men, women can win them over.
For this, they use politeness and the art of pleasing. Th ey can also use some expressions of emotion, such as crying, which the rules of society allow them to have. Women can use their tears to control men, disarming them with their "tears of exasperation" (Ant, 7: 304). Th e end of this showing (or pretending) of emotions is the refi nement of men's rude manners, so, it is a moral purpose.
Women are allowed in society to show their emotions through tears, contrary to men, who can only be excused crying if they do not make any noise. However, these feminine tears have a moral purpose, and it is not contrary to morality.
Women are also more polite in social interaction. Kant surely is a moralist who condemns all lies and false promises. However, this does not imply that we should always be completely sincere to the point of being rude. Even in the text "On a Supposed Right to Lie for Love of Humanity, " he acknowledges that we should tell the truth only if we are asked for it. In the same way, we should not make false promises. It is a sign of character neither to break one's promise nor to speak an untruth intentionally (Ant, 7: 294).
However, in social life man should abstain from always telling the truth, when he is not asked to do so. Politeness is also an art of dissimulation and Kant attributes total sincerity to uncivilized people. To know how to be an actor is a sign of education and refi nement: Collectively, the more civilized men are, the more they are actors. Th ey assume the appearance of attachment, of esteem for others, of modesty, and of disinterestedness, without ever deceiving since nothing sincere is meant: Persons are familiar with this and it is even a good thing that this is so in this world, for when man plays this roles, virtues are gradually established, whose appearance had until now only been aff ected. (Ant,7: 151) In this civilized art of deceiving, women are better than men. Th ey are trained from early in life to smile, to please, and not to reveal their inner secrets, while men are easy to fathom (Ant, 7: 304). Th is aptitude is not immoral, but helps to develop politeness and a good social life. A woman put a veil around her secrets, as well as the secrets of nature. Her modesty and sensitivity to shame show that capacity to veil the coarsest secrets of sexuality: Sensitivity to shame is a secrecy of nature addressed to setting bounds to a very intractable inclination, and since it has the voice of nature on its side, seems always to agree with good moral qualities even if it yields to excess.
. . . But at the same time it serves to draw a curtain of mystery before even the most appropriate and necessary purposes of nature, so that a too familiar acquaintance with them might not occasion disgust, or indiff erence at least, in respect to the fi nal purpose of an impulse onto which the fi nest and liveliest inclinations of human nature are graft ed. Th is quality is especially peculiar to the fair sex and very becoming to it. (Observations, 2: 234)

Women and passions
One of the main criticisms against Kant has come from feminist philosophy.
According to this critique, some ingredients that are important to female identity, such as emotion, love, empathy, and cooperation, are not in consideration in Kantian ethics. Sally Sedgwick expresses this criticism when she says that "because moral agency on the Kantian view is a function of acting from reason rather than from feeling, it is said to refl ect features more of male than of female identity. " 3 She, however, supports Kant against the critique of a misunderstanding of human psychology. What Kant is saying is not that we are-or should be-cold people without any feeling or that in our meaningful relationships feelings should not play an important part. What he is claiming is that empirical motives cannot be the foundation of moral worth.
Kantian and anti-Kantian feminists usually agree in one point: for Kant women are more emotional than men. 4 Kantian feminists, such as Marcia Baron, support that the ideas of reason, autonomy, and freedom are the foundation of women liberation and should be cherished by women as important tools for their autonomy. She claims that while Kant "certainly was not thinking about women and the moral outrageousness of the roles into which they traditionally have been cajoled or forced, his ethical theory is far more able to provide the tools for challenging the roles than many other theories. " 5 However, women were excluded from this idea of reason when they were considered mainly emotional beings. She also accepts that Kantian virtues are mostly male virtues. Kant could be blamed for by "the tendency to equate male virtue with generic human virtue. " For Baron, then, the problem with Kant is not only that "he didn't recognize that women are full-fl edged rational beings, but also that he has too narrow-too masculine-a picture of the virtuous person. " 6 Marcia Baron does not deny that women are mostly emotional in Kant's assessment, but she tries to respond to the feminist critique, by denying that all aff ective is bad in Kant's view.
Another way to refer to this problem is to challenge the rationality of Kantian morality as the only possible way to be moral, showing that women can have a diff erent voice regarding morality, a voice that cherishes care and emotions. Th is view was fi rst argued by Carol Gilligan and gave rise to the ethics of care. 7 Allen Wood follows the same line of reasoning when he claims that to say women are more emotional is not to regard them as generally inferior to men considered as moral agents, because for Kant "it is one thing to have a temperament that makes the moral life more diffi cult and quite a diff erent thing to be worse as a moral agent. " 8 Anti-Kantian feminists also criticize his idea of autonomy as independence, because this shows the isolation of his morality, which is not appropriate for women. 9 Pro-and anti-enlightenment feminists agree that women are mainly emotional in Kant's view; however, I think that there is not a defi nite position that women are more emotional than men in the Kantian texts. I will sustain the opposite, that according to Kant  One can understand that in this sense: when one is taken by a strong emotion of sadness, it is better to express that and come back to a normal state of excitement. Laugher can help digestion, because it is always an exercise of the muscles which are used for digestion.
