Kant’s Groundwork of the Metaphysics of Morals

                        Portia in The Merchant of Venice

What a ‘groundwork’ of moral philosophy can and cannot do

To avoid disappointment, readers of the Groundwork are well advised to keep in mind the very specific nature of Kant’s project. What does he intend to achieve? Which questions does he not even try to address?

   Let us start with what not to expect from a Grundlegung. The word can be used to describe the activity of laying the foundations of something or, when this is graced with success, its result.¹ We would therefore expect the Groundwork of the Metaphysics of Morals to contain the principles of another, distinct philosophical project. Indeed, that is how Kant describes the task that lies ahead in the Preface. The first five or so pages contain a careful discussion of the nature and necessity of a future metaphysics of morals, and it is only towards the end that Kant turns to the prior, ‘critical’ task of grounding this novel discipline. The slim volume is clearly part of the foundational project that preoccupied Kant during the 1770s and 1780s.

   If the Groundwork does not claim to be a complete guide to ethical theory or moral life as a whole, it would be a mistake to try to reduce Kantian ethics to this book, even if we judge it to be Kant’s most profound or influential contribution to moral philosophy.² The Groundwork leaves many questions to be addressed in later works, not just the new ‘Metaphysics of Morals’ itself. Kant occasionally mentions the idea that what we might call ‘applied’ ethics – moral psychology or ‘anthropology’, a project he never carried out – must follow a metaphysics of morality. There is also the idea of a fully fledged ‘Critique’ of pure practical reason, which settles fundamental problems that fall outside the narrow scope of the Groundwork. These projects are valuable in their own right, but they are not essential to the task of laying the foundations of ethics. We are not yet in a position to discuss, for instance, the comprehensive classification of first-order moral commands, the theological implications of the sum of all that is good, the unity of theoretical and practical reason, the casuistry of ethical conflict, the precise mechanism of moral motivation (and so forth). The purpose of the Groundwork is a more modest one: Kant ‘merely’ seeks to identify and firmly to establish the highest principle of moral volition.

   There are also questions Kant does not intend to address in any of his works on moral philosophy. Most significantly, he does not intend to overthrow everyday morality. This is not due to some sentimental attachment to a particular moral code. Kant’s reasons for thinking that common, pre-philosophical moral thought cannot be radically mistaken are largely ethical. Human beings must have access to moral truth to be responsible agents at all. Non-culpable ignorance renders attitudes of praise and censure invalid, and moral commands might be weakened to mere recommendations that apply on prudential grounds. If, as Kant argues throughout the Groundwork, moral action must be done ‘for the sake of the law’, all moral agents must have access to this law. He cannot dismiss common moral views and those who hold them as amoral. Kant is not immune to sceptical influences, but he takes some challenges – such as the twin threats of empiricism and physical determinism – more seriously than others. The Groundwork is not, therefore, an unbiased enquiry into what the grounding of morality might be, or whether there are moral principles at all. The truth of reflective common morality is the default position. Kant even makes what he considers the everyday notion of moral value the starting point of his enquiry: the only absolutely good thing is a good will (Ⅳ 393.5–7).

   Kant’s intentions have been misunderstood ever since the Groundwork was first published in 1785. In the preface of the Critique of Practical Reason, he says that a critical reviewer of the Groundwork – in fact G. A. Tittel in his little commentary Über Herrn Kant’s Moralreform, pp. 15–16 – `hit the mark better than he himself may have intended when he said that no new principle of morality is set forth in it but only a new formula’ (Ⅴ 8 fn.). Kant rejects the idea of a novel principle as preposterous:

For who would want to introduce a new principle of all morality and, as it were, first invent it? Just as if before him the world had been ignorant or in thoroughgoing error about what duty is. Whoever knows what a formula means to a mathematician, which determines quite precisely what is to be done to execute a task and does not let him miss it, will not take a formula that does this with respect to all duty in general as something that is insignificant and can be dispensed with.(V 8 fn.)

   In Kant’s ethics common understanding has a much more positive role to play than in his theoretical philosophy. In the Metaphysics of Morals of 1797 Kant goes so far as to christen sound reason an ‘unwitting metaphysician’ (Ⅵ 206.23). Pre-philosophical moral thought can get us started on the task of a metaphysics of morals, even if it cannot finish it. By contrast, Kant famously complains about the ‘lazy method’ (bequemes Mittel) of appealing to common understanding in speculative philosophy (Prolegomena, Ⅳ 259.12) because in that branch of human inquiry our natural prejudice in favour of sensibility obstructs the shift of perspective that is required for the metaphysics of nature to become a proper science: the ‘Copernican turn’ of the first Critique. It makes metaphysics of nature possible by allowing for a priori cognition of objects because the objects of knowledge themselves turn out to depend on our cognitive faculties (B xvi). Thus even if, by virtue of its foundational character, the latter parts of the Groundwork contain certain philosophical technicalities and intricacies, it is still closer to everyday thought than, for instance, the Critique of Pure Reason, which at the time had been widely condemned as impenetrable (and in fact still is).

