Socratic Wisdom and the Freedom of Self-Sufficiency

Like the Stoics and other Socratics, Kant identifies the ideal of virtue with the sage (der Weise) (MS 6: 383; CPR A569/B597).Footnote 1 This is unsurprising when we consider that Kant equates wisdom (Weisheit) not with intelligence or prudence, but rather with the ability to make morality one's end: the consciousness of duty as the incentive of one's action is the ‘principle of wisdom’ that makes a person a ‘practical philosopher’ (MS 6: 375n; cf. 405, 441). Hence Kant equates the ‘doctrine of wisdom’ with the proper content of moral philosophy itself (5: 163) and the requirements of duty with the ‘rules of wisdom’ (Toward Perpetual Peace 8: 370).Footnote 2 Kant might seem to depart from this Socratic ideal of wisdom by endorsing the Christian ideal of holiness. Indeed, Kant identifies the proper ethical ideal not with the Stoic sage, but rather with the sage of the Gospel, that is, Christ.Footnote 3 But Kant also tells us that the ideals of wisdom and holiness are ‘identical objectively and in their ground’ (KpV 5: 11n). The ideal of holiness differs from that of wisdom merely insofar as it ‘deprives the human being of confidence that he can be fully adequate to it, at least in this life’ (KpV 5: 127n; cf. Moralphilosophie Collins, 27: 251–2).

Kant's debt to the Socratic ideal of the sage is especially clear in his remark about wisdom in the Metaphysics of Morals:

Kant does not elaborate on this remark, but he clearly means to invoke some familiar ethical ideals associated with the sage of the Socratic tradition.Footnote 4 Indeed, all of the characteristics of the sage that Kant mentions here have ancient precedents.Footnote 5 In this tradition, what the ordinary person thinks is valuable in wealth, power, etc. is found, in reality, only in the life of the sage. Even bodily health and integrity are not truly valuable, while bodily harm and even death are not truly harmful; only the health of the soul, virtue, is valuable and an authentic benefit to its possessor. Similarly, being a freeman or even a tyrant is not truly valuable, and being a slave (in the legal sense) is not truly harmful. Indeed, what the ordinary person thinks is valuable about the tyrant's life is in fact illusory: his unruly desires make him dependent on other people and hence a kind of slave—in a spiritual rather than a legal sense. The sage, on the other hand, cannot be controlled or manipulated by others since he aims only at virtue: he does not value what others can take away from him.

This tradition suggests that the passage from the Metaphysics of Morals should be taken to be endorsing the following contrast. The person who finds satisfaction only in the fulfilment of his sensible desires is not in control of his own fate: he can try his best to ensure that his sensible desires are fulfilled, but their fulfilment depends, ultimately, on the course of nature or the whim of other human beings. But since virtue, by contrast, requires no fulfilment of such sensible desires, nothing can hinder the virtuous sage from satisfaction in the pursuit of his end. We can call this freedom, the Stoics’ eleutheria, the ‘moral freedom’ of the sage.Footnote 6

The Kantian account of such moral freedom is closely connected to Kant's more well-known (and well-developed) account of moral goodness: the only thing that is unconditionally good is a ‘good will’, which is ‘not good because of what it effects or accomplishes, because of its fitness to attain some proposed end, but only because of its volition; that is, it is good in itself’ (GMS 4: 394). This can be compared, for example, to Epictetus’s denial that any ‘externals’ (ta ektos) are good and his definition of the good as instead ‘a certain disposition of our choice’ (prohairesis poia).Footnote 7 For Kant, this means that virtue is an end ‘sufficient for itself independently of nature’, such that the human being must ‘separate from this [end] all those ends whose possibility depends on conditions which can be expected only from nature’ (KU 5: 431; cf. 434).Footnote 8 That is, the moral end is concerned merely with good willing itself rather than accomplishing some effect in the world—an effect that can always be thwarted by forces outside our control. In short, unlike natural goodness, moral goodness ‘is created by us, hence is in our power’ (Reflexion, 7202 (1780–9), 19: 281).

For the Stoics, it is precisely this feature of good willing—that it is in our power—that makes the sage free. Epictetus, for example, emphasizes the importance of distinguishing what is in our power (or ‘up to us’: eph’ hêmin) from what is not:

That is, a person is free and ‘his own master’ when he treats all those external things not in his power as indifferent, as not ‘his own’, and cares only about what is truly his own, that is, only about his own willing or choice: for ‘who has any authority over these, who can take them away? Nobody can, any more than he could hinder a god’.Footnote 10 The sage cares only about what is his own, about his virtue, and thereby lacks the passions connected with ‘externals’ that put the fool at odds with other people and make him vicious.

Moral Freedom and Happiness

For the Stoics, this ideal of freedom is also connected with that of happiness or well-being (eudaimonia). Everyone seeks happiness, and happiness requires that we lack nothing that we want.Footnote 11 The happiness that everyone seeks is therefore attainable only for those who care only about what is in their own power, only about their own good willing; unhappiness comes not from external things, but from our caring about them.Footnote 12

Kant makes an important break from this Stoic view when he denies that the satisfaction attained through virtue could itself count as happiness.Footnote 13 For Kant, happiness is a physical well-being that requires the satisfaction of sensible desires. The satisfaction accompanying virtue is, for Kant, thus merely ‘an analogue of happiness’ (KpV 5: 117); such satisfaction can offer us, at best, some consolation if we have sacrificed happiness for the sake of virtue. For Kant, even the virtuous person seeks physical well-being as something good. Indeed, one cannot have attained the highest human good unless one achieves both virtue and physical well-being. Kant's view is, of course, that the value of physical well-being must be subordinated to the value of a morally good disposition. But physical well-being must not be considered a mere ‘preferred indifferent’, as the Stoics supposed.Footnote 14

One consequence of Kant's rejection of this aspect of Stoic doctrine is that he cannot claim that the sage literally cannot be harmed. Hence when Kant says in the remark from the Metaphysics of Morals that the sage ‘is capable of suffering loss neither by chance nor by fate’, he must mean that the sage cannot lose what is most important to him, namely his virtue, though he may very well lose his happiness, which he also counts as something good (but cf. KpV 5: 60).

