The Taxonomy of ‘Race’ and the Anthropology of Sex: Conceptual Determination and Social Presumption in Kant

Abstract

This chapter argues that the different treatment of the concepts of; ‘race’ and ‘sex’ in Kant’s philosophy shows how the first was a philosophical problem that required justification while the latter was—and usually still is—taken for granted. It shows how, through his writings on race, Kant defended an historical conception of the natural genus. Refusing the natural distinction between ‘genus’ and ‘species’, Kant’s philosophical determination of the concept of ‘race’ then attempts to introduce, below these, a new terminal category in zoological taxonomy that justified, for him, the positing of significant classificatory differences between the human ‘races’ even if they all belonged to the same natural genus/species. In contrast, the concept of sex or gender (Geschlecht) is never subject to any philosophical determination but treated as an unquestioned basis for the justification of social and political differences between men and women.

This chapter aims to clarify the difference between the theoretical status of the concept of Rasse (race) and that of Geschlecht (sex) in Kant’s philosophy, and to consider the significance of that difference for feminist critique. Until the late-eighteenth century the dominant meaning of Geschlecht was still linked to ancestry, but it also meant ‘race’ (or ‘people’), ‘generation’, ‘genus’ and ‘sex’. Kant was influential in disambiguating these meanings, insisting on more specific words for most of them: Volk (people), Rasse (race), Gattung (genus) and Geschlecht (sex or gender).¹ Most important, for Kant, was the philosophical ‘Determination of the Concept of Race [Rasse]’ (the title of an essay he published in 1785). This chapter will show that the context for Kant’s philosophical determination of the concept of race is the problem of the status of systems of natural classification, or the problem of justification of a natural system of nature—that is, the problem of the justification of the usage of the (hitherto logical) terms ‘genus’ and species’ as natural taxonomical categories (categories that name fixed levels or ranks within a hierarchical system of the classification of nature). The central argument in this chapter is that Kant attempts to address this problem through his theory of the human ‘races’, and that, especially in the 1770s and 1780s, this theory aims to establish ‘race’ as a natural, taxonomical category. In so doing Kant introduced a new terminal category in zoological taxonomy; that is, he added ‘race’, as the lowest rank, to the hierarchical taxonomy (including class, order, genus and species) that Linnaeus had recently established. Thus, although Kant maintained (contra the polygenecists of his day) that all humans comprised one genus, and that the ‘races’ were not separate species, his introduction of the sub-specific concept of ‘race’ as a taxonomical category still justified, for him, the positing of significant natural, ‘biological’ differences between the human ‘races’.

In contrast, the concept of sex or gender (Geschlecht)² was not and could not be a taxonomical concept for Kant (as indeed it is not and cannot be for modern biology—as a classificatory category it is of a quite different kind to the categories of ‘genus’ or ‘species’, for example). Further, unlike ‘race’, ‘sex’ is never subject to any philosophical determination in Kant’s work. Unlike his discussions of ‘race’, which are explicitly philosophical and aim to contribute to the life sciences of his day, Kant’s discussions of Geschlecht are—like his discussions of national characteristics—part of his empirical anthropology, and they concentrate on the psychological and social characteristics of men and women. Whereas the concept of ‘race’ requires and is given a philosophical and scientific justification, ‘sex’ is assumed as a natural given that goes without saying. So what is the significance of the different theoretical statuses afforded to the concepts of race and sex in Kant, and of what relevance is this to us today?

The chapter begins with an account of the problem in natural history that provides the context for Kant’s theory of race—the problem of the relation between the logical and the natural categories of genus and species—and shows why neither the transcendental logic of the first critique nor any transcendental principle of purposiveness (in the first and third critiques) are adequate to the solution of this problem. Part Two shows how the development of a solution, in the form of a natural category of the genus, was proposed by Kant in his essays on race, and argues that the result of these essays is the postulation, for the first time, of what was, for Kant, a properly taxonomical and natural concept (‘race’) for the classification of human difference. Part Three contrasts this with the anthropological treatment of the differences between the sexes, demonstrating the extent to which ‘sex’ is presumed as given rather than the subject of any philosophical or other theoretical treatment. This chapter suggests, in conclusion, that the different discussions of ‘race’ and ‘sex’ in Kant demonstrate that the different conceptual and philosophical, as well as social histories of these terms must be a part of any adequate analysis of their contemporary manifestations. ‘Intersectional’ analysis quite rightly attempts to acknowledge and understand the relations between different axes of oppression and discrimination in the effort to avoid ‘race’ and ‘gender’, for example, being seen as “mutually exclusive categories of experience and analysis.”³ But this does not mean that the different histories of different types of discrimination are not important, and to understand these different histories we need to understand the different histories of the concepts on which they are based.⁴

1 From Logic to Nature: The Problem of the Natural Genus

Part of the scientific context of the Critique of Pure Reason and the critical philosophy more generally is the problem of the status of systems of natural classification. In the seventeenth and eighteenth centuries various ‘methods’ for the classification of the natural world were proposed, mainly and most influentially in botany, but also sometimes extending to all of the kingdoms of nature. Botanists and other naturalists aspired to add to the local catalogues of flora and fauna universal or global catalogues that required systems, rather than just ad hoc methods, of classification. Although almost everyone agreed that all systems of classification were, at some level, artificial, they were all motivated by the ideal of a natural system and sought to approximate it as far as possible.

