Kant’s Derivation of Imperatives of Duty

Abstract

On the currently dominant reading of the Groundwork, Kant’s derivation of ‘imperatives of duty’ exemplifies a decision procedure for the derivation of concrete duties in moral deliberation. However, Kant’s response to an often-misidentified criticism of the Groundwork by G. A. Tittel suggests that Kant was remarkably unconcerned with arguing for the practicality of the categorical imperative as a decision procedure. Instead, I argue that the main aim of Kant’s derivation of imperatives of duty was to show how his analysis of the form of moral judgement is indeed presupposed in the four types of moral imperative that philosophers of his time recognized.

1. Introduction

Contemporary readers of Kant are almost unanimously united in the assumption that Kant’s famous derivation of ‘imperatives of duty’ in Section II of the Groundwork introduces the Formula of Universal Law (FUL) and the Formula of Humanity (FH) as ‘tests’ or ‘decision procedures’ for the derivation of concrete duties. ¹ In this way, Kant’s Groundwork holds the promise of being both objective and practicable: because the categorical imperative abstracts from all contingent, empirical facts about human nature, it is applicable everywhere and at all times; and because the categorical imperative allegedly provides a decision procedure for the derivation of concrete duties, we expect that it will be useful in guiding our moral deliberation. Precisely, this promise of an objective and practicable foundation for normative ethics has made Kant’s ethics attractive to many readers.

However, Kant’s own engagement with his contemporary critics complicates this standard assumption about the Groundwork’s famous derivation of imperatives of duty. An early reviewer had criticized the Groundwork for providing a principle that is ‘unfit for practical use’ in determining which actions are right and thus concluded that the Groundwork did not provide any new principle for the ‘reform’ of ethics. In his response, Kant claimed that his reviewer had hit the mark better than he might have realized because the Groundwork simply never intended to provide such a ‘new principle’ for the reform of moral philosophy, but merely to capture the formula of our common moral knowledge. When we read Kant’s reply together with the original criticism, it becomes clear that Kant himself was remarkably unconcerned with arguing for the practicality of the categorical imperative as a decision procedure for the derivation of concrete duties. ²

Kant’s response should make us open to reassessing the standard assumption about the Groundwork’s examples of derivations of imperatives of duty and to reconsider what their purpose might be. I argue that Kant attempted to show, in his derivation of imperatives of duty, how his analysis of moral judgement and its fundamental principle underlies the four types of moral imperatives that philosophers of his time recognized and that outlining a decision procedure for the derivation of concrete duties was not his main concern in section II of the Groundwork. In other words, Kant’s infamous ‘derivation of imperatives of duty’ is best understood literally as a derivation of the traditionally accepted types of imperatives, rather than a derivation of duties.

Unfortunately, the most important translation of Kant’s work into English still today misidentifies the author of the commentary mentioned above. ³ Moreover, the actual commentary, authored by Gottlob August Tittel, has until recently not been translated into English. ⁴ At first sight, both the misidentification and the lack of an English translation might appear to be of historical interest only. But if we take a close look at Tittel’s actual criticism and keep in mind that Kant arguably meant to demonstrate how the categorical imperative underlies the four standardly recognized types of moral imperatives, we can also better appreciate Kant’s response.

Revising our assumptions about the purpose of Kant’s derivation of imperatives of duty in the Groundwork also implies a break with a long tradition of Kant scholarship. The most influential examples of this tradition are no doubt Onora O’Neill’s Acting on Principle, ⁵ Christine Korsgaard’s well-known ‘Kant’s Formula of Universal Law’, ⁶ and John Rawls’ Lectures on the History of Moral Philosophy, in which he characterizes the categorical imperative as providing a ‘procedure by which that imperative is applied to us as human beings’ (2000: 162). ⁷ Today, the standard assumption of this tradition still permeates introductory texts and popular philosophy. ⁸ So far, only a handful of scholars have deviated from this standard assumption about the Groundwork, and most deviations have been far less radical than the textual evidence would allow. Moreover, the relatively few deviations so far rely on philosophical resources outside the Groundwork for a full motivation. By contrast, I argue that the Groundwork taken together with Kant’s rejoinder allows for the deviation. ⁹ To my knowledge, until now only Ido Geiger (2010) has argued in print that Kant’s four examples of duties in the Groundwork’s discussion of the FUL are not meant to provide examples of a universalization test. ¹⁰ But since this pathbreaking departure from the standard assumption also still omits a full explanation of Kant’s likely reason for going through the traditional four kinds of imperatives with both the FUL and FH, my discussion below intends to fill this gap.