Rather than claiming with the Stoics that we can satisfy our aspiration for happiness by turning away from sensible desire and towards virtue, Kant takes the view that our ability to satisfy our aspiration for happiness is extremely limited. To that extent, Kant treats happiness in a way analogous to how the Stoics treat the external things. For Kant, happiness is an ideal of the imagination rather than reason since knowing what would make one happy would require omniscience (GMS 4: 418–19).Footnote 15 Indeed, Kant even suggests that the more we try to devise a plan for our happiness, the unhappier we become (e.g. GMS 4: 395–6).Footnote 16

Kant notes this difference between the ends of virtue and of fulfilling sensible desire quite early in his career:

Kant claims something he here later comes to reject: that the satisfaction arising from moral goodness is a secure source of physical goodness in this life, namely happiness.Footnote 17 But the pessimism about our ability to make plans for happiness persists in Kant's mature view and suggests a continued affinity with the Cynic's notion that the simplicity associated with a virtuous life offers the best chance for happiness.Footnote 18

Unlike the ancients, who saw happiness as something within our control to attain, Kant takes the Christian view that no human exertions can give us complete control over our own happiness; that control belongs ultimately to God, and our hopes for happiness therefore rest not in this life, but rather in the happiness rewarded for virtue in the next. The best we can do is aim to be worthy of happiness and hope happiness will follow (KpV 5: 129). Kant strikes a particularly gloomy note in his lectures: ‘Here on earth no human being can be happy. Maybe somewhere else’ (Moral Mrongovius II, 29: 604_.).Footnote 19 Kant is thus led to the very un-Stoic position that we cannot attribute happiness to the sage. Accordingly, ‘happy’ is conspicuously absent from the list of attributes of the Kantian sage given in the Metaphysics of Morals.

But despite this important difference with respect to the ideal of happiness, Kant agrees with the Stoics on the central point we are concerned with here: that only in virtue does rational choice operate solely within its own domain, that only in virtue is reason self-sufficient such that nothing can hinder it from attaining satisfaction. Kant captures the central point of the Socratic account of moral freedom when distinguishing, along these lines, the moral demands of virtue from the prudential demands of politics:

Kant is even willing to call the satisfaction arising from virtue ‘moral happiness’ as long as we strictly distinguish it from the ‘physical happiness’ attained by satisfying desires (Religion, 6: 67, 6: 75n; cf. MS 6: 387). Reason is not capable of achieving the satisfaction of happiness, but instead is ‘capable only of its own kind of satisfaction (Zufriedenheit), the satisfaction of fulfilling an end which in turn only reason determines’ (GMS 4: 396)—namely the end of willing according to the moral law. And this moral happiness, like the Stoics’ eudaimonia, is a happiness that does not depend on nature or other people. The human being cannot achieve the complete self-sufficiency imagined by the Stoics, but following moral maxims still results in an ‘intellectual contentment’, namely a ‘contentment with oneself, which in its strict meaning, always designates only a negative contentment with one's existence, in which one is conscious of needing nothing’ (KpV 5: 117–18; second emphasis added). Virtue produces a consciousness of independence from inclinations ‘and so too from the dissatisfaction that always accompanies them’, an independence that is even ‘analogous to the self-sufficiency (Selbstgenugsamkeit) that can be ascribed only to the Supreme Being’ (5: 118).Footnote 21 It is this feature of the moral end that justifies Kant's claim in the passage from the Metaphysics of Morals that only the virtuous person is, although not happy, ‘capable of suffering loss neither by chance nor fate, since he is in possession of himself, and the virtuous person cannot lose his virtue’, which in turn gives content to the description of the sage as ‘rich’, ‘a king’, and especially as ‘free’.

Evil Self-Sufficiency?

Perhaps it is necessary to forestall a certain objection here on Kant's (and, indirectly, the Stoics’) behalf. An objector might admit that the end of virtuous willing is always in my power to attain, whereas an end that involves fulfilling a certain sensible desire, for example, my end of attaining wealth, can never be completely within my power to attain. The objection is that using that distinction to arrive at a substantive result regarding self-sufficiency seems like mere verbal trickery. I can say I am in control of attaining the end of being virtuous only because that end is merely to will in a certain manner as opposed to actually willing to achieve any particular result. And no matter what my end is, I can always be said to will that end in a certain manner, even if it is, for example, selfishly. So why can I not say, similarly, that willing selfishly is my end and that in this case too it is completely within my power to attain this end? And why can I not say, therefore, that my selfishness in pursuing wealth makes me free? The thought motivating the objection is this: it seems to be a virtual tautology to point out that I am in control of the manner of my willing, and it therefore seems meaningless to say that I achieve some special kind of freedom through virtue, which is only one possible manner of willing.

