1Recently, people in many countries have witnessed the rise of social movements such as “Black Lives Matter” and “#MeToo.” Victims of racial discrimination and sexual harassment have raised their voices to remind us of the persistence of discriminatory practices in our societies. But are we not committed to moral universalism and the attribution of equal moral worth to all human beings? Obviously, our endorsement of these moral claims does not protect us from racist or sexist attitudes and behaviour. Immanuel Kant is one of the most eloquent defenders of moral universalism – or so it has long seemed to many Kant scholars and friends of Kant’s normative moral theory. Indeed, Kant argues that we, as rational and moral agents, must submit to the categorical imperative which, in one of its formulae, reads as follows: “so act that you use humanity, whether in your own person or in the person of any other, always at the same time as an end, never as a means” (4: 429).1 To this day, many moral philosophers refer to Kant when they speak of the “dignity” of all human beings and their equal right to be recognized as having moral worth, regardless of their ethnicity, gender, sexual orientation, age, state of health, religious beliefs, socio-economic status and education.2 Kant’s influence extended far beyond academia. The first amendment to the German constitution, for example, is clearly committed to a moral universalism of Kantian inspiration. It says: “Human dignity shall be inviolable. To respect and protect it shall be the duty of all state authority.”3 According to this amendment, every human being has the right to seek asylum in Germany or to apply for German citizenship.4 It is intended to put an end to all forms of discrimination against people on the basis of morally irrelevant criteria. The authors of the German constitution wanted to ensure that crimes against humanity like those committed in Germany under the Nazi regime would never be repeated.
- 5 See 4: 314-315. On Kant’s misogynism, see Wilson 2022: chapter 9.
2However, the moral universalism that we commonly ascribe to Kant today seems to be more inclusive than the one the historical Kant had in mind. That Kant was a misogynist is no secret. For example, he excluded all women from active citizenship.5 But Kantians generally dismiss his unfortunate misogyny as nonsense and deny that it affects the authority of his normative moral theory. The opening of the category of rational agent to both men and women is an easy step beyond Kant’s views.
- 6 On the difficulty of recognizing both moral and political duties of conformity to the law see Frick (...)
- 7 The same applies to his lectures on anthropology, physical geography and history. Kant held those l (...)
- 8 Here, I rely on the Oxford English Dictionary which defines racism as the “prejudice, antagonism, o (...)
3When Kant scholars praise Kant’s moral universalism, they mainly have in mind three of his books, namely the Groundwork of the Metaphysics of Morals (1786), The Critique of Practical Reason (1788), and the Metaphysics of Morals (1790). In these books, Kant speaks of human beings as equals, as rational agents and as subjects of objective and necessary moral – or political – duties.6 But in the years when he was writing his major works on normative moral and political theory, he also wrote other texts in which he turned his attention to the fact that human nature can take many different forms. Not only are there men and women, adults and children, healthy people and people suffering from illness, including mental illness; there are also ethnic differences between people. Kant claims that not only gender, age, and the state of mental health can affect the status of an individual human being as a rational and moral agent. The same is true of an individual’s ethnic identity. In his Conjectural Beginning of Human History (1786), On the Use of Teleological Principles in Philosophy (1788), and Determination of the Concept of Human Race (1785), Kant expresses racist views, views that reveal his prejudice and discrimination against people based on their membership of a particular racial or ethnic group.7 Indeed, he expresses the belief that members of a particular racial or ethnic group, namely white Europeans, are naturally superior to others because of some innate characteristics.8
- 9 For a detailed and careful analysis of this debate see Lu-Adler 2023: 36-43.
4For many committed Kantians, his openly racist claims and his defence of white moral supremacy pose a major challenge. How could Kant, the eloquent defender of moral universalism, hold such views? Did he not realize that they were incompatible with his moral universalism? These questions have given rise to a controversial debate. Participants in this debate can be divided into two groups, namely those who read Kant as an inconsistent moral universalist and those who read him as a consistent moral particularist and white moral supremacist. Members of both groups share the view that Kant’s moral universalism and his racist moral particularism are incompatible.9 Pauline Kleingeld has suggested a widely welcomed way out of this inconsistency; she has defended the claim that the old Kant underwent a radical change of mind; while he held undeniably racist views for most of his life, views that were moreover incompatible with his doctrine of universal human equality, he changed his views in the 1790s. According to her, Kant’s late works, especially his Toward perpetual peace (1795), are consistent with his commitment to universal human equality beyond all contingent differences between individuals and peoples (Kleingeld 2007). In this text, Kant develops the utopia of a perpetual peace between nation-states that respect each other’s freedom and right to peaceful existence, just as individual citizens of such states respect each other’s freedom and right to independence from each other’s domination.
