1. Introduction
The assumption, obviously, is that we have a principled, non-question-begging way to demarcate what is central from what is peripheral to [Immanuel Kant’s] philosophy, and a similarly principled way of showing how [his] racial views (and, of course, their implications) fail to penetrate this inner circle. (Mills Reference Mills and Valls2005: 175)
The Anglophone discussion of Kant’s view of race has its origins more than twenty years ago, and the important remarks of Charles Mills above remind us how this debate often turns on how scholars attempt to distinguish what counts as importantly ‘Kantian’ or not.Footnote 1 However, race is currently becoming one of the most prominent topics in Kant scholarship.Footnote 2 This is due, in part, to the fact that several of Kant’s writings endorse racist hierarchies, even well into the Critical period of the Groundwork. Consider published claims during Kant’s Critical period on Native Americans, who according to Kant are ‘incapable of cultivation [Cultur]’ and stand ‘far below the negro … who after all occupies the lowest of the remaining grades we have called racial difference’ (UTP, 8: 176).Footnote 3 As Kant states in his Physical Geography course, which he taught for four decades and which was published late in his life, ‘Mankind is in its greatest perfection in the race of the whites’ (PhyG, 9: 316). He writes in a parenthetical remark in unpublished Anthropology reflections, ‘[Native Americans] and Negros cannot govern themselves. Thus they are only good as slaves’ and in the body of the text that ‘All races will be wiped out … except for the white’ (Reflexionen zur Anthropologie, Refl 1520, 15/2: 878). While such claims in a mathematician might be viewed as mere personal prejudice, in a moral philosopher these statements cannot be easily swept under the rug (Elden Reference Elden, Elden and Mendieta2011).
While these racist remarks are not new to scholars, novel findings, such as the discovery of the infamous genocidal Reflexion 1520 and Kant’s apparent support for chattel slavery, both in his reflections on anthropology which he taught for decades to students, and as noted in his financial investments, make Kant’s racism much more visible. For example, even if Kant’s late Anthropology discussion of race appears devoid of the racism of his earlier writings, Steve Naragon’s scholarship on Kant’s investments late in life indicate that Kant most likely benefitted from racialized slave labour:
We know that about one-fourth of Kant’s invested wealth was with a local sugar refinery – the first in Königsberg – and because the raw sugar being processed in this refinery came from the West Indies, it was probably the product of slave-labor. It is also probable that Kant was fully aware of the nature of these investments, since he was on familiar terms with one of the owners of the refinery, the banker Friedrich Conrad Jacobi. (Naragon Reference Naragon2025c)
As a result, the incompleteness of the scholarship is taken advantage of by political actors.Footnote 4 The increased interest in the race debate is evidenced not only by the increased quantity and quality of scholarly publications on the issue but also by increased public attention to Kant’s reprehensible racial ideas among readers of the German and Anglophone public. To motivate our emphasis on the Kant-race debate and its increasing importance, consider recent op-ed publications in The Conversation and Medium, indicating that the debate has gained more traction as of late.Footnote 5
Despite the increased attention, scholars disagree about what to do with and about Kant’s racism. The guiding question in the literature now is not whether Kant was a racist (of course he was, at least for most of his life), but how Kantians are to make sense of his ideas in light of those racist views. Further, the debate about Kant’s views on race have become more complicated over the past few decades, and as the example above – as well as work from commentators like Huaping Lu-Adler (Reference Lu-Adler2023) – show, the simple question as to whether Kant was a racist has developed into more complex questions regarding the economic, social, and cultural impact of his writings.
While Kant scholarship is currently in the midst of a philosophical debate about how to treat his Rassenschriften – whether to treat them as essential to his philosophical project, as personal and easily dismissed prejudices, or as substantive philosophical beliefs he later altered – we suggest that underlying this disagreement is a more fundamental methodological disagreement.Footnote 6 As we attempt to show in this paper, there are value-laden methodological assumptions guiding the debate on Kant’s racism and its implications for his philosophy. More specifically, we claim that the demarcations between various classifications of textual objects given (and sometimes not given) by commentators – that is, what ‘counts’ as a legitimate source to substantiate a particular reading of Kant’s views on race in this debate – are often tacitly assumed rather than argued for.
This debate about the methodological choices commentators make that guide them in making sense of Kant’s writings on race is not hidden, but no one has offered an overview or taxonomy of the different methodological arguments and presuppositions at work behind the scene. The purpose of the present paper is not to resolve the Kant/race debate. Rather, we mean to show that assumptions within it have not been clearly stated, which has muddled the debate and prevents it from progressing. We coin the term ‘priority principle’Footnote 7 to articulate and categorize a set of methodological approaches that undergird much of the disagreements on Kant’s philosophy of race.Footnote 8 Though we focus on this striking topic in Kant scholarship, these priority principle methods are often tacitly employed by commentators on other historical philosophers as well. This includes scholarship on Nietzsche, Foucault, and Hobbes among others.Footnote 9 As a result, our taxonomy should not only assist those in Kant scholarship but also historically focused philosophy more generally.
It has been pointed out by Lu-Adler that some of the disagreements between scholars concern how they weigh and prioritize Kant’s texts, that is, how they take them to be evidence of a particular reading of Kant (Lu-Adler Reference Lu-Adler2023: 2–6). For example, should racist pre-Critical texts be de-emphasized in our reading of the late Anthropology, even if they were authorized by Kant for re-publication? These methodological disagreements about how to prioritize texts have deep implications regarding Kant’s philosophy of race, Kant scholarship, and philosophical methodology more generally. If, for instance, we only privilege the Critical ‘core’ (the three Critiques), Kant’s racism may be easily ignored. If we prioritize a specific time period, such as Kant’s Critical period, then Kant’s racist comments in Observations on the Feeling of the Beautiful and Sublime can more easily be ignored, but Kant’s 1788 ‘Teleological Principles’ must be dealt with. If we prioritize published over unpublished work, then we must give equal weight to, at minimum, the three essays Kant published on race and perhaps even his late Physical Geography. Overall, these methodological differences allow scholars to justify including or excluding certain texts as (il)legitimate. They therefore enable the attribution of different philosophies to Kant. Until these methodological problems are addressed out in the open, there is no clear path forward in philosophical discussion between methodologies. As a result, debate on specific issues, such as Kant’s racism, remains at an impasse. In short, before the philosophical debate can be resolved, we need to be clear about methodological presuppositions that inform those philosophical debates.