In this way, my argument also joins recent commentators who have emphasized the importance of understanding Kant’s Groundwork as containing a defense of ‘immediate’ common moral knowledge (Callanan 2019; Grenberg 2013; cf. Sticker 2016). Since this recent literature on the method and structure of the Groundwork has so far omitted a detailed discussion of Kant’s aim of the ‘derivation of imperatives of duty’ in light of his response to G. A. Tittel, my argument provides additional support to the reading of the Groundwork as providing a defense of ‘immediate’ or ‘unreflected’ common moral knowledge. Indeed, I believe that despite its invaluable contributions, recent scholarship pointing to Kant’s defense of ‘immediate’ common moral knowledge retains a lacuna as long as it does not provide an explanation of the purpose of Kant’s ‘derivation of imperatives of duty’ in section II of the Groundwork. For if the standard view of Kant’s derivation as exemplifying a decision procedure was correct, then it would imply that either Kant’s moral philosophy is much more reformist than he claims, ¹¹ or that ordinary ethical life typically consists in applying universalization tests, or possibly both. ¹²

Revising our assumptions about Kant’s derivation of imperatives of duty in section II of the Groundwork also carries consequences for Kant scholarship in general. This passage is commonly viewed as the most explicit passage in which Kant derives concrete duties from the categorical imperative by means of a decision procedure; by contrast, his other references to universality and humanity are not usually taken to be straightforward examples of the application of such a test. Consequently, if Kant is not concerned with the introduction of a decision procedure in section II of the Groundwork, Kant scholarship would have to look to other parts of Kant’s opus for paradigmatic examples of the role of the categorical imperative in moral deliberation, and this may ultimately require us to rethink the practical purpose of Kant’s moral philosophy more generally.

In section 2, I present Tittel’s criticism of the Groundwork and Kant’s response in the preface to the Critique of Practical Reason. In section 3, I argue that both Kant’s Groundwork as well as its historical-philosophical background provide some pushback against reading Kant’s infamous ‘derivation of imperatives of duty’ as examples of a decision procedure; instead, I argue that Kant’s infamous four examples in the Groundwork attempt to show that his analysis of the structure of moral judgement underlies the general types of imperatives already recognized by his contemporaries. In section 4, I argue that the same is true of Kant’s discussion of the FH in section II of the Groundwork. I conclude with some promissory remarks about the upshots of my reading.

2. No ‘new principle’ but a ‘determinate formula’

Less than a year after the 1785 publication of the Groundwork, a book-length commentary by G. A. Tittel reproached Kant for his failed attempts to ‘reform’ moral philosophy by providing an a priori foundation for a theory of right action. In the course of his criticism, Tittel covered a whole range of topics that are still echoed in contemporary criticisms of Kant: for instance, Kant’s insistence that a goodwill is the only thing of unlimited moral value, prior to happiness (Tittel 1786: 7–13, 37–8), the talk of ‘pure’ practical reason (pp. 17–29), Kant’s conception of obligation (pp. 29–32), as well as his obscure terminology and ‘mysticism’ (p. 4). Importantly, Tittel also argued that Kant’s ‘highest principle of morality’ would be practically useless. For, he claimed, the categorical imperative is entirely formal, and thus empty, and since it is empty, it would not generate any concrete duties unless it was supplemented with a calculation of consequences and with theoretical knowledge of which consequences would further people’s happiness. Tittel continued to charge Kant with having smuggled just such empirical considerations into his discussion of the categorical imperative in the derivation of imperatives of duty (pp. 32–7). Thus, Kant allegedly failed to give moral philosophy a practically useful a priori foundation. Elaborating on his criticism, Tittel asks: ‘Is the entire Kantian reform of ethics indeed to confine itself to just a new formula?’ (p. 35). ¹³ And he continues:

For what help should an empty formula be that one has made to look purely rational, but that is unfit for practical use unless one supplements it with material of experience? It cannot help that Kant now says ‘all duties depend upon this one principle of morality, as far as the kind of obligation, not the object of their action, is concerned’. ¹⁴ For one clearly saw in these examples [the four derivations of imperatives of duty] that the point of obligation, or rightfulness, must be determined only through the nature of the object and its relation to the human being’s nature, through its consequences. The categorical imperative (so that I say it once and for all), together with its imprinted stamp of pure reason, appears to me to be a veritable plaything (Spielwerk). ‘Act so that you can will your maxim to become universal’ may sound pretty and new. But if one searches for meaning, not words, then it can mean nothing else but: ‘Act so as you think it good and beneficial, if all other beings (that can act so) acted this way’. Thereby a prior, and separately determined, concept of goodness and benefit is necessarily presupposed; or the entire formula does not make sense. I do not learn what is good from the formula; instead, the formula merely tells me: ‘Do that which is good (generally beneficial)’. Not a single step am I taken further by the amended, seemingly new formula: ‘Act so that you can will that your maxim become universal’ than by one of the oldest known principles: Act so that perfection be furthered by your actions and maxims (taken in their entirety and universality). I still must first abstract the concept of the good and perception from the stuff of experience (empirically) through collection and comparison of cases. (Tittel 1786: 35–7)

In summary, Tittel complains that Kant’s ‘new principle’ does not provide us with a practically useful theory of right action – of determining which actions are right and which actions are wrong. Allegedly, the categorical imperative does not allow us to derive duties without presupposing empirical content and thinking about consequences, and once we have seen that Kant’s ‘new principle’ smuggles empirical content into its test, we must conclude that his ‘new principle’ has failed to reform the moral discipline by giving it an a priori basis. For by itself, it is entirely useless as a decision procedure. If we apply Kant’s alleged test, we either do not get any answer at all, or we must presuppose empirical content, which we can only get by comparing different actions and their results. But in the latter case, Kant’s formulations of the categorical imperative do not do the promised work; instead, we must rely on consequentialist considerations.