Pessimism and optimism in Kant’s moral theory

We have already caught a glimpse of the striking combination of pessimism and optimism that confronts readers of the Groundwork. Kant is extremely confident about the cognitive and affective moral capacities of human beings, but he is also very sceptical about the actual moral quality of their conduct. Let us discuss these two convictions in turn.

   The ease with which we are supposed to apprehend token moral truths stands in contrast with the endless complications of empirical cognition. For the latter, I must collect, arrange and process data, a procedure susceptible to all sorts of error. These uncertainties also affect instrumental – i.e. technical and especially prudential – reasoning, which makes use of empirical knowledge. By contrast, ‘to see what I have to do in order that my volition be morally good’, Kant declares towards the end of Section I, ‘I do not need any far-reaching acuteness’:

Inexperienced with regard to the course of the world, incapable of being prepared for whatever might come to pass in it, I ask myself only: can you really will that your maxim become a universal law? (Ⅳ 403.18–22)

   Kant’s optimism is not confined to moral cognition. To defend the universal authority of the moral law he must also make sure that we have a motive at our disposal that is always sufficiently strong to produce the action we recognise to be right. After all, ought implies can. A standard cannot otherwise be categorical, i.e. independent of merely subjective motivational conditions.⁵ At this point Kant makes some concessions to the frailty of human beings. We do not possess what he calls a ‘perfect’ or ‘holy’ will that effortlessly acts as the moral law bids. We merely possess ‘pure’ will – later called Wille as opposed to the faculty of choice or Willkür – which is governed by the laws that a metaphysics of morals must investigate. In short: not all, but part of our faculty of volition is pure. It is exposed to two forces: reason and inclination; and although in matters of conflict we must side with reason, we cannot make inclination go away. It can at best be conditioned to support our rational, especially moral projects. We can never become completely good. Reason is in essence the same in all of us, and so are the commands of morality, which depend on reason alone. Inclination, by contrast, displays huge variations. That is why moral action can be more or less difficult subjectively at different times, and why moral behaviour comes more easily to some human beings than to others. Kant is not blind to the diversity of humanity. He merely thinks that, for egalitarian reasons, human nature should not feature prominently in normative ethics.

   In sum: the moral option is available to all agents, even if owing to natural inclination it cannot be the only option. As a result, all grown-up human beings capable of using their rational faculties are responsible for their moral failings. They could have behaved morally if they had chosen otherwise. In the second Critique, the voice of practical reason is said to make even ‘the boldest offender’ (den kühnsten Frevler) tremble with awe (Ⅴ 80.1–2). In the Groundwork, Kant goes so far as to say that, when confronted with shining examples of virtuous conduct, ‘the most hardened villain’ (der ärgste Bösewicht) desires to be a moral man; and painful though it is for him to mend his ways, he can bring this about (Ⅳ 454.21–9).⁶

   If all seems well on the prescriptive side of moral philosophy, Kant has much less faith in our capacity to detect the actual moral value of our actions. We know what we ought to do; we are convinced that we can act morally when moral action is required; owing to the influence of inclination we are rather less certain whether we are going to do it; and in retrospect, if on the face of it we have done the moral thing, we can never be sure that we did it for the right reasons and not for selfish ones, i.e. that the act was done from and not merely in conformity with duty. Kantian morality commands that we take the right attitude in action, not just the performance of the right act. An action is morally good only if it proceeds from a subjective principle or ‘maxim’ that is fit to be a universal law. But the moral quality of action remains obscure. We know human actions, our own and those of others, only as they appear to us in experience, and the regularities they display are part of the causal process of nature. Free actions do not surface as such.