Charles Mills lays out the use of the priority principle on the topic of race and puts forward a suggestive critique. As Mills puts the problem:
[E]ither Kant’s racial views do not affect his philosophy at all (the extreme position),Footnote 10 or they do not affect it in its key/central/essential/basic claims (the more moderate position). The assumption, obviously, is that we have a principled, non-question-begging way to demarcate what is central from what is peripheral to his philosophy, and a similarly principled way of showing how the racial views (and, of course, their implications) fail to penetrate this inner circle. (Mills Reference Mills and Valls2005: 175 [emphasis added])
In the broadest sense, the first step of the priority principle, which we formulate below, presupposes we can demarcate between what is ‘core’ and what is ‘peripheral’ evidenceFootnote 11 in Kant. In recent scholarship, the core/periphery dichotomy is very much alive but is often referred to as the ‘quarantine approach’.Footnote 12 Put formally:
1.1. The priority principle
-
1. There is a clear and distinct demarcation between two types of evidence, core and peripheral.
-
2. Core evidence should be prioritized over peripheral evidence.
-
3. Peripheral evidence, if used at all, should only be used as a supplement to the complete evidence found in the core.
Propositions 2 and 3 can be found in a wide array of Kant scholarship, while 1 is an often unstated but necessary premise. To weigh in on which evidence is core and which is peripheral requires presupposing that the demarcation can be clearly made. Some commentators such as Lu-Adler have attempted to transcend the core/periphery binary of the priority principle. We suggest that it will be very difficult to do so without further clarity about how the core/periphery distinction is being used (and, indeed, it is still employed today),Footnote 13 and our proposed taxonomy will aid in this process.
In this paper, we want to challenge premise 1. There are unacknowledged differences in methodological presuppositions undergirding a veneer of agreement on key issues in Kant studies. While many scholars agree, for example, that writings from the Critical period should be prioritized in any sensible interpretation of Kant (in other words, a value-laden presupposition that texts from this period ought to be prioritized as important), the appearance of agreement conceals the fact (as we shall see) that they date the Critical period differently, and they consider what counts as a publication during this period differently as well. Given this, scholars may therefore in fact end up disagreeing substantially on key interpretative issues in Kant but for primarily methodological reasons.Footnote 14 The key issue we shall focus on in this paper concerns how commentators grapple with Kant’s writings on race, and how productively engaging in that debate requires that we be more careful about the value-laden nature of those methodological assumptions.
While some scholars have weighed in on the core/periphery debate (Mensch Reference Mensch2014, Reference Mensch2017), we are the first to explicitly articulate a taxonomy of these arguments. Indeed, while our emphasis is primarily methodological, exploring these variants and problems associated with them can itself be philosophically interesting, especially insofar as it reveals tensions and hidden commitments in the various substantive positions in that race debate.Footnote 15 The most commonly used methodological approaches, which we explore in section 2, employ the priority principle but use different demarcation criteria to separate core from periphery. This generates different sets of demarcated evidence. Then, in section 3, we home in on a text of interest in the Kant-race debate – namely the Physical Geography – to reveal tensions Kant scholars face as a result of differing applications of the methods discussed in section 2. We focus on the Physical Geography in particular because of its oddness in the race debate, and for its uniquely ambiguous status as a publication, being half authorized by Kant, half carelessly thrown together. Finally, we conclude by suggesting that a focus on individual texts provides a lesson for moving forward in the debate.
2. Kant scholars on race: a methodological difference
The tacit use of the priority principle has guided much of Kant scholarship. This is nowhere more apparent than on the topic of race, where methodological concerns are gaining traction (Lu-Adler Reference Lu-Adler2023: 2–6; Huseyinzadegan Reference Huseyinzadegan2018; Huseyinzadegan Reference Huseyinzadegan2022a: 653). Given different methodological approaches, for example, one scholar might hold Kant to be a consistent inegalitarian (a racist), while another may view him as a universal egalitarian (not a racist), and yet another that he is an inconsistent universalist (his racism is in tension with his philosophy). By choosing a methodology, one chooses the Kant one finds, one might say, ahead of time. It is therefore philosophically important that we concern ourselves with methodological questions. The priority principle is applied to the Kant-race debate in what is sometimes called the ‘quarantine approach’. The quarantine approach is a response to scholars such as Bernasconi, Eze, and Mills, who emphasize Kant’s racist comments (Bernasconi Reference Bernasconi and Bernasconi2001, Reference Bernasconi, Lott and Ward2002; Eze Reference Eze and Faull1994; Mills Reference Mills and Valls2005).Footnote 16 This approach isolates Kant’s racism, so it does not ‘infect’ his (core) philosophy. In general, scholars have not noticed that there are several distinct methodologies employed here.
The first application of the priority principle we wish to discuss attempts to demarcate and then prioritize what should count as Kantian philosophy ‘proper’. Robert Louden articulates the position that Kant’s views on race are regrettable ‘private prejudices’, but ‘Kant’s theory is fortunately stronger than his prejudices, and it is the theory on which philosophers should focus’ and ‘race seems weightless in Kant’s larger system’ (Louden Reference Louden2000: 105).Footnote 17 We refer to the interpretative strategy where commentators charitably prioritize and deprioritize elements of a philosopher’s writing on the basis of what they take to be his central philosophical principles as the living theory approach, that is, what commentators take to be the ideal form of the philosopher’s system or thought, charitably updated in light of current contexts and values so as to not be limited by the originator’s private prejudices.Footnote 18 We can formalize this position as follows.