In his preface to the Critique of Practical Reason, Kant responds to Tittel’s criticism in a striking footnote. In that preface, Kant tells the reader that his second Critique ‘presupposes, indeed, the Groundwork of the Metaphysics of Morals, but only insofar as this constitutes preliminary acquaintance with the principle of duty and provides and justifies a determinate formula of it;* otherwise, it stands on its own’ (CPrR, 5: 8, emphasis added). In the footnote concerning this ‘determinate formula’, Kant remarks as follows:

A reviewer who wanted to say something censuring [the Groundwork] 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. But who would even 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 of what duty is or in thoroughgoing error about it. But whoever knows what a formula means to a mathematician, which determines quite precisely what is to be done to solve a problem 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. (CPrR, 5: 8)

On first sight – and especially to a reader unfamiliar with Tittel’s commentary – this footnote might seem innocuous. After all, few present-day readers of Kant’s ethics would follow Tittel’s description of Kant’s ethics as attempting to introduce a new principle into ethics. For Kant presents the categorical imperative as the form of pure practical reason. That is, Kant presents the categorical imperative not as a new principle but as the underlying principle that was there all along in our practical judgements, and that his philosophy supposedly, for the first time, unearths in its precise form. Therefore, this footnote might not appear to demand much attention.

But on a second look, Kant’s response becomes more interesting. As we saw, Tittel’s reason for denying that the categorical imperative would be a ‘new principle’ was that he found it practically useless for the derivation of concrete duties. So why would Kant be happy to go along with this criticism if the purpose of the derivation of imperatives of duty in the Groundwork had been to exemplify a useful decision procedure for the derivation of concrete duties? If we read Kant’s derivation of imperatives of duty as the introduction of a decision procedure, his response to Tittel should puzzle us.

Unfortunately, there are still few discussions of Kant’s response to Tittel. ¹⁵ But his response is quite significant in the context of the present-day reception of Kant’s ethics. For Kant here explicitly addressed someone who shared what I have called the standard assumption about Kant’s Groundwork – namely that its examples of ‘derivations of imperatives of duty’ were supposed to demonstrate the application of a decision procedure – and then criticized Kant’s book for not providing a principle that could be put to use as a test for the determination of concrete duties. And in this light, Kant’s response to Tittel’s criticism is striking: in his footnote, Kant decidedly did not insist on the practicality of the formulas of the categorical imperative presented in the Groundwork as providing a decision procedure for the derivation of duties, nor did he provide new examples of such a derivation or point to another passage elsewhere that might provide such examples. Instead, Kant’s response was to question the reasonableness of the very idea that moral philosophers should attempt to introduce any ‘new principle’ of morality, as if people before had not known what duty is.

One may object to this reading that even if Kant was apparently not interested in insisting on the practical usefulness of the categorical imperative in his response to Tittel, he might have responded to this worry indirectly elsewhere. While it is important to distinguish Tittel’s criticism of uselessness from the conceptually separate claim that Kant’s ethics must be given a specifically consequentialist reading in order to be practical, Tittel joins these two charges together. Thus, if Kant had responded to the charge of implicit consequentialism elsewhere, might such a response not provide an indirect response to Tittel’s charge of uselessness ¹⁶ and thus vindicate the standard reading of the Groundwork? I believe that this would be a premature conclusion. For even if Kant had such a response to the charge of implicit consequentialism elsewhere, he explicitly decided to respond to Tittel’s separately. And what is at issue in Tittel’s ‘no new principle’ objection and Kant’s response is whether or not the formulas of the categorical imperative as presented in the Groundwork were meant to exemplify a decision procedure to be actively applied by agents to guide them in their moral deliberation and choice. Thus, even if Kant wanted to eventually provide a practically useful decision procedure elsewhere by way of answering the charge of implicit consequentialism, this would not change his response regarding the Groundwork.

Apart from providing a theoretical insight into the form of common moral judgement, what else, if anything, might his formula then be able to do? After all, Kant apparently believed that anyone who ‘knows what a formula means to a mathematician’ will understand its significance: it purportedly ‘determines quite precisely what is to be done to solve a problem’. And is this not just another way of saying that the formulas of the categorical imperative should be used as a decision procedure? If one were to read this remark with an emphasis on the ‘what is to be done’ without regard to the final formulation ‘to solve a problem’, one might indeed get the impression that Kant must be talking about a test or decision procedure that tells us what to do. However, this would be to read Kant’s passage too selectively. Instead, the crucial question is what kinds of problems Kant might have had in mind here. Mathematical formulas help solve problems, so what problems does moral philosophy help solve? If Kant had in mind theoretical problems, which the formulas of the categorical imperative help us solve, they are presumably the philosophical questions he addresses in the work he was writing, the Critique of Practical Reason, and the later Metaphysics of Morals. By contrast, if Kant was thinking about practical problems, what might he have had in mind? Strikingly, Kant did not believe that figuring out what to do would be a practical problem requiring philosophical knowledge of the formulas of the categorical imperative. As Kant remarks repeatedly, he believed that anyone, even children, could judge ‘without hesitation’ what is right to do without engaging in any technical decision procedure or having been taught the formulas of the categorical imperative. ¹⁷ By contrast, the central problem Kant identifies as providing a practical reason for doing moral philosophy is the difficulty of overcoming rationalizations against prior (common) moral knowledge in favour of inclinations. ¹⁸ As Kant points out in section I of the Groundwork, ‘[t]he human being feels within himself a powerful counterweight to all the commands of duty, which reason represents to him as so deserving of the highest respect…’. Because of this aspect of human nature, Kant adds that ‘from this there arises a natural dialectic, that is, a propensity to rationalize against those strict laws of duty and to cast doubt upon their validity, or at least upon their purity and strictness, and, where possible, to make them better suited to our wishes and inclinations…’. And according to Kant’s remarks in the Groundwork, it is this natural dialectic that gives a practical purpose to moral philosophy:

In this way common human reason is impelled … on practical grounds themselves, to go out of its sphere and to take a step into the field of practical philosophy, in order to obtain there information and distinct instruction regarding the source of its principle and the correct determination of this principle in comparison with maxims based on need and inclination, so that it may escape from its predicament… (G, 4: 405)

Importantly, this account of the practical difficulties that require moral philosophy for their solution also matches Kant’s formulation of the four examples of the derivation of imperatives of duty. For the latter conspicuously take the form of agents already aware of what they ought to do, but who are in danger of rationalizing away their moral knowledge in favour of one selfish inclination or another. ‘Someone feels sick of life because of a series of troubles that has grown to the point of despair’, so much so that the person just has enough strength left to put themselves to reflect about their situation. ‘Another finds themselves urged by need [durch Not gedrungen] to borrow money’. This is a person in despair (Not), not someone unclear about the morality of truthfulness. The third is tempted by the comfort of an easy and pleasurable life as opposed to the hassle of developing excellence in their natural gifts. Only the fourth seems to be doing alright for themselves and is merely tempted by a desire to ignore the hardship of others in order to enjoy their own fortune unimpeded (G, 4: 421–3). In none of these examples did Kant choose a merely puzzled agent; rather, they all struggle against the temptation to rationalize away their common moral knowledge.

Still, one might object that, surely, the analogy with a mathematical formula solving a problem must entail a test or decision procedure. Taking one of the (if not the) best-known mathematical formulas, the Pythagorean theorem, as an example, readers are frequently tempted to think of Kant’s analogy between a formula of duty and a mathematical formula in the following way. The formula of duty might stand to duty as a² + b² stand to c². On this view, a² + b² would be analogous to the formulas of the categorical imperative and c analogous to duty. On this interpretation of the mathematical analogy, the categorical imperative would provide a decision procedure for the determination of a duty, as the root of a² + b² determines c. However, this way of interpreting the analogy would be misleading. For a² + b² is only part of the formula, the formula being a² + b² = c². In the analogy with the Pythagorean theorem, duty is not analogous to c; rather, duty is analogous to a right triangle. Thus, in Kant’s analogy comparing a formula of duty to a mathematical formula, a² + b² = c² would be analogous to the categorical imperative, and duty analogous to a right triangle. Consequently, the mathematical analogy does not support interpreting the categorical imperative as an algorithm for determining duty as a² + b² determines c². How exactly a formula can be practical depends on the respective formula and the problems in question. But here, the mathematical analogy ends. Which problems the formulas of the categorical imperative can help us solve depends on what Kant believed made moral judgement difficult, not on mathematics. Consequently, Kant’s remark that a formula ‘determines quite precisely what is to be done to solve a problem’ does not entail that he would have been concerned with exemplifying a decision procedure in the Groundwork.

3. Kant’s derivation of imperatives of duty

If Kant’s own response to Tittel shows him as remarkably uninterested in defending the practical usefulness of the formulas of the categorical imperative as a test or decision procedure, how has the standard reading first become the uncontroversial assumption among readers of the Groundwork? Due to its popularity, it has so far received little explicit defense. But two things conjointly make the standard reading so intuitively plausible on first sight: Kant’s metaphorical remark about the fundamental principle of moral judgement being a ‘compass’ in section I and Kant’s terminology of ‘derivation’ in section II. I briefly consider the former, before discussing Kant’s terminology of ‘derivation’ and offering an alternative reading.

After arriving at a first formulation of the categorical imperative in the Universal Law formulation in the first section of the Groundwork, Kant tells us that:

Common human reason also agrees completely with this in its practical appraisals and always has this principle before its eyes. … [T]o this problem, whether a lying promise is in conformity with duty, I ask myself‘: would I indeed be content that my maxim (to get myself out of difficulties by a false promise) should hold as a universal law (for myself as well as for others)? and could I indeed say to myself that everyone may make a false promise when he finds himself in a difficulty he can get out of in no other way? Then I soon become aware that I could indeed will the lie, but by no means a universal law to lie … Thus, then, we have arrived, within the moral cognition of common human reason, at its principle, which it admittedly does not think so abstractly in a universal form but which it actually has always before its eyes and uses as the norm for its appraisals. Here it would be easy to show how common human reason, with this compass in hand, knows very well how to distinguish in every case that comes up what is good and what is evil, what is in conformity with duty or contrary to duty, if, without in the least teaching it anything new, we only, as did Socrates, make it attentive to its own principle; and that there is, accordingly, no need of science and philosophy to know what one has to do in order to be honest and good, and even wise and virtuous. (G, 4:403–4, emphasis added)