   However, Kant’s scepticism about moral value extends beyond agnosticism. As inclination is dear to us in a way morality is not, he is very suspicious about the actual moral quality of people’s character. Individual actions mostly coincide with duty, and everyone can be morally good, but very few of us actually are. Most human beings get their moral priorities wrong. This pessimism is manifest throughout the examples in Section Ⅰ of the Groundwork (Ⅳ 397–9). Shopkeepers commonly treat their customers decently; but they do so because they care about their reputation, not on moral grounds. The anxious care people take to preserve or advance their lives is likely to be grounded in self-love, not in moral principle. Beneficent action is frequently the effect of our natural sympathetic tendencies, not ethical conviction. Kant argues that any action motivated by inclination, rather than reverence for the moral law, lacks distinctive moral value, no matter how amiable the inclination in question may be. Moreover, human beings display a worrying tendency to conform the normative standards of action to their own desires, and to flatter themselves that they are more moral than in fact they are. They often wilfully mistake prudential regret for the nagging voice of conscience (Collins, ⅩⅩⅦ 251.16–17). Innocence is easily corrupted – which is why moral philosophy is needed at all (Ⅳ 404–5). Unsurprisingly, non-sophisticated people seem to be rather better assessors of their own moral worth than clever intellectuals. Kant’s spirit, as he puts it in the second Critique, ‘bows before the common man’ (Ⅴ 77.1–5).⁷

   Those who find the Kantian conception of moral value objectionably narrow should bear in mind that it is a consequence of his strong egalitarian convictions, which sentimentalism or virtue theory cannot accommodate. Kant considers it unrealistic to pretend that human beings can attain moral perfection, or that obligatory action is always pleasant or beneficial to the agent. These are elements of the human condition that a philosopher should not try to explain away. But we can at least presuppose that everyone is endowed with the same capacity to be moral, and create a level playing field in that respect. Morality must be about action, about what is up to us, not about the distribution of natural favours. There is a rather pithy handwritten note in which he says that ‘it can be required of human beings that they act as the law commands; but not that they do so gladly’ (R 8105, ⅪⅩ 647). In the Critique of Practical Reason, Kant makes it quite clear that sympathetic feelings are often welcome, amiable, desirable, beautiful. They can under certain conditions be good objectively, all things considered. But they are not morally good (Ⅴ 82.18–25). A happy, well-rounded character is an ideal that lies beyond the sphere of Kant’s conception of morality.⁸

The character of moral duty

Kant’s late work on the philosophy of religion contains some stunning examples of his confidence in common moral consciousness. In Religion within the Limits of Reason Alone he goes so far as to claim that conscientious moral judgement cannot err. The voice of conscience, which is our internal moral judge, can serve as a ‘guiding thread’ (Leitfaden) in matters of doubt. Kant introduces a new practical principle ‘that does not stand in need of a proof’: that we ought ‘to venture nothing where there is a danger that it might be wrong’ in Pliny’s Latin: quod dubitas, ne feceris (Ⅵ 185.24–5).⁹

   To illustrate this point Kant focuses on the possibility of conflict between the commands of biblical faith and morality. His example is that of an inquisitor who condemns a decent citizen to death for his alleged heresy. Kant assumes that the death sentence is unjust. Yet the inquisitor did not make an innocent mistake. He ‘consciously did wrong’ because ‘we can always tell him outright that in such a situation he could not have been entirely certain that he was not perhaps doing wrong’ (Ⅵ 186.28–30). In other words, he is violating the new criterion of moral permissibility to use his sense of guilt as a guiding thread in moral matters, i.e. never to do anything which he is not completely certain is right. It is incumbent on the agent ‘only to enlighten his understanding in the matter of what is or is not duty; but when it comes, or has come, to a deed, conscience speaks involuntarily and unavoidably’ (Ⅵ 401.14–16).

   Note that the inquisitor’s conflict is practical but not moral. Kant is so confident that the inquisitor’s sense of right and wrong could not be silent on the matter of whether he may kill an innocent person for religious reasons because he sees two distinct and quite unequal forces at work: revealed religion and pure practical reason. Kant does not question the sincerity of the inquisitor’s faith; but no-one can ever be certain, as is morally required, that historical religion justifies the destruction of an innocent human being. Similarly, in the Critique of Practical Reason, the unconquerable voice of conscience is said to support our judgement that physical determinism does not suffice to undermine morality and responsibility (Ⅴ 98.13–28). Again, the two opposing forces – nature and morality – are different in kind. Kant does not seem to envisage that we are torn between two courses of action for moral reasons. He makes no provisions for genuine moral dilemmas, where no option is unambiguously right or all options are equally problematic.