2.1 The priority principle (living theory approach)
-
1. There is a clear and distinct demarcation between Kant’s living theory and his personal prejudice.
-
2. Kant’s living theory should be prioritized as evidence over his personal prejudices.
-
3. Evidence of Kant’s personal prejudice, if used at all, should only be used as a supplement to explain Kant’s living theory.
As Robert Bernasconi writes about this heuristic, ‘The “real Kant” apparently is not the historical Kant but, rather, the author only of his “central philosophical principles”. The real Kant is defined not by texts so much as by select ideas contemporary Kantianism finds valuable’ (Bernasconi Reference Bernasconi2003: 16). Which texts and principles represent the living Kantian Philosophy vary from scholar to scholar.Footnote 19 Those who employ this method often deprioritize the empirical writings, such as Kant’s remarks on anthropology, physical geography, and his personal prejudices. Micheal Wolff, for example, claims Kant’s Physical Geography is not ‘pure philosophy’ (reinen Philosophie) and can therefore be disregarded (Wolff Reference Wolff2020; see also Boxill Reference Boxill and Zack2017; Vial Reference Vial and Vial2016). This approach often results in the ‘marginalization of other works [including] … Kant’s philosophical essays on race’ (Sandford Reference Sandford2018b: 188). Alternatively, if a scholar wants to include a work usually considered marginal (like his anthropological works), they argue that that work is philosophically important and central to articulating his theoretical philosophy, sometimes even appealing to Kant’s own prioritization. Kleingeld, for example, writes ‘Kant himself saw his race theory as significant’ (Kleingeld Reference Kleingeld2007: 579; see also Wilson Reference Wilson2006).
The second application of the priority principle, which we call the periodization approach, holds that we should prioritize works written during Kant’s Critical period. For ease of discussion, we will stipulate a division of Kant’s periods as follows: Pre-Critical (1745–1770), Silent Decade (1770–1781), Critical Period (1781–1791), and Post-Critical (1791–1804).Footnote 20 Unless otherwise noted in brackets, our use of these periods refer to these years.Footnote 21 As we will articulate below, scholars’ disagreements (both tacit and explicit) on dating and populating these divisions have led to much confusion. This is particularly the case with ‘the Critical period’, since it functions as part of methodological assumptions. When it comes to the topic of race, Samuel Fleischacker summarizes the promise of focusing on the Critical period: ‘[O]ne might have hoped that the onset of Kant’s Critical period in 1781 would have led him to reject racism. It is Kant’s transcendental philosophy, not his early flirtations with moral sentimentalism, that we should expect to be egalitarian’ (Fleischacker Reference Fleischacker2023:7, 7n12). We formalize this kind of methodological presupposition as follows.
2.2 The priority principle (periodization approach)
-
1. There is a clear and distinct demarcation between the Critical works and the non-Critical works.
-
2. The Critical works should be prioritized as evidence over non-Critical works.
-
3. Non-Critical works, if used at all, should only be used as a supplement to Critical works.
In practice, this approach usually results in prioritizing the three Critiques (1781–1790), the Groundwork for the Metaphysics of Morals (1785), and The Metaphysics of Morals (1797),Footnote 22 as well as a handful of other texts that vary from scholar to scholar (Marwah Reference Marwah2022). The difficulty in defending Kant against accusations of racism with this approach is that Kant published a number of pieces on race ‘[d]uring the full bloom of’ and ‘well into the critical period’ (Hoffmann Reference Hoffmann2016: 63; Larrimore Reference Larrimore2008: 342n2; Marwah Reference Marwah2022: 629), which scholars say is ‘especially disappointing’ (Fleischacker Reference Fleischacker2023: 8–9). The debate, particularly spearheaded by Pauline Kleingeld, then concerns if, or when, Kant abandoned aspects of his theory of race. This debate develops into questions about whether Kant abandoned his racism in the early Critical period (read 1780s) (Muthu Reference Muthu2003: 182–5), ‘middle of the ‘Critical period’ (read: early 1790s) (Kleingeld Reference Kleingeld2007: 575), late Critical period (read mid/late 1790s) (Kleingeld Reference Kleingeld, Flikschuh and Ypi2014; 45), or after the Critical period (read 1798 forward) (Bernasconi Reference Bernasconi, Elden and Mendieta2011: 300).Footnote 23
An essential part of the approach employed in this debate, as seen in Kleingeld for instance, involves the dating of works containing racism one might normally think of as Critical or post-Critical, such as Physical Geography, as actually being pre-Critical (and thus illegitimate) (Kleingeld Reference Kleingeld2007: 586, Reference Kleingeld, Flikschuh and Ypi2014: 45n4). Alternatively, scholars who take his early work seriously argue that Kant’s early racist works, normally considered pre-Critical or from the Silent Decade, should actually be considered Critical or ‘proto-Critical’Footnote 24 because they influence or ‘infect’ the philosophy of the Critical period.Footnote 25
This approach has two significant problems that lead to confusion. First, there is little to no agreement about what years count as ‘Critical’. Some argue it starts as early as the late 1760s or 1770s,Footnote 26 and some suggest it extends to Kant’s death in 1804,Footnote 27 while we have stipulated 1781–1791 for clarity. While it may seem as if it is very clear whether a work is Critical, in practice, scholars disagree about the specific dates of each period but, since they often use ‘Critical period’, these disagreements go unnoticed. Thus, two scholars might appear to agree to prioritize the Critical period, but one scholar means 1781–1791 and another means 1760s–1804.