Kant's defence against this objection will rest, of course, on the fact that the end of virtue is unique in this regard. While it is certainly true that virtuous willing always involves some specific ends, for example, increasing my neighbour's welfare, willing virtuously is not merely one manner among others of pursuing ends that aim at satisfying sensible desire. It is instead a manner of willing in which I pursue particular ends only because it is virtuous to do so and not at all because those ends serve to satisfy sensible desire. Thus, if increasing my neighbour's welfare is my end only because that end would satisfy some desire of mine, then it is impossible to will that end virtuously. Conversely, if my end is amassing wealth, then I may pursue this end selfishly, but I do not pursue this end simply because doing so is selfish: I am interested in the wealth and what it buys, not in selfishness. And in this sense, selfish willing cannot in any sense be called my end.

One might object further that I could, in fact, pursue wealth simply because I want to be selfish so that I achieve my end so long as I have willed selfishly and thus regardless of whether I actually attain that wealth. In that case, I would not be acting on any incentive rooted in sensibility, but rather on a disinterestedly or purely evil incentive, as we could say. But this rejoinder concedes the main point at issue: that we can coherently distinguish a self-sufficient agent from the more ordinary agent who can achieve his end only with the cooperation of nature and other people. Nevertheless, by allowing the possibility of an evil self-sufficiency, this rejoinder does threaten to undermine the claim that virtue is unique in bringing moral freedom. But this is less of a threat than it appears since Kant does not claim that self-sufficiency is what makes the virtuous disposition choiceworthy or good. Moreover, Kant has a direct response at his disposal: the kind of evil self-sufficiency imagined by this rejoinder is not any human kind of evil but is rather a diabolical (teuflisch) evil. It might be logically possible to attain self-sufficiency through evil, but that does not mean that there is any comprehensible motive for anyone to act out of a disinterested attraction to evil.

Kant does describe a kind of human approximation to diabolical evil: evil arising from a firm and self-consciously chosen principle, that is, from what Kant calls ‘character’. Such a character suggests a kind of ‘strength of soul’ analogous to that of virtue since the person of character is concerned with acting on principle rather than agreeable results and thus with his own willing rather than simply the expected effects of his willing (Anthropology, 7: 292–3). But Kant claims that even such a character would remain bound, even if indirectly, to sensible inclinations and thus to the hope for their satisfaction: to be truly principled in one's action would be to be interested solely in the action itself and not at all in the result or ‘object of the action’ (see GMS 4: 413n); and for Kant we take such an immediate interest in the action itself ‘only when the universal validity of the maxim of the action is a sufficient determining ground of the will’ (4: 460n). This means that for Kant an evil character is not even a character in the strict and proper sense: ‘character requires maxims that proceed from reason and morally practical principles’ (Anthropology, 7: 293).Footnote 22 The human being never chooses evil wholly on principle: he ‘never sanctions the evil in himself, and so there is actually no malice arising from principles, but only from the forsaking of them’ (ibid.). Strictly speaking, then, even ‘great crimes’ arise not from strength of soul but rather from ‘the force of inclinations that weaken reason’ (MS 6: 384). Although Kant criticizes the Stoics for lacking an account of positive evil, he nevertheless follows the Socratic view that the human being aims at evil sub specie boni and thus does not ‘incorporate evil qua evil into his maxim to serve as an incentive’ (Religion, 6: 37; cf. KpV 5: 59).

So the proper response to the objection that an evil will could also be self-sufficient, and in that sense free, is (1) that such a will would have to be not merely evil, but rather disinterestedly and hence diabolically evil and (2) that diabolical evil is not possible, at least for human beings. Kant seems to go even further than this second point when he claims that it is ‘absolutely impossible’ for a free being to renounce the authority of the moral law, where the apparent implication of this claim is that a diabolical evil is not just foreign to humanity but rather a self-contradictory concept (Religion, 6: 35). But Kant's response to the objection does not rely on that (obscure) conceptual point since all it needs is the weaker claim that human evil is never disinterested or wholly principled. To be sure, it would be difficult to establish even that weaker claim with certainty (let alone necessity) absent the kind of conceptual point just noted. But we can still attain the certainty appropriate to the case. In the first place, recall that Kant thinks that the possibility of principled good action cannot just be assumed—but must instead be grounded in a conception of a supersensible moral world—precisely because such action does not base its incentives on any inclinations or desires. But we lack any corresponding grounds for supposing the possibility of principled evil action. Second, we can show that the idea of a diabolical evil is dispensable by accounting for the various kinds of moral evil that we encounter around us and in ourselves in terms of people's indulgence of their sensible desires. Kant's account of the passions can be read as offering us a view of how a perverted subjective conception of the good can bring more than a mere lack of virtue, but rather this kind of positive evil. Once that promissory note is paid, we will be able to respond more adequately to the present line of criticism: immoral willing, even (relatively) principled immoral willing, always aims at an end whose attainment is always very uncertain, depending as it does on chance and the whim of other people; moral willing is unique in aiming at something completely within our control, and hence unique in bringing moral freedom.