   As the case of the inquisitor illustrates, Kant’s conception of duty lacks many of the unpalatable connotations that the word might evoke in today’s readers. It is important to keep this in mind. What human beings ought to do is not grounded in their social rank or station. After all, the moral law is universal. It is a result of one’s status as a rational being amongst others. Moreover, in the last consequence we freely impose the law of duty upon ourselves – which is the definition of Kantian autonomy. In the late Conflict of the Faculties, Kant explicitly argues that the command of a superior is not valid automatically. To apply, it must be freely judged to be right (Ⅶ 27.27–30). It is a symptom of the perversity of National Socialism that Adolf Eichmann, in his Jerusalem trial, sought to justify his part in the slaughter of millions of Jews with reference to Kant’s moral philosophy. Eichmann said that for a long time he was not just obeying orders, he was acting for the sake of the law – a law most certainly not his own – until finally he abdicated moral judgement to his superiors altogether.¹⁰ But he could and should have known that he was not doing his duty.

A priori and a posteriori: the grounds of action

The opposition of the a posteriori (that which is grounded in the natural world of experience) and the a priori (the rational, which is not) pervades the whole of Kant’s philosophy. Failure properly to distinguish between the two is just as pernicious in the practical realm as it is in the theoretical.¹¹ The opening statement of the 1787 edition of the Critique of Pure Reason (B 1) has close parallels in Kant’s moral theory. In either field, sensibility comes first in the temporal order, but is not sufficient to accomplish the task at hand: the generation of knowledge or action. ‘All our cognition’, Kant states, ‘begins with experience’, and so do all our actions. For how else should our cognitive, or practical, capacities be ‘awakened into activity [Ausführung]’? But it does not follow that either knowledge or action ‘arises from’ or is a mechanical product of empirical factors. In either case, the a priori addition of absolute spontaneity, i.e. activity of the self, is necessary to bring about the desired result.

   This model of interaction between sensibility and reason is less conspicuous in the practical sphere than in the theoretical, but closer inspection reveals it to be a constant theme also in Kant’s philosophy of action. For example, in the second Critique Kant maintains that inclinations ‘always have the first word’ (Ⅴ 146.34, summarising Ⅴ 74.8–15). This tendency can also be detected in the examples that Kant uses to illustrate the first variant of the categorical imperative in Section Ⅱ of the Groundwork. The first illustration concerns an unfortunate man who is tired of life. Yet he is ‘still so far in possession of his reason that he can ask himself whether it would not be contrary to his duty to himself to take his own life’ (Ⅳ 421.25–422.3). Inclination does not automatically translate into action. He can still reflect on the courses of action open to him and decide in the light of rational considerations. In the second example, someone finds himself ‘urged by need to borrow money’ and he knows well ‘that he will not be able to repay it; but he sees also that nothing will be lent him unless he promises firmly to repay it within a determinate time’. Kant says that this person ‘is inclined’ to make such a promise, but he has still ‘enough conscience’ to ask himself whether it is not perhaps ‘forbidden and contrary to duty to help oneself out of need in such a way’ (Ⅳ 422.15–20). The third and fourth examples, regarding our duties not to neglect our talents and to help those in need, follow the same pattern. Inclination has the first word but it need not – and often must not – have the last.

   If sensibility is insufficient to produce either knowledge or action, something over and above the empirical is required to complete the process: they must both rest on principles that are subject to rational evaluation. In action, it is up to us to reject the pretensions of inclination. We can conform the subjective principles from which our actions proceed (maxims and rules) to objective principles of reason (imperatives). We would like to give in to our natural desires; but we are still free to do the right thing. Moreover, if knowledge and action are capable of rational justification, the grounds of theoretical and practical principles must be a priori. The a priori nature of ethical norms is borne out by the fact that, as in the case of knowledge, morality involves an element of necessity (see Ⅳ 389.11–13); but if Kant is right, necessity cannot be encountered in experience. Experience merely informs us about the way things are, not the way they ought to be (see B 3). Kant takes the divide between (natural) is and (moral) ought very seriously indeed.

Kant’s method: analytic, synthetic and the need for a ‘deduction’

Kant is not exceptional in his professed reliance on commonly held moral beliefs to disclose their underlying principle. In the history of moral philosophy, even those who reach substantially different conclusions usually fail to come up with a plausible alternative. J. S. Mill is convinced that there is a ‘tacit influence’ on common moral judgement of an objective standard that can be revealed by philosophical means: the principle of utility (Utilitarianism Ⅰ.4); and Aristotle’s professed method in the Nicomachean Ethics consists in investigating reputable opinions or ἔνδοξα (Book Ⅶ, 1145b 2–7).¹² Starting with anything other than the views of common moral consciousness would expose Kant’s project even more to the sceptical worry that his ethical theory is just a figment of a particularly lively philosophical imagination.

