Kant’s Account of Affects and Passions

Please do not quote or copy this material without prior permission from the author.

Direct all questions and comments to frierspr@whitman.edu

            Recently, Kant’s moral philosophy has undergone something of a revival, and transformation, at the hands of philosophers who seek to rehabilitate Kant in the face of Aristotelian challenges.  One form that this transformation has taken is the introduction of “emotions” into Kant’s moral theory as important and acceptable influences on moral action.  Unfortunately, this introduction of “emotion” into Kant’s moral theory is often done without sufficient attention to the fact that Kant uses a variety of different terms with quite distinct meanings to refer to aspects of human psychology that we might call “emotions.”  Among these, the most significant are Gefühl (feeling), Affect (which I will just translate as “affect”), and Leidenschaft (passion).  None of these terms maps exactly onto what one means in English by the term “emotion.” [1]   In this paper I will focus on explaining and distinguishing two particular sorts of emotion in Kant, affects and passions.   In particular, I will show why both affects and passions are considered hindrances to morality – unlike “feelings,” which can even aid morality – and also why affects reflect merely a lack of morality, whereas passions are evil.

            The best known account of the passions and affects is from the Metaphysics of Morals.  There, Kant says the following:

Affects and passions are essentially different from each other.  Affects belong to feeling insofar as, preceding reflection, it makes this impossible or more difficult.  Hence an affect 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 in the use of one’s understanding coupled with the strength of agitation of one’s soul [Stärke der Gemüthsbewegung] is only a lack of virtue and, as it were, something childish and weak, which can indeed coexist with the best will.  It even has one good thing about it: that this tempest quickly subsides.  Accordingly a propensity to an affect (e.g., anger) does not enter into kinship with vice so readily as does a passion.  A passion is a sensible desire that has become a lasting inclination (e.g., hatred, as opposed to anger).  The calm with which one gives oneself up to it permits reflection 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 to take up what is evil (as something premeditated) into its maxim.  And the evil is then properly evil, that is, a true vice. [2]  

In this passage, Kant mentions three related differences between affects and passions.  First, affects “belong to feeling” whereas passions are “desires.”  Second, affects reflect merely a “lack of virtue” whereas passions are “properly evil.”  Finally, affects “quickly subside” whereas passions are “rooted deeply.” 

            In this paper, I draw from Kant’s published Anthropology and his lectures in both anthropology and ethics to elucidate the connection between the first point, that affects belong to feeling whereas passions belong to desire, and the second, that affects are merely a lack of virtue whereas passions are properly evil.  I will also have something to say about how each of these points connect to the third point, the different durations of affects and passions.

            Kant considers both affects and passions “illnesses of mind.” [3]   Before turning to the particular sorts of illnesses they are, it is important to be clear on how a healthy mind functions for Kant.  In particular, since affects and passions interfere with normal processes of choice, it is important to briefly clarify what goes into a normal choice for a human agent.  The account offered here is taken from the introduction to Kant’s Metaphysics of Morals (6: 211-14).  According to this account, feelings lead to inclinations, which lead to interests, which lead to a choice, which leads to action.  One can distinguish five stages in normal action. 

(1) One has certain feelings of pleasure and/or pain. 

(2) These experiences of pleasure and pain give rise to inclinations – either immediately, as in the inclination to escape a painful situation, or mediately, through the anticipation of potential pleasures and pains. 

(3) These inclinations become interests insofar as the agent judges the pleasure or pain to connect with certain actions.  One must actively take an interest in one’s inclination.

(4) These interests lead to a choice – what Kant calls a “desire” – when the interest is considered as something to be realized by the agent.  This involves the agent viewing the object of interest as both capable of being brought about and as something to be actually brought about through her activity.  She must, in other words, decide to act for an end.

(5) This choice becomes action when the agent, through a further feeling, is actually motivated to act on the choice.