Further, even if there is agreement on the years of the Critical period, there is also substantial disagreement about how to date particular works.Footnote 28 For example, one might consider a work Critical if it was written during the Critical period but published much later. On the other hand, one might consider a work Critical if it was written early but published during the Critical period. It is even conceivable that scholars may argue that a work written in the pre-Critical period and published in the post-Critical period is actually Critical because Kant lectured on it or authorized it for publication during the Critical period. This lack of consistency means that even if scholars agree a work is racist, there are often disagreements about whether or not it should be considered ‘Critical’ because they date it differently. Kleingeld’s discussion of Physical Geography (1802) as being actually pre-Critical is a good example of this (Kleingeld Reference Kleingeld2007: 574n4).
The final approach most commonly employed in the literature argues that the core of Kant’s philosophy is what he actually published, and so the unpublished work represents the periphery. Lu-Adler writes about this approach (an approach she does not endorse), ‘whether one should follow the usual practice of privileging Kant’s published writings over his lectures depends on how one understands racism. If it is only a question of whether Kant is racist, then his publications should certainly serve as the more authoritative source of interpretation’ (Lu-Adler Reference Lu-Adler2022b; see also Marwah Reference Marwah2022: 616). We see an example of this in a recent article from 2024 where commentators state, ‘In Kant’s published works, discriminatory racial statements occupy a very small part of the text. This [means] he was not a racist’ (Govedarica and Roljić Reference Govedarica and Roljić2024: 85). We formalize this historical position on publication status as follows.
2.3 The priority principle (publication approach):
-
1. There is a clear and distinct demarcation between Kant’s published work and his unpublished work.
-
2. Kant’s published work should be prioritized as evidence over his unpublished work.
-
3. Unpublished works, if used at all, should only be used as a supplement to the published work.
If texts such as Physical Geography are considered published, then Kant is undeniably a racist.Footnote 29 However, if they are unpublished, or not ‘really’ published, then these racist claims can be seen as peripheral and more readily dismissed.
Thus far, we have identified the priority principle and three plausible sub-variants used in the literature. While it is possible for different methods to converge, it is more useful to demonstrate how they can render contrary judgements about particular texts. This should press us to investigate tensions between commentators who employ disparate methodological strategies.
3. Focusing on particular texts: Physical Geography
These three methods are employed in Kant scholarship, but why is it important to distinguish between them? After all, all three methods lead commentators to agree that there are central (core) and marginal (peripheral) texts and that we should prioritize the central over marginal texts.Footnote 30 However, interesting differences appear when we ask about which texts belong to which categories. Let us illustrate with a look into one text. When it comes to Physical Geography (1802), scholars disagree whether it is a legitimate Kant work and our articulation of methodological strategies can demonstrate why.Footnote 31 For example, some scholars using the living theory approach prioritize it, scholars using a periodization approach prioritize only some manuscripts from which it draws, and scholars using a publication approach deprioritize it altogether. So, we have a text where scholars fundamentally disagree about the place of Physical Geography in Kant’s work, and fundamentally this disagreement is methodological. We explore each of these in turn.
Some scholars using a living theory approach prioritize this text because it is essential to understanding Kant’s use of teleological principles (Clewis Reference Clewis2016), the relationship between metaphysics and experience relating to space and causal sequence (Cooper Reference Cooper2019; Zammito Reference Zammito, Eigen and Larrimore2006: 82–3), the relationship between geographical space, mental space, and cognitive maps (Elden Reference Elden2009; Richards Reference Richards1974), and recently even the Critical project itself (Costantini Reference Costantini2024). In fact, some scholars hold that Kant is claiming – in order to answer philosophical questions about metaphysics, morals, and religion – one must refer to anthropology (‘transcendental anthropology’), which itself finds its foundation in physical geography (Elden Reference Elden2009: 15; Frierson Reference Frierson2013: 11, 90; Larrimore Reference Larrimore2008: 361).Footnote 32 It is argued that the philosophical topography put forward in Physical Geography does not just figure into human knowledge but is ‘part of its very structure’ (Malpas and Thiel Reference Malpas, Thiel, Elden and Mendieta2011: 195).Footnote 33 David Harvey even argues, ‘Kant considered that geography (together with anthropology) defined the conditions of possibility of all knowledge’ (Harvey Reference Harvey2000: 534). Physical Geography is, therefore, not marginal but central to Kant’s philosophical project. Race plays a prominent role in Kant’s Physical Geography and is therefore important to explain his transcendental and Critical project, with particular focus on how his racial understanding of teleology informs his third Critique (Gray Reference Gray2012; Zammito Reference Zammito1992).Footnote 34 That is, counterintuitively, one can use the living theory approach to argue that, in order to understand Kant’s theoretical philosophy, one must prioritize Physical Geography as a legitimate and not merely marginal philosophical text.
Scholars using the periodization approach hold that we should prioritize the Critical period over earlier and later periods. The Physical Geography was published after the Critical period (1802) and the announcement for the course was published well before the Critical period (1757), so strictly speaking neither of these can be prioritized as Critical works. Both Kleingeld and Muthu date Physical Geography as pre-Critical or ‘early’ and therefore illegitimate as a post-1780s or 1790s Critical work (Kleingeld Reference Kleingeld2007: 575n4; Muthu Reference Muthu2003: 182–4). However, Kant did lecture on physical geography during the Critical period (until at least 1796), which suggests he may have endorsed such views throughout the Critical period (Clewis Reference Clewis2016: 318; Bernasconi Reference Bernasconi, Elden and Mendieta2011: 307). Stuart Elden even states, ‘Kant gave lectures on this topic from 1756 until 1796, and his comment that only he could read the text in 1798 indicates that he never ceased to revise it over that period’ (Elden Reference Elden2009: 7). Of the thirty-six known manuscripts relating to his physical geography course (many of which are student transcripts), five of these likely date from the Critical period,Footnote 35 two are still in existence,Footnote 36 and there are transcriptions of others. A periodization approach could prioritize unpublished manuscripts relating to Physical Geography during the Critical period while at the same time deprioritizing the pre- and post-Critical published texts relating to physical geography. One might prioritize the Dönhoff Manuscript ofPhysical Geography because, as Robert R. Clewis states, ‘[The Dönhoff manuscript] dates from 1781 or 1782 and hence belongs to the official Critical period’ (Clewis Reference Clewis2016: 329; see also Stark Reference Stark2009: 107–9). Thus, a periodization approach might deprioritize both publications relating to physical geography; one set because they were pre-Critical (e.g., Kant Reference Kant1775; see also Kleingeld, Reference Kleingeld2007: 574n4), and the other because it is post-Critical (Kant Reference Kant1802).Footnote 37 Those who use this approach may, however, prioritize those texts on physical geography during the Critical period such as the Dönhoff Manuscript.