   What is rather more remarkable than Kant’s starting point is the division of labour assigned to the three sections of the Groundwork. Kant briefly discusses his research method towards the end of the Preface. Sections Ⅰ and Ⅱ are declared to proceed ‘analytically’, while Section Ⅲ is said to be ‘synthetic’ (Ⅳ 392.17–22, cf. Ⅳ 445.7–8). In particular, the analytic sections are devoted to the identification or discovery of the highest principle of morality and its variations, which is then to be confirmed or justified in the final synthetic section (Ⅳ 392.3–4). Kant expects readers of the Groundwork to be familiar with the analytic/synthetic distinction from his theoretical writings (see Ⅳ 420.14–17). How is it employed in his moral philosophy?

   The obvious reason why Kant considers the justification of morality to be in some sense synthetic is the following. A priori principles are either analytic or synthetic. The supreme principle of morality, if a priori, cannot rest on analytic foundations because the analysis of concepts helps us to understand them better but cannot establish their reality (see Ⅳ 420.18–23). Analytic judgements develop or clarify given concepts without assessing their validity – which is precisely what Kant does with the concept of duty in Section Ⅰ and, in a more roundabout way, in Section Ⅱ. The concept of God by necessity points to his perfection or necessity; but this does not justify the assumption that there is a God. Analysing the concept of a ‘bachelor’ will reveal such a creature to be an eligible man who has never been married; but a woman interested in the existence and whereabouts of bachelors would be ill advised to confine her efforts to conceptual analysis.

   That there are bachelors is an empirical judgement, unlike the equally synthetic judgement that God exists, or that human beings are subject to duties. As experience cannot vouch for our right to use concepts like God and duty, Kant is worried that our use of synthetic a priori principles may not be justified. In the case of some concepts it is rather doubtful whether they can be applied to reality at all, i.e. whether there is something that corresponds to them, and whether we say anything meaningful when we use them. In the Critique of Pure Reason, Kant mentions the examples of fortune and fate (Glück, Schicksal, A 84/B 117). His twelve categories are exposed to this suspicion because they are pure concepts of the understanding and as such not rooted in experience. To use Kant’s term, borrowed from the legal literature of his day, synthetic a priori principles stand in need of a ‘deduction’:

Jurists, when they speak of entitlements and claims, distinguish in a legal matter between the questions about what is lawful (quid juris) and that which concerns the fact (quid facti), and since they demand proof of both, they call the first – that which is to establish the entitlement or legal claim – the deduction.(A 84/B 116)

   In the last section of the Groundwork, Kant therefore intends to demonstrate, as far as is possible, that we are entitled to apply to human action the concept of duty as developed in the first and second sections. The principle of duty – the categorical imperative now in its final, most metaphysical variant: the principle of autonomy – must be both synthetic and a priori. But like fate and fortune, the concept of duty might be no more than an ‘empty concept’ (Ⅳ 421.12), a natural and understandable idea to which nothing corresponds in reality. For all we know, human beings could be incapable of moral action, which would turn the analytic sections into the literally academic project of developing a fantastical concept.

   This kind of philosophical concern should not be confused with the scepticism of the amoralist, who cannot see the point of moral action at all. Kant’s problem is the worry of someone who is well disposed towards morality but cannot understand it. There is nothing in the world of experience – an otherwise reliable source of matching concepts and reality – that confirms the existence of duty. We can point to a bachelor when we see one, but we cannot point to a free human action done from a sense of duty, just as we cannot empirically identify an act of providence. Experience tells us about matters of fact, not norms, imperatives or values.¹³ Yet if the world of experience cannot be the source of the authority of the moral law, what is? Or, as Kant himself puts it, ‘whence’ does the moral law obligate or ‘bind’ us (Ⅳ 450.16)?

   There is now another sense in which the categorical imperative is a ‘synthetic’ practical principle (Ⅳ 420 fn.). Moral commands are characterised by the fact that the action commanded is independent of – not contained in or entailed by – any ends we want to pursue in action. A synthetic practical proposition tells the agent to do ‘something new’, just as a synthetic theoretical proposition provides us with new information beyond that which is contained in a specific concept (e.g. that a ball is blue, as opposed to round; or that the will is free, as opposed to a kind of causal power). By contrast, something in accordance with an analytic practical principle follows from a given end and does not constitute a separate action in its own right. When I add hot water to ground coffee beans to make coffee, I do not both add hot water and make coffee. According to the technical rules of coffee making, making coffee consists in, amongst other things, adding hot water to ground coffee beans. If you observe me as I pour water into a cafetière, my action can be explained with reference to an identifiable end or desire. The imperative I act on is hypothetical. If the human will was perfect like the will of God and unaffected by the obstacles put in its path by inclination, moral action would follow in a similar fashion. Unfortunately, we do not possess such a perfect will. The problem with human morality is that it does not rest on an antecedently given end that we wish to realise.