Another expression of feeling which is good for health is weeping: "A widow, who, as the saying explains, will not allow herself to be comforted, that is, who does not want to know how to dry up the fl ow of tears, unknowingly or unintentionally takes care of her health" (Ant, 7: 262). However, this eloquent expression of sadness is only allowed in the feminine sex. If men are moved to the point of crying, their expression should be more discrete: "Th is expression of tenderness, as a weakness of the sex, however, must not permit the male involved to be moved to shedding tears, but only to have tears in his eyes" (Ant, 7: 263).
Th ese minor comments about aff ects in women do not imply that they are more emotional than men. Kant claims that they have more moral feelings, but is defi cient or more subtle in women. Regarding sex, women seem to have a natural self-control that is absent in the masculine gender. Inversely, they are endowed with modesty and with a natural capacity to refrain from men's bold initiatives. Allen Wood is one of the few commentators who acknowledge that for Kant men are more emotionally vulnerable than women. In a remark about marriage, he claims that while "men take advantage of women through their greater physical, intellectual, social, and economic power, " "women take advantage of men through their manipulation of the man's emotional vulnerability and lack of self-control. " 10 He also points to men's lack of selfcontrol. Th e fact that Kant claims that women are less rational does not imply that they are more emotional than men.
Even women's vanity is less harmful than the masculine one. If all vanity should be considered bad for morality, at least women's vanity is not as dangerous as men's vanity: it is only a beautiful fault. Men's vanity, on the other hand, takes the form of the three main passions: lust for honor, lust for power, and lust for money. A woman's vanity, on the contrary, manifests itself in the enjoyment of adornments and dressing, which is not meant to attract the attention of men. Kant believes that women dress for their own sex, while men dress for women, "if this can be called dressing, it goes so far as not to shame his wife by his clothes" (Ant, 7: 307). Women are more inclined to some aff ects than men, but these are harmless aff ects, which are even good for the health (Ant, 7: 262). complementary to those of men (having patience rather than tolerance, being sensitive rather than responsive, saving rather than acquiring). " 11 Contrary to Rumsey, I consider that it would not be fair to say women are led by their emotions, because the Kantian portrait of the fairer sex is not painted with the bold colors of passions and aff ects. Women are rather represented as endowed with a subtle morality, where we can observe the mild color of moral feelings, such as sympathy, or virtues such as benevolence. Th is is certainly not a morality of principles, but indubitably a beautiful morality.
Patrick Frierson is one of the few commentators who are sensitive to a positive aspect of a beautiful morality of women. He admits that although "Kant's account might seem to preclude virtue in women, " they are capable of virtue, but a beautiful one. Since virtue is also diffi cult for men, "whereas few men will attain sublime virtue, women are well equipped for beautiful virtue. " 12 Frierson also admits that the apparent weakness of women helps the moral progress of human being: "Th e apparent weakness and timidity of women ends up becoming one of the driving forces behind cultural and even proto-moral progress in the human species. " 13 Mari Mikkola has also showed that Kant's view on women is less unsettling than the one attributed to him by feminists and has challenged the view that he claims women are morally defi cient. 14 However, Frierson claims that the Kantian interest in women as a driving force toward morality changed from the Observations to the mature critical philosophy. According to him, Kant changed from an empirical and sentimentalistic moral theory in the 1760s to a more rigorous rationalist morality in the Groundwork and later works.
I disagree with Frierson in two aspects. First, I think that Kant never supported a sentimentalist view about morality. In the Observations he only acknowledges that women may have a morality, even if they are not capable of principles. Second, a later text such as the Anthropology also indicates that women may play a role in the cultural education of mankind through a beautiful morality.
Kant clearly attributes a moral role to women. Even if they do not fi t an ideal of a morality of principles as individuals, their role in society will certainly help to build a more civilized world, which is a condition to morality.
Th e liberal tradition that goes back to Kant has divided feminists between the ones who think that autonomy and rationality are essential for women, and the ones who think these are mainly male ways of reasoning and should be replaced by another way of thinking. Martha Nussbaum is a supporter of the view that the enlightenment ideas of autonomy and freedom are positive aspects of the Kantian doctrine that can be valuable for the emancipation of women. She also claims that many elements of liberalism are important for women in their struggle to become full citizens and human beings that could totally enjoy their capacities.
In the book Sex and Social Justice , Nussbaum clarifi es what she calls liberalism: "When I speak of 'liberalism, ' then, I shall have in mind, above all, the tradition of Kantian liberalism represented today in the political thought of John Rawls. " 15 According to her, this tradition brings a twofold intuition that are essential for feminism: fi rst, that all human beings are of equal dignity and worth; second, that the primary source of this worth is a power of moral choice within them, consisting of the ability to plan a life in accordance to one's end.