Much more could be said about this account of action, but this summary suffices to make sense of the ways in which affects and passions disease the soul.  Kant’s most general and consistent criterion for differentiating passions from affects relates to the different faculties that these “illnesses of mind” infect.  He says in the published Anthropology,

The inclination that can hardly, or not at all, be controlled by reason is passion.  On the other hand, affect is the feeling of a pleasure or displeasure at a particular moment, which does not give rise to reflection (namely the process of reason whether one should submit to it or reject it). [4]

In his lectures on anthropology, Kant again makes specific reference to desire and feeling: “Agitations of soul are twofold, affects and passions . . ..  Affect is a feeling through which we come out of composure [of soul].  Passion is rather a desire that brings us out of composure.” [5]   And this distinction between desire and feeling is also the distinction to which I have drawn attention in the Metaphysics of Morals.  In the next two sections, I will look at affects and passions in turn.  I will show how the distinction between feeling and desire accounts for the differences between Kant’s moral estimations of passions and affects, and how these differences are connected to the particular ways that each alters normal processes of choice.

a) Affects.

            Affects disrupt the normal process of choice because feeling completely dominates.  When this occurs, one is either incapable of action or acts purely on the basis of feeling, without any reflection at all.  Thus Kant says that

Affect is a surprise through sentiment, whereby the composure of soul . . . is suspended.  Affect therefore is precipitate, that is, it quickly grows to a degree of feeling which makes reflection impossible (it is thoughtless). [6]

Under the influence of an affect, one does not choose in the ordinary sense.  Feeling simply leads directly to action (not, of course, intentional action). [7]   The affect itself takes control.  Thus one goes directly from the first stage of choice (feeling) to the final stage (action) without any intermediate inclination, interest, or choice. 

            Kant illustrates this bypassing of choice with several examples.  I will mention two of the most illuminating.  The first example of an affect bypassing choice is shock. [8]   In his lectures on anthropology, Kant explains that

one sees a child fall into the water, who one could save, however, through a small aid, but one is so shocked that one thereby cannot do anything.  Shock anaesthetizes someone such that one is thereby totally incapable of doing anything. [9]     

Shock leads to an incapacity to act.  The strength of the feeling of shock prevents the normal process by which various feelings, like sympathetic pain, become desires and thereby bases on which one can act.              In contrast, a feeling of concern, however intense, that does not incapacitate will issue forth in an inclination to help, which may be taken up as an interest, made the basis of a choice, and thus lead to action.

            A very different example is rage. [10]   Though the affect of rage might incapacitate, it more often leads to rash action.  The feeling bypasses the normal process of choice and issues forth in action.  It is important to remember that even when one acts according to choice (step 5, above), that action is always effected through feeling.  In this case, feeling motivates directly, without any intervening choice. [11]   However, since the action is based on feeling, as soon the feeling passes – either because the object is no longer present or simply because the object no longer arouses the feeling – the impetus to action passes.  Thus “what the affect of rage does not accomplish quickly will not be accomplished at all.” [12]   Affects similar to rage in this respect include sympathy and courage insofar as these rise to the level of affects.  These affects lead one to act immediately on the basis of feeling, without any deliberation at all, but once the object of feeling ceases to provoke, the impetus to action vanishes as well.  In contrast to rage, normal anger that does not rise to the level of affect will give rise to an inclination, say, to do harm to the object of one’s anger.  This inclination may then become an interest, which one weighs against others in deliberation about whether or not to do harm.  The affect of rage, in contrast to this feeling of anger, prevents this reflection.

Because feeling bypasses choice altogether, it is proper to say that “this weakness . . . is only a lack of virtue . . . which can indeed coexist with the best will” (6:408).  One might murder someone out of rage, but one’s will is not involved.  One might fail to provide easily provided assistance to someone is desperate need, but only because one’s will is inaccessible during that time.  Acting or failing to act due to affects is morally similar to acting or failing to act when one is asleep.  In both cases one is morally unconscious, so to speak, rather than morally corrupt.  Thus affects hinder the development of virtue only in the negative sense.  One is not made less virtuous because of them, but one does not progress in virtue.