Moral Freedom and Transcendental-Practical Freedom

It is important that this ‘moral freedom’ of virtue not be confused with the kind of freedom required for moral accountability. There is no need to see Kant, any more than the Stoics, as claiming that I somehow relinquish responsibility for my actions to outside forces when I pursue sensible desires. Kant can be taken to mean instead that I cannot attain the end that I set for myself through my own exertions alone if that end involves the fulfilment of sensible desires. Since the end of virtuous willing is the only end that can be attained without the cooperation of external forces, it is also the only end that is completely within my control to attain—which of course is not to say that virtue is therefore easy to attain. It is a separate question whether I am in control of willing the end that I do. For the Stoics and Leibniz we can be considered in control of willing the end that we do as long as that willing follows from our own nature or character. For Kant, this is a necessary but not a sufficient condition for moral responsibility. Kant insists that we must, in addition, be responsible for that character itself. Our freedom therefore must be not merely the practical freedom of acting on reasons rather than sensible impulses, but also the transcendental freedom of complete independence of all external causes.Footnote 23 On either view, merely having this control over willing an end cannot be equated with willing an end that makes one free in the sense of being self-sufficient.Footnote 24

This allows us to see that an account of moral freedom, even a Kantian account of moral freedom, can be developed without any commitments to Kant's distinctive metaphysics of freedom. In fact, we can find in Kant's early ethical notes and lectures thoughts on moral freedom that are continuous with his mature views on the topic.Footnote 25 It would be rash to assume that these early thoughts concerning moral freedom offer an alternative to the mature account of (what we can call) transcendental-practical freedom.Footnote 26 Indeed, Kant continues to endorse a roughly Stoic account of moral freedom in his mature work even though he labels Leibniz's Stoic-influenced account of free will as a ‘wretched subterfuge’. The mature Kant does, of course, hold that we can speak of virtue and vice only where the agent can be considered practically free and even transcendentally free. But this account of transcendental-practical freedom represents Kant's account of the conditions of moral responsibility. The account of moral freedom is independent of this set of problems since it is not concerned with our ability to adopt ends or, more generally, with our responsibility for our practical character and its effects; it is concerned rather with the different moral and pragmatic implications of aiming at different ends and, in particular, with our ability to attain our adopted ends. In fact, Kant's account of moral freedom had better be independent of his account of transcendental freedom: if it is not, then an immediate problem would arise about how we could be responsible for our immoral actions.Footnote 27

Both kinds of freedom concern a kind of control over our actions. But transcendental-practical freedom concerns the control we have over our own ultimate choices or practical character, whereas moral freedom concerns our control with respect to the attainment of our end and, by extension, the attitude of mind in choosing an end that makes one's satisfaction independent of nature and other people. The two kinds of freedom are related in the following way: I am in control of attaining my end and thus have moral freedom only when my end falls solely within the scope of my power of transcendental-practical freedom. That is the Kantian way of putting the Stoics’ point that moral freedom requires we care only about what is ‘up to us’.

Inner Freedom as the Moral Freedom Distinguishing Moral from Legal Conduct

Kant's discussion of the ‘inner freedom’ of virtue in the Metaphysics of Morals might seem to undermine this tidy distinction between the transcendental-practical freedom possessed by all human beings and the moral freedom attained with virtue. Kant defines ethical duties as ‘duties of inner freedom’ (6: 406) and explains: only a free aptitude (freie Fertigkeit, habitus libertatis) to act in accordance with the law counts as virtue since if the aptitude is a habit (Angewohnheit, assuetudo) ‘it is not one that proceeds from freedom and therefore not a moral aptitude’ (6: 407). Indeed, ‘if the practice of virtue were to become a habit (Gewohnheit), the subject would suffer loss to that freedom in adopting his maxims which distinguishes an action done from duty’ (6: 409). When Kant speaks here of the loss of a person's ‘freedom in adopting his maxims’ it seems most natural to take him to be referring to a loss of transcendental-practical freedom, that is, control over our own choices or practical character. But although this is clearly the most natural reading of the passage considered in isolation, it is also highly problematic: if only an action done from duty is the result of transcendental-practical freedom, then only an action done from duty will be something for which we are responsible; and the implication of this, of course, is that we would lack such freedom when we adopt an evil maxim and hence that we would not be responsible for moral evil. This is not only an unwelcome consequence; it also contradicts Kant's guiding principle that the human being's power of choice can be determined only by an incentive that is freely incorporated into his maxim (e.g. Religion, 6: 24; cf. MS 6: 320n).Footnote 28 Moreover, even the passage under consideration clearly implies that actions that do not arise from a free aptitude would still be ‘free actions’ (6: 407).

To understand Kant's intent in speaking of ‘inner freedom’ here, it is essential to note the context: a discussion of the distinction between the doctrine of right and the doctrine of virtue and hence the distinction between merely legally good and authentically morally good willing (in which morality is the sufficient incentive determining the will to action). This context—along with Kant's repeated allusions in the surrounding text to the Stoic doctrines of freedom, self-mastery, and apathy—suggest that the ‘inner freedom’ described here is nothing other than what we have called ‘moral freedom’. The concept of moral freedom serves to distinguish ethical duties from duties of right since however much one's conduct respects the outer freedom of others (and is thus dutiful by the standards of right), the principle of one's conduct will not be consistent with an inner, moral freedom unless one does one's duty from duty. Only then does one adopt a maxim without a view to results that make one's satisfaction dependent on external circumstances and especially on other people; for only then does one have an aptitude that ‘proceeds from freedom’ in the sense that the standard for correctness of the maxim lies wholly within (transcendental-practical) freedom itself, namely in the law of freedom (the moral law).