9
Evil and passions 1

Aff ects and passions
If we disregard, for the moment, the diff erence in how they are related to objects, we fi nd that both aff ects and passions are considered illnesses of the mind, because both aff ect and passion hinder the sovereignty of reason.
However, the former is less harmful than the latter. Th is can be demonstrated if one compares anger (aff ect) with hate (passion). Anger intensifi es quickly and subsides in an equally instantaneous manner. Hatred, because it is a passion, does not allow for such control.
Since the passions can be coupled with the calmest refl ection, one can easily see that they must neither be rash like the emotions, nor stormy and transitory; instead, they must take roots gradually and even be able to coexist with reason. (Ant, 7: 265) Passions are more closely related to the will; nevertheless, this does not imply that they can be brought under greater control by reason. Th e inverse is suggested, namely, that they "take root" in reason and coexist with rational decision. Curiously the irrational aspects of aff ects make them preferable to passions. And Kant uses many medical metaphors to stress just this distinction: an aff ect is an intoxicant that causes a headache while a passion is a poison that causes a permanent illness (Ant, 7: 252), aff ect is a delirium (7: 266) or a "stroke of apoplexy" (7: 252), while passion "works like consumption or atrophy"(7: 252) or an illness that abhors all medication (7: 266), passions are "cancerous sores for pure practical reason" (7: 266) to which the physician of the soul could only prescribe palliative cures (7: 252). Th e metaphorical bundle of infi rmity of emotions speaks to their degree of evil. Aff ect, the least dangerous of the "illnesses of mind, " is related to weakness which can still coexist with a good will: Aff ects belong to feeling insofar as, preceding refl ection; it makes this impossible or more diffi cult. Hence an aff ect is called precipitate or rash ( animus praeceps ), and reason says, through the concept of virtue, that one should get hold of oneself. Yet this weakness is the use of one's understanding coupled with the strength of one's aff ects is only a lack of virtue and, as it were, something childish and weak, which can indeed coexist with the best will.
(TL, 6: 408) Aff ects make the work of understanding more diffi cult. If one has a weak understanding united with a strong aff ect, one momentarily loses control. But such a lack of control is not, properly speaking, a vice, but, as already discussed, a lack of virtue. In the Religion , this loss of control is called the frailty ( fragilitas ) of human nature, and consists in taking the moral law as the objective ground of action. However, it lacks suffi cient subjective force when compared to inclinations (Rel,6: 30). Passions, on the other hand, are beyond the weak adjectives of "childish, " because they are not just signs of weakness, but of true evil: A passion is a sensible desire that has become a lasting inclination (e.g., hatred , as opposed to anger). Th e calm with which one gives oneself up to it permits refl ection and allows the mind to form principles upon it and so, if inclination lights upon something contrary to the law, to brood upon it, to get it rooted deeply, and so take up what is evil (as something premeditated) into its maxim. And the evil is then properly evil, that is, true vice. (TL, 6: 408) Unlike aff ects which are temporary emotions, passions are characterized as lasting inclinations. Evil is connected to refl ection and to the will's formulation of maxims based on emotions. While an aff ect constitutes a subjective incentive that opposes a maxim, a passion may form principles for action. If one is overtaken by anger, he may be unable to act on maxims which he refl ectively acknowledges to be the right ones. But if the passion of hatred is present, the agent may choose maxims against the moral law, a choice that is classifi ed as a third degree of evil-that is, perversity of the human heart, in the Religion within the Boundaries of Mere Reason (Rel,6: 30). Th e passion of hatred, for instance, can lead someone to premeditate a murder. One could also murder someone based on a momentary uncontrolled aff ect. Even if the wrong action is the same, the latter is based on a discrepancy between the force of emotion and the will; the former is based on a will that has chosen to act according to a nonmoral maxim. Th at is why Kant says passions are more harmful to freedom than aff ects: One can also easily see that passions do the greatest harm to freedom; and if aff ect is a delirium, then passion is an illness which abhors all medication.
Th erefore, passion is by far worse than all the transitory aff ects which stir themselves at least to the good intention of improvement; instead, passion is an enchantment which also rejects improvement. (Ant, 7: 265) Th e evil character of passions comes from two features. First, a passion presupposes a maxim of the subject and is associated with the purposes of reason. It implies that maxims are a kind of distortion and perversion of reason.
Second, passions are never completely satisfi ed, thus they are labeled by the word "mania" ( Sucht ), meaning that they become an obsession about their never totally conquered object. Th at is why Kant supports that no physical love can count as a passion. Only the refusal of the object of the love can turn the aff ect of love into a passion of love.