The publication approach, by contrast, demarcates between the published and unpublished work and then prioritizes the former. Although Physical Geography was ‘published’ in 1802, Kant did not send this to a publisher but rather gave a number of manuscripts to one of his students (Friedrich Theodor Rink) to edit and publish under his name. Kant had, earlier, abandoned the idea of publishing it (Anth, 7: 122n) but eventually felt compelled to publish because a pirated or bootleg copy was published by Gottried Volmer in 1801 without his permission (Corr, 12: 372). As Erik Watkins in the introduction to the Cambridge edition states:
Despite his consistent and considerable attention to physical geography, however, Kant did not submit a manuscript to a press for publication, as was the case for almost all the works that are currently referred to as publications by Kant … Rink’s version of the Physical Geography … is neither a document that Kant himself wrote nor a reliable indicator of what Kant said in his class … Kant’s Physical Geography is truly Kant’s in a sense that is quite unlike that of any of his other publications given that there is no straightforward and unequivocal sense in which it can be taken to represent his actual views. (Watkins Reference Watkins2012: 434, 436)
In line with this, Werner Stark suggests we should think nothing of this publication, that it would have been impossible for Rink to make it into a ‘satisfactory publication’ (befriedigende Publikation) (Stark Reference Stark, Elden and Mendieta2011: 82; see also Stark Reference Stark and Bohr2019: 24). Stark explicitly states, ‘We know that Kant did not publish a book with the title Physical Geography. Not he himself, but other people produced two books with this title’ (Stark Reference Stark, Elden and Mendieta2011: 69 [emphasis added to ‘did not publish’]). Relatedly, Harvey remarks, ‘[a]ll manner of other excuses [for Kant’s racist claims in Physical Geography] can be manufactured: [such as] Kant never revised the materials for publication’ (Harvey Reference Harvey2000: 533–4). Other Kant scholars have argued that Physical Geography is not ‘really’ a publication by Kant. Otfried Höffe states, for example, ‘“Physical Geography” is not an autograph work by Kant’ (Höffe Reference Höffe2020). Wolff, in one of the strongest rejections, claims, ‘“Physical Geography” cannot be described as a work by Kant’ (Wolff Reference Wolff2020). As recently as 2024, Govedarica and Roljić claim, ‘Looking impartially at the facts surrounding the works that Kant published, especially the fact that he never published a book entitled: “Physical Geography,” we can conclude that this subject – and everything he lectured about in that course (including racist comments he reportedly made) – were not philosophically relevant to Kant’ (Govedarica and Roljić Reference Govedarica and Roljić2024: 82).
By focusing on a particular text, Physical Geography, we have seen how scholars might use different methods to come to different conclusions about Kant on race. The living theory approach can be used to prioritize Physical Geography, a periodization approach may reject the published versions of Physical Geography but prioritize those manuscripts dating to the Critical period, and the publication approach has been used to reject Physical Geography altogether. If it were simply the case that these three different methods – when applied to a particular text of Kant’s – generated three distinct recommendations of priority, this would not be particularly interesting or philosophically important. However, not only do they generate three different recommendations, but also three opposite recommendations, which any scholar should find very worrisome. We briefly explore these problems below.
3.1 Flipping the script: living theory approach
While many of the scholars listed above argue that Physical Geography is part of the ‘core’ of Kant’s philosophy, other scholars using the same living theory approach reach the opposite conclusion that it is merely peripheral. As Louden points out, ‘most scholars simply do not believe that Kant’s geography is as important as his ethics, logic, metaphysics, theology, and anthropology’ (Louden Reference Louden2014: 252). According to Harvey, many Kantians dismiss the geography because it is ‘irrelevant’, ‘not to be taken seriously’, or ‘lacks interest’ (Harvey Reference Harvey2000: 532). Harvey also states, ‘the content of Kant’s Geography is nothing short of an intellectual and political embarrassment’ (Harvey Reference Harvey2000: 532). Droit mentions that reading the geography ‘comes as a real shock’ because it is ‘an unbelievable hodge-podge of heterogeneous remarks, of knowledges without system, of disconnected curiosities’ (Droit Reference Droit1999: v). In this sense, one might see geography as simply practical and not theoretical, making up ‘The empirical detail that inhabits the categories of abstract thought’ (Elden Reference Elden, Elden and Mendieta2011: 9). In The Conflict of the Faculties, Kant seems to confirm this by not putting geography in the category of disciplines dealing with ‘pure rational knowledge’ but with mere ‘historical knowledge’, which is explicitly excluded from ‘pure philosophy’ (CF, 7: 28). We see these critiques pop up in Louden’s work where he writes, ‘The anthropology and physical geography lectures are thus not primarily intended as further contributions to Kant’s critical, transcendental philosophy program’ (Louden Reference Louden2000, 65; see also Wilson Reference Wilson2006: 20).