For Kant—and in many histories of natural history still—the difference between ad hoc, artificial methods of classification and a natural system is exemplified in the different approaches of Linnaeus and Buffon. Linnaeus’s ‘sexual system’ of the classification of plants and his general zoological taxonomy picked out specific, static characteristics of organisms as the basis for classification. This is a practical and logical but abstract and, in some sense, arbitrary method—Kant called it a ‘school system’ for memory (ODR, Ak2: 435). Buffon, on the other hand, attempted to base his natural historical studies on what he called ‘physical truths’, on the real relations obtaining between individuals of the same species. Most famously, this gave rise to the idea that members of the same species were to be identified as such if they could produce fertile offspring—what can be called the ‘interfertility criterion’. Buffon thus introduced an historical, genealogical dimension to natural history and Kant was clear that only in this way could the real, natural relations between organisms be determined. To a great extent, much of Kant’s philosophical work on living nature (as opposed to planetary systems and the forces of nature) can be seen as an attempt to provide a thorough philosophical grounding for Buffon’s work in natural history, according to Kant’s conviction that scientific knowledge is systematic knowledge—that is, that there can be no purely empirical science. Effectively, the idea of a natural system needed to be justified in such a way that its logical, classificatory categories were at the same time natural (or ‘real’) categories—that is, the relation between the logical and the ‘natural’ categories of ‘genus’ and ‘species’ needed to be justified. This is well known. What is less well-known but crucial to understand Kant’s approach to the problem is the central role that would be played by the category of ‘race’ and its distinction from and relation to the categories of genus and species. Although the word ‘race’ had been used before, Kant was the first to provide a philosophical, systematic and (as we would now say) ‘biological’ theory of race, as a category of natural classification.⁵

The more general terms of the problem of the natural system were inherited from Aristotle. The intimate relationship between ontology and logic in Aristotle determines this, for him, as a problem of definition. A species is defined through its genus (what it shares with its congeners) and its differentia (what marks it out from them). The ‘species’, ‘eidos’, is also for Aristotle the individual form, and the definition of the essence of the individual thing and of its species are the same.⁶ The aim of definition is to arrive at what commentators call the ‘infima species’, the lowest species—that species which is not a genus for anything else.⁷ Only when we have arrived at the (infima) species, can we give a true definition of something, which is not just an explanation of the meaning of a word but tells us what it is to actually be that thing, what the ‘essence’ of that thing is. This is the origin of the idea that the ‘species’ is the basic, that is the lowest, unit in modern taxonomy (sub-divisions like ‘variety’ or ‘form’ having no taxonomical status).⁸

When Aristotle deploys the logic of genus and species in his zoological works, the definitions of various animals (including what we now call mammals, birds, fish, some insects, and invertebrates like worms) are implicit in his descriptions (he rarely gives actual definitions) and they are not contentious because they follow the common names (just as the folk designation of common animals—horse, cow, dog, and so on—is not contentious today). Aristotle groups these kinds of animals in different ways, according to what is being investigated in the larger groupings, which means that particular animals might fall into different groups at different times. For example, when the topic is blood, the sanguineous animals are distinguished from the non-sanguineous animals, which would put humans and chickens in the same (sanguineous) group. But when the topic is generation the oviparous animals are distinguished from the viviparous, which would separate humans and chickens. Even so, ‘chicken’ or ‘human’ remain as ‘infima species’, to the extent that they cannot be further subdivided into natural groups (although they can be divided in other ways, for example, into tame and wild; male and female). These subgroups may become the topic of philosophical problems for Aristotle (as the distinction between male and female does)⁹ but the species group remains the unit of essential definition.

However, when natural history subsequently took as its task the classification of living things to a level of detail that would go beyond and indeed contest folk taxonomies, the differentiation and the identification of species could no longer be taken for granted. This problem first arose in relation to the classification of plants, and for the naturalists of the sixteenth to the eighteenth century (the period of the first scientific attempts at classification) it could be characterized as an empirical problem. Arguments tended to concern the identification of the salient characteristic or characteristics upon which a system of classification ought to be based. Because of the sheer number of known species, the genus became the most important category in the search for a natural system. Although Andrea Cesalpino, for example, had asserted the centrality of the genus in classification in the sixteenth century, it was Linnaeus who cemented the idea both in his theoretical statements and in his binomial system of classification, still in use today, giving a species its genus name first: “Every genus is natural, thus created in the very beginning; hence one should not arbitrarily and on account of some theory or other rudely split it or join it to another one.”¹⁰ According to Linnaeus, in 1756, the ‘character’ of a genus is its definition. Its ‘essential’ character provides the most ‘proper’ and ‘peculiar’ feature of the genus.¹¹

But the problem remains, how to know the ‘essential’ character? In 1696 the naturalist John Ray had said that “[t]he correct and philosophical division of any genus is by essential differences. But the essence of things are unknown to us. Thus, in place of these essential characters, characteristic accidents should be used.”¹² For Ray and many naturalists after him it was the ‘fructification’ parts of the plant that would be the basis for the identification of the genus. Linnaeus sought to make this method of classification as complete and unequivocal as possible by breaking down the ‘fructification’ parts analytically into 31 elements. Many of these were based on intuitive distinctions (for example that between flower and fruit) and easily identifiable parts of plants (at least for the plants that he was observing at that time). But, as Scott Atran says: “there is no explicit account given in Linnaeus’s many philosophical aphorisms of the metaphysical principles that justify his assertions that all genera are natural, that each genus is determined by a unique fructification, and that the fructification characters of genera constitute the foundations of a true science of botany.”¹³ That is, despite the intuitive, ‘folk-biological’ plausibility of many of Linnaeus’s divisions and the usefulness of his system of the division of plants into genera, there is no philosophical justification for these intuitions. The logical genus-species relation has become fixed in what are now thought of as the ‘natural’ categories of genus and species. The binomial designation of each species is a statement of one such relation. But in the absence of any philosophical justification the suspicion remains that it still might be a merely logical, artificial system, albeit a very useful one. This is an important part of the context in which the problematics of Kant’s first and third Critiques arise (as discussed below). When the logical terms ‘genus’ and ‘species’ are, as it were, embedded in nature, becoming categories of nature, the relation between logical form and empirical content becomes a philosophical problem, or requires philosophical explanation and justification, even if natural history can and did progress without a solution.