            Still Kant does claim that affects are to be prevented and cured. [13]   One may wonder why one would have any moral concern about affects if they do not reflect any moral failing.  One obvious reason, of course, is that insofar as one is morally obligated to promote the happiness of others, one should prevent those affects that will inhibit this end.  A more important reason for Kant, however, is that affects undermine self-mastery, which is a condition of moral agency.  In his lectures on ethics, Kant explains that

The self-regarding duties are the conditions under which alone the other duties can be performed. But what is the subjective condition of the performance of duties to oneself?  Here is the rule: Seek to maintain self-mastery; thou wilt then be fit to perform thy self-regarding duties. [14]

At the end of this section of his lectures, Kant highlights affects and passions as two important hindrances to self-mastery.  It is clear from the account of affects I have offered how they interfere with self-mastery.  One is completely unable to regulate one’s conduct while under the influence of affect.  But this account of affect also seems to undermine the importance of constant self-mastery for being dutiful.  If being under the occasional, or even frequent, sway of affect is consistent with a good will, one might wonder why affect would be a moral problem at all.  To be sure, one is not virtuous during the instant in which one is affected, but one might still be virtuous overall and be influenced by affects often.

I do not have the time in this talk to give a full account of the moral danger of affects, but one passage from the Metaphysics of Morals provides a least of hint of how that account would go. Kant says, “virtue is always in progress . . ..  If it is not rising, [it] is unaviodably sinking.” [15]   In Kant’s mature account of the good will, offered most fully in the Religion, all good human wills begin from a condition of radical evil and are therefore characterized by constant struggle against the evil in one’s own volition.  Insofar as one is under the influence of affects, one rests in a fight that should be incessant, and vice is able to maintain a foothold in one’s will.  Even if one is not evil because of the actions one performs under the influence of affects, those actions can hinder one’s capacity to develop a good will in the way that one should.  Needless to say, this account needs more elaboration.  For now we will move on to the passions, which Kant considers to be direct threats to morality.

b) Passions.

Passions do not bypass the normal process of choice and action.  They pervert it. [16]   Under the influence of passion, feeling still gives rise to inclination (step 2).  But whereas normally reason intervenes in making that inclination a choice (steps 3-4) by weighing it against other inclinations and perhaps also the moral law, in the case of a passion one immediately takes an interest in one’s inclination and chooses to pursue its end.  Feeling still leads to inclination, interest, choice, and then action, but all of these steps are immediate, without full control by reason. 

Although they do not allow full rational control, however, passions still allow reflection in a certain sense.

Since the passions can be coupled with the calmest reflection, one can easily see that they must neither be rash like affects, nor stormy and transitory; instead, they must take roots gradually and even be able to coexist with reason. [17]  

Passions allow reflection in the sense that they depend upon the understanding transforming a mere inclination into an interest and thereby into an object of choice.  Thus they persist even when the initial feeling is gone.  They do not, however, involve the comparison of the particular inclination with one’s other inclinations or one’s interest in morality.  

Inclination that hinders the use of reason to compare, at a particular moment of choice, a specific inclination against the sum of all inclinations, is passion. [18]

In a state of passion, reflection is necessary in the sense that one deliberately chooses to make a particular object of inclination the goal of one’s activity.  But this reflection is not a self-conscious choice of this inclination over some particular others, or of this inclination over the moral law per se.  Passion undermines consideration of those other interests.  Still, deliberation is important and sets passions apart from affects.  When one acts on a passion, one chooses.  Kant reiterates the extent to which passions are rational when he claims that

Passion always presupposes a maxim of the subject, namely, to act according to a purpose prescribed for him by his inclination.  Passion is, therefore, always associated with the purposes of reason, and one cannot attribute passions to mere animals any more than to purely rational beings.  (173; 7:266) [19]

Passion cannot be attributed to mere animals precisely because it always involves action on a maxim.  When one acts out of a passion, one endorses one’s inclination as a reason.  In fact, one endorses one’s inclination as the overarching and supreme reason for action.  And on that basis, one acts.  But one does not choose one’s inclination over and against any particular other inclinations, or the sum of inclinations, or the moral law.  Rather, one intentionally disregards those other interests without consideration. [20]   In that sense, though passion is reflective, and hence cannot be ascribed to animals, it is not fully rational.