5Huaping Lu-Adler, in her book on Kant, Race, and Racism, has undermined the assumption underlying this debate by arguing convincingly that there is no logical contradiction between Kant’s moral universalism and his racist claims (Lu-Adler 2023). In what follows, I will briefly recall Kant’s racist claims and then offer an equally brief analysis of Lu-Adler’s argument. (I.) I will then distinguish between three different assumptions underlying racist beliefs and attitudes and ask which of these assumptions Kant seems to endorse. Whether we can dismiss Kant’s racist claims as nonsense and conclude that they do not affect the core of his normative moral theory depends on the answer to this question. (II.) Next, I will turn to Kant’s rationalist, formalist and seemingly universalist normative moral theory as he develops it in his Groundwork and in the Critique of Practical Reason. I will focus on his account of moral deliberation and the complex role he assigns to maxims in this process. I will argue that Kant grants the rational agent the exclusive power to determine the meaning of (her or) his action. I will then return to the questions raised in II. and argue that there is one kind of racist attitude that Kant’s normative moral theory does not help us to avoid. (III.) This claim can be challenged, especially in the light of what Kant says about the duties of virtue in the Metaphysics of Morals. I will attempt to meet these challenges and conclude that Kant’s account of the duties of moral agents does not provide a sufficiently strong bulwark against the effects of structural racism on our beliefs and actions. (IV.)
6In his Determination of the Concept of Human Race (1785), Kant says:
We know with certainty of no other hereditary differences of skin colour than those of the whites, the yellow Indians, the Negroes, and the copper-red Americans. (8: 93)
- 10 See on the extensive audience of these lectures see above, footnote 7, and Lu-Adler 2023: 58, footn (...)
7He admits that the colour of a person’s skin “might appear very insignificant to some” (8: 93). Yet he does not hesitate to assert that skin colour functions as an indicator of the degree of “perfection” achieved by the members of an ethnic group. As we know from his Lectures on Physical Geography (1802),10 he argued that
[h]umanity has its highest degree of perfection in the white race. The yellow Indians have a somewhat lesser talent. The Negroes are much lower, and lowest of all is part of the American races. (9: 316)
8In the text following this passage, he distinguishes between “savages” and “civilized people.” First, he explains the differences between the two in terms of the way they spend their childhood; the “greater strength” of the savages is due “to the free movement they are allowed in childhood”– leaving open the possibility that this difference might be due to different cultures of education. But he then goes on to claim that people of different skin colours also differ in the degree of their “perfection” (9: 316). On the scale of perfection, white people, or rather adult white, mentally healthy (and male – but that is not my main point here), Europeans, occupy the highest level and copper-red Americans the lowest.
- 11 For a detailed account of the discussion of this seeming inconsistency, see Lu-Adler 2023: 36-43.
9His doctrine of differences in “perfection” between human beings of different races goes back to his Observations on the Feeling of the Beautiful and Sublime (1764) where he already asserted that members of different races have attained different levels of perfection, and that black people in particular are incapable of attaining the level of perfection that only white people can – and have – attained: black people cannot attain that level even if they “have been transported elsewhere from their countries”; and he continues: “although very many of them have been set free, nevertheless not a single one has ever been found who has accomplished something great in the art or science or shown any other praiseworthy quality” (2: 253). This argument reveals a view of non-white people that seems to be at odds with Kant’s moral universalism. Indeed, many Kant scholars believe that Kant’s moral views, his moral universalism and his racist moral particularism, are incompatible.11
- 12 See Kant’s 9: 94-99, especially § 10, and 9: 143-144, especially § 106, and Lu-Adler 2023: 13ff. Th (...)
10However, Huaping Lu-Adler argues convincingly that there is no such inconsistency; indeed, she understands his racist views as an integral part of his philosophical system. To defend her claim, she draws on conceptual tools from Kant’s theory of logic. Of great importance to her argument is her distinction between what she calls ‘real abstraction’ and what Kant calls “logical abstraction”; she uses this distinction to specify the sense and reference of the terms Kant uses when he talks about human beings.12 On the one hand, he speaks of “humanity” or the “idea of a human being,” and on the other of “human beings” in general. Modern readers may overlook this distinction, but it is crucial. “Humanity” and the “idea of a human being” are generic terms, terms for the species “human being” as opposed to other species of animals such as “horses,” “tigers,” or “ducks.” Generic terms are based on real abstraction. “Humanity” refers to the human species as the only animal species endowed with reason. Rationality is the defining characteristic of human beings, the characteristic that distinguishes the biological species of human beings from other biological species of animals. The real abstraction is the abstraction from the fact that there are human beings who are not rational. In contrast to the generic term “humanity” (the term Kant uses in the formula of the categorical imperative quoted above), Kant uses “human beings” as a general term for the aggregate of individual human beings, that is, for all the members of the biological species of human beings. These individuals differ from each other in many ways. The general term “human beings” is based on logical abstraction, i.e. on bottom-up processes of increasingly general classifications of individual objects. While “human beings” is the most general term for human beings, there are various less general terms, such as “men” and “women,” or general terms for classifying human beings as belonging to one of four races, namely black, red, yellow and white human beings (see Lu-Adler 2023: 45-47).
- 13 Here, Lu-Adler relies mainly on passages from Kant’s Anthropology, including 7: 329-330.