Here, Kant speaks metaphorically of an innate principle as a ‘compass’ that we always have before our eyes that makes moral judgement possible; and yet, importantly, we do not make use of this principle by consciously thinking about it in its abstract form. ¹⁹ On the standard reading of the Groundwork, Kant is here alluding to an innate formal principle that also serves as a decision procedure in moral deliberation, which he would then exemplify in his derivations of imperatives of duty. However, it is important to keep in mind that such an innate principle, the form of moral judgement, need not necessarily figure in agents’ conscious deliberation – as Kant makes clear, common moral judgement does not usually ‘think so abstractly’ through some universalization test. ²⁰ And correspondingly, examples of common moral judgement conforming to this principle may not be meant to exemplify a decision procedure for practical deliberation but ‘to confirm at each stage of a very abstract argument about an a priori rational principle that we are in fact tracking morality’ (Herman 2011: 51). Importantly too, we should not presuppose a reading of Kant’s compass-metaphor that is already influenced by previous assumptions about the ‘derivation of imperatives of duty’ in section II; in light of Kant’s response to Tittel, it is worth reconsidering his infamous derivations themselves.

According to the standard assumption about the Groundwork, section II reiterates the highest principle of morality that Kant had just excavated from common moral knowledge in the form of the Formula of Universal Law: ‘act only in accordance with that maxim through which you can at the same time will that it become a universal law’ (G, 4: 421) and then continues to demonstrate how this formula of his highest principle of morality helps us figure out what is truly right and wrong. This demonstration allegedly takes the form of a derivation of four duties from the categorical imperative: the duty not to take one’s own life, the duty not to lie to others, the duty to cultivate one’s talents, and finally, the duty to help others in need. And at first sight, the wording of section II of the Groundwork indeed suggests this ambitious reading. Immediately after introducing the Formula of Universal Law, Kant says:

Now, if all imperatives of duty can be derived from this single imperative as from their principle, then, even though we leave it undecided whether what is called duty is not as such an empty concept, we shall at least be able to show what we think by it and what the concept wants to say. (G, 4: 421)

The first half of the sentence clearly talks about deriving ‘imperatives of duty’. And after a quick digression into the right ways of formulating the categorical imperative, Kant then announces: ‘We shall now enumerate a few duties in accordance with the usual division of them into duties to ourselves and to other human beings and into perfect and imperfect duties’. (G, 4: 421) – followed by his famous discussion of the four canonical duties of his discussion of morality. Hence, on first sight, it might seem that Kant’s main concern here is to demonstrate how the categorical imperative might enable us to derive concrete duties from it.

But before jumping to old conclusions, it is helpful to keep some distinctions in Kant’s terminology in mind: Kant here talks about deriving imperatives of duty, and then says that he will demonstrate this by enumerating (herzählen) ‘a few duties’. In Kant’s terminology, imperatives are distinguished from both duties and obligation. The clearest statement of this distinction can be found in Kant’s introduction to the Metaphysics of Morals: ‘An imperative is a practical rule by which an action in itself contingent is made necessary’ (MM, 6: 222). By contrast, ‘[o]bligation is the necessity of a free action under a categorical imperative of reason’ (6: 222). And ‘[d]uty is that action to which someone is bound. It is therefore the matter of obligation…’ (6:223). In other words, a duty is the specific action type that an imperative commands (something that it is right to do), ²¹ and an ‘imperative’ is the representation of the action as something to be brought about. ²² But we need not look ahead to the Metaphysics of Morals. Already in his discussion leading up to the ‘derivation of imperatives of duty’, Kant makes clear that imperatives are specific ways of representing action. There, he repeatedly points out that only through representations can there be practical judgement, and that these representations are properly called ‘imperatives’. More precisely, ‘imperatives’ are the linguistic formulations that attempt to express the representation of a necessary (free) action. This is perhaps most clearly expressed in the following passages:

The representation of an objective principle, insofar as it is necessitating for a will, is called a command (of reason), and the formula of the command is called an imperative. (G, 4: 413, emphasis added)

Now, all imperatives command either hypothetically or categorically. The former represents the practical necessity of a possible action as a means to achieving something else that one wills (or that it is at least possible for one to will). The categorical imperative would be that which represented an action as objectively necessary of itself, without reference to another end. Since every practical law represents a possible action as good and thus as necessary for a subject practically determinable by reason, all imperatives are formulae for the determination of action that is necessary in accordance with the principle of a will which is good in some way. (G, 4: 414, emphases added)

With these terminological distinctions in mind, it is possible to think about another purpose than the provision of a test or decision procedure for Kant’s derivation of imperatives of duty: Kant’s aim in section II of the Groundwork is to argue that all representations of actions as intrinsically good share in the form of one single highest principle of morality, the categorical imperative. Thus, Kant first claims that ‘all imperatives of duty can be derived from this single imperative [the categorical imperative] as from their principle’. ²³ Then, he enumerates a handful of examples of duties (action types) that he believes his readers will readily agree to be morally obligatory in order to substantiate his claim. And finally, he attempts to demonstrate that despite their many differences, the respective imperatives through which practical reason represents the necessity of actions always share the same form – whether people describe them as perfect, or imperfect, towards others or oneself. ²⁴