            The fact that passions are perversions of choice, that they are reflective even if not rational, has two important implications.  It makes passions extremely difficult to cure, and makes them properly evil.  Passions are more difficult than affects to cure because one infected with a passion resists cure.  If one frequently acts impulsively from rage, when the rage subsides, one will regret one’s action and seek to protect against future outbreaks in the future.  But passions persist [21] as stable frameworks of choice according to which one never wants to be rid of the passion. [22]   Thus Kant says

If affect is a delirium, then passion is an illness that abhors all medication.  Therefore, passion is by far worse than all those transitory affects, which stir themselves at least to the good intention of improvement; instead passion is an enchantment which also rejects improvement . . ..  Passions are cancerous sores for pure practical reason, and most of them are incurable because the sick person does not want to be cured and avoids the dominion of the principles by which alone a cure could be effected. [23]

Later he reiterates this theme:

Affect produces a momentary loss of freedom and self-control.  Passion surrenders both, and finds pleasure and satisfaction in a servile disposition.  But because reason does not desist from its summons to inner freedom, the unfortunate victim is suffering under the chains from which he cannot free himself, because they have already grown into his limbs, so to speak. [24]

For one infected with passion, the passion works its way into the very freedom by which one might be able to free oneself from it.  Hence passions are incurable, as Kant says, because the sick person does not want to be cured. [25]

            The incurable nature of passions is one reason that they are considered worse than affects. The other and more important reason is that they are morally evil.

Passions are not, like emotions, merely unfortunately moods teeming with many evils [Übeln], but they are without exception evil [böse]. [26]   Even the most well-intended desire if it aims (according to matter) at what belongs to virtue, that is, to charity, is nevertheless (according to form), as soon as it changes to passion, not merely pragmatically pernicious, but also morally reprehensible. [27]

Because passions do not preclude choice but reflect a perversion of it, they involve the will.  And they involve a will that deliberately rejects all interests, including the moral law, in favor of the pursuit of an inclination.  To set any inclination, however kind-hearted, above the moral law is evil.  So all passions are utterly incompatible with having a good will.  They are “properly evil.”

            Affects and passions both hinder morality, and both interfere with normal processes of rational choice.  Affects infect “feeling” to bypass choice, and are thus merely a lack of virtue.  Passions infect the faculty of “desire,” or choice, itself.  They thus pervert choice and are evil.



[1] Thus the feeling of pain at being kicked by someone is a feeling, but I don’t think we would want to call it an “emotion.”  One’s anger at that person is also a feeling, but one that rightly belongs to emotion.  Similarly rage is an affect that we would call an emotion, but so is shock, which we might be less likely to consider an emotion.

[2]   Metaphysics of Morals, 6:407-8.

[3]   Anthropology 7:251.

[4] Anthropology 155; 7:251, my emphasis.

[5] Lectures on Anthropology, Ak. 25:589, my emphasis.

[6] Anthropology 156; 7:252.

[7]   Throughout this account of affects, I will use “action” to refer to a person’s behavior in the world.  This precisely does not, in the case of affects, involve agency, hence does not involve an “actor” in the strict sense.

[8]   Though Kant gives several examples of affects in the Anthropology, they often obscure the difference between feeling and affect more than in his lectures.  Thus, examples will be taken primarily from the lectures.

[9] Lectures on Anthropology, 25:592 

[10] See too KU  273; 133, where Kant describes anger as an “affect of a VIGOUROUS KIND.”