   It is tempting to recast the problem of the possibility of synthetic practical principles in the terms of a prominent contemporary debate: how are external reasons possible? Or perhaps: how can my rational faculty create a new incentive not initially contained in my motivational set?¹⁴ In slightly more old-fashioned terms: how can reason, just by itself, motivate? Kant’s concept of duty makes sense only if it can.

   The first Critique recognises the need to ‘deduce’ synthetic a priori judgements quite generally.¹⁵ Even the concepts of space and time deserve a ‘transcendental elucidation’ – despite the fact that space and time, which form the ground of the synthetic a priori principles of arithmetic and geometry, are involved in experience. The two concepts rest on pure forms of intuition and are therefore not empirical in origin (B 40–1, B 48–9). The twelve pure concepts of the understanding or ‘categories’, the application of which makes experience possible in the first place, require a fully fledged transcendental deduction, even though they are at least indirectly confirmed by experience (A 84 ff./B 116 ff.). If so, it should be obvious why the position of moral concepts like ‘duty’ and ‘autonomy’ is so precarious. They are a priori in origin, but they have no possible link with experience at all. They are not even, like the categories, capable of indirect corroboration; and what is worse, experience appears to confirm that all human action is subject to natural laws and therefore, by definition, not free.

   The parallel between the theoretical and the practical that was introduced in the previous section is now complete. Kant resumes the argument quoted above as follows. There remains, he says,

a question which at least requires closer investigation, and one not to be dismissed at first glance, whether there is any such cognition independent of experience and even of all impressions of the senses. One calls such cognitions a priori, and distinguishes them from empirical ones, which have their sources a posteriori, namely in experience.(B 1–2)

The story of the Groundwork

When the Groundwork appeared in print in the spring of 1785, it was Kant’s first published work devoted exclusively to the subject of moral philosophy. But Kant was not a newcomer to the discipline.¹⁶ By the mid-1780s, he had been planning to write a book on the foundations of ethics, entitled ‘Metaphysical Principles of Practical Philosophy’, ‘Critique of Moral Taste’ or ‘Metaphysics of Morals’, for at least twenty years.¹⁷ In February 1767, J. G. Hamann told Herder that Kant was working on a ‘Metaphysics of Morality’, which unlike previous ethical theories was meant to investigate the question of ‘what man is, rather than what he ought to be’ (Ⅳ 624); and on 9 May 1768 Kant wrote to Herder expressing his hope that he might complete a ‘Metaphysics of Morals’ by the end of that year (X 74, No. 40 [38]). But, like his theoretical philosophy, Kant’s ethical theory soon changed beyond recognition. He abandoned the idea that a metaphysics of morals should be a descriptive, psychological study of human nature. By the early 1770s, the ‘first grounds’ or ‘pure principles’ of morality had become part of the new critical project of exploring ‘The Boundaries of Sensibility and Reason’, published as the Critique of Pure Reason a decade later.¹⁸ In the Groundwork, Kant denounces the empiricist project of moral enquiry as at best irrelevant and at worst pernicious. Pure normative moral theory must precede moral psychology or ‘anthropology’.

   The idea that a dual metaphysics of morals and nature should, in that order, follow the critical foundations of transcendental philosophy remained a constant feature of Kant’s philosophical ambitions. It is first mentioned in a letter to Marcus Herz in late 1773 (X 145.20–2, No. 79 [71]). However, on more than one occasion Kant changed his mind as to how much of a ‘critical’ preparation was needed to ground the moral part of the metaphysical system. Moral philosophy was very much part of the critical enterprise in the early 1770s but was then discarded, either because Kant realised that he had enough work on his hands with laying the foundations of the metaphysics of nature, or because he thought that the later Critique of Pure Reason provided a sufficient foundation of both parts of the twofold metaphysics. At the time of composition of the Methodology (see A 841/B 869), Kant did not seem to feel the need of a grounding of metaphysics, moral or natural, other than the 1781 Critique itself, which after all makes room for freedom and responsibility.¹⁹ Yet a rudimentary Critique of Pure Practical Reason was published in 1785 under the title of a Groundwork of the Metaphysics of Morals. The second edition of the Critique of Pure Reason, which for a while was meant to revert to the original plan and cover moral as well as speculative philosophy, followed in 1787; the Critique of Practical [sic!] Reason as an independent publication in 1788; and by 1790, Kant deemed an additional third ‘Critique’ necessary to complete the foundation of the dual metaphysical system: the Critique of Judgement.²⁰

