The Rationality of Love: Benevolence and Complacence in Kant and Hutcheson

1. Kant’s General Division of Love

Towards the end of Part One of Religion Within the Boundaries of Mere Reason, Kant discusses whether it is possible either for an evil person to become good or a good person to become evil. In this context, Kant argues that self-love, even if it is adopted as the principle of all our maxims, is not necessarily the source of all evil in human beings and then proceeds in a footnote to disambiguate the concept of self-love:

Words that can admit of two entirely different meanings often long delay conviction on even the clearest grounds. Like love in general, self-love too can be divided into that of benevolence and that of complacence (benevolentiae et complacentiae), and both (as is self-evident) must be rational. (6:45n, translation modified)

Although this is the only place in all of Kant’s published writings where he explicitly divides the concept of love in general into either benevolence or complacence, the distinction is implicitly working in the background of nearly all his discussions of love. This has been painstakingly illustrated by Pärttyli Rinne in his recent book Kant on Love (2018). To mention just a few representative examples of the ubiquity of the distinction: Kant employs it in the context of self-love (see, e.g., 5:73, 6:45n, and Rinne 2018: ch. 1), love of neighbour (27:417 and Rinne 2018: ch. 4), love of God (6:145, 6:182, and Rinne 2018: ch. 3), as well as love of friendship (27:680 and Rinne 2018: ch. 5). To be sure, Kant uses the term ‘love’ in a variety of different ways and in numerous distinct contexts,⁷ and it is true that the distinction between benevolence and complacence does not arise in a select few of these contexts.⁸ Nonetheless, Rinne is correct that “Kant consistently uses or implies” (2018: 6) the distinction throughout his writings and, as a result, this “general division” is “a key for understanding love in Kant” (2018: 2 and 168–69).

The distinction between benevolence and complacence is not Kant’s own, however. Most notably, the distinction plays an essential role in the philosophy of one of Kant’s most important British predecessors, Francis Hutcheson.⁹ Hutcheson introduces the distinction between benevolence and complacence in his An Inquiry into the Original of Our Ideas of Beauty and Virtue as follows:

The Affections which are of most Importance in Morals, are Love and Hatred: All the rest seem but different Modifications of these two original Affections. [. . .] Love toward rational Agents, is subdivided into Love of Complacence or Esteem, and Love of Benevolence: And Hatred is subdivided into Hatred of Displicence or Contempt, and Hatred of Malice. (Hutcheson 2008: 102)

Kant owned the first German translations of both Hutcheson’s Inquiry and Essay (see Warda 1922: 50) and, as has been illustrated by others in detail, these works were profoundly influential on the development of Kant’s moral philosophy (see esp. Henrich 2009; Schmucker 1961; Schwaiger 1999; and Walschots 2015). Somewhat surprisingly, however, their very similar distinctions between love of complacence and benevolence have not yet been explored in detail.¹⁰ Compare Hutcheson’s distinction in the Inquiry quoted above with the following passage from the Kaehler lecture notes from Kant’s course on moral philosophy from the 1770s:¹¹

All love is either love of benevolence or of complacence. Love of benevolence consists in the wish and inclination to promote the happiness of others. Love of complacence is the capacity that we have to show approval at the perfections of others. (Kant 2004: 285)

While not identical to Hutcheson’s distinction, the similarity is striking and the reasons for thinking that Kant was influenced by Hutcheson in this regard are strengthened if one compares the first German translation of Hutcheson’s Inquiry with Kant’s original text, where it is clear that not only are their respective distinctions conceptually similar, but Kant appears to have adopted Hutcheson’s terminology as well, namely Liebe aus Wohlgefallen for love of complacence and Liebe aus Wohlwollen for love of benevolence (cf., e.g., Hutcheson 1762: 142 and Kant 2004: 285).

In the next section I outline Hutcheson’s understanding of benevolence and complacence in more detail to then utilize his conception of complacence in particular as a guide for better grasping Kant’s own understanding of the concept, which I lay out in Section 3. As we will see, both Hutcheson and Kant conceive of love as rational in an important way.

