The apparent kinship between Stoic and Kantian ethics is often observed by philosophers – albeit more often in passing than with full scholarly attention.Footnote ¹ Philosophers considering Kant’s relation to ancient ethics typically take Aristotle to be the most obvious and fruitful point of reference, not the Stoics. This default orientation may be partly attributable to the Aristotelian influence on contemporary virtue ethics: to think about virtue, at least since John McDowell’s (Reference McDowell1979) ‘Virtue and Reason’, has been to think through Aristotle in the first instance. Contemporary Kantian ethics has been influenced by these developments, especially as it has moved away from narrowly deontological concerns.Footnote ² The default orientation may also be partly attributable to the simple fact that Plato and Aristotle figure in the foundation of any philosophical education in our own time, whereas Hellenistic philosophy remains more of a specialist interest.Footnote ³ Although in recent years we have seen more scholarship tracking the legacy of Hellenistic philosophy in early modern philosophy, inquiry on the Stoics’ significance for Kant is only just beginning to gain traction.Footnote ⁴

Even if ancient philosophy for us may consist chiefly of Plato and Aristotle, the same does not follow for Kant. Histories of philosophy produced in the seventeenth and eighteenth centuries followed the models of ancient doxographies, according ‘Stoicism … its proper place as one of the ancient schools of philosophy with no prejudice against it’ – which we see, for example, in Jakob Brucker’s five-volume Historia critica philosophiae (Leipzig 1742–4), among other works (Ierodiakonou Reference Ierodiakonou1999:3). Moreover, we find various signs that Kant’s thoughts turn, in the first instance, to the Hellenistic schools, and especially the Stoics, when ‘the ancients’ are adduced. One example of this can be found in his endorsement of the ‘ancient’ division of philosophy into physics, ethics, and logic at the start of the Groundwork (4:387): the division is not obviously found in Plato or Aristotle, nor is it even generically Hellenistic. Rather, it is distinctively Stoic.Footnote ⁵ Another example, of course, is that Kant has Epicurus and the Stoics represent opposing conceptions of eudaimonia as the highest good in the ‘Antinomy of Practical Reason’ of the Critique of Practical Reason (5:115ff.). Moreover, a considerable amount of Stoic thought was presupposed in the German rationalist tradition, and Kant’s immediate intellectual milieu. For example, providential natural teleology in a recognisably Stoic vein was commonly presupposed in long-running German debates about human progress and the ‘destination’ or ‘determination’ of the species (die Bestimmung des Menschen), particularly in contributions from Christoph Wieland, Moses Mendelssohn, and Kant himself.Footnote ⁶

Yet such facts about Kant’s intellectual climate do not entail that he would have simply absorbed his Stoicism by intellectual osmosis.Footnote ⁷ Nor should we assume, as some have, that Kant would have been reliant on German translations of classical works, such as Christian Garve’s rendering of Cicero’s De officiis.Footnote ⁸ Kant excelled in Latin at the Collegium Fridericianum, where students read a lot of Cicero, including De officiis in the sixth year; Kant also, of course, wrote his early work in Latin, and gave formal addresses in Latin at least as late as 1786.Footnote ⁹ Nor should we adopt Allen Wood’s (Reference Wood, Denis and Sensen2015b:120) dim assessment of Kant’s knowledge of ancient philosophy, which he supposes to have come largely second-hand, through histories such as Brucker’s and ‘Cicero’s Latin popularisations’. Several points must be made in reply. First, a diet of pure Cicero would not be so meagre for an education in Hellenistic philosophy: Cicero remains, for us, a central and indispensable source of Hellenistic philosophy.Footnote ¹⁰ Second, the diet was not pure Cicero. As a student, Kant and two friends supplemented the school curriculum by reading classical texts outside of class (Kuehn Reference Kuehn2001b:48–9). Reinhold Jachmann, Kant’s friend and biographer, testifies that Kant had ‘fully absorbed the entirety of Greek and Roman classical literature’, and continued to read classical texts – especially Roman – with ‘much relish’ into old age (Groß Reference Groß1912:137–8).