   The composition of the Groundwork of the Metaphysics of Morals is shrouded in mystery. As Hamann reveals in a letter to Kant’s publisher, J. Fr. Hartknoch, in January 1782, Kant returned to working on a ‘Metaphysics of Morals’ soon after the publication of the Critique in the previous year (Ⅳ 625). However, we possess some evidence to the effect that Kant’s intention to ‘issue [a] groundwork in advance’ (Ⅳ 391.17) was influenced by the publication of Christian Garve’s annotated German translation of Cicero’s De officiis in 1783. Kant had held Garve, a ‘popular philosopher’ at Leipzig, in high regard. He was hoping to recruit Garve for the critical cause and was therefore disappointed to learn that he was the author of a scathing anonymous review of the first Critique published in the influential Göttingische Anzeigen von gelehrten Sachen in January 1782. It had, admittedly, been abridged, edited and not at all improved by J. G. H. Feder, who was professor of philosophy at Göttingen. The Prolegomena represents Kant’s reply. When after a conciliatory exchange of letters between Garve and Kant (X 328–33, No. 201 [184], and X 336–43, No. 205 [187]) the original review was published in 1783 in the Allgemeine deutsche Bibliothek, Kant still had little reason to be impressed. It was again Hamann who, in a letter to Scheffner in February 1784, reported that Kant was working on a ‘Counter-Critique’ (Antikritik) of Garve’s ‘Cicero’ that was, as a matter of fact, intended as a retort against the unabridged review of the Critique (Ⅳ 626). It is difficult to say whether Hamann’s testimony is credible.

   Kant would have been upset by Garve’s new publication – even if the two men had not previously come into conflict – by Garve’s blatant, uncritical eudaemonism as well as the lack of systematic rigour of his supplementary Philosophical Remarks and Treatises, rather than his translation of Cicero’s three books On Duties, with which Kant had long been familiar in the original Latin.²¹ In other words: if it is true that ‘Garve’s “Cicero”’ inspired Kant to turn his attention to the foundations of moral philosophy it was probably Garve’s work, rather than Cicero’s. Yet by the end of April 1784 Kant apparently decided to abandon the plan of writing a response to Garve in favour of a short, foundational ethical treatise – a prodromus or ‘forerunner’ of moral philosophy, as the ever-prolific Hamann calls it in his letters (Ⅳ 627). If for a while Kant still intended to attach a direct reply to Garve as an appendix to the Groundwork it did not find its way into the final version.²² Kant sent the manuscript of the Grundlegung to Hartknoch in September 1784. It was published, after some delay at the printer’s office, at the Easter book fair of 1785. Kant received his first copies on 8 April of that year.

   It is difficult to say how much of Kant’s response to Garve’s ‘Cicero’ was in the end incorporated into the Groundwork.²³ The fact that Kant mentions neither philosopher should not, of course, be taken as evidence that he did not intend them to be his targets – Kant rarely refers to his most prominent opponents by name, and he may well have thought that he succeeded in elucidating common moral thought while Garve, who pretends to do the same on Cicero’s behalf, failed.²⁴ Moreover, there are some striking similarities between the two projects. At a rather superficial level, Cicero divided his brief treatise on duty into three sections (or ‘books’), and so did Kant. More interestingly, the Groundwork – particularly Section Ⅰ – contains a plethora of allusions to ancient themes: Kant rejects an ethics of social status, the moral sufficiency of a desire for honour and, above all, the identification of happiness with the highest good. The first variant of the categorical imperative – the formula of universal laws of nature – is clearly intended as a sensible restatement of the Stoic thesis that we should strive to live in accordance with nature. But this is where similarities end. Kant did not need Garve’s translation to remind him of the Stoic principle, which was still popular with eighteenth-century thinkers like Wolff and Baumgarten; and the other variants are hardly directed against Cicero. Moreover, the Groundwork is too complex, even as a piece of philosophical rhetoric, to be inspired by two second-rate philosophers. Kant adapted the notion of a moral commonwealth or ‘kingdom of ends’ from Leibniz, and it had been in place as an ideal long before Kant wrote the Groundwork. Most importantly, the main innovation of the Groundwork, Kant’s theory of morality as autonomy, can hardly be reduced to a reaction to Garve.²⁵