This philosophical problem is addressed in Kant’s lectures on logic, which present Aristotle’s account of the logical genus-species relationship almost unchanged. Kant makes the same point repeatedly: “A conceptus communis is called genus [das Genus, die Gattung] in regard to the concepts that are contained under it, but species [Kant uses the Latin term] in regard to the concepts under which it is contained. Thus, e.g., man is genus in regard to males, etc. [;] but in regard to rational being, man is again species. … The conceptus infimus … will be a species that is no longer a genus” (LO, Ak24: 240).¹⁴ Each time Kant stresses that the designation of a conceptus communis as a genus or a species is relative; that is, the same thing may be either genus or species, depending on its relation to what he calls its ‘superior’ or ‘inferior’ terms: “Genus and species [Kant uses the Latin terms] do not in themselves make any distinction among concepts, then, but only in the relation of concepts” (LO, Ak24: 911). This is how Aristotle uses the terms both in his logic and in his zoology.

In the Blomberg Logic (from the early 1770s, so predating the development proper of the critical philosophy), having noted the relativity of genus and species, Kant also says that the

subordination of concepts … can occur both logice and realiter. Logical subordination consists in the fact that I take that which is common to many concepts and thereby form for myself a universal concept, under which I can subordinate the individual representations. In this way I make for myself various genera [Latin] and I subordinate the species [L] and individua to them. Real subordination, however, consists in the fact that I actually combine concepts with one another, so that not only is one contained under the other, but instead they also cohere as cause and effects. (LO: 208/Ak 24: 260)¹⁵

What, if anything, is the relation between, on the one hand, the distinction between logical and real subordination of species to genera and, on the other hand, the distinction between general and transcendental logic in the Critique of Pure Reason? This question suggests itself because the problem of the relation between the logical and the real looks very much like the kind of problem that the transcendental analytic is designed to solve.

From the standpoint of the critical philosophy, we might say that ‘logical subordination’ belongs to pure general logic: “a canon of the understanding and reason, but only in regard to what is formal in their use, be the content what it may, (empirical or transcendental)” (A 53/B 77).¹⁶ General logic gives rise to a “merely logical criterion of truth” (A 59/B 84) concerned with the validity of the merely logical form of cognitions, without regard to their truth content. General logic deals with logical subordination in the form of the syllogism, regardless of the truth of the propositions involved.

Transcendental logic, on the other hand, is concerned with those concepts that can be related to objects a priori, and which indeed make it possible for there to be objects for us. These are, of course, the pure concepts of the understanding, or the categories. As the causal relation is categorial it is not a great stretch to say that the ‘real subordination of concepts’—in which, as Kant says, “not only is one contained under the other, but … they also cohere as cause and effects” (LO, Ak24: 260)—has for its condition the elements of transcendental logic, and in particular the categorial relation of causality. But the problem is to understand the ‘real’ subordination of species to genera in a system of nature; that is, the possibility for that subordination to be known as more than a merely logical relation between genera and species that we “make for ourselves”, for it to be something like objective knowledge of relations between determinate species and genera. Transcendental logic only affirms that such knowledge is a real (not just a logical) possibility, but the concepts of genera and species are not—and cannot be—thereby determined as real, natural categories.

Towards the end of the transcendental dialectic and in its Appendix, Kant discusses the genus-species relationship explicitly, in the context of the justification of the regulative use of the ideas of pure reason, a use which allows the merely contingent aggregate of the understanding’s cognition to become systematic, which is to say properly scientific. Here he suggests that the concepts of natural substances (pure earth, pure water, pure air) are ideas of pure reason that allow students of nature [Naturforscher] to reduce all of the various materials to these—in effect—generic classificatory categories (A 645–646/B 673–674). Similarly, reducing the various laws and principles of nature to more general laws and principles (the ‘parsimony of principles’) is not merely a principle of economy for reason but “becomes an inner law of its nature.” The logical principle of rational unity among rules presupposes a transcendental principle “through which such a systematic unity, as pertaining to the object itself, is assumed a priori as necessary” (A 650–651/B 678–679).