2. Hutcheson on Benevolence and Complacence

In the introduction to Treatise II of the Inquiry, Hutcheson claims that his intention is to argue two things: 1. that we perceive an immediate goodness and experience a particular kind of pleasure when we contemplate certain actions “by a superior Sense, which I call a Moral one,” and 2. that what excites us to virtuous action is neither the desire for pleasure, nor the influence of reward and punishment, “but an entirely different Principle of Action from Interest or Self-Love,” namely benevolence (Hutcheson 2008: 88). Important for my purposes in this paper is that Hutcheson casts these two positions as directly corresponding to two kinds of love, namely ‘love of complacence’ and ‘love of benevolence’ respectively. Love of complacence¹² is a “Good-liking” (2008: 102) that is “entirely excited by some moral Qualitys, Good or Evil, apprehended to be in the Objects; which Qualitys the very Frame of our Nature determines us to love or hate, to approve or disapprove, according to the moral Sense” (2008: 103). Love of benevolence, on the other hand, is our disinterested, that is, not self-interested, desire for the happiness of others and it is one of human nature’s primary motives for action. Let us consider these two kinds of love in further detail.

Beginning with love of benevolence, Hutcheson argues that human nature has two “ultimate desires” (see, e.g., 2008: 228), both of which are a kind of love: self-love and love of benevolence. These are ‘ultimate’ desires because their objects, namely one’s own advantage (self-love) and the happiness of others (benevolence), are desired for their own sake and in themselves or “with no farther view” (Hutcheson 2002: 140). As Hutcheson argues in detail in the Inquiry (see Hutcheson 2008: 112f.), benevolence is the motive of virtue: only actions flowing from purely disinterested benevolence are virtuous, which means that both self-interested actions and any seemingly benevolent actions that are in fact ‘interested’ are not worthy of moral approval and thus are morally permissible at best.

The moral approbation of benevolent actions is where love of complacence comes into play. As mentioned, love of complacence corresponds to Hutcheson’s stance on the ultimate origin of our ideas of moral good and evil, which for him is the moral sense. Senses, for Hutcheson, are the faculties responsible for the production of sensations, which are perceptions that are raised in the mind involuntarily such that “the Mind in such Cases is passive, and has not Power directly to prevent the Perception or Idea, or to vary it at its Reception, as long as we continue our Bodys in a state fit to be acted upon by the external Object” (2008: 19). Sensations themselves are not only passively received perceptions, rather they are often, or perhaps always, accompanied by feelings of pleasure and displeasure as well; indeed, Hutcheson says “[t]here is scarcely any Object which our Minds are employ’d about, which is not thus constituted the necessary occasion of some Pleasure or Pain” (2008: 8).¹³ Hutcheson argues that human beings possess more than just five external senses on the following grounds: “When two Perceptions are entirely different from each other, or agree in nothing but the general Idea of Sensation, we call the Powers of receiving those different Perceptions, different Senses” (2008: 19). One such sense is the moral sense, which Hutcheson argues we must posit to explain the fact that we perceive moral good as something distinct from natural good. A natural good is what is advantageous, that is, what is in our own advantage or self-interest, examples being a fruitful field (2008: 89), houses, lands, gardens, strength, and wealth (see 2008: 85), and the attitude we take towards them is the desire to possess them. Moral good, on the other hand, is that which is good independently of what is in our personal interest, examples being kindness, friendship, generosity, and benevolence (see 2008: 90), and the attitude we take towards such goods is approbation, love, and admiration (see 2008: 89). Hutcheson argues that we would not be able to distinguish between these two distinct kinds of goods, and thus not be able to distinguish between the goodness of a fruitful field and the goodness of a benevolent friend, for example, if we did not possess a specific sense that makes it possible to have these two different ideas of goodness. Not only this, but, as Hutcheson says in an early edition of the Inquiry, if natural and moral good were the same, then “[w]e should have the same Sentiments and Affections toward inanimate Beings, which we have toward rational Agents, which yet every one knows to be false” (Hutcheson 1738: 111). Since this is not the case, there must be a sense which makes this possible, and this is the moral sense.