3.2 Flipping the script: periodization approach
One way to prioritize texts is based on whether or not those texts are considered Critical and therefore form the core of Kant’s philosophy. This method usually dismisses or de-prioritizes both pre- and post-Critical works. As Erik Watkins states, ‘the traditional line, encouraged by some remarks Kant himself made late in his life, is to dismiss his pre-Critical writings as pre-Critical, that is, as not reflecting his Critical position, which alone are of interest’ (Watkins Reference Watkins2003: 26). Relatedly, Kant’s post-Critical work is often rejected as ‘the product of Kant’s senility’ (Hall Reference Hall2015: 4). Dating where Kant’s senility started is a matter of debate but some, like Arthur Schopenhauer, discount even The Metaphysics of Morals (1797) as a work of senility. Schopenhauer states, ‘Kant’s whole theory of law is a strange tangle of errors, one leading to another, and he attempts to establish the right of property through first occupation. I can explain this only by Kant’s feebleness through old age’ (Schopenhauer Reference Schopenhauer and Payne1969: 336).
This priority based on periodization encounters three problems. First, scholars do not agree on the dating or division of periods, nor even the number of periods that should be used.Footnote 38 Second, even if they agree on period dates, they do not necessarily agree on which texts belong to which periods because they date them in different ways. Finally, even if scholars agree on the division of periods and dating demarcation criteria, they can still prioritize both the pre-Critical and post-Critical over the Critical period. This can either be a flat-out rejection that the Critical work is the most important, or, more contextually, that the best way to understand the Critical period is to prioritize one of the non-Critical periods.
Again, even if scholars agree we should prioritize the Critical work and agree on dating periods, they still disagree about which works belong to which periods. For example, Kant authorized parts of his Physical Geography course to be made into a text, Physical Geography, which was published in 1802. While it was published after the usual Critical period dating (1781–1791), it was based on lectures Kant gave throughout the Critical period. Physical Geography was based on two manuscripts. One was prepared by Kant in 1757–59 (sometimes referred to as an-rink 1 or Diktattext), while the second one was based on a set of student notes from 1774 (an-rink 2). Further, according to the Dohna Lectures (a late set of manuscripts compiled by students in Kant’s Physical Geography course), Kant’s presentation of race still contained essentialist racist claims as late as 1792. So there exists an ambiguity where scholars can date Physical Geography as a pre-Critical, Critical, or post-Critical text depending on which aspect one wants to emphasize.
Kleingeld argues we should date texts to when they were written and prepared by Kant or more specifically to when they can be ‘regarded as reflecting the views Kant held’ at that time (Kleingeld Reference Kleingeld2007: 574n4). Kleingeld, therefore, dates Physical Geography (1802) as a work not reflecting the views he held in the 1790s because ‘quite a few [specific passages] go back to the pre-Critical period’ (ibid.). Physical Geography (1802) was edited and compiled by others in the 1790s and Kant was not involved, so it does not date from the Critical period. She then claims, if Physical Geography is not a text from the 1790s, then there is no evidence of racism in the mid-1790s (which she considers Kant’s middle Critical period). Part of her evidence is that she dates the Anthropology (1798) to the late 1790s and claims it does not contain any racism where the older manuscripts from the courses do. The Anthropology (1798) also has a textual origin tracing back to Kant’s Physical Geography course.Footnote 39 However, unlike Physical Geography (1802) which was just old texts re-edited by others, Kant prepared the manuscript for the Anthropology himself in the 1790s. Kleingeld claims that ‘in the published version of the Anthropology there is no discussion at all of the supposedly different “characters” of the races’ as evidence that ‘Kant had given up his description of the different races as having very different “characters”’ (Kleingeld Reference Kleingeld2007: 590). This is simply false, since in the published Anthropology (1798), we find the following in a note: ‘Because of his inborn lifelessness, the Carib is free from this arduousness [of boredom]. He can sit for hours with his fishing rod, without catching anything; thoughtlessness is a lack of incentive for activity…’ (Anth, 7: 233n). This ‘inborn lifelessness’ of particular peoples is certainly a racial feature. However, based on Kleingeld’s own criteria of priority, she cannot deprioritize Anthropology, because it was written by Kant himself during the late 1790s and can therefore be ‘regarded as reflecting the views Kant held’ at that time (Kleingeld Reference Kleingeld2007: 574n4). Kleingeld specifically uses Anthropology (1798) as an example of his non-racism, despite the problematic note quoted above.Footnote 40
Finally, if the exact date of the Critical and non-Critical periods are agreed upon and how to date each work is agreed upon, this does not completely solve the problem. The periodization approach also has a built-in instability. Even if everyone agrees we should focus on the Critical period, this does not mean we must prioritize texts from the Critical period. Some scholars argue that it is ‘only’ by understanding the post-Critical work that we can understand the Critical work because the post-Critical work is an elaboration’ of the Critical work (Hall Reference Hall2015: 1–10). Others point out that it is only by understanding Kant’s pre-Critical work that we can understand their anticipation of the Critical work (Schönfeld Reference Schönfeld2000). Some argue that aspects of the pre-Critical work are philosophically and scientifically interesting for their own sake, even in relation to Kant’s views on race,Footnote 41 and not of mere instrumental value for making sense of the Critical work (Schönfeld Reference Schönfeld2000: 114, 123–4). Finally, a few suggest the Silent Decade is perhaps ‘The most important period of Kant’s intellectual development’ (Washburn Reference Washburn1975: 167). So even if a scholar agrees that the Critical work is the most important, they sometimes prioritize non-Critical work like Physical Geography over the Critical work to better articulate Critical work.Footnote 42
3.3 Flipping the script: publication approach
Let us look briefly at how someone using the publication approach might use different demarcation criteria to argue that Physical Geography (1802) is a legitimate text that should be prioritized. The range of criteria used by scholars within each approach is complex. Within the publication approach, for instance, we observe a number of different sub-demarcation criteria scholars consider. These include texts that were actually published or published during Kant’s life time, or were authorized by Kant, or involved Kant in editing, outlining, etc., or were in Kant’s own voice; and in each case there is the matter of degree to which they were made public. Oddly enough, there can be substantial disagreements even with a single sub-demarcation criterion. There is enough substantial disagreement within the authorization criterion, for example, to provide sub-sub-demarcations including: legally authorized, intentionally authorized, written/authorized, dictated/authorized, edited/authorized, and reviewed/authorized.