This presupposition is also evident (although implicit) in “the principles of the philosophers”: “That all the manifoldness of individual things does not exclude the identity of species [Art]; that the several species must be treated only as various determinations of fewer genera [Gattungen], and the latter of still higher families [Geschlechtern], etc; that therefore a certain systematic unity of all possible empirical concepts must be sought insofar as they can be derived from higher and more general ones: this is a scholastic rule or logical principle, without which there could be no use of reason” (A 651–652/B 679–680). In reducing, for example, “all main salts to two main genera [Hauptgattungen]”, reducing the several species [Arten] of earth to fewer species with the idea in mind that they might reduce to a single genus [Gattung], it is presupposed that “the unity of reason conforms to nature itself” (A 653/B 681)—that is, that these are not just logical genera of our own making, but objectively real or natural genera:

If among the appearances offering themselves to us there were such a great variety … that even the most acute human understanding … could not detect the least similarity … then the logical law of genera would not obtain at all, no concept of a genus … would obtain. … The logical principle of genera therefore presupposes a transcendental one if it is to be applied to nature … According to that principle, sameness of kind [Gleichartigkeit] is necessarily presupposed in the manifold of possible experience … because without it no empirical concepts and hence no experience would be possible. (A 653–654/B 681–682)

The logical principle of species [Arten], which prescribes to the understanding that it should seek variety despite agreement of things under the same genus [Gattung] must similarly presuppose its transcendental counterpart (A 654/B 682).

To the extent that the transcendental principles of genera and species are presupposed not just logically, but as being in conformity with nature, they might seem to speak to the problem of the real or natural genera and species in a system of natural classification, where ‘genus’ and species’ and indeed ‘subspecies’ would cease to be merely relative logical terms and would name instead taxonomical ranks. But this is not the case. The transcendental principles, although they make the logical relation of genera and species possible, they only do that. They make it possible for us to seek relations of genera and species in nature and to produce classifications—indeed this is necessary for science—but they do not, and in principle cannot, help us to identify whether any such system of classification is the natural system. Thus, despite Kant’s sophisticated philosophical apparatus, he is in the same position in 1787 as Ray in 1696 and Linnaeus in 1756. The system is really a method in all but name. The categories of genus and species are still without any philosophical justification as properly natural, taxonomic categories.

This problem is addressed head on the in the Critique of the Power of Judgment, in the context of the identification of the peculiarity of reflecting (as opposed to determining) judgment. Reflecting judgment requires a principle: “The principle of reflection on given objects of nature is that for all things in nature empirically determinate concepts can be found” (CJ, Ak5: 211). Here Kant also explicitly explains that this is not a merely logical principle but the “condition of the possibility of the application of logic to nature”. It is “a principle of the representation of nature as a system for our power of judgment, in which the manifold, divided into genera and species [Gattungen und Arten], makes it possible to bring all the natural forms that are forthcoming to concepts (of greater or lesser generality) through comparison.”

This is distinguished from the “transcendental system in accordance with a priori concepts (the categories)” because with its principle the reflecting power of judgment “must further assume for this purpose that nature in its boundless multiplicity has hit upon a division of itself into genera and species [Gattungen und Arten] that makes it possible for our power of judgment to find consensus in the comparison of natural forms” (CJ, Ak5: 213). The principle of reflecting judgment (that is, the principle of purposiveness) makes it possible, for Kant, to distinguish in principle between a merely logical classification (the ‘school system’ for memory) and what he variously calls a physical or a natural system. But when Kant says that nature “has hit upon a division of itself into genera and species [Gattungen und Arten]” there is still no justification to understand ‘genus’ and ‘species’ as natural taxonomical categories, and concerning the possibility of a natural classification the move from the logical to the real subordination of species to genus still remains to be made. For in this case it is not enough to say that the scientist can act ‘as if’ these are natural categories, because this is not enough to distinguish an artificial from the natural system (the user of the artificial system operates precisely ‘as if’ their categories were natural). The lack of explanation or justification for the real subordination of species to genus, and the problem of the identification of the natural genus, is visible in the gap between the epistemic principle of purposiveness (the condition of possibility for systematic, scientific knowledge) and the ontological specificity of living beings. Kant’s commitment to the idea that living beings, which cannot be accounted for with mechanistic explanations, are each ‘natural ends’ (CJ, Ak5: 370–376) acknowledges this ontological specificity. But it does not say anything about the natural relations between living beings and thus cannot on its own help with the problem of the identification of the natural genus.

2 Kant and ‘Race’ as a Taxonomic Category

To see the concept of genus, at least, being used in an unambiguously natural or taxonomic way, with an accompanying justification, we need to look outside of the Critiques, to Kant’s essays on race. These essays are explicitly concerned with the problem of the natural classification of the human. Kant’s aim in all of his discussions of race, as he makes increasingly clear, is to determine the nature of the ‘classificatory difference’ [Klassen-Unterschieden, or klassichen Verschiedenheit] (DCH: 146)¹⁷ between the races in a natural, rather than merely logical way.

The 1775/1777 essay “Of the Different Races of Human Beings” opens with the distinction between the ‘school division’ and the ‘natural division’ of genus [Gattung]. “The natural division into genera and species in the animal kingdom is grounded on the common law of propagation, and the unity of the genus is nothing other than the unity of the generative power.” Thus, it is a version of Buffon’s rule that defines a physical (that is, natural) genus: the genus is an historical category (all members are offspring from the same line) and its members are interfertile. (Kant’s shift from the concept of ‘species’ to ‘genus’ here is important, as we shall see later.) As Kant puts it (using a different terminology than Buffon) the natural division “concerns [genealogical] stems (Stämme), which divide the animals according to relationships in terms in generation”. Accordingly, all human beings belong to the same natural genus [Naturgattung] because they consistently beget fertile children with one another, no matter what other differences there are between them (ODR, Ak2: 429, 430).