The moral sense is therefore what makes it possible for certain objects to raise the ideas of moral goodness and evil in our minds. These objects, whether internal or external to the mind, are benevolent actions, affections, or characters.¹⁴ Sensing such things also involves experiencing a particular kind of pleasure and Hutcheson argues that, as a result of this experience, we judge such objects to be morally good.¹⁵ Not only this, but because the moral sense is a sense, that is, a passive faculty, it necessarily approves or disapproves of them when it is presented with these particular objects. Our approval or disapproval is therefore immediate when presented with the appropriate object (see, e.g., Hutcheson 2008: 88). The most important part to stress about Hutcheson’s doctrine of the moral sense for the purposes of this paper is that the approval carried out by the moral sense, namely the ‘liking’ we have for certain objects and the pleasure associated with our perception of them, is one and the same with what Hutcheson calls love of complacence. Love of complacence is therefore the ‘good liking’ or approval we have of benevolent actions, affections, or characters and the pleasure associated with it.

Hutcheson casts the distinction between benevolence and complacence as a distinction between a perception and a desire: love of complacence is a sense perception accompanied by a feeling of pleasure and is a kind of approval, whereas love of benevolence is a desire, namely the disinterested desire to bring about the good of another (see Hutcheson 1738: 135 and 2008: 223). To be noted is that both kinds of love can only take place between rational agents. In the case of complacence, only rational agents possess the moral sense and thus are capable of making judgments of moral approval (see, e.g., Hutcheson 2008: 103 and 197–98),¹⁶ judgements which, of course, can only be made about the actions, affections, and characters of other rational agents. In the case of benevolence, Hutcheson is clear that morally good or evil action cannot be a mere external motion of the body but must “flow from some Affection toward rational Agents,” namely “toward God or Man” (2008: 101). Benevolence, as an affection, is thus intentional (see 2008: 102), and both has rational agents as its object and can only be performed by rational agents as subject, namely beings who can intentionally and disinterestedly promote the good of others (2008: 103). This means that sexual desire, for example, does not qualify as love because it “is only Desire of pleasure, and is never counted a virtue” (2008: 102), that is, it is not an intentional action. One could therefore say that love is rational, for Hutcheson, for two reasons: first, because both benevolence and complacence have rational beings as their object and, second, they can only be experienced by rational agents, that is, agents who possess certain capacities, namely the moral sense and the capacity for intentional action.

I argue in the next section (§3) that Kant’s understanding of love of complacence shares important features with Hutcheson’s version of the concept. Clarifying this will allow me to argue in the subsequent section (§4) that both benevolence and complacence are rational, for Kant, in a sense similar to Hutcheson, namely in the sense that they can only be experienced by rational agents who possess certain higher cognitive capacities.

3. Kant on Love of Complacence

Kant’s clearest definition of love of complacence in his published writings is in the footnote from the Religion quoted at the beginning of Section 1. After making the distinction between benevolence and complacence and then briefly discussing love of benevolence, to which I turn in the next section, Kant focuses on love of complacence and distinguishes between two different kinds:

A rational love of complacence in oneself [Eine vernünftige Liebe des Wohlgefallens an sich selbst] can either be understood in the sense that we feel complacence [wohlgefallen] with those maxims, already mentioned, that are aimed at the satisfaction of natural inclination (so far as this end can be attained by complying with them); and then it is identical to love of benevolence towards oneself; one is pleased with oneself [gefällt sich selbt] just as a businessman who has done well in his business speculations rejoices over his good discernment because of the maxims he adopted in them. But the maxim of self-love, of unconditional complacence in oneself (independent of gain or loss resulting from action), would be the inner principle of a contentment only possible for us on condition that our maxims are subordinated to the moral law. No human being, to whom morality is not indifferent, can have complacence in oneself, or even be without a bitter displicence [Mißfallen] in oneself, if they are conscious of such maxims that do not conform to the moral law within themselves. (6:45–46n, translation modified)