Th ere are also other feelings that can either be an aff ect or become a passion. Besides love, Kant gives the example of ambition. An ambitious person, besides his own ends, usually wants to be loved by others; however, if he is passionately ambitious, he can be hated by others and even run the risk of becoming poor, because his passion makes him blind. If ambition, however, remains as an inclination, it will be compared to other inclinations and will not ruin the ambitious man. Th at is why Kant declares that "inclination, which hinders the use of reason to compare, at a particular moment of choice, a specifi c inclination against the sum of all inclinations, is passion" (Ant, 7: 265).
Passions do not operate like aff ects, making the subject momentarily incapable of acting according to what he has decided to do. An agent taken by an incontrollable aff ect may act against the maxim she has decided to follow, which may lead to irrational actions, going beyond what one call rational agency.
On the contrary, passions may form maxims of action, which highlights their evil disposition. Actions from passions belong to the realm of rational agency; however, they do not follow prudential reasons. Th is is the case of the ambitious man. If ambition is only an inclination, one can ground maxims of action in ambition, which will lead to the conquest of what the ambitious man desires.
When ambition as passion ground maxims of action, since passion is a mania ( Sucht ), it can lead to the opposite of what is desired. A blind ambition, such as Lady Macbeth's lust for power, can lead to the opposite of what is desired. She desperately wanted her husband to be king, but she ended up causing his death.

Natural and social passions
Kant divides passions into natural and social ones: natural passions are called "burning passions, " for example, the inclinations for freedom and sex; and the social passions are called "cold passions" and are ambition ( Ehrsucht ), lust of power ( Herrschsucht ), and greed ( Habsucht ) (Ant, 7: 272-75). Th e passion of freedom should not be understood as a rational desire to determine the will in an autonomous way; rather it is a desire not to depend on other people: "Whoever is able to be happy only at the option of another person, feels that he is unhappy" (Ant, 7: 268). It is a natural desire, a desire to keep others far away, and to live "as a wanderer in the wilderness. " It is a natural desire, not a rational one and comes from the desire not to depend on anyone, which belongs to the natural man before "public law protects him"-that is, in the state of nature.
Th e most dangerous passions, however, are not the innate, but the acquired ones, which arise from culture. In the Religion , Kant states that the evil principle of human nature resides in passions, "which wreak such great devastation in Aft er all these negative features imputed to passions, we should ask: Can virtue be a cure to passions? Th at the evil of passions is worse than the evil of aff ects can be attested by many passages in the Religion . Kant also cites the Bible in his own words: "We have to wrestle not against fl esh and blood (the natural inclinations) but against principalities and powers, against evil spirits" (Rel, 6: 59) in order to assert that evil does not reside in sensible incentives.
In the Religion , Kant maintains that inclinations are good and that evil should be sought for in a rational principle. Th is position seems to contradict the Anthropology from a Pragmatic Point of View , where both aff ects and passions are considered illnesses of mind. We should try to solve the apparent contradiction between the Religion and the Anthropology . If we correctly distinguish the purpose of the texts, we see that while in the Religion Kant is concerned with the source of evil, which cannot be placed in the natural realm; in the Anthropology he is merely trying to explain emotions. In the latter, it is correctly shown that both aff ects and passions may impede the will, either as stormy feelings that hinder the accomplishment of the action based on a moral maxim, or by grounding the choice of the maxim. However, in both the Religion and Anthropology the worst evil resides in a rational principle, not in a natural one, thus, even in the Anthropology , passions are thought to be more dangerous than aff ects. Aff ects can be the cause of weakness, but passions are the cause of true evil.

How to heal passions
Th e extirpation of aff ects is not Kant's central goal and he even claims that the extirpation of inclinations would "not only be futile but harmful and blameworthy as well" (Rel,6: 58). However, it is his invariable position that we should extirpate passions, since they are not natural feelings or inclinations.
In his analysis of emotions and evil in Kant, Michael Rolf correctly argues that, for Kant, "all passions are evil, and that all passions are social in content, " but Kant "does not claim, and in fact he explicitly denies, that aff ects are evil, at least in the sense that passions are evil. " 2 He considers that "aff ects, in contrast with passions, are not evil in the way passions are because they lack what makes passions evil, namely, a maxim opposed to the moral law. " 3 In order to win the battle against this principle of evil, one should fi nd its cause. If men search the circumstances that lead them to evil principles, they will fi nd out that they are not related to their raw nature, but to the corruption of the will that one man produces over others. If a man considers himself to be poor, he does so "only to the extent that he is anxious that the other human beings will consider him poor and will despise him for it" (Rel,6: 94).
In their works about evil, both Allen Wood 4 and Sharon Anderson- Gold 5 call attention to the fact that evil in Kant has its source in our social condition.
Since evil originates from social relations, fi ghting against the evil of passions implies an eff ort to build a new society that could counteract passions.