In the essays on race and works in anthropology and physical geography, the distinction between the logical and natural forms of classification and of genera map broadly on to a distinction between description of nature (giving rise to logical, school systems, exemplified for Kant by Linnaeus) and natural history (proceeding from the principle of purposiveness, giving rise to a natural system). In “Of the Different Races of Human Beings” Kant writes that description of nature describes things as they are now, whereas history of nature would explain what they once were and how they arrived at their present state:

Natural history, which we still lack almost entirely, would teach us about the changes in the shape of the earth, likewise that of its creatures (plants and animals)[,] that they have undergone through natural migrations and the resultant deviations [Abartungen] from the prototype of the stem genus [Stammgattung]. It would presumably trace a great many of seemingly different species [Arten] to races [Rassen] of the same genus [Gattung] and would transform the school system of the description of nature, which is now so extensive, into a physical system for the understanding. (ODR, Ak 2:435)

In the ‘school system’ the logical concepts of genus and species have come to designate natural groups, but they are not yet the concepts of a natural or real species or genus. The unity of a natural genus (the natural belonging together of its members, rather than their logical assembly) has only one possible ‘natural cause’, as Kant says in “On the Different Races”: their common descent, which he sometimes refers to as their having a common stem (Stamme). For Kant, an animal genus does not contain different species because the very meaning of ‘different species’, for him, is difference of origin, which means, precisely, belonging to a different Stammgattung (ODR, Ak2: 430). Here Kant’s text is indeed confusing, because while it seems to distinguish terminologically between genus (Gattung) and species (Art) the definitions of the two terms are effectively the same. Nothing that comes from a different genealogical stem (where ‘stem’ is not a technical term, but rather a indication of ancestry¹⁸) can be called different in genus or species, meaning that the difference between the concepts of genus and species is difficult to determine. However, Kant characteristically uses Gattung where we might have expected Art.¹⁹ For example, the unit of special creation is the genus:

For animals that are so different from each other that just as many different creations would be necessary for their existence might indeed belong to a nominal genus [Nominalgattung] (in order to classify them according to certain similarities) but never to a real genus [Realgattung], which absolutely requires at least the possibility of descent [Abstammung] from a single pair. (DCR, Ak8:102)

In order to explain how the single genus of humankind is divided into, not different species, but different races, Kant speculates that organisms contain ‘germs’ or ‘seeds’ (Keime) and ‘natural predispositions’ (natürliche Anlagen), which determine the development of the organism. Any given species will contain, in some sense, all of the germs the expression of which determine all possible variations of that species. However, only those germs are developed which furnish the species with the characteristics needed to flourish in the environment in which it finds itself. Members of a species which have migrated to colder climes, for example, will develop the germs (if they have them) for shaggy coats, whereas those that remained in the warmer parts will not (ODR, Ak2: 434). The differences expressed in the development of these different germs may give the impression of new species, but they are in fact, different ‘variations’ [Abartungen] and ‘races’ [Rassen] of the same genus. This is also true of human beings:

The human being was destined [bestimmt] for all climates and for every soil; consequently, various germs and natural predispositions had to lie ready in him to be on occasion either unfolded or restrained, so that he would become suited to his place in the world and over the course of the generations would appear to be as it were native and made for that place. (ODR, Ak2: 435)

According to Kant, as humans spread across the earth they formed different groups in which different germs developed in response to the difference climates. The development of these germs in response to the environment somehow ‘suffocated’ those germs that did not—and could not now—develop. These groups, according to Kant, can be divided into the four ‘races’ of the human genus [Gattung]; all the classificatory lower varieties can be derived from these (ODR, Ak2: 432). In some English translations Kant’s use of the term Gattung here is rendered as ‘species’, but this obscures one of the main points in his theory of race. Kant uses the term Gattung (genus) to refer to the taxonomic level at which all humans are unified and uses the term Rasse (race) to refer to the taxonomic level at which they are differentiated. This means that there is, in Kant’s work, no place for an intermediate ‘species’ division in the natural history of the human. (In “Determination of the Concept of a Human Race” he writes that the different classes (Klassen) of the human genus may only be called ‘races’ (Racen) and not ‘species’ (Arten): “The class [Klasse] of the whites is not distinguished from that of the blacks as special species [Art] within the human genus [Menschengattung], and there are no different species [Arten] of human beings” (DCR, Ak8: 99–100)²⁰).

It is important to remember that Kant’s theory of germs and natural predispositions, which many commentators take be his main contribution to the life sciences of his day, is developed specifically to produce a natural classification of the human genus, that is, its division into four distinct ‘races’. Thus, it is the project of the natural classification of the human genus that gives rise to the account of the natural concept of a genus, an account which can then be generalized to the whole of the animal kingdom. This project is intimately related to the critical philosophy, even though the reception of Kant’s work has tended to isolate his essays on race from his philosophy, assigning them to the category of anthropology.²¹ The central point in Kant’s explanation of the relationship between stem genus and variation or race is the requirement to use a teleological principle of judgment. This is most explicit in the last (1785) essay on race, as the title, “On the Use of Teleological Principles in Philosophy”, suggests. In this essay he replies to some of his critics by emphasizing that the teleological principle is the most important aspect of his claims about race, and thus that it is in relation to this that the success or otherwise of his claims must be judged.