For Kant, this ideal of the freedom that distinguishes morally good from merely legally good willing is personified in its purest form by Jesus. On Kant's telling, the Jewish theocracy is a kind of metaphor for mere legality: it was a government instituted for the purpose of the veneration of the moral law, but its subjects wished ‘to be ruled through rewards and punishments in this life’, such that the laws were ‘in part indeed ethical but only inasmuch as they gave rise to external coercion’ (Religion, 6: 79). This people was ‘ripe for a revolution’ when Jesus begin teaching—not only because they suffered the ‘evils of a hierarchical constitution’, but also because their ‘slavish mind’ (Sklavensinn) became unsettled and more reflective due to the influence of ‘the Greek philosophers’ moral doctrines on freedom’ (6: 79–80; cf. 126ff.).Footnote 29 But Jesus displayed a wisdom ‘even purer than that of previous philosophers’—a wisdom he proved by refusing the devil's bargain: to wield power over the whole earth in exchange for an inner subservience to the devil (6: 80–1). Instead, he endured the most extreme suffering and death without relinquishing his inner commitment to the good principle. This death could be considered a defeat solely in physical terms since that very physical defeat exhibits a holy will that cannot be defeated even by the greatest imaginable rewards or punishments (6: 81).

If we remove its ‘mystical cover’, this story has a purely rational meaning not dependent on any historical event (nor on the metaphor of two persons, good and evil, outside the human being):

That is, good willing is possible only insofar as we act solely for the sake of morality without a concern for punishments and rewards. In that case, we remain subject to physical defeat, but not moral defeat. The slave is the one who is concerned only with rewards and punishments. Thus, Kant remarks that Christianity, properly understood, seeks to instil not a servile obedience to moral obligation, but rather to instil a moral obedience to the law that arises from ‘the liberal way of thinking, as distant from a slavish mind as it is from unruliness’ (The End of All Things, 8: 338).

In the passages from the Metaphysics of Morals discussing ‘inner freedom’, Kant is particularly concerned to counter the view that virtue could be defined as ‘a habit of morally good actions steadily acquired through practice’ (6: 383). That definition is deficient since it fails to account for the specific difference between right and ethics. According to the doctrine of right, ‘one can begin with the end and seek out the maxim of actions in conformity with duty’ (6: 382). Through habituation, one can acquire a maxim of actions in conformity with duty without any change at all in one's ultimate end (e.g. the maxim of honest dealing from prudence). Thus Kant remarks in the Religion that virtue in the legal sense (virtus phaenomenon) can be acquired gradually and through ‘long habituation’ without the ‘slightest change of heart’ (6: 47)—which would instead require a ‘revolution in the disposition of the human being (a transition to the maxim of holiness of disposition)’ (6: 48).

Moreover, Kant thinks that acting on a habit of this sort amounts to a forfeiture of the prerogative to act on the basis of one's own moral judgement and signals instead a willingness to be led by prevalent customs and the expectations of others. The lesson Kant draws from the account of a habitual virtus phaenomenon is thus that moral education requires a cultivation of the pupil's own moral judgement (6: 48). And in the Lectures on Pedagogy, Kant expands on this point, telling us that the kind of discipline which merely creates a habit of good behaviour is such that ‘the pupil will become a human being who cares only how he can get on well in the world and is good or bad depending on what he finds most conducive to that end’ (9: 480–1). But what is important for our purposes is how Kant characterizes the alternative: ‘one must see to it that the pupil acts from his own maxims, not that he does the good, but that he does it because it is good’ (9: 475). Only if the pupil learns to act from his own maxims can he avoid learning, for example, to be truthful because that is what other people expect of him. In the latter case, there is a sense in which truthfulness would not be his own end, not something he does on principle; truthfulness would rather be merely an effect of his willing that serves (more or less adequately) his own further prudential ends. (These ulterior ends are often hidden—even from the agent himself.) Such a person would be, at best, like the mere ‘imitator (in moral matters)’ who lacks a character of his own and thus lacks ‘originality in the way of thinking’ (Anthropology, 7: 293). For an agent to have a character indicative of authentic virtue, by contrast, ‘he must have his own will (ein eigener Wille), i.e., he must not let himself be led by others’ (Anthropologie Reichel, 25: 1356); that is, he must think for himself rather than conforming to the expectations of others.Footnote 31 Virtue requires enlightenment over our self-incurred ‘minority’ (What is Enlightenment? 8: 35).Footnote 32 In this way, Kant's account of moral education aims to fulfil the ideal he finds in Rousseau's Émile: ‘that education be free and also make a free human being’ (Remarks on the ‘Observations’, 20: 167). Only actions chosen in this spirit of freedom could be morally as opposed to merely legally good.

All of this gives us a way to understand Kant's puzzling claim that ‘if the practice of virtue were to become a habit, the subject would suffer loss to that freedom in adopting maxims which distinguishes an action done from duty’ (6: 409). Read alone, this claim seems to imply that only the adoption of a moral maxim exemplifies the control over our own willing or choosing that is required for moral responsibility, i.e. transcendental-practical freedom. But the context in which this claim appears allows us to see that Kant is warning against the attitude of mind at the basis of a mere virtus phaenomenon: the agent does not choose his maxim with an attitude of moral freedom marked by an interest in the action itself; the agent instead chooses the maxim either from simple conformism or because he hopes such a conformity will serve some further purpose. In either case, he places his aspirations in something beyond his control, making himself dependent on things and especially other people. And this reading seems to be confirmed in Kant's remark on inner freedom: after describing the moral aptitude as ‘free’, he immediately proceeds to identify inner freedom with having a noble (erecta) rather than a base or slavish character (indoles abiecta, serva) (6: 407). In these terms, the person who aims at merely lawful behaviour chooses his maxim in a servile spirit (with an indoles servilis)—always seeking to gain rewards or to evade punishments—rather than in a noble spirit of a free man (indoles erecta, ingenua).Footnote 33 The person of character is concerned with his action itself (i.e. with the maxim of his action) and not with rewards and punishments; hence he acts as a free person who cannot be hindered or coerced by others. Thus: ‘The less a human being can be coerced by natural means and the more he can be coerced morally (through the representation of duty), so much more free he is’ (6: 382n; cf. 381).Footnote 34 Kant invites a further comparison here between the inner freedom of virtue and the moral freedom of the Socratic sage by linking the servile character with passions and affects and the noble character with a healthy condition of the person who has overcome these, thereby achieving a ‘tranquil mind’ (6: 407–9) (the ataraxia of the ancients). And it is in Kant's account of the passions that a more definite account of moral slavery emerges, one that encompasses not just a lack of virtue, but positive evil.