There is much to talk about here, but the first thing to notice about Kant’s discussion of love of complacence in this footnote is that he uses the term in a sense broadly similar to Hutcheson, namely, to denote a kind of satisfaction or pleasure in an object. In the limited secondary literature that mentions love of complacence, scholars rightly highlight that this kind of love involves feeling pleasure in relation to the representation of an object. However, many scholars go on to construe complacence as a strictly pathological phenomenon, in line with the broader tendency in the literature to interpret Kant’s understanding of love more generally as solely a matter of feeling. Thus, Christopher Arroyo claims that “the love that delights [love of complacence] is a sensible, pathological emotion” (2016: 586). Similarly, Eleni Filippaki says that complacence is “the sensory pleasure or ‘delight’ we take in the perfections of others” (2012: 32–33), and Ina Goy claims it is “an empirical reference of love to the object” (2013: 193).¹⁷ Kant’s understanding of love of complacence is much more complicated than this, however, and in this section I offer a comprehensive interpretation of the concept that is informed by both his published and unpublished writings. I suggest that, similar to Hutcheson, Kant’s conception of love of complacence is best understood as a kind of approval in relation to two sorts of objects, and that the pleasure involved is of two distinct kinds.

One of the most important aspects of complacence that Kant clarifies in the above passage is that we experience the pleasure of complacence in relation to two distinct objects: 1) maxims that aim at satisfying natural inclination, and 2) maxims that conform to the moral law. In the first case and sticking to the example of complacence for oneself for the time being, love of complacence is being pleased with the extent to which our maxims promote our own happiness. In the second case, love of complacence is being pleased with the extent to which our own maxims are moral. Kant further clarifies these two kinds of complacence in the context of love of others in the Kaehler notes, where he distinguishes between benevolence and complacence in almost the exact same way as in the Religion footnote mentioned above:

All love is either love of benevolence or of complacence. Love of benevolence consists in the wish and inclination to promote the happiness of others. Love of complacence is the capacity [Vermögen] that we have to show approval [Beyfall zu beweisen] to the perfections of others. This complacence can be sensible or intellectual. [. . .] The love of sensible complacence is a pleasure [Gefallen] in sensible intuition, from sensible inclination [. . .]. The love of intellectual complacence is more difficult to comprehend. Intellectual complacence is not difficult to imagine, but the love of intellectual complacence is. Which intellectual complacence does inclination bring about? The good dispositions of kind-heartedness. (Kant 2004: 285)

This passage is important because it provides a number of details that are absent from the Religion footnote. First, it states explicitly that love of complacence is the capacity to show approval on the basis of perfection. Second, the passage further clarifies that this approval is of two kinds, defined here as either ‘sensible’ or ‘intellectual’ complacence, based on the kind of object approved: if we approve of sensible inclinations, that is, those directed towards happiness, then we experience sensible love of complacence. Intellectual love of complacence, on the other hand, is experienced based on the perception of the good dispositions of kind-heartedness, in this case in the context of others. It is important to note in this context that we do not have direct access to the true nature of our own inclinations and dispositions, let alone those of others, according to Kant’s doctrine of motivational opacity.¹⁸ However, we can of course infer the probable dispositions of others, for example, based on their outer behaviour.¹⁹ Thus, complacence is experienced when we show approval and feel pleasure when we take either ourselves or others to have good inclinations or a good disposition on the basis of inference, even though we are incapable of knowing such things directly.

An important point to highlight is that the pleasure involved in love of complacence is of two distinct kinds, depending on the object we are pleased with. As he states in the passage from the Kaehler notes quoted above, the pleasure we experience in sensible love of complacence is sensible pleasure. Kant makes it clear, however, that the ‘pleasure’ we experience when our maxims conform to the moral law is not pleasure strictly speaking, but is rather a kind of “contentment” (6:45n) that he elsewhere calls “self-contentment [Selbstzufriedenheit]” (see, e.g., 5:117). Self-contentment is distinct from a positive feeling of pleasure in that it is the ‘negative satisfaction’ we experience when we act on the basis of freedom, that is, when we lack the burdens and frustrations that necessarily accompany the inclinations (see, e.g., 5:118).²⁰ Thus, a distinct kind of ‘enjoyment’ is tied to perceiving that our maxims conform to the moral law, as opposed to the sensible pleasure tied to perceiving that our maxims satisfy natural inclinations.