11What is constitutive of an individual living being’s membership of the human species is its “phylum” (“Stamm”) which, as Lu-Adler points out, “contains certain germs (Keime) and natural predispositions.” But “these germs and predispositions developed differently under the various material conditions (climates) in which the early humans lived for a long time” (2023: 46). Underlying this is Kant’s monogenism, the view that all human beings have the same ancestors, namely Adam and Eve. The descendants of this first human couple, all carrying the same human germs, spread out over the world, and settled in different climatic zones. There, they developed into four different races, each adapted to a different climate. Although all human beings carry the same germs, it is not the case that all human beings have gone through the same evolutionary process. Quite the contrary. The germs need certain external conditions to germinate, and, as Kant argues, it was only in Europe that humans found conditions favourable to the development of rational and moral capacities. Nevertheless, it took the descendants of Adam and Eve who migrated to Europe many generations of “hard work and intergenerational transmission of enlightenment” to develop the rational and moral capacities characteristic of adult, mentally healthy, white European males (Lu-Adler 2022: 47). Members of other races, though having the same germs, were neither so lucky nor so skilled. They did not develop to the level of Europeans – unless they did not develop at all.13
12Now, on the basis of her distinction between generic terms of human beings, such as “humanity” and the “idea of a human being,” and general concepts of humans such as “human beings,” Lu-Adler convincingly denies that there is any inconsistency between Kant’s moral universalism and his defence of white male moral supremacy. Her point is that when Kant speaks of human beings as bearers of reason and equal moral worth, or as agents bound by the moral law, he is speaking of human beings as specimens of the human species, paradigmatically embodying what only human beings are capable of. On the other hand, when he asserts that members of different human races have attained different degrees of perfection, and that some of them lack not only rationality but also the capacity to acquire rationality and thus the status of rational and moral agents, he speaks of individual human beings, categorized by means of various more or less general concepts. While the human species is defined as an animal species endowed with reason, not every individual member of the human species is equally rational. As Lu-Adler puts it: “What is true of the human species as a whole is not on that account true of all individual human beings” (2023: 45).
13To illustrate, compare the generic term “horses” as used in “Horses are good runners” with the general term “horse” as used in “My horse has recently been injured and can no longer run fast.” The generic term is used in claims that define the species; only a specimen or paradigmatic sample, such as a healthy adult horse, exemplifies the property mentioned in the definition. But the definition of horses as a species is compatible with claims about particular horses that lack the property that defines the species (namely the property of being a good runner), and so could not serve as a paradigmatic sample. I can say that “horses are good runners” and that “my own horse cannot run fast” without contradicting myself.
14The crux of Lu-Adler’s argument for the claim that Kant’s moral universalism and his racist moral particularism are not incompatible is that Kant, in his moral theory, speaks exclusively of human beings who are rational agents; in his writings on anthropology, physical geography and human history, he makes it explicit that only adult, mentally healthy, white Europeans meet the necessary and sufficient conditions for being rational agents. These chosen few, according to Kant, are the specimens or paradigmatic samples of the biological species of humanity. In fact, this sample is unfair in two ways: it does not represent a statistical majority of human beings; nor is its defining characteristic, namely the possession of reason or rationality, exclusive to the members of this sample. After all, according to Kant, human beings share the characteristic of being rational with divine beings, including God.
- 14 One piece of evidence for this claim is provided by psychologists Kenneth and Mamie Clark’s 1947 “d (...)
15As I mentioned earlier, racism is commonly defined in terms of beliefs and attitudes that reflect certain types of prejudice and discrimination against people because of their membership of a particular racial or ethnic group. Many white citizens of European liberal democracies believe themselves to be committed to moral universalism and the equal moral worth of all people; accordingly, they deny that they hold racist prejudices or discriminate against non-white people. They fail to recognize that some of their beliefs and attitudes nevertheless reveal their commitment to white supremacy: They do not welcome black people as neighbours; and they cannot hide their surprise when they see black people among the CEOs of a company; they ask non-white people where they come from or whether they speak the national language of the respective country, implying that non-white people cannot be indigenous to a white-dominated society or fluent in the language spoken by its members. As Charles Mills has pointed out, racial categories and common practices of racial discrimination affect the self-understanding of the racialized groups; this applies not only to those who are discriminated against, but also to those who enjoy presumed ethnic superiority and the privileges that go with it. Mills reminds us that the assignment of racial identity “influences the socialization one receives, the life-world in which one moves, the experiences one has, the worldview one develops – in short […] one’s being and consciousness” (Mills 1998: xv; emphasis in original). Indeed, non-white people tend to expect to be treated as inferior by white people; they are therefore particularly sensitive and vulnerable to racist discrimination; white people do not share this vulnerability; they simply assume they are normal and are often unaware of the privileges that only white people enjoy.14
- 15 For a recent account of intersectionality and its history see Hancock 2016.
16Racist beliefs and attitudes can take many different forms. People can be discriminated against because of the colour of their skin. But the German Nazis did not discriminate against Jews because of the colour of their skin. Belonging to the white race does not necessarily protect people from racial discrimination. Moreover, different criteria for racial discrimination are not mutually exclusive: someone can be discriminated against both for being black and for being Jewish. This phenomenon is commonly known as “intersectionality”.15 Other differences between racist beliefs and attitudes concern the presumed explanation for the racist discrimination against members of particular groups. A racist may discriminate against members of particular racial or ethnic groups on the assumption that (a) they do not belong, or do not fully belong, to the human species, that they are subhuman; (b) that, while they do belong to the human species, they are not fully rational, they are incapable of self-management, and therefore need tutelage or political government from white people; or (c) that while they are human and fully rational and responsible agents, their lack of education and knowledge, their laziness and consequent poverty, their lack of social and moral commitment, and their propensity to crime inevitably place them at the bottom of the socio-economic hierarchy of the society in which they live.
- 16 The distinction between a metaphysical or biological kind of racism on the one hand and a socially (...)