In the chapter "Radical Evil" of the book Political Emotions , Martha Nussbaum also stresses the social feature of human evil in Kant. She says, "Th e fact that we are animals is not the primary source of our moral diffi culty" and Kant's "key contention is plausible: the tempter, the invisible enemy inside, is something peculiarly human, a propensity to competitive self-love, which manifests itself whenever human beings are in a group. " 6 Th e raw nature of men, although able to produce strong inclinations that are diffi cult to master, does not lead to the corruption of the human heart.
Kant is unequivocal in asserting that only association of men is able to produce pure evil: Envy, addiction to power, avarice, and the malignant inclinations associated with these, assail his nature, which on its own is undemanding, as soon as he is among human beings . Nor it is necessary to assume that these are sunk into evil and are examples that lead him astray: it suffi ces that they are there, that they surround him, and that they will mutually corrupt each other's moral disposition and make one another evil. (Rel,6: 94) Th is claim is unambiguous: the inclinations are not by themselves the source of evil, nor are our aff ects. Th e passions of envy, addiction to power and avarice are awakened by interaction with other human beings, even if there is no bad behavior from others. Human beings are not evil because they are corrupted by already wicked persons. Ordinary social interaction makes human beings evil, because this interaction awakes comparison between people. Kant also claims that comparison is the source of this social evil: men feel that they are poor because they compare themselves to others, and the fear of being despised or dominated produces the evil passions of ambition and greed.
Nussbaum agrees with this very pessimistic Kantian viewpoint: "Even when people are well fed and housed, and even when they are reasonably secure with respect to other prerequisites of well-being, they still behave badly to one another and violate one's other rights" (Nussbaum 2013, p. 167). Nor is evil a matter of social teaching: "Kant is surely right when he suggests that people require no special social teaching in order to behave badly, and indeed regularly do so despite the best social teaching" (Nussbaum 2013, p. 167).

Is virtue enough to heal evil?
Could virtue also be considered a cure for evil? If evil comes from the weakness of the will, virtue can help to strengthen the weak will. Weakness is the fi rst degree of the propensity to evil: it refers to the case in which one has a weak will and is infl uenced by a strong aff ect, and quickly loses control. However, such lack of control is not, properly speaking, a vice, but a lack of virtue.
Virtue, as strength, could work as a cure for aff ects, because these are temporary outbursts of feelings. As Kristi Sweet highlights, "Th ere are numerous ways in which Kant defi nes virtue, and virtue itself is manifold in its constitution, perhaps fi rst in Kant's understanding of it is that it is strength. " 7 Like strength, it can work against inclinations and aff ects that make it diffi cult to maintain our resolve. Virtue implies abiding with the principle of moral law, but it also requires fortitude in keeping our decision to follow the moral law.
Could virtue be a cure for the third degree of evil, or malignity? Recently some authors have pointed out that virtue can be the cure for all evil. Michael Rohlf states that "in general, virtue is the strength to comply with moral maxims in the face of our propensity to evil, understood as our tendency to prefer the satisfaction of inclinations, " and the education for virtue "will promote not only a good heart and the adoption of fundamental moral maxims, which together constitute the intelligible character of virtue, but also the strength of will to comply with those maxims in the face of our propensity to evil. " 8 However, since the evil of passions is connected to society, this education for virtue can only fully occur in a society based on the idea of virtue. Only a social remedy can overcome these cancers of pure practical reason.
If evil is social, the only way to overcome the evil of passions is through a community based on the ideal of the moral good. Virtue in the sense of individual strength is insuffi cient to accomplish this task without setting up a society, which will rule over passions.
Th e social solution to evil is clearly stated in the following quote: Inasmuch as we can see, therefore, the dominion of the good principle is not otherwise attainable, so far as human beings can work toward it, than through the setting up and the diff usion of a society which reason makes it a task and a duty of the entire human race to establish in full scope. For only in this way we can hope for a victory of the good principle over the evil one. (Rel,6: 94) Th is society is not juridical-civil society, but an ethico-civil society which can coexist with the former. While a juridical-civil, or political society, is the relation of human beings to one another under public juridical laws, an ethical-civil society is one in which they are united under the laws of virtue alone, without being coerced. Th ey can coexist and be composed of the same members.
An association of human beings merely under the laws of virtue, ruled by this idea, can be called an ethical and, so far as these laws are public, an ethico-civil (in contrast to a juridico-civil society), or an ethical community.
It can exist in the midst of a political community and even be made up of all the members of the latter (indeed, without the foundation of a political community, it could never be brought into existence by human beings). (Rel,6: 94) Kant points to an ethical community, which is the embodiment of virtue and of moral principle. Th is is not a political society, since even a perfect civil society could not overcome passions and therefore defeat true evil by itself.