Before going through the four examples in his argument, Kant also tells us about his motivation for this argument. He thinks it important to ‘enumerate a few duties in accordance with the usual division of them into duties to ourselves and to other human beings and into perfect and imperfect duties’ (G, 4: 421, emphasis added). The infamous distinction between perfect and imperfect duties does not originate with Kant but with the writers of the Natural Law tradition. ²⁵ Its most likely origin is the work of Hugo Grotius, and it was taken up in the works on moral and political philosophy by such writers as Pufendorf and Wolff, with whose work on natural law Kant was familiar. ²⁶ Traditionally, this distinction was supposed to signify a distinction between duties that should or should not be enforced. And at the time of the publication of the Groundwork, it appears to have been an open philosophical question how this traditional distinction between perfect and imperfect duties could be explained and justified. ²⁷ This distinction might easily lend itself to the thought that the representation of the necessity of an action by practical reason differs correspondingly. Given the philosophical context, Kant’s remarks in the second part of the Groundwork on his ‘derivation of imperatives of duty’ must be read as acknowledging this debate about different kinds of duties and rejecting the thought that the will’s representation of actions as to be done would differ correspondingly. ²⁸

But in what sense is this argument that all practical representations of actions as intrinsically good share a common form a derivation of imperatives of duty? I take it that Kant uses the term ‘derive’ (ableiten) quite consistently to mean simply the drawing of a conclusion from a principle: ‘every syllogism is a form of derivation [Ableitung] of a cognition from a principle’ (CPR, A300/B357). Thus, when Kant uses the term ‘derive’ here at G, 4: 421, ²⁹ he likely means that there can be arguments in which the formula of the categorical imperative serves as a major premise, and from which we can draw conclusions about the form of other commonly accepted imperatives of morality. A helpful analogy might be to compare Kant’s argument in the ‘derivation of imperatives of duty’ with some hypothetical argument about formal logic. Kant claims that the form of the commonly accepted moral imperatives just is the form of the categorical imperative. His argument proceeds by taking four representations of actions (maxims) commonly thought contrary to duty and shows how these four representations also contradict the form of the categorical imperative. By analogy, we might claim that the forms of certain commonly accepted logical functions in fact share in the form of another logical function – take, for instance, A ∨ B, ¬B → A and (P | P) | (Q | Q). In an argument resembling Kant’s ‘derivation of imperatives of duty’, we might try to show how contradicting one logical function also means contradicting the other logical functions. And the conclusion we might draw is that the different functions in fact share a specific form. In a similar way, I believe, Kant wants to show that the form of the categorical imperative is indeed the form underlying all representations of actions as intrinsically good. ³⁰

At this point, one might object that surely Kant’s central concern must have been to provide a decision procedure in section II of the Groundwork because he talks about a ‘derivation’ after the conclusion of his notorious four examples. In most contemporary English translations, Kant is there recorded as saying: ‘These are a few of the many actual duties, or at least of what we take to be such, whose derivation [Ableitung] from the one principle cited above is clear’ (G, 4: 423–4). But as these translations usually point out, this is not what Kant says in the original. In the first edition of the Groundwork, Kant says ‘[t]hese are a few of the many actual duties, or at least of what we take to be such, whose division [Abtheilung] from the one principle cited above is clear’ (Kant 1785: 57). Starting with Gustav Hartenstein in Kant 1838, some editors – and most English translators ³¹ – have decided to replace Kant’s Abtheilung (division, or classification) with the term ‘derivation’. ³² Unfortunately, the contemporary German terms for division (Abteilung) and derivation (Ableitung) are anagrams, which might predispose contemporary readers to assume a mistake in the original printing – however, they were not during Kant’s time. In Kant’s time, the spelling contained an additional h (Abtheilung) (which is the spelling of the original editions of the Groundwork as well as in volume IV of the Akademie Ausgabe, first published 1903). This alone makes it rather unlikely that the original wording would have been a mistake. Moreover, the context of the passage does not obviously support reading ‘derivation’ instead of the original ‘division’. In this passage, Kant argues that maxims that violate perfect duties cannot even be thought as universal laws of nature, whereas those maxims that violate only imperfect duties could be thought as universal laws of nature but cannot be willed as such universal laws. As Allen Wood has pointed out, this makes it sound like Kant is more concerned with the classification of distinct types of duties in this passage than with a potential derivation of duties (2017: 26–7). ³³ Moreover, even on the assumption that Kant really meant to write ‘derivation’, it would still be far from obvious that Kant’s reference to a derivation of duties could not be understood as shorthand for the derivation of types of duties as they are commonly recognized as indicated by the foregoing discussion of the Groundwork. In short, the sentence following the infamous four examples is not good evidence in favour of the standard assumption.