17While the first two types of racism are based on false assumptions about the metaphysical or biological differences between white and non-white people, the third type relates to a social reality shaped by socio-economic hierarchies and the pervasive practice of white supremacy, such as we find them in all European societies. Liberal democracies are no exception.16 This latter type of racism is present in attitudes and behaviours that reveal what is commonly referred to as “structural racism.” Much of the racism that persists in the beliefs and attitudes of citizens of liberal democracies is of this third in kind.
- 17 See Eze 1997 and Mills 1997; see on this topic also Mills 2005.
18The question is which of these putative explanations for racial discrimination against non-whites we have reason to attribute to Kant. Kant’s distinction and ranking of the four different human races might lead one to suspect that he holds the view that non-white people are not fully human. This is the reading that Eze defends. According to him, Kant regards non-white people as subhuman, “humanoid entities” (a term used by Charles Mills), who, on the scale of ranking living organisms, are somewhere between non-rational animals and proper human beings.17 But Eze overlooks the distinction between the generic term “humanity” and the general term “human beings in general,” the importance of which Lu-Adler has pointed out. Given Kant’s monogenism and his claim that the germs of humanity are present in all individuals who belong to the biological species of human beings, I hesitate to follow Eze’s reading. What Kant denies is that non-white people could ever be specimens or paradigmatic examples of humanity; he does not deny that they are human in the sense of belonging to the biological species of human beings.
- 18 See for example 8: 303-306. On Kant’s views on women see Wilson 2022: chapter 9.
- 19 Wilson already argues for a certain analogy between Kant’s claims on women (in comparison to men) a (...)
- 20 For those who want to keep the whole of Kant’s philosophical system in view, dismissing his racist (...)
19Given Kant’s distinction of four human races and his ranking of them according to their perfection and perfectibility, it seems to me to be much more plausible – even though no less objectionable – to attribute to Kant the view that non-white people, though members of the human species, are not fully rational and are therefore less inclined than members of the white race to engage in the hard work necessary to reach the highest level of perfection. However, since white people have reached a high level of perfection in a historical process of civilisation, of increasing rationality and learning to submit to the authority of laws, this view is compatible with the idea that non-white people will also reach this level at some point in the future – even though Kant denies that some non-white people have any capacity to make such progress. If this is how we read Kant’s racist claims, then we might conclude that they are somewhat analogous to his claims about women. Women are human beings, and they are not devoid of all rationality. But they are more guided by emotion than by reason and its laws and are therefore inferior to men; they need men’s guidance.18 By analogy, non-white people are human, and they are not devoid of all rationality. But they are less inclined than members of the white race to exercise control over their emotions, to submit to the authority of laws, and to allow their rationality to dominate their desires or inclinations, and are therefore inferior to the latter.19 Given this analogy, we might conclude that Kant’s racist claims are of minor importance for how we should read his moral theory, that they cannot undermine the authority of that theory, that we can dismiss these claims as nonsense and open the category of rational and moral agents to non-whites as we have opened it to women. Both men and women, and both white and non-white people, can have the status of rational and moral agents, and when they have that status, they have equal moral worth and dignity.20
20There is, however, a third way of understanding racist beliefs and attitudes, according to which racism is structural in nature, manifested in a social reality of both individual and institutional discrimination against members of an ethnically or racially identified group. The question is whether a rational agent who follows the moral law as Kant explains it is necessarily immune to beliefs and attitudes that are constitutive of structural racism. As I will argue, this question must be answered in the negative. There is, then, a kind of racism that affects the authority of Kant’s moral theory; and it cannot be dismissed as easily as the versions of racism discussed earlier.
21Before defending this claim, I would like to draw attention to another problem that arises within the framework of Kant’s rationalist account of moral deliberation and moral agency. This problem challenges his commitment to moral universalism as it is commonly understood as much as his racist claims do. There is the question of how an individual’s age and state of mental health affect her or his capacity to reason, to recognise the authority of the moral law, to exercise control over her or his desires and inclinations, and to fulfil her or his moral – or political – duty. Young children and the severely mentally impaired, whatever their gender or ethnic identity may be, do not have the status of rational and moral agents; we do not expect them to do their moral duty; nor do we hold them responsible for their actions. Contemporary Kantians generally assume that all people have dignity and equal moral worth, regardless of their membership of a racial or ethnic group, of their gender, their age, or their mental health. This is how they understand Kant’s moral universalism. As Oliver Sensen has pointed out, “[i]t is a common view in the literature that Kant conceives of human dignity as an inner value all human beings possess”; and “[h]uman beings simply have this value in virtue of being human” (2009: 309 and 310).
- 21 Kant himself departs from his understanding of dignity as the metaphysical status of a rational and (...)
- 22 Attributing this alternative kind of dignity is not limited to members of the biological species of (...)