Kant not only was formed on the study of classical texts, but returned to them throughout his life. Jachmann reported that Kant studied Seneca in particular ‘for the purpose of his practical philosophy’ during the final years of his teaching (Groß Reference Groß1912:138) – that is, from about 1790 onwards. We may reasonably suppose that Kant’s Stoicism was a Roman one, which he accessed largely through the direct study of Latin sources, such as Cicero and Seneca. What about Epictetus? The extent to which Kant encountered Stoicism through Greek sources is somewhat less clear:Footnote ¹¹ Jachmann remarks that Kant ‘appears not to have studied all of the Greek works in the original’ (Groß Reference Groß1912:138; my emphasis). Yet students at the Collegium read the entire New Testament in the fourth and fifth years of study, before being introduced to classical Greek texts in the sixth. Thus Kant had considerable grounding at least in koinē Greek, the language of Epictetus.

Kant owned in his personal library a copy of the philosophical works of Seneca (Halle 1762) containing all of the essays in ethics, including the letters of consolation, but excluding the letters on ethics to Lucilius.Footnote ¹² Here a word of caution is in order. Kant’s personal library was relatively small, may have largely been comprised of presentation copies, and, at any rate, excludes many works of uncontested significance for his philosophical development. (We find, for example, not a single work of Rousseau.) There is a collection of Cicero’s speeches, but none of his philosophical works, apart from Garve’s rendering of De officiis (Breslau 1785). Yet Kant makes precise reference to the philosophical works of Cicero repeatedly: for example, the gouty Stoic of the second Critique (5:60) is none other than Posidonius, as he appears in a story told in Tusculan Disputations (2.61); Chrysippus’ bracing remarks likening the soul of a pig to salt, so as to preserve the flesh for our use, appear in a 1796 essay (Ak. 8:413) duly attributed to De natura deorum (2.160); and Cicero’s distinctive deployment of lines from Terence’s The Self-Tormentor (Heautontimōroumenos) as a motto for cosmopolitan duty are invoked in the same spirit by Kant in the Metaphysics of Morals (6:460) and a 1793 set of lectures (VM-Vigilantius 27:677).Footnote ¹³ Kant, of course, had ready access to books apart from those he owned. His first salaried position was as ‘sublibrarian’ of the Schloßbibliothek, which ‘basically amounted to the university library’ in Königsberg; the newfound financial security allowed Kant to move in 1766 into the large house of his publisher, Johann Kanter, and ‘borrow all the books he wanted and take them up to his apartment’ (Kuehn Reference Kuehn2001b:159–60).

Garve’s translation of and commentary on Cicero’s De officiis provides the context of an anecdote that has influenced how some scholars have traditionally approached questions about the significance of Stoicism for Kant. In February 1784, Kant’s friend J. G. Hamann wrote in a letter to J. G. Herder that Kant was apparently ‘working on a reply to Garve’s Cicero’.Footnote ¹⁴ Although Hamann could only have been referring to Kant’s Groundwork for the Metaphysics of Morals, it remains far from clear how that work could consist of such a reply. In his 1935 Kant und die Ethik der Grieschen,Footnote ¹⁵ Klaus Reich argued that each of Kant’s three ‘formulas’ of the moral law could be traced to ideas in De officiis. Although the assessment of Reich’s interpretation has been mixed,Footnote ¹⁶ the anecdote certainly invites us to reconsider how Kant’s appreciation of Stoic thought in a Roman context may have had some impact on his conception of the formulas of the moral law. Brad Inwood’s Chapter 1 in this volume examines formula as a Roman legal term deployed by both Cicero and Seneca, in an extended sense, in their writings on ethics. One key idea is the distinction of formula from a universal rule or law: the formula, Inwood explains, is a pretrial statement articulating ‘what is at stake, the point to be settled by argument’ in the case at hand. Cicero and Seneca each use the notion in an extended sense as a point of reference for determining appropriate action.

Although Inwood leaves open how this account of the Roman-Stoic conception of formula may inform Kant’s Groundwork, the background is promising for advancing Kant scholarship. For in fact, despite Kant’s listing just three formulas of the moral law (G 4:436.8) – the formula of universal law of nature, the formula of humanity, and the formula of autonomy – Kant scholars regularly take there to be four formulas, adding to this list the ‘formula of universal law’, and sometimes five, taking there to be a formula of ‘the kingdom of ends’ as well. But Kant does not generally refer to what Kant scholars call the ‘formula of universal law’ – that is, ‘I ought never to act in such a way that I could not also will that my action should become a universal law’ (4:402.8–9) – as a formula. Rather, he refers to what he introduces with those words as the ‘principle’ (Princip) of a rational will (4:402.7). Recalling this principle later in the imperative mood (‘act only in accordance with that maxim through which you can at the same time will that it become a universal law’, 4:421.7–8, my underscore), he calls it simply the categorical imperative. He proceeds to identify its three formulas, explaining in the end that these formulas serve to ‘represent’ the ‘principle of morality’ in such a way as to bring the rational ideas at stake in this law ‘closer to intuition … and thereby to feeling’ (4:436.12–13). Inwood’s chapter on the Roman Stoic conception of ‘formula’ thus provides resources for further examination of Kant’s distinction between the law and its three Formeln.Footnote ¹⁷