[11]   It is important to distinguish the case of action from affect with a similar sort of action characterized by the example of Korsgaard’s wanton.  (See Sources of Normativity.)  The wanton is someone who chooses to endorse every one of his desires.  He feels like attacking someone who gave him a funny look, so he decides that he will.  He feels like relaxing for a beer, so he decides that he will.  By contrast, one who attacks another out of affect never even endorses his feeling.  He just acts on it. 

The distinction might be made clearer by saying that for one who acts according to affects, there is no first person perspective.  The first person perspective about which Korsgaard is concerned is precisely the perspective of reflection that affects bypass.  And it is because this reflection is bypassed that affects are not signs of an evil will.  Whereas the wanton is evil (in Kant’s sense) because he endorses his raw feelings, one acting from affective rage is momentarily animal, because they do not act (strictly speaking) at all.

[12] Anthropology 7:252.

[13] Anthropology 156; 7:251

[14] Lectures on Ethics p. 138 (Infield edition).

[15] 6:409.

[16]   Here it is worth pointing out that Kant’s account of the passions is different in his early lectures.  At that time, Kant still seems to have thought of desire as more akin to feeling than to choice.  As a result, passions, like affects, are thought of as bypassing rather than corrupting choice. Kant also thought of passions, in his early lectures, as only slightly longer lasting than affects.  Neither was considered truly lasting.  (See 25: ref.)  This changed by the time Kant published discussions of the passions and affects in the Metaphysics of Morals and Anthropology.  By then, the faculty of desire in a rational being just was the empirical faculty of choice.  Hence an out of control desire – a passion – is a choice that is out of control.  This led to a different treatment of the passions and a more clear differentiation between affects and passions in Kant’s later writings.

[17] Anthropology 172; 7:265.

[18] Anthropology 172; 7:265.

[19]   This is a marked shift from Kant’s view in his early lectures, where he claimed that both passions and affects are characteristic of humans qua animal.  See 25: 616-17.

[20] Hence Kant says in the Anthropology: “The ambition of a person may always be an inclination whose direction is sanctioned by reason; but the ambitious person desires, nevertheless, to be loved by others also; he needs pleasant relations with others, maintenance of his assets, and so forth.  But if he is, however, passionately ambitious, then he is blind to those other purposes that his inclinations also offer to him.  Consequently he ignores completely that he is hated by others or that he runs the risk of impoverishing himself through his extravagant expenses.  This is foolishness (making one’s partial purpose the whole of one’s purpose) which even in its formal principle smacks reason right in the face”  (7:266; 173-74).  On the one hand, one’s inclination is sanctioned by reason, but on the other hand, one simply ignores, rather than rejects, the disadvantageous or immoral effects of one’s passionate actions.

[21] This is one area in which Kant seems to change his views on passions during the course of his lectures.  In early lectures, he claims that “both affects and passions are agitations of the soul and not lasting conditions” (25:589).  Later, however, he insists that while “affect is as it were a storm, passion rather takes time” (25:1526).  And in the published Anthropology, Kant points out that “since the passions can be coupled with the calmest reflection, one can easily see that they must neither be rash like the motions, nor stormy and transitory; instead, they must take roots gradually and even be able to coexist with reason” (7:265; 172)

[22] Sometimes Kant may seem to suggest that one would wish to be rid of one’s passion, as when he explains that “sometimes people wish for affect, but never for passion.  For . . . passion rules and is burdensome for a human being” (25:1526).  Given what he says elsewhere about one who is infected by a passion, however, it is best to read this as a claim about those who are not yet passionate, and who thus would not want to become passionate, rather than about those who already are.

[23]   Anthropology, 172-73; 7:265-66.

[24]   Anthropology, 174; 7:267.