22The historical Kant, however, ascribed dignity primarily to those human beings who have the status of rational and moral agents, that is, to the specimens or paradigmatic examples of what it means to be human: “dignity” is the status “of a rational being, who obeys no law other than that which he himself at the same time gives” (4: 434). Human rational and moral agents have dignity by virtue of being rational – that is, by virtue of being specimens of humanity, not by virtue of belonging to the biological species of human beings. This leaves open the question of whether and to what extent dignity – and moral worth – can be ascribed to those human beings who do not have the status of rational and moral agents. When contemporary Kantians ascribe dignity and moral worth to all human beings, regardless of whether they are rational and moral agents, they must rely on a different notion of “dignity.” The dignity they have in mind allows for degrees.21 What constitutes this extended dignity is notoriously unclear. Contemporary Kantians rely on it to define the extension of the class of beings that rational and moral agents must respect because of their moral worth; this class includes all members of the human biological species, if not all sentient and vulnerable beings.22
23As I said earlier, my claim is that a rational agent who follows the moral law as Kant explains it is not necessarily immune to beliefs and attitudes that are constitutive of structural racism. Why is this so? To answer this question, I will briefly outline what I take to be the main tenets of the historical Kant’s normative moral theory as he developed it in the Groundwork and in the Critique of Practical Reason.
- 23 See the title of a famous text of his, namely On a supposed right to lie from philanthropy (Mensche (...)
- 24 See for example 4: 389 and 400.
24One of Kant’s central claims is that rational moral agents should not let themselves be guided by feelings of benevolence or the “love of mankind,”23 but should follow the moral law in the form of the categorical imperative. To do this, they should engage in a process of moral deliberation. Kant denies that this deliberation is a matter of weighing costs against expected benefits – Kantian rational agents do not maximize their utility. Moreover, moral agents must rely on purely rational means of deliberation, namely on analytic judgments and tests of the logical consistency of various assumptions. Otherwise, their conclusion that an action is morally permissible, and therefore in accordance with the moral law could not be infallible.24 Another requirement is that the deliberating agents must ignore all those features of the rational and moral agents to whom their actions are addressed that are contingent and have no bearing on their status as rational and moral agents. If we dismiss Kant’s claims about women and non-white people as nonsense and recognise all adult and mentally healthy people as rational and moral agents, it follows that rational and moral agents must be gender-blind and colour-blind when considering what to do and whether what they intend to do is morally permissible. But how can rational moral agents who are presumed to be blind to all contingent differences between them, engage in decisions and actions that reveal structural racism? Would not the prescribed blindness inevitably save them from such a reprehensible attitude?
- 25 See 4: 433. On Kany’s account of rational agents as legislators in the ideal moral world see Fricke (...)
25Kant gives several formulations of the categorical imperative. In the formulation I quoted earlier, it prescribes to act in such a way that “you use humanity, whether in your own person or in the person of any other, always at the same time as an end, never merely as a means” (4: 429). In another formulation, it prescribes always to act “in such a way that I could also will that my maxim should become a universal law” (4: 402). The first formula reminds us that, in many cases, when we perform an action, we are addressing another rational agent. It asks us to recognize this other agent as a being like ourselves, an equally rational agent, and an agent who, like ourselves, is a legislator in the ideal moral world. Kant speaks of this world as of the “kingdom of ends”; in this ideal moral world, all citizens are free and they always act in accordance with the moral laws.25 The second formulation gives us a more precise account of the deliberative process by which we come to a decision about what to do in a given set of circumstances: following this process, we recognize an action that we can perform in a given set of circumstances as the right and morally permissible action, and we do so with certainty, that is, by relying on purely logical reasoning.
26In the universal-law formula of the categorical imperative, Kant uses the term “maxim.” Given the importance of maxims in an agent’s moral deliberation, Kant says surprisingly little about what he understands by a maxim. As Patricia Kitcher has pointed out, he could rely on his readers being more familiar with the concept than we are today, since Christian Wolff had already explained it in his German Metaphysics (see Kitcher 2003). Nevertheless, Kantians have spilled a lot of ink trying to explain what maxims are. Kant defines them as “subjective principles” (5: 19; see also 4: 400, footnote). They are principles for action, principles on which an agent can rely to decide which course of action to take in a given set of circumstances (see Fricke 2008). Unlike the purely formal categorical imperative, maxims have content. They specify what a particular agent has learned in the course of her or his life about sources of pleasure and pain, and about certain kinds of achievement that promise to increase her or his happiness. Maxims are motivational forces; they can determine an agent’s will. They derive their motivational power from an agent’s self-love and natural desire to be happy. But “happiness” is a very vague concept; as such, it does not provide much practical guidance. To learn what would make her or him happy, what to do for reasons of self-love, an agent must rely on her or his experience. What has made her or him happy? What has made her or him unhappy? What could she or he learn from others about sources of happiness and unhappiness?
27Rational agents do not only want to be free from pain, from hunger, thirst, cold, and fear. They also want to have the means to provide for themselves and their families, to protect themselves from harm now and in the future, and to enjoy social prestige. So they set themselves long-term goals such as getting an education, becoming rich, or achieving high social status. All of this informs the content of the maxims that agents endorse. As Patricia Kitcher has put it, “agents come to be stocked with maxims for judging actions,” thereby making themselves “rationally self-interested agents” (see Kitcher 2003: 224). And when agents come to a crossroads and rely on maxims to decide which way to turn, the actions chosen turn out to be those that promote the agents’ happiness. So maxims are devices for assigning value to certain kinds of actions (see Kitcher 2003: 224). But this value is not really moral in nature, it is a value deriving from the agents’ self-love.