Chapters 2 and 3 consider questions broadly related to the implications of the standard German rendering of the Stoic notion of officium, or appropriate action, with Pflicht – duty – as, for example, in Garve’s translation of De officiis. One difficulty for this rendering is that, for Kant, ‘duty’ refers to what is strictly required, whereas officium, as appropriate action, has a much wider significance. After all, Kant distinguishes duty both from what is contrary to duty or prohibited, and from what is discretionary or permissible, as long as it does not conflict with duty.Footnote ¹⁸ Iakovos Vasiliou (Chapter 2) argues that the Stoic category of media officia, or intermediate appropriate actions, is akin to what is permissible and discretionary in a Kantian scheme. Jacob Klein (Chapter 3) argues, by contrast, that the relevant parallel between Kant and the Stoics does not lie in kindred deontological categories, but rather in their agreement that the value of good action has its sole source in the character of the agent.

The next set of chapters concern virtue and eudaimonia. The broad historical context of the papers from Katja Vogt (Chapter 4) and Stephen Engstrom (Chapter 5) concern the legacy of Socrates as the source of the thesis that virtue is knowledge, and thus that there is only one virtue (wisdom). Vogt considers Kant’s explicit rejection of the latter thesis (‘there is only one virtue and one vice’, MS 6:405), which he most plausibly associates chiefly with the Stoics. Yet since Kant himself typically speaks of ‘virtue’ in the singular, the remark is puzzling. Vogt argues that Kant’s conception of virtue in the singular can be traced to his view that there can be only one single motivation of virtue – morality itself – and explains that Kant nevertheless recognises a plurality of virtues, which may be distinguished by the plurality of ends (as the object, or matter, of the will) that a virtuous person would adopt. As these ends are heterogeneous, she suggests, Kant does not endorse a strong view of virtue as one. Engstrom examines the conception of wisdom that stems from the Socratic identification of virtue with knowledge, arguing that there is a parting of the ways, in the development of this Socratic inheritance, between Aristotle (and, later, Kant), on the one hand, and the Stoics, on the other. The parting of the ways is traced to differing conceptions of the object of the knowledge in which such wisdom consists: for the Stoics, this object is nature as divine creation; for Kant, it is practical reason’s own self-awareness through the moral law, which is manifest in ordinary moral understanding as conscience.

Ian Blecher (Chapter 6) examines Kant’s remarks about the dignity and sublimity of virtue in Stoic ethics: the self-sufficiency of the Stoic sage presupposes a standard of conduct that is beyond human capacity. Certainly one of Kant’s most persistent criticisms of the Stoics was that their ethical ideal of the sage can only be conceived as having transcended human nature: we can trace the charge from his early lectures on moral philosophy through to at least the 1788 Critique of Practical Reason.Footnote ¹⁹ Michael Vazquez (Chapter 7) situates Kant’s anti-eudaimonism in the context of his immediate intellectual milieu (especially Christian Garve’s neo-Stoicism), and explains that the ‘error of subreption’ (vitium subreptionis) familiar from Kant’s catalogue of traditional metaphysical error in the first Critique is redeployed for a criticism of Stoic ethics in the second Critique.

Jens Timmermann (Chapter 8) frames his chapter around another episode in the modern reception of Stoic ethics – namely, Adam Smith’s view that Stoics place the chief ethical value in the evaluative perspective from which a virtuous person acts, rather than in any result of those actions. Timmerman considers how Stoic and Kantian ethics might, respectively, be resourced to reply to the ensuing objection that they overlook much that is valuable in human life as a result. He points to the ways in which Kant, with his appeal to two distinct sources of human motivation, is arguably better resourced to reply to this objection, but concludes by noting the irony that Kant, who consistently charged Stoic ethics with upholding an impossible ethical ideal for human beings, is arguably exposed to the same charge himself.