The crux of the matter, for Kant, is this. The single most important heritable characteristic that, for him, determines the classificatory difference of the races is skin color. Skin color is, he claims, the only characteristic that is unfailingly hereditary, the only thing that we unfailingly inherit from both of our parents. In the 1785 essay this becomes the claim that race itself is an “unfailing hereditary peculiarity which justifies the division into classes” (TelP, Ak8: 165). Although the development of certain germs and predispositions in humans, leading to the development of the different races, most probably had, as its efficient or occasioning cause, the different climates into which humans migrated, the “predispositions to all this classificatory difference must have lain necessarily in the germs of a first single stem, so that the latter would be suitable for the gradual population of the different regions of the world” (DCR, Ak8: 98). These first predispositions, implanted in the one stem are purposive; the germs develop purposively, (TelP, Ak8: 168, 175–176)²² just as any living organism develops purposively, as the third Critique later teaches. To insist on the teleological ground of explanation in relation to race poses no fresh burden on us because, according to Kant, we already have to suppose that organized beings are purposive as regards the “preservation of their species” (TelP, Ak8: 169). Mere empirical observation of the differences between human beings will not give rise to a natural, taxonomic category: “what [we] need in order to decide whether there is a real or merely a nominal affinity among the creatures will not present itself to [us] on its own” (TelP, Ak8: 164). Thus for Kant ‘race’ is not a merely empirical concept but one well-grounded in reason, proposed to explain a hereditary peculiarity with the postulation of a purposive cause, that is, an end (TelP, Ak8: 163).

Extrapolating from this, it must also be the case that a natural system of genera and species could also not be grounded merely empirically. The construction of a natural system finds its logical possibility in the transcendental principles of genera and species (from the first Critique). According to Kant, the natural division within this system concerns genealogy “which divide the animals according to relationships in terms in generation” (ODR, Ak2: 429). As we have seen, Kant relies heavily on Buffon’s interfertility criterion to determine the meaning of ‘genus’, and equates belonging to the same genus with common ancestry. How to decide, he asks, whether the wolf, the fox, the jackal, the hyena and the house dog, which are so many (logical) classes of four-footed animals, are so many different species or rather different races of the same genus? That is, how do we make this logical division into a natural division? The answer is: if they share the same ancestry then they belong to the same genus and are races within this genus, but if one must presume for them each a different ancestry then they are different species (DCR, Ak8: 99–100).²³ We must presume for them a different stem-genus because they are not interfertile; therefore they are different species, not different races.

However, it is a peculiarity of the origin of Kant’s theory in the problem of the human ‘races’ that his account focusses not on the natural relation between genus and species, but the relation between genus and that variation or subspecies (Abartung) that in relation to human being he calls a ‘race’.²⁴ Because difference in species is defined as a relation to a different ancestral genus, we can gather that the relation between a natural species and a natural genus must similarly be a relation of descent. But as Kant’s definition of a natural (or physical) genus is Buffon’s fertility criterion Kant cannot provide a separate, determined definition of a species, and indeed he more than once claims that, in natural history (as opposed to description of nature) genus and species “are not distinguished as such” (DCR, Ak8: 100).²⁵

But of course, ‘genus’ and ‘species’ do not mean the same thing for non-Kantian natural history and Kant’s conflation of the terms is precisely what allows him to determine the concept of race as taxonomic category. It is now ‘race’, and not ‘species’, that is the terminal category of zoological taxonomy; that is, ‘race’ is now the lowest level of significant ‘biological’ classificatory difference.²⁶ The true mark of this terminal category is not interfertility, or being able to fulfil the natural end of generation—that is the mark of a natural genus. The true mark of a race is the unfailing inheritance of a purposive peculiar characteristic. For Kant, wolf, fox, dog and so on would be different species because (not being interfertile) they have different stem genera. In principle, these animals could also be divided into classificatory races, in Kant’s sense, although he does not ever suggest it himself.²⁷ The monogenetic argument of Kant’s classificatory theory affirms the unity of the human genus and denies that the races are different species. This is an important rebuttal to the idea of polygenesis (the idea that the different ‘races’ are separate, and specially created, species of humankind). But Kant’s commitment to monogenesis (the belief that humans belong to the same natural genus) does not exclude the specification of a new, for him properly taxonomic category—the terminal taxonomical category of human being—on the basis of which he claims a strong classificatory and natural difference: ‘race’.²⁸

3 Kant and the Social and Moral Anthropology of Sex

The case is quite otherwise with the distinction between the sexes. None of Kant’s many discussions of women appear in his writings that aim to contribute to the emerging life sciences; rather, they belong to his anthropological speculations and his moral and political writings. Of course, these are all linked; but in none of Kant’s writings do we find any attempt at a scientific account of sex difference, directed at a scientific community.

Kant’s most extended comments on woman per se, and the constellation of sex-race-nation, are to be found in his anthropological writings.²⁹ Here his discussions of race and of sex do share one important feature. As groups of natural, living entities, men and women are to be understood first and foremost as natural ends (rather than mechanically explicable phenomena). Kant’s discussions of living beings elsewhere (in the third Critique, especially) emphasize purposiveness in the natural development of individual organisms, a development which, he believes, cannot be explained merely mechanically. But his discussions of women (men as a group are not half as much discussed)³⁰ are concerned not with the development of each individual according to the internally purposive principles of the organism but with women’s ‘purpose’ in another sense, in relation to the species—their reproductive raison d’être. To some extent all living creatures have reproduction as a ‘purpose’ in this sense, but when it comes to women this purpose is, for Kant, determining.