28Rational agents, when arriving at a crossroads and deciding where to go from there, do not toss a coin; nor do they take the path they expect to be free of obstacles; instead, they rely on a maxim as a principle of rational choice. Maxims are principles, that is, they are general rules, rules of life (“Lebensregeln”) as some scholars have called them (see Bittner 1974 and Höffe 1977). They are long-term intentions, i.e. intentions that specify goals to be achieved, rather than short-term intentions which are highly specific and can only guide an agent in circumstances that will never recur. For example, an agent may endorse the maxim to treat his mother generously on her birthday and at Christmas each year; but the opportunity to treat her by taking her to a particular concert will never arise again. Therefore, the intention to take her to that one concert is not a maxim.
- 26 On the rationality of an agent’s maxims see also Pollok 2017: 260.
29Nevertheless, an agent who comes to a crossroads and chooses an action, will, when asked why she or he has made this choice, answer the question by making explicit the maxim that guided her or his choice. Going to a particular concert is not in itself the rational thing to do. It becomes the rational thing to do if the agent has chosen to go to that concert by relying on a maxim she or he has previously endorsed. Maxims provide reasons for action. Rational agents do things for reasons, and the reasons for the decisions they make are maxims.26 Finally, maxims are subjective principles in the sense that their motivational force is limited to a particular agent. While all rational agents act under the influence of self-love and while they all want to be happy, what they take their happiness to depend on may vary from one agent to another. Different agents endorse different maxims.
30The second formula of the categorical imperative requires that an agent, before relying on a particular maxim to make a choice in a given set of circumstances, should subject that maxim to a particular kind of test, commonly called the “universalizability test.” She or he should ask whether she or he “could also will that [this] maxim should become a universal law.” The question of how exactly to proceed for testing a maxim is controversial. What I take to be crucial for answering this question are two Kantian claims: on the one hand, his claim that rational agents, along with divine rational agents, are legislators of the moral laws that uniformly guide the actions of all citizens of the kingdom of ends, the ideal moral world in which it is impossible to violate the moral laws; and on the other, his claim that rational deliberation should rely exclusively on analytic claims and claims about logical (in)consistency. Thus, the agent should ask whether a particular maxim she or he has previously endorsed could be among the moral laws that govern the citizens of the ideal moral world. Given that the agent will not be alone as a legislator, the question is whether all the other citizens of the ideal world, who are equal moral legislators to the agent, have any reason to reject this maxim as a universal law in this world.
31To answer this question, the agent does not go round asking other people what they would want the ideal moral world to be like. Such a device would introduce contingent and therefore uncertain elements into the process of moral deliberation. Instead, the agent focuses on the agent to whom she or he is addressing the action chosen under the guidance of the maxim. Who this other agent is, what her or his contingent and therefore morally irrelevant properties are, does not matter. It can be anyone, as long as she or he is a rational agent and thus a co-legislator in the ideal moral world. The agent then imaginatively assumes that this other agent – a co-citizen and co-legislator in the morally ideal world – wills that among the universal laws of the ideal moral world there be one that has the content of the agent’s maxim. But that is not all the agent does. The agent also relies on the same maxim to ascribe certain expectations or volitions to this other agent, namely those expectations and volitions which this other agent must necessarily endorse if she or he is to assume the role of the addressee of the agent’s action. This ascription takes the form of an analytic judgment. For example, if I make a promise to Jill and if Jill accepts the promise, Jill takes on the role of the promisor. This allows me to attribute to Jill the expectation that I will keep my promise and the volition that I will perform a certain action in the future, namely the action I promised to perform.
- 27 My reading of Kant’s account of practical deliberation is to a large extent in agreement with the o (...)
32What remains is to ask: given the role of the other agent as the addressee of my action, and the expectations and volitions that I must ascribe to this agent, and given the assumption that this agent wills that there is a universal law in the ideal moral world which has the content of my maxim, do I ascribe to this agent volitions that are consistent? The answer to this question is a matter of pure logic. If the other agent’s volitions are consistent, then I can conclude that my maxim could be a moral law in the ideal moral world, and that the action I choose to take in reliance on that maxim is morally permissible. Otherwise, if the volitions I ascribe to the other agent turn out to be inconsistent, I must conclude that my maxim could not be a moral law in the ideal moral world, and that I, as a moral agent, I must abandon this maxim once and for all. For example, if I make a false promise to Jill, on the basis of my maxim that I may make a false promise in an emergency, I must reason as follows: Jill, as the addressee of my promise, expects and wants me to do what I have promised. But I must also assume that Jill wills that it is a universal law that people may make false promises in cases of need. So she must want me to make a false promise because I am in need. She must therefore will that I should not do what I have promised her; but, as the promisor, she must will that I should do what I promised her. So she must will that I do, and that I do not, what I have promised. And these volitions of her are inconsistent. I must conclude that my maxim cannot be a law in the ideal moral world. I must eliminate it from the stock of my maxims. I cannot avoid this conclusion, for it follows with certainty from the premises. A maxim that turns out to be a possible law in the ideal moral world still serves as a means for assigning value to a particular action; but this time, the value in question will be moral, and thus objective, rather than related to an agent’s self-love and thus merely subjective.27
33One of the conclusions to be drawn from Kant’s account of maxims and their universalizability is that an agent, by relying on a maxim for making her or his choice in a given set of circumstances, determines the meaning of her or his action. Kant scholars agree that the actions that agents take in given circumstances do not in themselves reveal which maxim provided the reasons for their choices. Indeed, different maxims may provide different reasons for performing the same action. Kant’s famous shopkeeper provides an example. The shopkeeper chooses to be honest with his customers. But his honesty may be motivated either by prudential reasons of self-love or by moral reasons. Only in the latter case, if he acts in accordance with the maxim of always submitting to the moral law, is his honesty truly moral (see 4: 397). Agents, when asked why they did what they did, reveal their reasons, the maxims that guided their choices, and thereby determine the meaning of their actions.