Affects, Passions, and Reason

For the Stoics, the contrast to the sage who is free because he aims at nothing other than his own virtuous disposition is the ordinary foolish person whose passions (pathê/perturbationes) subject him to forces outside of his control. These passions not only make the ordinary person unhappy and psychically unhealthy, they also make him unjust. In light of Kant's endorsement of the ideal of the sage, it is not surprising that he also endorses the Stoic principle of apathy, ‘namely that the sage must never be in a state of affect (im Affect), not even of sympathy (Mitleid) for his best friend's misfortune’ (Anthropology, 7: 253; cf. GMS 4: 398).Footnote 35 But for Kant, this Stoic principle is inadequate insofar as it underestimates what must be overcome to attain this practical ideal: the Stoics ‘send forth wisdom against folly, which lets itself be deceived by inclinations merely because of carelessness, instead of summoning it against the malice (of the human heart)’ (Religion, 6: 57). Mere ‘folly’ would presumably bring a mere lack of virtue—akin to what Kant calls the mere ‘weakness’ or ‘frailty’ of human nature in following its good maxims—rather than evil proper, which indicates a positive corruption or depravity (6: 29–30, MS 6: 407–8).Footnote 36

Although Kant highlights the insufficiency of the principle of lack of affect, he nevertheless marks out the whole topic of ethical duties in a way reminiscent of the Stoic ideal of lack of passions: the inner freedom of virtue presupposes not just lack of affect, but also ‘dominion over oneself’ (Herrschaft über sich selbst, über sich selbst Herr sein) which Kant equates with ‘dominating one's passions’ (Leidenschaften beherrschen) (MS 6: 407–8). And although we can infer from his criticism of the Stoics that Kant thinks that passions, when properly understood, indicate a kind of perversity or corruption that the Stoics (as heathens) failed to notice, Kant's own account of the passions remains faithful to the Stoic doctrine in the following important ways: a passion (1) is not a merely natural error, but rather arises from our reason and hence (2) is something for which we are responsible and that (3) is associated with vice and (4) moral slavery.

Kant describes passions and affects in terms of a kind of corruption of the deliberative process. An affect is ‘the feeling of pleasure or displeasure in the subject's present state that does not let him rise to reflection (the representation by means of reason as to whether he should give himself up to it or refuse it’ (Anthropology, 7: 251); it ‘makes itself incapable of pursuing its own end’ and even ‘makes us (more or less) blind’ (7: 253). A passion, for its part, is the ‘folly’ of ‘making part of one's end the whole’ (7: 266). Moreover, a passion is an inclination ‘that excludes mastery over itself’ (Religion, 6: 28n) and ‘that can be conquered only with difficulty or not at all by the subject's reason’ (7: 251); indeed, passions are ‘for the most part incurable because the sick person does not want to be cured and flees from the dominion of principles, by which alone a cure could occur’ (7: 266; cf. KU 5: 272n). All of this might seem to suggest that affects and passions bring a forfeiture of the control over choices characteristic of transcendental-practical freedom.

We might say in this regard that the person in the grip of an affect or passion is like the drunken person: although he cannot deliberate properly while drunk, he did choose to drink knowing the potential for bad behaviour that could result, and he is thus responsible for what he does while drunk (Metaphysik der Sitten Vigilantius, 27: 559). On such an account, we can say that even the drunk person retains his transcendental-practical freedom: he remains the free cause of all that results from his drunkenness even while he is drunk; he is the ‘author’ and ‘efficient first cause’ of the action; ‘i.e., the determining ground of the action can be sought nowhere else in nature’ (27: 558–9).Footnote 37 This is not surprising once we recall that, for Kant, the kind of control characteristic of transcendental-practical freedom is not in any case an ability to disrupt the temporal order of events, including the psychological order of deliberation. In particular, transcendental-practical freedom is not the control of our deliberation from one point in time to the next. To that extent, Kant is in agreement with the compatibilists—including, significantly, Stoic compatibilists. For this reason, it would be a mistake to think that the deliberative disruption engendered by the passions and affects could be a closing off of future free choices: that future choices are closed off by the past is a general feature of the temporal order, not something pathological.Footnote 38 Hence when Kant says that it is in our power right now to do as duty requires despite all sensible inclinations to the contrary (e.g. MS 6: 380), he does not mean that the present is indeterminate with respect to the future. He means rather that our practical character is itself something for which we are responsible since it is something freely chosen; it is our intelligible character, which, in turn, is the ground of the temporal order of our actions (CPR A541–46/B569–74; Reflexion, 5612, 18: 253). And part of that character is the influence we allow the affects to have. But the basic point can be made without appeal to Kant's account of timeless agency: the affects prevent reflection on the part of the agent only because the agent's own character allows them to have this effect. Hence Kant says regarding natural inclinations as obstacles to virtue: ‘it is the human being himself who puts these obstacles in the way of his maxims’ (MS 6: 394; cf. GMS 4: 457–8).