The distinction between the purposive unfolding of an individual organism and the individual’s purpose in relation to the species is related to the distinction, explained by Jocelyn Holland, between the concepts of Zeugung (generation of a new individual, which unfolds according to its inner purpose) and Fortpflanzung (reproduction of the species).³¹ Kant discussed women’s purpose (or nature’s purpose for women) primarily in relation to the latter. In the Lectures on Anthropology (Friedländer) it is said that women “are only made for preserving the species” (LAn, Ak25: 675). In Anthropology from a Pragmatic Point of View, the task is defined as the investigation into what the “free-acting being makes of himself,” rather than the physiological investigation of “what nature makes of the human being” (An, Ak7: 119). But in its discussion of ‘woman’ Kant says that “one can only come to a characterization of this sex if one uses as one’s principle not what we make our end, but what nature’s end was in establishing womankind” (An, Ak7: 305).

This appears to rule woman out of the purview of pragmatic anthropology entirely, and Kant’s remarks on ‘character’ seem to confirm this. In the twofold presentation of the human being, “what the free-acting being makes of himself’—the proper subject matter of pragmatic anthropology—is concerned primarily with ‘character’ as a moral attribute, or as the individual’s capacity for morality. ‘Character’ consists in what is ‘peculiar to the higher powers of the human mind”. It depends not on “drives and desires, but rather solely on the manner in which he [the human being] modifies these”, how the human being uses his powers and talents (LAn, Ak25: 437–438). Character is the subjective rule of the “higher faculty of desire,” manifest in the ability to follow objective (moral) rules (LAn, Ak25: 438) it is “the employment of our power of choice to act according to rules and principles” (LAn, Ak25: 630). This is contrasted with ‘temperament’, which is “the proportion of sensible feelings and desires … With temperament we do not act in accordance with principles and dispositions as with character, but according to inclinations” (LAn, Ak25: 636). Accordingly, an unfortunate temperament can go together with a good character, to the extent that the latter is precisely the ability to master the former. According to Kant, “[w]ith human beings, character is the main thing” (LAn, Ak 25:648). However, in women it is “not exactly in keeping with their nature to have a character at all” (LAn, Ak25: 631). Women may have good sentiments, but this can never be ‘brought up’ to good character; for her feelings of honour must take the place of principles (LAn, Ak25: 1170).

Kant also says that, as well as the reproduction of the species, woman’s purpose is “the cultivation of society and its refinement” (An, Ak7:306). This is only evident in ‘the refined nations’ where woman is to be seen in her ‘developed state’, (LAn, Ak25: 699)³² where the ‘art’ in the feminine nature allows woman to rule over the strength of the masculine. This at least acknowledges a social and properly human role for woman, but it is still not ‘character’. Woman’s ‘art’ is a natural disposition towards learned behavior, compensation for her lack of power. (LAn, Ak25: 697–698)³³ And as Inder S. Marwah has argued, woman’s lack of moral character (having her ends determined by nature rather than freedom) in Kant’s descriptions of her proper virtues and capacities is not merely empirically conditioned but required by Kant’s account of humanity’s moral advancement. Imperfect human beings must be awoken to their moral orientation (one is not born with but must develop receptivity to the moral law) and women’s role is to civilize and ‘moralize’ the men whose moral character will thus be enabled to overcome even the worst temperament.³⁴

Although, properly speaking, women have no character, Kant can and does describe their characteristics. This is part of ‘anthropological characterization’ more generally—as Cohen says, “an analysis of human varieties according to four criteria: person, sex, nation and race.”³⁵ ‘Character’ (what the human being makes of themselves) is part of the larger ‘characteristic’, which includes the discussion of natural aptitudes and temperaments (sanguine, melancholic, choleric and phlegmatic), in relation to all persons, to men and women, to nationalities (or ‘peoples’, Volk), and to races.³⁶ In this respect there is a certain commonality of treatment of sexes, nations and races and—the theory of race being very largely absent from Anthropology From A Pragmatic Point of View and the Lectures on Anthropology—the impression is easily given that ‘sexes’, ‘nations’ and ‘races’ are the same kinds of category for Kant. But if we ask what, in each case, is the support for or the determination of each of these concepts, their status varies considerably. In fact, strictly speaking, of these only ‘race’ has any status within Kant’s philosophy as a determined concept. The precise meaning of and justification for the scientific concept of ‘race’, in the context of natural history and the emerging life sciences, stands in stark contrast to the presumption of the vernacular categories of ‘sex’ and ‘nation’, even if the restriction of the use of ‘Geschlecht’ to name the former is historically notable.³⁷ Kant recognizes that the concept of ‘race’ needs to be justified as a natural category; that justification having been completed (as far as he is concerned) the roles of the different races in his account of the development and progression of humankind stands on a theoretical basis (albeit we can see that it is both flawed and objectionable). But the category of ‘sex’ (Geschlecht)—its use in the anthropological writings is not such that it deserves to be called a concept³⁸—is merely presupposed as an item of common linguistic currency and reference.

Robert Bernasconi has argued that by the last decade of the eighteenth century the word ‘race’ itself was frequently associated with Kant when it came to determining its precise sense; indeed, “[t]he Kantian lexica show that, at that time, the word race had come to be seen as Kant’s word—a technical term that required explanation.”³⁹ Bernasconi shows that Mellin’s Encyclopädisches Wörterbuch der Kritischen Philosophie (1799) and the one-volume summary published while the Encyclopedia was being prepared (1798) contained long exegeses of Kant’s concept of race, marking it out as (for Mellin) one of the most important and original of his concepts. Krug’s general philosophical dictionary (1827 and 1833) also devoted considerable space to it, or even associated it as a philosophical concept exclusively with Kant.⁴⁰ In contrast, Geschlecht in the sense of sex is completely absent as a philosophical term in these dictionaries. Unlike ‘race’, ‘sex’ could be—and was—taken for granted. The characterization of the sexes and the identification of their specific roles, qualities and virtues needed no grounding in a determination of the concept of ‘sex’ itself. ‘Sex’ appears in Kant’s work, and elsewhere, as a natural and social given. And it is not just sex difference itself that is presumed, but all of the sex-specific capacities and virtues and the sex-specific social and political statuses that follow from them.⁴¹