34This brings me back to the question I posed at the beginning of this section, namely whether a rational agent who follows the moral law as Kant explains it is necessarily immune to beliefs and attitudes that are constitutive of structural racism. My answer to this question is in the negative. In order to defend this claim, I will rely on another of Kant’s examples for morally permissible and good actions, namely on the example of charitable donations. The agent contemplating such a donation must be aware of the socio-economic differences between people in his social environment. How should a potential donor identify the recipients of her or his donation? Within the framework of his formalistic moral theory, Kant does not raise this question. For the moral permissibility of a charitable donation does not depend on whom the donor chooses as the recipient of her or his donation. It depends solely on the maxim on which the donor bases her or his decision to donate, and on its universalizability and suitability as a law in the morally ideal world. The awareness that the donation will be costly, that it will prevent the agent from buying a ticket to the theatre, should not prevent her or him from making the donation (see 5: 23). The permissibility of making the donation does not depend on its actual consequences.
35However, when it comes to the persistent structural racism of the rational moral agent, the question of who she or he chooses as the recipient of her or his charitable donation may be central. It may well be the case that a white donor assumes that non-white people are most in need of charitable donations because they tend to be at the bottom of the socio-economic hierarchy. Thus, the white donor, confident in the moral permissibility and goodness of his charity, may choose a non-white recipient and, without being aware of it, let his choice be guided by the belief that non-white people are poor and in need of charity. The non-white recipient, socialized in a society where structural racism prevails, may feel humiliated; after all, she or he is vulnerable to such acts of seemingly well-intentioned discrimination. But the white donor, confronted with his having humiliated the non-white recipient of his donation, can defend himself by pointing out that he meant well and that he was far from humiliating anyone. As I have argued, Kant gives the rational and moral agent the monopoly on determining the meaning of his action. Unintended side effects cannot diminish the moral value of an action; nor can such effects change the meaning of an action. Kant’s moral theory, as developed in the Groundwork and in the Critique of Practical Reason, is a theory of agents’ moral convictions; it is not consequentialist in nature. It affirms the power of agents, including the power to determine the meaning of their actions and to reject any moral criticism ex post, that is, criticism which is informed by the actual consequences of an action, including its side effects.
- 28 I would like to thank the three anonymous reviewers of a previous version of this paper who request (...)
36Kantians, however, may not be convinced by this argument. They will point out that my conclusion, namely that a moral agent committed to the moral law as Kant explains it is not necessarily immune to structural racism, is based on a mere selective reading of his writings on moral matters, and that I have neglected what he says on the subject in the second part of his Metaphysics of Morals, namely in his Doctrine of Virtue.28
- 29 In the light of Lu-Adler’s distinction between the generic term ‘humanity’ and the general notion o (...)
37In the Doctrine of Virtue, Kant makes claims about the moral duties of citizens of a state. The first duty of these citizens is to obey the laws of the state. Within the framework set by these laws, they have two additional moral duties: They must promote their own “perfection” and “the happiness of others” (6: 386). Citizens cannot promote the happiness of another citizen unless they recognise his “legitimate claim to respect from his fellow human beings” (6: 462). Accordingly, “to be contemptuous of others […], that is, to deny them the respect owed to human beings in general, is in every sense contrary to duty; for they are human beings” (6: 463, Kant’s italics).29 Nevertheless, he admits that “at times one cannot […] help inwardly looking down on some in comparison with others […]; but the outward manifestation of this is, nevertheless, on offense” (6: 463, Kant’s italics). Charitable donation is one way of promoting the happiness of others, and it should be done in accordance with the requirements of respect.
38In fact, in the Doctrine of Virtue, the question is no longer just whether the donor’s maxim could be a moral law in the morally perfect world. The question is also how the donor should address the recipient of her or his donation. Kant is aware of the possibility that a donor might choose a person who is not in need as the recipient of her or his donation. However, according to Kant, the moral challenge arising from such an error is not on the side of the donor, but rather on the side of the recipient of the donation. He claims that morality requires us not “to accept favors […] [we] could do without; and […] not be a parasite or a flatterer or […] a beggar” (6: 436). Furthermore, Kant considered that the recipient of a charitable donation might feel humiliated by the fact that the donor had chosen her or him for that role:
[…] we shall acknowledge that we are under obligation to help someone poor; but since the favour we do implies that his well-being depends on our generosity, and this humbles him, it is our duty to behave as if our help is either merely what is due to him or but a slight service of love, and to spare him humiliation and maintain his respect for himself. (6: 448-449)
39What Kant is recommending here is that the virtuous agent performs his duty in such a way that others do not thereby feel humiliated by him.