However, this model of responsibility is more appropriate for our failure to subdue our affects than for governing our passions. Kant remarks about the passions: ‘It is also easy to see that they do the greatest damage to freedom, and if affect is drunkenness, then passion is an illness that abhors all medicine’ (7: 265–6). At times, Kant seems to suggest that the most important difference between passions and affects in this regard is that passions are more permanent: an affect is a feeling that can quickly subside, whereas a passion is ‘a sensible desire that has become lasting inclination’. And it seems to be this permanence that is morally problematic: a passion is calm and hence ‘allows the mind to form principles upon it and so … to incorporate what is evil (as something intentional) into one's maxim. And then the evil is properly evil, that is, a true vice’ (MS 6: 408).

This longevity or incurability of the passions signals a state of mind that is relatively immune to revision in light of favourable or unfavourable results. In that regard, the passions mimic the firm principles of character. And this is closely connected to a more fundamental feature of passions that distinguishes them from affects: the passions are rooted in practical principles that make the evil ‘intentional’.Footnote 39 Affects are either a substitute for a properly resolute maxim of virtue or else they disrupt the implementation of otherwise good maxims. (In the latter case, this disruption is still something for which we are responsible, something that can be traced to a higher maxim.) A passion, on the other hand, is not something that we merely allow to interfere with the operation of reason: far from being a weakness with respect to the inclinations, a passion makes one ‘blind’ to other sensible incentives whose satisfaction would be needed for happiness (7: 267). A passion instead results from a corruption of reason itself. Kant explains, for example, that physical love is not a passion since

‘Accordingly, the outbreak of a passion can be imputed (zugerechnet) to the human being’ (7: 269). Some passions have a more direct connection with maxims than others. Thus, although Kant says that all passions presuppose a maxim, he also tells us that the acquired passions of culture, unlike the innate passions of nature, ‘are not connected with the impetuosity of an affect but with the persistence of a maxim established for certain ends’ (7: 268).

The kind of deliberative disruption that Kant seems to have in mind here (‘making part of one's end the whole’) recalls the Stoics’ own account of the cognitive error embodied in the passions. For the Stoics, it is an unhealthy passion to consider something like money to be good (rather than merely a ‘preferred indifferent’). To consider such an end good would be to consider it choiceworthy in all circumstances, and such a passion therefore stands opposed to virtue, which alone is choiceworthy in all circumstances. In terms closer to Kant's, we can say that, although having an end of attaining money contributes to the end of happiness and can even be consistent with the end of morality, to make the possession of money the whole of one's end is to exclude the end of morality (not to mention happiness). Hence Kant remarks that ‘making part of one's end the whole’ is something that ‘directly contradicts the formal principle of reason itself’ (7: 266). According to the Stoics, we are responsible for what we do as a result of our passions since our passions are an expression of our own practical character.Footnote 40 And Kant, by associating passions with maxims, can be seen to be presenting an analogous view. Regarding this ‘folly’, Kant remarks:

The Moral Slavery of the Passions

The connection of passions with maxims explains why we are responsible for our passions, but it does not yet explain why passion should be associated with finding ‘pleasure and satisfaction in a slavish mind’. Affects are also incompatible with wisdom and the self-mastery required for inner freedom. The agent under affect lacks ‘wisdom’ and is instead a ‘fool’ (Menschenkunde, 25: 1121). Such an agent cannot be said to be in control of his fate since he is not even able to pursue the end that he himself thinks would give him satisfaction. And Kant suggests in lectures that this loss of control allows other people to assume control:

With a passion, by contrast, this relinquishing of power to the other is something intentional. The passionate person does not place his aspirations in an end that his reason can attain all by itself; instead, he places his highest aspirations under the control of other people and thereby becomes their slave. And it is in this pathological relation to others that true evil can be found.

Kant makes the surprising claim in this connection that passions are not desires for any particular things; they are rather desires ‘concerned merely with the possession of means’ (7: 270). Moreover: ‘Properly speaking, passions are always directed only by human beings to human beings, not to things’, such that the passions ‘can be satisfied only by them’ (7: 268, 270). Kant explains in lectures: ‘The cause of this is that the human being is the principal means to the satisfaction of all inclinations’; ‘our inclinations are directed to nature, but by means of human beings’ (Menschenkunde, 25: 1141–2). Kant mentions the inclinations of (outer) freedom and sex along with the desire for vengeance as ‘passions of natural (innate) inclination’.Footnote 41 Kant does not treat this class of passions in a systematic way, but more relevant to the origin of evil are the passions based on ‘acquired’ inclinations, that is, inclinations that ‘proceed from human culture (Cultur)’. The passions arising from this inclination are ambition, lust for power, and greed (Ehrsucht, Herrschsucht, and Habsucht):Footnote 42

The passions based on the desire for honour, power, and money concern ‘the inclination to the ability to have a general influence over other people’, an inclination that ‘comes closest to technically practical reason, that is, to the maxim of prudence (Klugheitsmaxime)’ (7: 271; cf. Anthropologie Petersburg, 25: 1141–2). These passions are thus closely connected with prudence, and in particular with what Kant calls ‘worldly prudence’ (Weltklugheit): ‘skill in influencing others so as to use them for our purposes’ (GMS 4: 416n). Yet when this inclination is elevated into a passion, our happiness is actually thwarted. The problem is rooted in the fact that the passions of acquired inclinations thus ‘consist in valuing the mere opinions of others regarding the worth of things as equal to their real worth’ (7: 270). To take an important example, it is natural and prudent that we should want to be held in esteem by others. But when we become ambitious, we take our whole happiness to consist in our social status; we take our whole worth to consist in the opinions others have of us. And insofar as we place our aspiration for happiness in a comparison with others, we simply give ourselves new reason to be unhappy.