Kant’s Groundwork: synopsis of the argument

Preface

1 Classification of the disciplines of philosophy, according to their subject matter and mode of cognition (IV 387–8)

2 Why pure moral philosophy, or a ‘metaphysics of morals’, is necessary (IV 388–91)

3 The project of grounding a metaphysics of morals (IV 391–2)

Section I: Transition from common to philosophic moral cognition of reason

1 On the unconditional value of a good will (IV 393–4)

2 A morally good will, not happiness, is the natural purpose of reason (IV 394–6)

3 Elucidation of the concept of duty by means of three propositions (IV 397–401)

   a The general concept of a good will must be made more determinate by analysing the concept of duty (IV 397)

   b Proposition 1: An action that coincides with duty has moral
       worth if and only if its maxim produces it by necessity, even
       without or contrary to inclination (IV 397–9)

   c Proposition 2: The moral worth of an action does not lie in the
       effect intended but rather in its maxim [to be judged by the
       standard of a formal principle] (IV 399–400)

   d Proposition 3: Duty is the necessity of an action from reverence for the law (IV 400–1)

4 The law of duty, general conformity to law as such, is the condition of a will that is good in itself (IV 402–3)

5 Concluding remarks: common and philosophic moral cognition of reason (IV 403–5)

Section II: Transition from popular moral philosophy to the metaphysics of morals

1 Preliminaries (IV 406–12)

   a The origin of the concept of duty is not empirical but a priori (IV 406–8)

   b On the limited value of exemplars in ethics (IV 408–9)

   c True and false popularity in moral philosophy (IV 409–10)

   d The primacy of metaphysics in moral philosophy (IV 410–12)

2 The doctrine of imperatives (IV 412–20)

   a The will as the capacity to act in accordance with the representation of laws (IV 412–13)

   b Imperatives necessitate an imperfect will to act in accordance with laws (IV 413–14)

   c Imperatives, hypothetical and categorical: skill, prudence, morals (IV 414–17)

   d How are all of these imperatives possible? (IV 417–20)

3 The categorical imperative (IV 420–1)

   a Derivation of the general formula of the categorical imperative from its concept (IV 420–1)

   b The general formulation (IV 421)

4 The first variant: universal laws of nature (IV 421–4)

   a The universal-law-of-nature formulation (IV 421)

   b Application of this formula to the four examples of duty (IV 421–4)

5 Interlude (IV 425–7)

6 The second variant: rational creatures as ends-in-themselves (IV 427–31)

   a Derivation of the ‘formula of humanity as the end-in-itself’ from the concept of a will (IV 427–9)

   b Application of this formula to the four examples of duty (IV 429–31)

7 The third variant: autonomy in a kingdom of ends (IV 431–6)

   a Derivation of the formula of autonomy from the other two (IV 431)

   b A universally legislative will is independent of all interest (IV 431–3)

   c Self-legislation, morality and the kingdom of ends (IV 433–4)

   d A moral being possesses dignity, not a price (IV 434–6)

8 Reflections on the variant formulations of the categorical imperative (IV 436–40)

   a The connection between the three variants of the categorical imperative (IV 436–7)

   b Review of the Groundwork so far: the good will and the formulations of the categorical imperative (IV 437–40)

9 The autonomy of the moral will (IV 440–4)

   a Autonomy and heteronomy (IV 440–1)

   b Division of ethical theories according to the principle of heteronomy (IV 441–4)

10 Transition to Section Ⅲ : How is a synthetic practical proposition possible ? (IV 441–5)

Section III: Transition from the Metaphysics of Morals to the Critique of Pure Practical Reason

1 The concept of freedom is the key to the explanation of the autonomy of the will (IV 446–7)

2 Freedom as property of the will of all rational beings (IV 447–8)

3 The interest attaching to the ideas of morality (IV 448–53)

   a Preparation of the ‘circle’: our consciousness of freedom and morality are not grounded in any conventional interest (IV 448–50)

   b The suspicion of a ‘circle’: freedom and morality (IV 450)

   c The escape: we step outside the circle when we consider ourselves members of an intellectual world (IV 450–3)

4 The ‘deduction’: how is a categorical imperative possible? (IV 453–5)

5 The extreme boundary of all practical philosophy (IV 455–63)

   a The problem of reconciling natural necessity and free will does not yet mark the extreme boundary of practical philosophy (IV 455–7)

   b We are conscious of our free will but cannot cognise or explain it (IV 457–9)

   c The inexplicability of the interest we take in morality is the outermost boundary of moral philosophy (IV 459–63)

6 Conclusion: Comprehending that we cannot comprehend morality (IV 463)