40The question is whether the precautions Kant urges citizens to take when promoting the happiness of other citizens provide a sufficiently substantial bulwark to keep structural racist beliefs and attitudes at bay. This is a difficult question to answer. Here, I can only express my scepticism. Kant speaks mainly of agents who pride themselves on being rational, a characteristic they share with divine beings. But divine beings are not vulnerable. Human agents who pride themselves on their rationality are not naturally the most competent judges of human vulnerability. How do they acquire the competence to approach others in a way that does not humiliate them? Recognizing their right to be treated with the respect due to them as rational agents, and as agents who do not engage in inconsistent volitions, may not be enough. We want a moral agent to be sensitive and empathetic to the vulnerability of others – whether they have the status of a rational and moral agent or not. While all human beings share certain vulnerabilities, namely those arising from their biological human nature (the vulnerabilities they share with non-human animals), they also differ in their individual vulnerabilities.
41Citizens are likely to rely on the Golden Rule to decide how to fulfil their moral duty to promote the happiness of other citizens. They assume that all citizens are rational agents and therefore not only morally equal, but also like each other. Thus, when citizens wish to promote the happiness of other citizens, they will draw on their own experiences of harm and humiliation and seek to avoid inflicting similar harm and humiliation on the recipients of their good deeds. However, white people are rarely aware of the vulnerability of non-white people to racial discrimination because they do not generally share this vulnerability (see Sullivan 2006). A white donor who chooses a non-white person as the recipient of his charitable donation, and addresses that person with due respect, may not be aware of the structural racism underlying his choice of recipient, the assumption that a non-white person is likely to depend on charity for her or his subsistence. The fact that the recipient of his donation feels racially discriminated against may well escape the white donor. If someone who has observed how the white donor gives his charitable donation to a non-white recipient takes the liberty of reminding the donor of his racial prejudices, the donor may bluntly deny having any such prejudices. After all, he made a charitable donation, and he did so respectfully; if the recipient still felt humiliated, that is an unfortunate side effect; but it was unintended and therefore not the donor’s fault.
42Since the rise of the social movements such as “Black Lives Matter” and “#MeToo,” the victims of racial discrimination and sexual harassment have spoken out about how often and how much they have suffered. White people, and white men in particular, finally have an opportunity to learn how to avoid discriminatory behaviour when dealing with others, including non-white people and women. Kant would probably have approved of these movements. After all, he sees the communication of victims’ suffering as a legitimate ground for arousing “active and rational benevolence” as a reminder of the “conditional […] duty […] of humanity” (6: 456). Victims, however, are usually unwilling to speak out and demand the empathetic attention of others. This requires a courage that many of them do not have. And Kant would not be Kant if he did not set strict limits on the role that empathic attention to the suffering of victims can play for citizens and their moral duty to promote the happiness of other citizens. Empathy with the suffering of others that is not under the rational control of an agent should be avoided: “[…] there cannot possibly be a duty to increase the ills in the world and so to do good from compassion” (6: 457).
43I have asked whether Kant’s views on different human races and his defense of white supremacy have the potential to undermine the authority of his rationalist moral theory and its apparent claims to moral universalism and the equal moral worth of all human beings. To sharpen the question, I have distinguished between three ways of understanding racial beliefs and attitudes. Given Kant’s monogenism and his endorsement of white supremacy, we have good reason to attribute to Kant the view that non-white people, while fully human, are inferior in their rationality and their capacity to commit themselves to the moral law and are therefore dependent on white people’s tutelage and political government. But this view is nonsense, just as his claims about women and their inferior rationality are nonsense. We can reject these nonsense claims, open the category of rational agents to people of all ethnicities and colors, and to both men and women. Then we can conclude that these claims do not affect the authority of his rationalist moral theory. There is, however, a third kind of racism, structural racism, which prevails among the citizens of European countries, even though they endorse moral universalism and the equal moral worth of all people. I have argued that the rational and moral agent as Kant sees (her or) him is not immune to the effects of structural racism on (her or) his decisions and actions. At the heart of this argument is the claim that Kant gives the rational moral agent a monopoly on determining the meaning of (her or) his actions. Accordingly, an action that was not intended to be discriminatory cannot in fact be discriminatory – regardless of how the person chosen to be the addressee of that action felt about it. This monopoly is an essential part of the privileges only white rational agents enjoy; without it, white supremacy might not be possible.
44One question remains to be raised. Should we reject Kant’s moral theory and remove his ethical writings from our reading lists? My answer is twofold. On the one hand, we should recognize the limitations of Kant’s rationalist and non-consequentialist ethics of conviction; an agent who decides what to do according to the categorical imperative and thereby determines the meaning of her or his action is not necessarily inclined to pay empathic attention to the contingent but nonetheless real vulnerabilities of those whom she or he has chosen to address with her or his action. This agent has no reason to listen to their complaints and to critically reassess her or his action in their light. White people tend to overlook the vulnerability of non-white people to the pain of racial discrimination because they are unaware of the privilege of not sharing this vulnerability. Thus, and given the persistence of structural racism even among people who are committed to moral universalism and the equal moral worth of all people, we should prefer an ethics that conceives of the moral agent not only as someone who follows a moral law, but also as someone whose first duty is to avoid doing harm to others and to pay empathic attention to their contingent vulnerabilities. On the other hand, we should not leave Kant’s moral writings off our reading lists. Studying Kant’s ethics of conviction and its limits can help us to become aware of the practices of white supremacy and the racial prejudices that our actions reveal, prejudices that, moreover, we remain unaware of as long as we understand our moral duty as Kant would have it.