Thus far the focus of my discussion has been on love of complacence for oneself and for others. Kant discusses complacence in a third context as well, namely in the context of love of God, which further clarifies the nature of intellectual love of complacence in particular. When Kant speaks of ‘love of God’ he can be referring to either the human being’s love of God or God’s love of human beings. In the case of the human being’s love of God, since God is not an object of the senses (5:83), it is impossible to have sensible love of complacence for him: God is incapable of happiness because he is not a sensible being (see, e.g., 28:808), so we cannot take pleasure in or approve of his ability to make himself happy.²¹ We can, however, have intellectual love of complacence for God, but because God is a perfect being whose will necessarily conforms to the moral law (see, e.g., 5:79), Kant says that our love of complacence for him is simply “complacence for the law” (6:182). When it comes to God’s intellectual love of complacence for human beings, God “makes his complacence [Wohlgefallens] depend upon the agreement of human beings with the condition of his love of complacence” (6:145–46, trans. modified), namely God’s love “is that of moral complacence of human beings so far as they conform to his holy laws” (6:145, trans. modified).²² God’s love of complacence for human beings is therefore similar to one human being’s intellectual love of complacence for another.

I am now in a position to briefly summarize Kant’s understanding of love of complacence. Complacence, for Kant, is the pleasure or satisfaction we experience on the basis of a being’s perfection and is of two kinds: 1. ‘sensible’ when we take pleasure in a being’s natural perfection, that is, their ability to obtain happiness, or 2. ‘intellectual’ when we take satisfaction in a being’s moral perfection. Love of complacence is not merely a feeling of pleasure or satisfaction, however. Rather, it is a feeling experienced on the basis of the perception of perfection, thus it might be better described as a kind of approval, as Kant himself calls it (2004: 285), similar to how Hutcheson understands complacence.²³ An important feature to be stressed is that both kinds of complacence are dependent on the presence of the object that incites them (see 6:402 and 5:275–76). Kant makes this claim in connection to the idea that this kind of love cannot be commanded (see, e.g., 6:402). As it is stated in Kaehler, for instance, “love of complacence cannot be universally commanded, in that nobody can have complacence where there is no object of approbation” (Kant 2004: 286 and 27:418). The idea here is that we cannot be commanded to approve something because this approval is dependent on there being an object present with the qualities we approve. Furthermore, and as mentioned in the introduction to this paper, Kant also says that a command to love is “an absurdity” (6:401) because, as a feeling, love is not under the control of our will. We have seen above that the object approved in sensible love of complacence is a maxim’s tendency to promote happiness (see Kant 2004: 285 and 27:417), and the object approved in intellectual love of complacence is the conformity of our maxims to morality, that is, the good will and purity of disposition. Kant’s understanding of love of complacence therefore has a further feature in common with Hutcheson’s, namely the approval of love of complacence is ‘immediate’ or ‘direct’ in the sense that it arises involuntarily when we (sensibly or intellectually) perceive the qualities we approve. Presumably for this reason, Kant says in the Metaphysics of Morals that “only love of complacence (amor complacentiae) is direct” (6:402, translation modified) and that complacence is “a pleasure joined immediately to the representation of an object’s existence” (6:402, emphasis added). In general, then, Kant’s conception of love of complacence denotes the pleasure or satisfaction we experience upon representing maxims that promote happiness or morality, whether in ourselves or in other rational beings.

Before concluding this section, there are three additional aspects of Kant’s understanding of love of complacence that deserve to be mentioned because they illustrate how the concept fits with other aspects of his philosophy more generally. First, Kant cautions that intellectual love of complacence for oneself in particular can be mistaken. This is what Kant calls ‘arrogance’ in the second Critique’s discussion of self-conceit and self-contempt: “All the inclinations together [. . .] constitute regard for oneself [Selbstsucht] (solipsismus). This is either the self-regard of love for oneself, a predominant benevolence [Wohlwollens] towards oneself (philautia), or that of complacence [Wohlgefallens] with oneself (Arrogantia). The former is called, in particular, self-love [Eigenliebe]; the latter, self-conceit [Eigendünkel]” (5:73, translation modified). As Kant goes on to clarify: “all claims to esteem for oneself that precede accord with the moral law are null and quite unwarranted because certainty of a disposition in accord with this law is the first condition of any worth of a person [. . .] and any presumption prior to this is false and opposed to the law” (5:73). Arrogance, then, is ‘unwarranted’ intellectual complacence with oneself, that is, the presumption of a good disposition before we have sufficient evidence that we possess it. The important point for my purposes, of course, is that Kant describes our ‘esteem’ for the morality of our disposition, even if unwarranted, in terms of love of complacence and he confirms in his discussion of self-conceit that it is our approval of the extent to which we are moral.