In addition, this ethical community is a community of virtue, although not individual virtue, but of a shared one. It is-as Kant  In a political community, political citizens are still in the ethical state of nature. Th e citizens cannot be coerced to enter an ethical state, but they can do it. Th is decision rest on the person's will, since the citizen of the political community remains free: Th e citizen of the political community therefore remains, so far as the latter's lawgiving authority is concerned, totally free: he may wish to enter with his fellow citizens into an ethical union over and above the political one, or rather remain in a natural state of this sort. (Rel,6: 96) Th e ineff ectiveness of political institutions Kant claims in the Religion that human beings cannot ground the overcoming of evil only in the development of political institutions. In order to attain their moral destination, they need to build an ethical community. He seems to have changed his mind about possible progression in history based on improving political institutions. In the Idea for an Universal History , he claims that "the greatest problem for the human species, to which nature compels him, is the achievement of a civil society universally administering right"(Idee, 8: 22). In the Idea , the just civil institutions are considered suffi cient to develop the aim of human nature and to accomplish our moral end. Paul Guyer remarks that there is already a shift from the text Idea for a Universal History (1784) to the appendix of Perpetual Peace (1795). He argues that in the fi rst, moral change happens through a natural process, while in the second Kant claims that only the exercise of human freedom can lead to the moral destination of man. 9 Mutchnik claims that in order to understand Kant's conceptual shift one must turn to the Religion (1793), "where the problem of radical evil receives its fullest expression. " 10 He criticizes among others, Allen Wood, who has based his interpretation of evil in Kant only on the Idea : "Interpreters like Allen Wood have found in Idea for a Universal History the key to understanding the social dynamics of the propensity of evil, tracing the roots of Kant's view to his thesis about unsocial sociability. " 11 Th e idea of unsociable sociability plays an important role in the Idea , as an explanation of how immoral inclinations or passions can engender a moral outcome. Th is unsociable propensity, Kant affi rms, "is this resistance that awakens all the powers of human being, brings him to overcome his propensity to indolence, and, driven by ambition, tyranny, and greed, to obtain for himself a rank among his fellows, whom he cannot stand, but also cannot leave alone" (Idee, 8: 20). Some commentators have found in the idea of unsocial sociability the main social evil. Kristi Sweet remarks, "Th ose who suggest that there is something in our unsociable nature that promotes evil are right. " 12 She goes further and associates this social evil with the unsociable sociability of human beings "evil and the principle of self-love in which it is embodied is profoundly anti-social.
Th is is highlighted in the way that unsociable sociability is expressed in one's desire to 'direct everything as to get his own way'" (Sweet 2013, p. 87).
In the Idea for a Universal History with a Cosmopolitan Aim (1784), the unsociable sociability is an antagonism that will overcome our initial unsociable nature: from a bad origin, we will obtain a good outcome: Th us happen the fi rst true steps from crudity toward culture, which really consists in the social worth of the human being; thus all talents come bit by bit to be developed, taste is formed, and even, through progress in enlightenment, a beginning is made toward the foundation of a mode of thought which can with time transform the rude natural predisposition to make moral distinction into determinate practical principles and hence transform a pathologically compelled agreement to form a society fi nally into a moral whole. (Idee,8: 20) In the Religion , on the other hand, Kant renounces to the idea of a possible moral outcome from immoral passions. Th ere is no possibility that passions left by themselves will fi nd their way to morality. In the Religion , Kant stresses another kind of evil, very diff erent from the unsociable sociability. It is not this tendency to run away from society in order to be lonely that leads to evil, but the passions that are aroused through comparison with others.
Th e Anthropology (1797) presents another way of overcoming our evil inclinations through the cultivation of a cultivated society.
Th e summary of what pragmatic anthropology has to say about the vocation ( Bestimmung ) of the human being is that he is destined ( bestimmt ) through his reason to live in a society of human beings, and in this society, through the arts and sciences, to cultivate himself, civilize himself, and moralize himself. (Ant, 7: 324) Unlike the radical optimist of the Idea , in the Anthropology , Kant acknowledges that there is evil in men, which "is an inclination to desire actively what is unlawful, although he knows very well that it is unlawful" (Ant, 7: 324). He also recognizes that passions are cancerous sores of reason and does not attribute any good property to them. However, some hopefulness still remains, since passions, even if they are sores for pure practical reasons, can be overcome through the cultivation of arts and sciences. Th is sociocultural development, not of the individual, but of the species as a whole will be able to counterbalance evil and accomplish the natural destiny of the species, allowing it to attain full rationality.
However, neither the radical historical optimism of the Idea nor the cultural confi dence of the Anthropology seemed to be enough to overcome evil. In the Religion there is a new condition for this development, the establishment of an ethical community, which is not guaranteed by the suggested cultivation of human being of the Anthropology , nor by the progress of history and political institutions of the Idea.