[25] Strictly speaking, Kant goes a little too far in these passages.  It is certainly true that one who is passionate cannot cure himself, so any principles of self-discipline that Kant might suggest would serve no purpose.  Moreover, one ought not force a passionate person to reject his passion, as that would be an interference in free choice, prohibited by the rights of even the most passionate person as long as they are not dangerous.  But one might be able to coax and bribe a passionate person into more rational reflection, which may break him out of his passion.  Kant’s example of social norms of conduct is a good example of how this might work.  One who is passionately attracted to another and seeks to marry her might be forced into a “respectable wedding” by the legitimate demands of the prospective spouse and the social norms of the time.  A condition of this wedding being performed, perhaps set down by whomever officiates, might be participation in pre-marital counseling sessions.  In the course of these the passionate lover will be unable to ignore, even if he can still disregard, the range of moral and prudential considerations that his passions previously obscured.  This is not a guaranteed cure, but it is a means by which one can, without violence or deceit, coax a passionate person out of passion.

This is related to another theme that Kant occasionally raises in the context of passion.   Because passions reflect a persisting irrational commitment to a particular inclination, they make one who is passionate easy to control.  Thus in his discussion of passions in the published Anthropology, Kant points out how various passions can be used to influence others: “ambition is a weakness of people, which allows them to be influenced through their opinions; lust of authority allows them to be influenced through their fear; and avarice allows them to be influenced through their own interest” (7:272; 179).  In his lectures on anthropology, he gives an even more interesting example.  There he says, rather strikingly, that

There are some passions, that are nice [gutartig] just because they are passions, and thus as passions have a higher level of worth than if they were mere inclinations, just as springing from reflection or principles and the act springs from passion acquires thereby greater worth.  Thus a wife does not like it when she is loved by her husband only out of duty from mature deliberation.  It is much more agreeable to her that he cares for her as her husband, and testifies to a genuine good will towards her.  But she still considers herself less happy, when the husband loves her out of reflection and not out of passion . . ..  (25:613)

At first, this seems to be a concession to proponents of the emotions who insist that some actions just ought to be done without reflection.  In particular, Kant seems here to agree with Williams’s (in)famous suggestion that one who rescues his wife because he recognizes that it is morally proper to favor her rescues her with “one thought too many.”  But Kant’s further explanation of why a wife values the passion of her husband shows not only why Williams’s argument is so dangerous, but why passions are such an important part of an anthropology that teaches how to use others to promote our interests.  Kant explains,

But she still considers herself less happy, when the husband loves her out of reflection and not out of passion, such that he cannot leave her.  The origin is this: passion is a means to rule others.  One who has passion can by means of it . . . be ruled, and thereby the wife, when her husband loves her out of passion, has power over him.  But if the husband loves merely out of inclination, so that he is not in love, then he is thereby harder for his wife to control . . .. (25:613)

Kant by no means suggests that one should love one’s wife out of mere reflection and cold deliberation.  A husband should love with an inclination that is endorsed on reflection.  But in some cases one’s wife might not want one to take that moment of reflection.  She might wish that one were hopelessly devoted to her, considering nothing higher than her happiness in one’s deliberation.  And Kant explains why she would want this.  It enables her to control her husband.  No rational reflection on his part interferes with her rule over him.  (Note the similarity of this explanation to Herman’s defense of Kant against Williams in ref.)

                This influence of a wife over her husband can, of course, be used for good.  Kant does not suggest that the wife is necessarily selfish in her desire to rule over her husband.   In fact, insofar as women play an important role in the refining of men, which ultimately promotes their development of good wills, this control may even serve a valuable purpose.  If a wife is able to control her passionate well, she might eventually wean him from his passion and help him become a truly good man.

[26]   Dowdell translates böse as “bad,” but given Kant’s clear use of it in the Religion and elsewhere to mean ‘evil,’ it seems appropriate to translate it as ‘evil’ here as well.  His claim in the next sentence that it “morally reprehensible” [moralisch verwerflich] provides further support for this change in Dowdell’s translation.

[27] Anthropology, 7:267; 174.