Second, in a few places Kant discusses the opposite attitude of intellectual love of complacence for the virtue of others, namely dislike of their vice. In the ‘Theory and Practice’ essay, for example, in the context of asking whether there are predispositions in human nature that ensure that the race will always progress towards the better, Kant says that if this were the case, then “we could still love the race, at least in its constant approach to the good; otherwise we should have to hate or despise it, whatever might be said to the contrary by the affections of universal philanthropy” (8:307). Kant goes on to add in parentheses that this universal philanthropy, if it is accompanied by hate or despising it on account of its lack of a tendency to the good “would then be at most only a love of benevolence, not of complacence” (8:307, translation modified). The implication here is that love of the human race’s potential ability to progress morally is love of complacence, and the opposite attitude would be hate or despising the race’s tendency to the opposite. This is why if we hate or despise the human race’s lack of a tendency to the good but still possess universal philanthropy, then this philanthropy would at most be benevolence, not complacence.

The above passage from ‘Theory and Practice’ makes a third and final aspect of Kant’s understanding of complacence apparent, namely that there is a distinction to be made between having intellectual love of complacence for another as an individual, on the one hand, and as a human being, on the other. In Kaehler, for example, Kant says:

But in the human being there is a distinction to be made between the human beings themselves and their humanity, thus I can have a complacence for the humanity, even though I have no complacence for the human being. I can also have such a complacence for the scoundrel when I distinguish the scoundrel and the humanity from each other, for even in the scoundrel there is a kernel of the good will, there is no scoundrel who could not comprehend and distinguish between good from evil and who would not wish to be virtuous. (Kant 2004: 286)

Kant’s reference to the good will in this passage indicates that he is talking about intellectual love of complacence, in that we are concerned with approving another’s morality, not their happiness. The point, however, is that even if we do not have intellectual love of complacence for another human being because we find them to be insufficiently moral, we can still have intellectual love of complacence for their humanity, that is, for the capacity they possess to be moral or the possibility of good will that remains in them.²⁴

With all of this now in hand, the next question to ask is: in what sense is complacence rational, for Kant? In the next section (§4) I argue that both benevolence and complacence are rational in the same way, namely they presuppose the possession of certain higher cognitive capacities, specifically the capacities to use concepts and make inferences. This will allow me to then consider, in the final section (§5), the implications of this conclusion for Kant’s understanding of love in general. I give reason to believe, contrary to the suggestion of much recent secondary literature, that the proper concept of love, for Kant, is not merely a pathological phenomenon, but also requires reason.

4. The Rationality of Love

I have argued in the previous section that Kant understands love of complacence as primarily a kind of approval and that it consists in taking pleasure or satisfaction in the perfection, whether natural or moral, of either oneself or another. On this interpretation, love of complacence “must” be rational, as Kant says (6:45n), for a very simple reason: only rational beings have the higher cognitive faculties that make it possible to take pleasure or satisfaction in concepts. More specifically and as we have seen, the pleasure or satisfaction involved in love of complacence is felt based on the representation of maxims (in either ourselves or others) that either serve our natural inclinations or are in conformity with the moral law, and Kant is clear that only rational beings can act according to a maxim, that is, “a subjective principle [Princip] of willing” (4:400n, my emphasis), where reason is the “faculty of principles [Principien]” (A299/B356).²⁵ In the first instance, therefore, love of complacence can only have rational agents as its object, because only rational agents act on maxims.²⁶ Additionally, however, complacence can only be experienced by rational agents as well because only rational beings can possess concepts like that of perfection or a maxim.²⁷ Furthermore, complacence requires the capacity to make inferences in that we must be able to infer our own maxims and those of others in the first place in order to then determine the extent to which they affect our happiness (sensible complacence) or are compatible with morality (intellectual complacence). According to the interpretation that I have offered in the previous section, then, complacence is rational in the sense that it can only be experienced by rational beings who possess these higher cognitive faculties, and because it can only have rational beings as its object.