A social solution, the ethical community, should supplement a historical and cultural solution to evil, since a civil political society, even the most perfect, will never attain it. Wood explains correctly how a moral community diff ers from every political community: Its laws cannot be statuses, derived from an arbitrary human authority, but must instead be purely moral laws, which recommend themselves to each man through his own reason. In addition to this, the very principle of a moral community of men will diff er from that of a political one. Th e legislation of every political or juridical state "proceeds from the principle of limiting the freedom of each to those conditions under which it can be consistent with the freedom for everyone. " 13 Th e laws of the political community are always coercive laws and a moral community should promote moral relations between its members. Good laws can compel men to an outward legality, but not to a real internal moral improvement of their character. Without a moral community, there could be an external conformity to the law, but we would never attain the full development of morality.
Allen Wood did not realize that a moral community is only necessary because evil in society is not unsociable sociability but pure evil, which will never be healed by the development of political and cultural history. However, he is right in explaining the necessity for a moral community to heal evil, because an outward legality is insuffi cient to attain the full development of morality.
Only an ethical community can overcome evil, because the roots of evil are social, and belong to passions that are stimulated through social interaction.
Political institutions are necessary, but not suffi cient conditions, because they can compel men to an external legality, but not to an improvement of their hearts.

Conclusion: An emotional Kant?
Kant has been criticized by many philosophers for not allowing any role at all for emotions in moral life. Bernard Williams even blames Kant's morality for this fl aw in contemporary ethics. In a celebrated chapter of Problems of the Self , entitled "Morality and Emotions, " he regrets that recent moral philosophy in Britain has not paid enough attention to the problem of emotions. According to him, British philosophers limit themselves to acknowledging emotions in "one of their traditional roles as possible motives for backsliding, and thus as a potentially destructive of moral rationality and consistency. " 1 Williams blames Kant's account of morality for that. In opposition to this, he struggles to dismiss several Kantian views about emotions, like the one according to which emotions are supposed to be too capricious and passively experienced, and only a product of natural causation. What Williams does not take into consideration is the fact that in Kant's morality there is a place for emotions, in the sense that it is not composed solely of a priori principles and that it should be completed by an account of emotions. In order to be in keeping with Kant's philosophy, one has to acknowledge that his moral anthropology deals with emotions in great length by analyzing the entire spectrum of human aff ects and passions.
I have shown so far that Kant's ethics accounts not only for a priori principles, but also for some empirical facts about human beings. One of these facts is that we have passions and aff ects that are diffi cult to control. Kant also shows that the construction of moral character, in the Aristotelian sense of choicemaking, is limited by the temperament of each individual. Among the four temperaments listed in the Anthropology (phlegmatic, melancholic, choleric, and sanguine), the fi rst one is the most likely to attain freedom of aff ects.
It is also not the case that Kant fails to account for individual character, as Bernard Williams argues. Th e idea that Kant's ethics does not leave room for the signifi cance of the personal in moral theory is misleading. Kant has a place for moral character. He only points out that moral character can be infl uenced by natural characters, called temperaments.
Temperament is related to physiological components, which make some temperaments too diffi cult to tame. Sometimes Kant states that we can diminish the disposition of a bad temper: "A hot temper can be controlled gradually by inner discipline of the mind" (Ant, 7: 260). However, while Kant allows for a possible reform of sensibility (Rel,, he also seems to accept that specifi c instances of aff ects are beyond the rational agent's control. Even if, through the inner discipline of the mind, someone can indeed decrease the number of specifi c instances of aff ects, some specifi c outbursts of emotions will still be beyond our control.  23 In the § 13 of the Discourse of Metaphysics , Leibniz states that "as the individual concept of each person includes once for all everything which can ever happen to him, in it can be seen, a priori the evidences or the reasons for the reality of each event, and why one happened sooner than the other. But these events, however certain, are nevertheless contingent, being based on the free choice of God and of his creatures. It is true that their choices always have their reasons, but they incline to the choices under no compulsion of necessity" (Leibniz, DM , § 13). In this statement, Leibniz tries to conciliate the necessity of each event with the free choice of creatures. First, Leibniz holds the position that the concept of an individual substance includes once for all everything which can ever happen to it. Th en, if someone knows the concept of an individual substance, one will be able to know everything that can be said concerning this individual. In this case, to know a particular substance is the same as the nature of a circle and derives all its properties from that. 3 I take "sympathy" as translation for Teilnehmung , instead of "compassion. " Aff ekt will be translated by aff ect, while Leidenschaft by passion. I will reserve emotions for a generic term that denotes moral feelings, aff ection, and passions. 4 Hume also doubts the existence of a creature, in which sympathy was completely absent, which he calls a "monster of fancy. " "It can be said that it does not have such a human creature, for whom the happiness of the others did not provoke pleasure (where it did not have place for envy or revenge), and the appearance of suff ering, pain. " D. Hume 2 In the Groundwork , Kant writes the following: "Suppose that the mind of this friend of man is clouded with sorrow of his own which extinguished all sympathy with the fate of others, but he still had the power to help those in distress, though no longer stirred by the need of others because suffi ciently occupied with his own and suppose that, when no longer moved by any inclination, he tears himself out of this deadly insensibility and does the action without any inclination, simply from duty, then for the fi rst time his action has its genuine moral worth" (G, 4: 398).