In distinction to Govedarica and Roljić, Höffe, and Wolff who all submit that Physical Geography should not be prioritized as published, some scholars hold it should be given a place because it was authorized. As Luise Fischer claims, ‘Since Kant authorized and mandated Rink to produce (and publish) the latter edition, Rink’s is considered (more) official’ (Fischer Reference Fischer2012). Elden states that the Rink edition of Physical Geography is ‘The semi-authorized version’ (Elden Reference Elden2009: 6). The argument here is, because Kant legally authorized Rink to publish Physical Geography, it is in some sense a legitimate publication. Bernasconi’s position, as summarized by Eduardo Mendieta, ‘relies on the fact that Kant allowed work that relied on early notes, lectures, formulations contaminated by racism to go to press, including anthropological writings and the authorized Rink edition of the Physical Geography’ (Mendieta Reference Mendieta2013). Lu-Adler also argues, ‘The fact that Kant legally authorized Rink to edit the Physical Geography … and had it published under his own name sufficed to establish its perceived authenticity’ (Lu-Adler Reference Lu-Adler2023: 7). Lu-Adler’s prioritization focuses on how Kant played a role in the public formation of racist ideology and therefore perceived legal authorization is enough to consider them published. The fact that Physical Geography was legally authorized and was perceived as intended to be public Footnote 43 is sufficient to consider this work published. Bernasconi also briefly argues that Rink’s Physical Geography can be used legitimately for ‘discussions of Kant’s influence’ (Bernasconi Reference Bernasconi, Elden and Mendieta2011: 298).Footnote 44
The publicness of Kant’s physical geography lectures of 1792 is brought up by Bernasconi, who writes: So ‘far as I am aware this lecture course [Physical Geography] is the only occasion where we know that Kant publicly condemned the slave trade, and he did so in the context of a reassertion of his hierarchical view of race’ (Bernasconi Reference Bernasconi, Elden and Mendieta2011: 304). In contrast, Marcus Willaschek – who also prioritizes Physical Geography (1802) – suggests that even if its authenticity [Authentizität] is disputed, the statements made therein are still in Kant’s ‘voice’ (a Voice demarcation criterion) (Willaschek Reference Willaschek2020). While some scholars claim that Kant’s lectures are not in his voice (Franzel Reference Franzel2013: 4), Willaschek claims Physical Geography is in his ‘voice’ because his racist claims are paralleled in his other publications (Willaschek Reference Willaschek2020). Thus, three different scholars – all using the publication approach and who come to the same conclusion – have different methodological reasons for their conclusions because they are using different demarcation criteria. Clearly, then, it is paramount for future commentators, especially those discussing Kant’s views on race, to carefully clarify and justify their interpretative criteria.
To summarize, the three methodological approaches we have discussed all employ the priority principle. Further, by using different strategies, scholars can come to radically different conclusions about whether to prioritize a text. Even when considering a single text and a single methodological approach, conclusions can diverge immensely given different demarcation criteria. One scholar can claim that Physical Geography is not a text by Kant at all. Another commentator, by contrast, can claim the same text to be authentically published. And lastly, even among those who agree on which texts are published, some may emphasize why it is published (Authorization, Publicness, Voice, etc.).
4. Further implications
The point we hope to make here is simple: the widespread employment of these approaches in Kant studies contain a mess of unstated disagreements and tacit pre-theoretical values. These tacit values are akin to unstated premises and inevitably lead to disagreements among scholars if we do not first turn our attention to our methodological presuppositions. If we do not first turn our focus to these methodological problems, even scholars who agree we should ‘prioritize the core works’ will inevitably disagree for methodological reasons. Our suggestion is that some of these unstated methodological assumptions, values, and presuppositions become apparent when we focus on particular texts. However, this can have unexpected and indeed paradoxical results. Namely, a work might be both theoretically central and marginal, both Critical and non-Critical, and, as we have noted and further demonstrate below, both published and unpublished.
Let us conclude with an example of how commentators could refine the publication approach in future research. Resolving disagreements over ‘publication’ is not simply a matter of getting better necessary and sufficient conditions of a text being published. ‘Published’ means many things in many different contexts. The simple demarcation between published and unpublished is confounded by Kant’s incredibly complicated textual history. For a few clear examples, consider Kant’s public defences (New Elucidations [1755], Physical Monadology [1756], Inaugural Dissertation [1770]), a public address Kant gave in either 1786 or 1788 (‘On Philosophers’ Medicine of the Body’ [1788]), a variety of public declarations and announcements that were published but are sometimes considered correspondence (Corr, 12: 359–72), a privately published letter to the mother of Kant’s student (Friedrich von Funk [1760]), Kant’s letter published (with some degree of authorization) as an appendix in Soemmerring’s ‘On the Organ of the Soul’ (1796), Kant’s ‘On the Propensity to Fanaticism and the Means to Oppose it’ which appeared anonymously in a book by Kant’s biographer Borowski, the partial publication of the first introduction to the third Critique by Kant’s student Beck (‘Erste Fassung der Einleitung in die Kritik der Urteilskraft’ [1794]), and his address ‘Concerning Sensory Illusion and Poetic Fiction’ given in 1777 which was actually handwritten on the published dissertation of Johann Gottlieb Kreutzfeld upon which Kant’s comments were delivered publicly.
Are these published? These are just a few examples where commentators in Kant studies, without clearly defined methodological principles and criteria, run into interpretative issues which inevitably muddle their debates. The more one digs into Kant’s textual history, the more one finds examples of texts which, according to one demarcation criterion, could be considered published but from another, unpublished. These cases, or counterexamples, can help us sharpen, clarify, and understand our often tacit value-laden methodological commitments and better understand our disagreements with others. Things are not so simple, and this is a problem for those making sense of Kant’s ‘views’ on race, which is the case study of the present paper.