This seems to show that in the period of Kant’s lectures on anthropology (1772–1789), and in their edited collation into Anthropology from a Pragmatic Point of View (1798), Kant could still take much for granted concerning traditional ideas about women and gendered attributes and roles (or bourgeois European ones, at least). Lettow has argued that it is “only when social change and the politics of equality—most obviously in women’s activism and criticism during the French Revolution—undermined traditional gender arrangements and hierarchies [that] the bodily differences between women and men became the object of a new cultural concern.”⁴² It is only after this that a more determined conception of a ‘natural’ biological sexual difference was articulated—a conception that would then be available as a justificatory reference point for claims about women’s ‘natural’ social and political place. Increasingly, a biological conception of natural sex difference could function as the basis for arguments for traditional arrangements, just as a biological conception of ‘race’—the origins of which lie in Kant’s justification of it as a natural taxonomic category—could function as the basis for racist and colonial arguments.

But the fact stands that in the late eighteenth-century Kant could take ‘sex’ for granted when the same could not be said of ‘race’. Although they both come to be articulated in biological terms there are important differences in their histories. Whereas the concept of ‘race’ was relatively new as a way of conceptualizing human diversity in the eighteenth century, ‘sex’—the word used to designate the distinction between male and female—was in use in its original Latin form by the second century CE and the alleged exclusive duality of ‘the sexes’ was taken for granted centuries before Kant. This is seen, for example, in the history of the controversies over whether plants have male and female. Whether the sex of plants was seen as an analogy with animals that made their being-male or being-female merely metaphorical, or whether that analogy was the basis for the claim that plants were literally sexed, the animal (including human) model for sex—what was taken to be the mere fact of sex difference—was an unquestioned ‘given’.

‘Race’—the category and the phenomenon that it allegedly named—had become a problem requiring philosophical and scientific justification by the 1770s because colonialism and the racist justification of the slave trade were already a political problem in the mid-eighteenth century. If ‘sex’ lagged behind in this regard it is because the challenge to traditional gender arrangements was not yet a political problem in the same way or to the same extent (it did not claim men’s attention in the same way). But even when the problematic of sex entered into the life sciences and philosophy the basic presupposition—that the duality of sex difference, the distinction between male and female, is a natural given—remained. ‘Sex’ in this sense is never justified as such either in Kant and the subsequent idealist and other philosophical traditions, or in the social and life sciences. ‘Sex’ is not a philosophically or otherwise theoretically determined concept, but the presumed basis for a politics of social-sexual differentiation, the presumed basis for various physiological and psychological claims about the differences between men and women and what stems from this, and the presumed basis for the metaphysics of sexual difference, the latter often cosmological in scope. The concept of ‘race’ in Kant’s sense emerged, became controversial, was debated and today has been rejected as a category of philosophical natural history—a way of accounting for human natural diversity.⁴³ ‘Race’—its postulation, justification and criticism—was played out as a topic in and for natural history and cognate areas, especially geography and anthropology. ‘Sex’, on the other hand, was simply taken as given, although what could be inferred or what followed on from this basic given would become the topic of extensive debate in medicine and psychiatry especially.

When it comes to the contemporary criticism of ‘race’ and ‘sex’, then, these different histories and the different standings of the concepts must be taken into account. The different theoretical status that these two concepts enjoy in Kant’s work is just one example of this. Robert Bernasconi has argued that in his theory of race Kant “was not primarily attempting to offer a classificatory system, although that became a preoccupation of later generations.”⁴⁴ But Kant’s argument does, in the end propose a new taxonomical framework, replacing ‘species’ with ‘race’ as a category of specifically human diversity and effectively proposing the latter as the basis for a slew of other kinds of difference—moral, political and intellectual. Although contemporary racism may not appeal directly to this ‘biological’ conception of racial difference it is an ineliminable part of the history of the concept of ‘race’ and continues to exert its influence in popular race thinking, despite there being, in fact, no significant biological reality corresponding to the idea of ‘race’.⁴⁵ The critique of the concept of race that is at work in contemporary racism is thus also the critique of the biologization of ‘race’ as a taxonomic category, because the ‘biology’ of racial taxonomy was always a disreputable sleight of hand. There is, on the other hand, some significant biological reality to the idea of sex—remembering that ‘sex’ is found across the animal and plant kingdoms. But this biological reality, even if we accept that it always refers to a distinction between ‘male’ and ‘female’, is not that of an exclusive and fixed duality. Rather, the biological reality of sex across the different kingdoms of life (mammals representing only a tiny fraction of this) is demonstrably one of exuberant sexual diversity—an array of different hermaphroditisms and frequent and multiple sex changes. The critique of the concept of sex grounding contemporary sexism is not the critique of this biological reality, but of the reduction of biological reality to a fixed and exclusive duality which is taken to determine the social and psychological being of men and women. This reduction and its presumed determining function biologizes the political categories of ‘men’ and ‘women’. But the critique of this biologization need not be a critique of biological reality per se.

This research was made possible by a Major Research Fellowship from the Leverhulme Trust (‘Sex Difference in Natural History’, MRF 2017-006). I am very grateful for the editors’ comments on a first draft.