Notably, though, Kant’s persistent charge that the Stoic normative ideal would have transcended human nature does not figure in the later (1792/3) Vigiliantius lectures on moral philosophy, that is, from the time when Kant was particularly absorbed in Seneca.Footnote ²⁰ The Roman Stoics were less interested in describing the rarer-than-a-phoenix sage, and more concerned with what it is to be a progressor – a non-sage, thus no less a ‘fool’ than anyone else, but nevertheless plausibly making progress towards virtue.Footnote ²¹ (Relatedly, Vazquez suggests that Seneca has a distinct conception of progressor-joy, which stands to Stoic pleasure or hēdonē as Kantian moral self-contentment, Selbstzufriedenheit, stands to happiness or Glückseligkeit.) So it seems that Kant, as he came to study Seneca closely in his later years, was thinking through a Stoicism that focused on the progress of ordinary human beings – and may have been less apt to rehearse the old complaint about their normative ideal going beyond what is humanly possible. Paul Guyer (Chapter 9) helps us appreciate a related development in Kant’s understanding of Stoicism, in connection with the topic of the highest good. Guyer argues that, in the works from 1790 onwards, Kant moves towards a conception of the highest good as an end to be realised by the natural human species, which brings him closer to Stoicism and further from Christianity than commentators have hitherto appreciated.

The final set of chapters takes up questions about Kant’s relation to Stoicism on topics to do with human feeling and ethical development. Stoic accounts of human development fall under the scope of the doctrine of oikeiōsis – a term that is variously rendered ‘appropriation’, ‘familiarisation’, or ‘orientation’. Stoics take nature to be governed by rational law – right reason – which they identify with the soul of Zeus (see, e.g., DL 7.88). As a result, nature is understood to be providentially arranged. One feature of this account concerns the constitution of animals. Animals are distinguished from plants as the living part of nature, according to Stoics; unlike plants, they must do certain things in order to develop fully as creatures of a certain kind. Hence animals are providentially equipped so that they are predisposed to act in completion-promoting, or appropriate, ways. Animals are thus predisposed to find fitting those actions and things that sustain them in their own constitution, and to find abhorrent what threatens such sustenance. The doctrine of oikeiōsis is concerned with this providential set-up, as the basis of all animal (including human) action. The story is complicated, of course, in the human case, since once we come into the use of reason, we can only act on our own initiative: then we can no longer be directly compelled by our predispositions, but merely oriented by them. Moreover, according to the Stoics, we invariably corrupt ourselves just as soon as we acquire reason – so that we are often mistaken, or misguided, in what we take to be the appropriate thing to do. Alix Cohen (Chapter 10) observes that Kant invokes a conception of orientation as necessary for at least our agency, which she examines against the Stoic precedent in the theory of oikeiōsis. Now, part of what belongs to the providential set-up of the human being is a readiness to find fitting those ways of acting that preserve one as one is meant to be qua rational: this idea is emphasised by Epictetus in his distinctive conception of aidōs as self-respect. In Chapter 11, I argue that Kant draws, in the Religion, on Epictetus’ distinctive pedagogical practices aimed at arousing aidōs in his students.

Nancy Sherman’s essay (Chapter 12) can be appreciated from two angles. Here is one angle: if we human beings invariably make ourselves bad in the course of our development, then our social worlds can only be rotten, as well. Centuries back we burned people at the stake in the town square; now we beat them to death for not fully covering their hair. The litany of our sick cruelty goes on and on. Sherman asks: Is there scope in Stoic and Kantian thought for morally grounded anger directed at such depravity? For Stoics famously argue that the sage will be apathēs – free of pathē, such as ordinary anger. And Kant appears to endorse, in several places, some version of the Stoic duty of apathy.Footnote ²² But surely – and this is the second angle on the topic – some ordinary emotions could promote one’s moral development, like Alcibiades’ shame for his own bad character.Footnote ²³ Sherman argues that Kant has scope for a conception of moral, or righteous, anger in his view of well-developing, and ultimately virtuous, character.

Rachel Zuckert (Chapter 13) proceeds from something like the first angle on the background to Sherman’s chapter: the mess of the human being’s moral situation, as it has been, as it is now, and as it is tending. Taking up Kant’s suggestion that his Idea for a Universal History with a Cosmopolitan Aim offers a ‘consoling prospect into the future’ (Idee 8:30), she considers Kant’s philosophy of history as a form of consolation writing, looking to Seneca’s consolation essays as his model.