5. Conclusion
The point we hope to hammer home is that, as scholars of canonical philosophers, we need to be more explicit about our methodological approaches to the texts we use. In the recent work Kant, Race, and Racism: Views From Somewhere, Huaping Lu-Adler has done an excellent job articulating her own methodological demarcation criterion including why and how those demarcations are important to her project on Kant (Lu-Adler Reference Lu-Adler2023). This has led to an interesting and productive scholarly debate about the justifications of methodology itself, which in many ways pushes us beyond the old core/periphery binary of the priority principle. Kleingeld (Reference Kleingeld2024), later on in this development, is not only defending her own substantive position in the race debate but also actively engages in critical methodological work to reject the methodology proposed by Lu-Adler. Lu-Adler has recently responded with further methodological clarifications (Reference Lu-Adler2025), including a methodological distinction between ‘core Kantian notions’ (quoting Kleingeld), and seems to endorse the racial ‘ideological core’ approach (quoting Baumeister) that passes through Kant.Footnote 45 Lu-Adler has effectively abandoned the core/periphery dichotomy focused on texts by an individual but has re-inserted a different core/periphery dichotomy with the ‘ideological core’ (populated by those texts that had a social impact in the construction of racist ideology). Whatever one’s position on Lu-Adler’s thesis, her self-reflexive methodological practices are exemplary and she has made significant progress in critiquing the old binary.Footnote 46 At the same time, as we have tried to show above, that core/periphery binary is alive and well among many Kant scholars, and part of the reason has to do with a lack of methodological clarity. Indeed, a core requirement of philosophical debate is clarity in methodological presuppositions.
As scholars, we can be clearer and more concise with regard to our methodology and demarcations. Although we do not necessarily advocate this here, we offer two possible suggestions that may help identify instabilities in the core/periphery dichotomy which could encourage scholars to transcend it.
Some methodological approaches are simply more appropriate than others to answer certain kinds of value-laden questions, and this should be expected.Footnote 47 There simply is no meta-methodological (or Kuhnian extra-paradigmatic) rule to determine once and for all which value-laden inquiries are best. Having multiple explicit and competing methodologies is indicative of healthy scholarly discourse. So one approach that may be fruitful could involve articulating conditions for an adequate methodological pluralism, though we lack the space to consider this further.
Another approach, which we believe to be promising to advance the Kant-race debate, but again we lack the space to defend here, involves focusing on texts individually, in conjunction with scholarly archival workFootnote 48 that treats texts explicitly as individual material and historical objects.Footnote 49 Although useful, even ‘texts’ and ‘editions’ are scholarly shorthand, which refer to critical editions, which themselves contain value-laden editorial decisionsFootnote 50 and often represent more than one single textual object or divide single objects into pieces.Footnote 51 For example, Kant’s personal copy of the Observations with his annotations about slavery were sent to his publisher on 18 September 1800.Footnote 52 Certainly, the Observations are published, but what about the annotations? Did Kant want them published? The 2012 edited collection of Kant’s Observations and Remarks focusing on Kant’s 1764 edition and his annotations is a good example of focusing on an individual textual object, but we could further ask: what reasons are there for these annotations to be prioritized or deprioritized and what values inform those reasons? By focusing our attention on individual physical manuscripts, the reasoning and contradictions between priority criteria become more obvious. Doing so may lead to a re-thinking of a presumed single core/periphery distinction.
Finally, consider that a recent work on the Opus Postumum pairs facsimile scans of primary documents with transcriptions as part of the ‘Opus Postumum Online Edition’.Footnote 53 Kant often wrote on top of published documents in the Opus Postumum. We can then ask: should these material documents be considered published or unpublished? Perhaps both in different ways? (See Figure 1.)
Figure 1. Screenshot of Opus Postumum, Convolut 1, pg 1. Opus Postumum Online Edition (Accessed 17 September 2025. (https://telota.bbaw.de/kant_op/edition.html#/C01/001))
Scholars analysing specific documents often find that publication history is more complex than our critical editions and methodological categories can easily handle. We tentatively suggest that by focusing on material objects in conjunction with critical editions, future commentators may bring to light tacit methodological assumptions, shortcomings, contradictions, and instabilities that may encourage us to transcend our methodological dogma of a single correct core/periphery distinction
In the final analysis, the point we want to make is not ‘which of these texts should scholars use?’ Rather, we wish to emphasize that methodological demarcation criteria are important yet rarely stated explicitly. Even if hermeneutics is more of an art than a science, hermeneutics is also guided by certain methodological considerations, and we should be suspicious of any meta-theoretical framework that prioritizes one kind of inquiry if it has not deeply considered its own methodological presuppositions. Importantly, we cannot have a productive hermeneutics until we articulate the methodological problem. What appear to be substantial philosophical disagreements are often artificially sustained by unacknowledged methodological disagreements that prefigure which ‘Kant’ we find, one might say, in advance. If your methodological choices only give you certain texts, then you can only get certain philosophies out of them. Thus far, we have often been too methodologically dogmatic. We hope the taxonomy of methodological approaches offered in this paper will inform new directions, not only for the curious case of Kant on race but also for Kant scholarship more generally.
Acknowledgements
We thank two anonymous referees for their helpful suggestions on how to refine our arguments. We are also grateful for feedback and encouragement from participants at the European Consortium for Political Research (ECPR), Kantian Political Thought Standing Group Panel, ‘Science and Racism – Kant’s “Anthropology” and the Need for Reorientation of Kantian Scholarship’ in Dublin of 2024. Finally, we thank Sam Badger for his suggestion regarding the constitutional originalism analogy.
Open access