Despite this, it might seem surprising that the passions would make one servile. After all, the root of the passions is our desire for just the opposite: to be free from the domination of others and to dominate them instead. But precisely because the passions are aimed, in this way, exclusively towards other people, the passionate person depends entirely and directly on other people for the satisfaction he seeks. Consequently, whereas the passions may seem to bring a kind of strength to their possessors, they are each in actuality a ‘weakness’ by means of which the passionate person can be exploited by others:

For example, Kant says of ambition: ‘One need only flatter this pride, and by means of the fool's passion one has him in one's power’ (7: 272). Although the desire for honour is rooted in the desire to control other people, when this desire becomes the passion of ambition, the ambitious person actually becomes subservient to others since his satisfaction depends entirely on their approval. And in that case, the ambitious person is just as subservient to the flatterer as the flatterer is to him. And this pathological social relation is not restricted to my relation to other human beings. If I view the authority of moral demands as resting on divine threats or promises, then I implicitly treat God as a person who is possessed by passions: I would represent God as possessed by ‘desire for honor and power (Ehr- und Herrschbegierde) combined with dreadful representations of power and vengefulness’ (GMS 4: 443).

The connection of these passions with a positively vicious disposition is fairly straightforward. In the Religion, Kant makes this connection explicit: it is the passions—envy, lust for power, and greed, in particular—that corrupt the human being's originally good predisposition (6: 93).Footnote 43 If I am possessed by one of these passions, then I seek status, power, and money not simply for their own sakes, but rather as a force by means of which I can control other people: I value these things ultimately only because other people value them and because I believe possessing them will therefore allow me to direct their free choice to ends that serve my own. Therefore, if I am possessed by a passion, I implicitly deny the dignity of the free, rational beings I seek to control, treating them not as ends in themselves (as ‘persons’), but rather using their free choice as a mere means for promoting my own happiness (as ‘things’).

The idea that the passionate person—the person who values wealth, honour, etc., as good without qualification—is a slave because he subjects himself to other people by means of his passions is already part of classical Stoic doctrine.Footnote 44 But the Stoics see the slavery to others engendered by the passions as merely an extension of our dependence on things.Footnote 45 In taking the passions to be essentially social dependencies, Kant shows his debt to Rousseau.Footnote 46 Like Kant after him, Rousseau sees the slavishness of the passions, especially ambition, as intimately connected with evil and the corruption of society. What Kant takes from Rousseau is the view that the real source of unhappiness and vice is not our natural needs (which we have already in the state of nature), but rather the pathological social dependence engendered by the passions.

Rousseau opens the first chapter of the first book of his Social Contract with the famous sentence: ‘Man is born free, and he is everywhere in chains.’ Less famous is the next sentence: ‘One believes himself the others’ master, and yet is more a slave than they.’Footnote 47 While Rousseau is certainly concerned with the problem of what Kant calls ‘outer’, political freedom, he is just as often concerned with the inner freedom attained by extirpating the slavish passions (especially ambition) and pursuing virtue. And like Kant, he sees the slavishness of the passions as intimately connected with evil and the corruption of society. The following passage is characteristic of Rousseau's view, and its anticipation of Kant's account is unmistakable:

Only Rousseau's emphasis on the role of property in the origin of these pathological social relations distinguishes this account of moral slavery from Kant's. Moreover, Rousseau, like Kant, connects the kind of inner slavery he describes in this passage with the passions and suggests that extirpation of the passions in the manner of a Stoic sage is a means of becoming a virtuous and a free man. It is true that Rousseau emphasizes, more than Kant, the ability of external laws to moderate these passions (by forcing citizens to act for the common good rather than for their own selfish good) and thereby to foster such a ‘moral freedom’ in the citizens subject to the law.Footnote 48 But Kant's account of history also holds out hopes that moral improvement could be encouraged by a just political order.

Rousseau is happy to speak about the ‘perfectibility’ of the human being without entering into the metaphysical debates about the nature of free will: Rousseau is a psychologist and moral theorist, not a metaphysician. He therefore leaves aside metaphysical questions about ‘free will’ when speaking about the slavery and freedom of man. This is already clear from the wholly social character of the dependencies engendered by the passions. For Rousseau, it is precisely these social dependencies—and not some property of the will that could be specified in abstraction from these relations—that lead to our unhappiness and moral corruption.

Unlike Rousseau, Kant places the metaphysics of free will at the centre of his entire philosophical project (e.g. KpV 5: 3–4). But his Rousseauian account of the social dependencies engendered by the passions highlights the fact that Kant, like Rousseau before him, does not see the slavery engendered by the passions as having anything to do with the metaphysics of free will: the passions do not imply a loss, or even diminishment, of our transcendental-practical freedom.Footnote 49 Their slavery is moral and thus neither one of physical captivity, nor metaphysical determination. The evil will does not lose its transcendental-practical freedom, but rather misuses it by slavishly offering itself to other wills to be used as mere tool. The virtuous will, by contrast, attains its own kind of freedom—not because it alone is fully free from necessity, but rather because it rejects a mercenary or slavish attitude towards its duties and instead fulfils its duties in a spirit of freedom that considers virtue to be its own reward. Only such a will has the self-sufficiency to attain satisfaction through its own activity alone.