In the Religion footnote where Kant argues that all love is either that of benevolence or complacence, he is relatively clear about the way in which benevolence is rational. Shortly after making the general division, Kant says that love of benevolence towards oneself amounts to wanting things to go well for oneself and, for that reason, it is “natural” because “who would not want that things always go well for oneself?” (6:45n). Kant then explains that this kind of love of benevolence towards oneself is rational only in the case where our end is our “greatest and most abiding well-being” and “the most apt means for each of these components of happiness are chosen” (6:45n). In other words, to have rational love of benevolence towards oneself is not only to want things to go well for oneself, but to have one’s best and long-term happiness in mind and to take the appropriate means to this end. Kant is clear that reason is only functioning instrumentally here, that is, it “occupies here the place of a servant of natural inclination” (6:45n). Nonetheless, that love of benevolence is rational in this way is significant, because wanting things to go well for yourself merely in the short-term or merely at any given moment and/or failing to choose the most efficient means to achieve one’s “greatest and most abiding well-being”, does not count as rational self-love, that is, self-love properly speaking.²⁸ It also illustrates that benevolence is rational in the same way as complacence: first, benevolence requires that we be able to infer the effects of our actions on our greatest and long-term well-being and, second, it requires the capacity to use concepts, especially that of happiness, which again refers to something relatively abstract, namely our ‘greatest and most abiding well-being.’ Thus, both benevolence and complacence are rational, for Kant, in the sense that they require at least two higher cognitive capacities: the capacity to make inferences and the capacity to use concepts.²⁹

That both benevolence and complacence are rational in this way does not preclude them from being feelings. In fact, that both kinds of love involve both reason and feeling is more consistent with Kant’s empirical psychology than the interpretations that suggest that love is merely or solely a pathological state. This is due to a little-known feature of Kant’s broader empirical psychology, namely that all feelings presuppose cognition. For Kant, there are three fundamental faculties of the human mind: the faculty of cognition, the faculty of desire, and the faculty of feeling (see, e.g., 5:177, 20:205–6 and 10:513–16). Each of these faculties has a ‘superior’ and ‘inferior’, that is, a sensible and intellectual, version depending on their ultimate source (see 7:140 and A15/B29). To take the faculty of feeling as an example, Kant clarifies in the Anthropology that feeling is sensible when “introduced [. . .] through sense” or “the power of imagination”, and intellectual when introduced “through representable concepts” or “ideas” (7:230). Indeed, according to Kant’s understanding of how these faculties interact, feelings are always preceded by a cognition of some kind, whether sensible or intellectual. Consider the following passage from the Metaphysik Mrongovius lecture notes from the early 1780s, where Kant clearly explains his view:

Pleasure precedes the faculty of desire, and the cognitive faculty precedes pleasure [. . .] we can desire or avert nothing which is not based on pleasure or displeasure. For that which gives me no pleasure, I also do not want. Thus pleasure or displeasure precedes desire or aversion. But still I must first cognize what I desire, likewise what gives me pleasure or displeasure; accordingly, both are based on the cognitive faculty. (29:877–78, translation modified)³⁰

The faculty of cognition is therefore the ultimate source of both feelings and desires, for Kant, and thus while feelings are distinct from and not reducible to cognitions, they nonetheless always presuppose cognition as a necessary precondition.³¹

That benevolence and complacence are feelings but remain rational in the way I have described above is therefore perfectly compatible with Kant’s broader empirical psychology: complacence is a feeling of pleasure or satisfaction based upon the representation, that is, cognition, of perfection (natural or moral),³² and benevolence is the “want,” that is, desire, for the “greatest and most abiding well-being” of other beings (6:45n), which, as a desire, presupposes both the cognition of, say, a human being in need, and a feeling that approximates the pleasure we have felt in the past from helping others.³³ Interestingly enough, Kant therefore distinguishes between benevolence and complacence in a way broadly similar to Hutcheson: for Kant, benevolence and complacence as mental states are a desire and a feeling, whereas for Hutcheson they are a perception and a desire.

To be sure, Kant not only claims that benevolence and complacence are rational, he also claims that “love in general”, that is, all love, is either one or the other, and he thereby implies that all love is rational. While a full investigation of the matter would require a thorough analysis of the many ways in which Kant uses the concept of love, I wish to conclude in the next section by citing some evidence which suggests that Kant does indeed conceives of all love, properly speaking, as rational in the sense that it requires the presence of certain higher cognitive capacities.