The reading of the Groundwork I have defended so far is further supported by Kant’s last sentence in the very same passage following the ‘derivation of imperatives of duty’. Kant there concludes his discussion of the four famous examples by pointing out that ‘so all duties, as far as the kind of obligation (not the object of their action) is concerned, have by these examples been set out completely in their dependence upon the one principle’ (G, 4: 424, emphasis added). Recall that in Kant’s terminology, an imperative is the representation of the necessity of an action, and the obligation is the subjective necessitation that results from the representation of the necessity of an action. Thus, by demonstrating that the traditionally distinguished imperatives of morality are all based on the same underlying principle, Kant also takes himself to demonstrate that the ‘kind of obligation’ is equally based on the same underlying principle: the categorical imperative. In other words, by showing that the representation of the necessity of an action has the same form in all instances of duties traditionally distinguished, Kant also believes himself to show that the kind of obligation an agent takes herself to be under depends in each case on the same form of representation. This indicates once more that the purpose of the ‘derivation of imperatives of duty’ was to show how moral imperatives of all classes share a particular form, rather than providing examples of a decision procedure.

4. Kant’s introduction of the Formula of Humanity

Kant’s introduction of the ‘Formula of Humanity’ (FH) in the Groundwork is the second passage apparently aimed at providing a test or decision procedure for the derivation of concrete duties from the categorical imperative. In the forgoing sections of the Groundwork, Kant had first argued that moral cognition concerns the representation of actions as good independently of inclination (a good will) and subsequently argued that the foundational principle of the capacity to represent actions in this way (the categorical imperative) could be captured adequately by the Formula of Universal Law. In the second section of the Groundwork, Kant proceeds with his analysis of this capacity to represent actions as morally good and to determine oneself to action through such representations. He then eventually concludes that such a capacity supposes that every rational being must represent her own existence, and the existence of every other rational being, as an end in itself (i.e., as capable of setting oneself ends autonomously). ³⁴ This fact, that rational beings must represent their own and others’ existence as an end in itself, must serve as ‘an objective principle from which, as a supreme practical ground, it must be possible to derive all laws of the will’ (G, 4: 429, emphasis added). He continues: ‘The practical imperative will therefore be the following: So act that you use humanity, whether in your own person or in the person of any other, always at the same time as an end, never merely as a means. We shall see whether this can be carried out’ (G, 4: 429). Kant then turns back to the examples of duties discussed earlier in his ‘derivation of imperatives of duty’ to see if ‘all laws of the will’ can indeed be derived from the new formulation of the categorical imperative offered here.

Once again, on first sight this might appear as an attempt to derive duties from the categorical imperative by means of another decision procedure – this time, from the FH. However, I believe that we should again be cautious about such an interpretation of the passage. For one, the fact that Kant moves through the same examples suggests that the point of his discussion of the FH is serving the same purpose as his discussion of the FUL in the derivation of imperatives of duty discussed above. Earlier, Kant attempted to demonstrate that the form of all four types of imperatives commonly recognized is that of a universal law; now, he attempts to show that the form of practical reason also entails a special status of those beings who represent actions through it. In each case, Kant can only demonstrate this by going through examples of each type of imperative commonly recognized by his contemporaries, in order to give some plausibility to the claim that he is really unearthing the form of all moral cognition.

Importantly too, note that Kant explicitly speaks about ‘laws of the will’. While it can be frustrating to have an author use the term ‘law’ in so many ways, we should be careful when conflating different kinds of laws with each other in Kant’s text. In this context, ‘laws of the will’ are the laws through which rational beings represent actions and ends as good: imperatives. Recall that earlier in the Groundwork, Kant had described the capacity to represent moral actions in this way:

The categorical imperative would be that which represented an action as objectively necessary of itself, without reference to another end. Since every practical law represents a possible action as good and thus as necessary for a subject practically determinable by reason, all imperatives are formulae for the determination of action that is necessary in accordance with the principle of a will which is good in some way. (G, 4: 414, emphasis added)

For the purpose of the present discussion, what matters is the connection between what Kant calls a practical law that represents a possible action as good and what he calls the imperative: practical laws of the will and imperatives are just one and the same thing. It is through certain laws that reason represents actions as necessary. The different kinds of practical laws are different kinds of imperatives.

Thus, when Kant goes through the four examples of duties again here, he is doing two things. First, he strengthens his previous claim that the different kinds of imperatives commonly recognized at his time share the same form. Second, Kant’s previous argument regarding the FUL in the derivation of imperatives of duty did not imply that it perfectly, or even sufficiently, captured the form of representation of moral actions. In his discussion of the FH, Kant further specifies the form that all moral representation of ends supposedly share: not only are all moral imperatives universal, but according to Kant they also concern the status of rational beings as ends in themselves.

5. Concluding remarks

In light of Kant’s reluctance to insist on the practical usefulness of the categorical imperative as a decision procedure, and in light of the historical and philosophical context of the Groundwork, we have good reasons to doubt the standard interpretation of Kant’s infamous derivation of imperatives of duty as providing examples of a decision procedure for the derivation of concrete duties in moral deliberation. On a closer look, Kant’s argument is concerned with showing that the categorical imperative, as the form of practical reason, underlies all commonly recognized types of imperatives. And in his response to Tittel in the preface to the Critique of Practical Reason, Kant explicitly replied to someone who adopted the standard reading of the Groundwork and criticized it for failing to provide a practically useful test for the derivation of concrete duties. In this light, his remark that Tittel understood his Groundwork better than Tittel himself seemed to have realized can help us better appreciate the purpose of Kant’s derivation of imperatives of duty. For if we read Kant’s derivation of imperatives of duty as demonstrating that his highest principle of morality really underlies all imperatives of duty commonly recognized, then Kant’s response makes perfect sense: there was no need to argue about the practicality of the categorical imperative as a decision procedure because the Groundwork just did not intend to provide a new principle ‘fit for practical use’ in this way.