9 Wiebke Deimling also claims that "Kant's taxonomy of aff ective states is more complex than the distinction his predecessors in the seventeenth and eighteenth centuries make. " Descartes and Hume, for example use "passion as their most basic term for aff ective states. Spinoza and Hutcheson use 'aff ect' and 'aff ection' respectively as their most general concepts. Kant decides to use these terms for more specifi c aff ective states to get a nuanced picture. " See Wiebke, "Kant's Pragmatic Concept of Emotions, " in Emotion and Value in Kant , ed. Alix Cohen (Cambridge: Cambridge University Press, 2014).
10 While I use aff ects and passions as translations for Aff ekten and Leidenschaft en , respectively, I use moral feelings as referring to respect ( Achtung ) and moral feeling itself ( Moralische Gefühl ). In the paper "Kant's Taxonomy of Emotions, " Kelly Sorensen relate these two feelings to a reason-caused desire. See Sorensen, "Kant's Taxonomy, " 109-12.
11 I use inclination here in a wide sense, referring to a wide range of empirical impulses, including instincts, aff ects, and appetites. Inclination can also have a specifi c meaning in Kant's thought, as the third degree of the faculty of desire. For a more detailed explanation about inclinations, see Wood, Kant's Ethical Th ought , and Louden, Kant's Impure Ethics .

12
In Anthropology , Kant refers to this kind of sympathy as sensitivity, a capacity to feel states of pleasure or displeasure with the happiness or misery of the others. Kant writes the following: "Sensitivity is not opposed to equanimity because it is a faculty and a power which either permits the states of pleasure or displeasure, or even keeps them from being felt. Sensitivity is accompanied by a choice. Sentimentality, on the other hand, is a weakness because of its interest in the condition of the others who could play the sentimentalist at will, and even aff ect that person against his will" (Ant, 7: 236).
13 Love of human beings also belongs to what Kant calls "concepts of what is presupposed on the part of feeling by the mind's receptivity to concepts of duty as such. " However, Kant denies that this kind of practical love is, properly speaking, an emotion, since we cannot have the duty to love someone, if love is a matter of feeling. One can only cultivate this practice of benevolence, but to call it love is "very inappropriate" (MS, 6: 401). 17 According to Collins's notes on Anthropology (1772-73) we observe the soul in a triple perspective, like soul (anima/Seele ), mind ( animus/Gemüth ), and spirit ( mens/ Geist ). Th e passive soul is called soul/anima ( Seele ); the reactive soul is mind/animus ( Gemüth ); when the purely active soul is spirit ( mens / Geist ). Th ese three instances of the soul refer not to three substances, but to "three forms of feeling alive. 20 According to Elster, some emotions, like love and anger, are characterized by a point of no return beyond which self-control is no longer possible, since we lose control before we become aware of the emotion. See Elster, Strong Feelings , 28-29. I would like to suggest both that this no return point is due to inertial physiological aspects, and that Kant was aware of this when he said that the man in anger should sit down to calm himself. Once anger is triggered, the moment one notices it, it is already too late to maintain self-control.
21 Th e feeling theory goes back to Descartes, Hume and James, "What is an Emotion, " in Th e Nature of Emotion , ed. M. Arnold (London: Penguin, 1968). Recently, the work of Zajonc emphasizes the noncognitive elements of the feeling theory. Zajonc, "Feeling and Th inking, " 151-75. . " Together with this commentary, there is a note that refers to Sextus Empiricus's ataraxy, mentioning that the diff erence between Aff ekt and Leidenschaft was unknown to ancient philosophy, and that this was the reason why they did not distinguish between apathy and ataraxy. See 25,2: 1353, note 199 a . In the Religion , the Stoic's apathy is a battle against inclinations (Rel,6: 58). Here, inclinations should be understood in the general meaning of incentives of sensibility. In the published Anthropology , apathy is considered the freedom from aff ects (7: 753), having the same meaning as the moral apathy of the Doctrine of Virtue.
6 Ant, 7: 253. Kant is ambiguous about sympathy. While in the Doctrine of Virtue , sympathy is considered a moral feeling; in the Anthropology it is considered an aff ect. In this quotation, sympathy is undoubtedly considered an aff ect, whose suppression will be good to a virtuous state of self-control.
7 Th e soul cannot have full control over its passions because "they are nearly all accompanied by some disturbance which takes place in the heart and consequently also through the blood and animal spirits. " Since they have this intrinsically bodily component, once an emotion is incited, there is no possibility of a total rational control of it. We could control the lesser passions, but not the more violent one, until the movement of blood and spirits has been ceased (Descartes, Th e Passions of the Soul , AT, XI, 364).