5. Conclusion: Is All Love Rational, for Kant?

As mentioned at the beginning of this paper, Kant claims that all love is a matter of feeling (6:401), and he thereby implies that all love, properly speaking, is pathological. As has been my focus in this paper, however, he also says that “love in general” can be divided into benevolence and complacence and that both “must be rational” (6:45n), which implies that love in general, that is, all love, is rational as a result. Can these seemingly conflicting claims be reconciled?

If we examine Kant’s use of the concept ‘love’ in contexts other than that of practical love (which is not a feeling and is thus not love, strictly speaking), on the one hand, and benevolence and complacence (which seems to comprise all love), on the other, we find several examples which not only suggest that he uses the term ‘love’ in such contexts only in a qualified sense, but that he reserves the concept of love, properly speaking, for a phenomenon that involves reason. In the Metaphysics of Morals, for instance, Kant says that benevolence as a mere pathological incentive or motive of action, that is, “unselfish benevolence towards human beings,” “is often (though very inappropriately) also called love” (6:401, my emphasis).³⁴ Similarly, Kant says that what he calls the “predisposition to animality in the human being,” which comprises the instinct of self-preservation, the sexual drive, and the social drive, “may be brought under the general title of physical and merely mechanical self-love, i.e., a love for which reason is not required” (6:26, my emphasis). Indeed, like Hutcheson (2008: 102), Kant argues in a number of places that sexual love, which he conceives in terms of the sexual drive (der Trieb zum Geschlecht), is not to be construed as love, properly speaking, because it is not rational. In the Metaphysics of Morals, for instance, Kant says that the sexual drive is only called love “in the narrowest sense of the word” and “cannot be classed with either love of complacence or love of benevolence” (6:426, trans. modified), thereby implying that the sexual drive is not rational. Kant confirms this in the ‘Conjectural beginning’ essay where he makes a contrast between the “animal desire” for sex, on the one hand, and love, on the other, and he states that “some dominion of reason over impulse” contributes to the transition “from merely animal desire [. . .] to love” (8:113), which makes it clear that the sexual drive is not love, properly speaking, because the sexual drive does not involve reason.³⁵ Taken together, the above passages suggest that, although the term love is often used to refer to things like the sexual drive and self-preservation in Kant’s writings, these things are not love properly speaking. Indeed, these passages suggest that just as love (properly speaking) is a matter of feeling, so is love (properly speaking) a phenomenon that always involves reason.

Establishing this claim in full is beyond the scope of this paper, but I take my previous conclusions to illustrate that the rationality involved in Kant’s concept of love as a feeling deserves to be taken seriously and investigated further. If Kant’s proper concept of love turns out to be rational in the way I have described, the implications are far-reaching, and I would like to conclude by mentioning just a few of the implications. First, even what Kant occasionally calls love as attraction (Anziehung) (see 6:470), which “bids friends to draw closer” and involves “intimacy [Vertraulichkeit]” (6:470), and which Helga Varden has recently called “affectionate love” (see 2020: ch.1), does not qualify as love, strictly speaking, because it falls under the heading of the social drive (der Trieb zur Gesellschaft) (see 6:26 and Varden 2020: 36) and thus does not require reason. Second and more importantly, we may need to revisit the place of Kant’s view within the history of philosophy more generally: as Ryan Patrick Hanley has argued, for instance, it is due to the role played by reason in Kant’s conception of practical love in particular that he “deserves to be regarded as one of the truly preeminent modern theorists of other-directed love” (2017: 135). If Kant’s conception of love as a feeling involves reason as well, then his proper concept of love might also set his view apart from the sentimentalist and theological conceptions of love that came before him (see Hanley 2017: 135). Third and finally, if Kant conceives of love, properly speaking, as a phenomenon that is both a feeling but nonetheless remains rational, then we may need to rethink whether Kant conceives of love as an emotion, if by emotion we mean something completely pathological or “pre-cognitive,”³⁶ as scholars like Arroyo have recently implied (see 2016). These are big questions, however, and each of them deserves their own extended treatment. At the very least, I hope to have illustrated that they deserve to be taken seriously and that my conclusions are a first step towards answering them.