If my exegesis of Kant’s derivation of imperatives of duty was correct, then we cannot focus entirely, or even primarily, on the Groundwork in order to understand if, and how, the formulas of the categorical imperative were supposed to figure in practical deliberation. A full discussion of the practical purpose of the formulas of the categorical imperative would go far beyond the scope of this article. Here, I merely want to gesture toward some larger consequences for how we may understand the practical purpose of Kant’s moral philosophy more generally.

If Kant believed that either (i) pre-philosophical, common moral knowledge knows right from wrong by applying a test or decision procedure along the lines of the formulas presented in the Groundwork, or (ii) philosophically informed moral reflection would proceed through such a decision procedure, then we should expect Kant to exemplify or describe this decision procedure somewhere. So far, Kant’s derivation of imperatives of duty in the Groundwork has commonly been viewed as the most explicit such example – but if this standard assumption is ill-founded, where else should we look?

The most important passage outside the Groundwork in which Kant might appear to describe a decision procedure for moral deliberation is the ‘typic’ of pure practical judgement in the Critique of Practical Reason (CPrR, 5: 69). As some authors have suggested, Kant there arguably responds to Pistorius’ charge of implicit consequentialism. And since Tittel’s objection was that Kant’s ethics could not be practical unless it availed itself of consequentialist considerations, one might arguably take the typic to be Kant’s implicit insistence on the practical usefulness of the categorical imperative. ³⁵ As mentioned above, this still leaves intact Kant’s response to Tittel as indicating that the Groundwork itself was not concerned with exemplifying a decision procedure – rather, this would indicate that Kant might have introduced such a decision procedure in the typic. However, it is controversial if the typic serves chiefly as exemplifying a decision procedure to be used by agents in practical deliberation, or an unconscious cognitive mechanism. ³⁶ Although I am sceptical of reading the typic as the introduction of a decision procedure, even this reading would benefit from my present argument. Taking the typic to describe a decision procedure suggests a break with the Groundwork, since in the second Critique neither the categorical imperative itself nor its formulas, but a type of the moral law, would provide the decision procedure (Zimmermann 2015: 439–40). However, if my reading of the Groundwork is correct, then proponents of this reading of the typic would have no need to posit any break in Kant’s ethics with respect to the derivation of imperatives of duty, for the former was simply not in the business of exemplifying a test or decision procedure.

Apart from this brief section of the second Critique, we might also look to the Metaphysics of Morals, Kant’s work on pedagogy, and the ‘Common saying’ essay. However, the reason Kant scholarship has focused on the derivation of imperatives of duty in the Groundwork for examples of how the categorical imperative ought to figure in moral deliberation is that few passages elsewhere provide straightforward examples of an agent applying a decision procedure. Despite Kant’s brief remark about the categorical imperative as allowing us to ‘test’ maxims for their universality in his introduction to the Metaphysics of Morals (MM, 6: 225), commentators have frequently noted the almost complete absence of the Formula of Universal Law in the work; ³⁷ and it has also been questioned whether Kant’s references to ‘humanity’ throughout the Doctrine of Virtue can plausibly be read as constituting derivations from the Formula of Humanity (Kuehn 2010: 23–7). Similarly, Kant’s ‘casuistical questions’ do not contain the application of any test or decision procedure. And Kant’s remarks about common moral cognition in his lectures on pedagogy, his ‘fragments of a moral catechism’, as well as his ‘Common saying’ essay do not portray agents as walking through any kind of decision procedure either. ³⁸ Finally, the same applies to Kant’s notorious essay ‘On a supposed right to lie from philanthropy’. There, Kant provides a brief paragraph that might appear like an explanation of why truthfulness is really a universal, unconditional duty; but even there, Kant’s actual aim is to argue that being truthful is not a duty to a specific person regarding a particular object but a duty to humanity in general (SRL, 8: 426). Consequently, if we discount the Groundwork as providing examples of a decision procedure for the derivation of concrete duties, then the passages where Kant seems to describe anything like the application of such a test are both remarkably rare and remarkably less straightforward than we might have expected. And given how rarely Kant describes anything resembling a decision procedure on the basis of the formulas of the categorical imperative outside the Groundwork, we may ultimately have to conclude that Kant was simply not interested in general in providing a new principle that would be ‘fit for practical use’ for the derivation of concrete duties – in other words, that for Kant the practical purpose of moral philosophy was not primarily in giving us a test or decision procedure.

Consequently, this revision of our background assumptions about Kant’s moral philosophy could open up a newly unprejudiced engagement with his remarks about the natural dialectic and the practical necessity of moral philosophy (see section 2, above). For if the practical purpose of moral philosophy is to help us overcome the natural dialectic, then we should not expect – and certainly must not presuppose – this therapeutic benefit of moral philosophy to function merely, or even primarily, in terms of a test or decision procedure.