Significance and Symbolism collects Timmons’ groundbreaking work on Kant's ethics spanning more than twenty years. Having the papers collected in a volume is not just convenient; it also helps us appreciate the immense importance of Timmons’ contributions to the field. Each paper is clear, insightful and well-argued. The collection ranges from topics that receive sustained attention from Kant interpreters (such as the problem of relevant descriptions) to under-explored issues in the analytical tradition of Kant scholarship (such as questions about the psychology of devilish vices).

The first chapter, ‘Necessitation and Justification in Kant's Ethics’, aims to explain why Kant considers the hypothetical imperative to be analytic while he takes the categorical imperative to be synthetic. Timmons’ discussion is mostly focused on trying to understand how imperatives can be analytic or synthetic, given their imperatival form. After all, Kant's definitions of analytic and synthetic seem to focus on judgements that have a subject/predicate form (CPR, B10). Timmons’ strategy is to focus on the relation of necessitation, given that according to Kant ‘oughts’, or imperatives, are ‘The representation of an objective principle, insofar as it is necessitating for a will’ (cf. G, 4: 413). Timmons argues that, once we understand necessitation, we can see that these imperatives can be reduced to descriptive statements with predicative form. In particular, to say that a subject is necessitated to adopt a certain maxim is to say that she would act on this maxim if she were ‘reasoning in a completely rational manner’ (p. 24). The proposed reduction then reads as follows (S is a subject and M is a maxim):

The most general form of the hypothetical imperative, which Timmons calls ‘The Principle of Heteronomy’ (a label I find a bit jarring since a fully autonomous agent would also act from this principle), is rendered as follows:

Timmons clearly succeeds in giving the principle of heteronomy a descriptive form, but when explaining why such a principle is analytic, Timmons seems to rely on our intuitions about fully rational agents:

However, we are asking why a principle of rationality is analytic and the answer ‘Because it is a principle of closure for a fully rational agent’ does not seem to take us far. We wanted an explanation of how we can know that this is the case simply by reflecting on our concepts, an explanation of why this principle is a part of our concept of rational agency, but we seem instead only to get the stipulation that it is. Of course, explanations must come to an end, but Kant does seem to provide an explanation of why the principle is analytic:

Kant argues that for the thought of an object as the effect of my volition, the thought of means is already contained; the subject term does not seem to be a general notion of practical rationality. Timmons argues that the similarly descriptive statement of the categorical imperative (under the formula of universal law) is synthetic given that ‘our ordinary notion of practical rationality … does not contain anything like Kant's universalizability constraint’ (p. 37). To establish the categorical imperative, we need a third term connecting rationality to this constraint, namely, freedom. Since freedom and the moral law are reciprocal concepts, this connection shows that the categorical imperative is a synthetic a priori proposition. Let me just briefly mention that it is not fully clear to me how this reading establishes that the categorical imperative is synthetic. If it turns out that our conception of a rational agent includes the conception of a being acting under the idea of freedom (under the laws of freedom), would that not show that the imperative is analytic after all? But Kant's famous explanation of why every rational agent must act under the idea of freedom seems to suggest that freedom is included in our conception of a fully (practically) rational agent:

Of course, in the end Timmons must be right that the connection is not analytic, but I wish that more were said in this excellent paper to explain why this is not so.

Chapter 2, ‘Decisions, Procedures, Moral Criteria, and the Problem of Relevant Descriptions in Kant's Ethics’, defends the view that the different formulations of the categorical imperative have different functions. In particular, Timmons argues that the Formula of Universal Law (FUL) and the Formula of Universal Law of Nature (FLN) should be understood as providing a decision procedure that guides the agent from a first-person perspective in making moral decisions, while the Formula of Humanity (FH) provides a moral criterion: it specifies the property of the action in virtue of which the action is (objectively) right. I find that this understanding of the function of the different formulas has immediate advantages: it seems counterintuitive to say that my refraining from harming you, for instance, is right in virtue of something about my maxim, rather than something about you. But, more interestingly, it provides Timmons with a compelling way to solve the infamous problem of relevant descriptions for the universalization test. It seems that, by describing an action in different ways, the same action could fail the universalization test under one description but ‘pass’ it under another. Timmons points out that a moral criterion also determines moral relevance: the facts about what makes an action right also determine which features of action are potentially relevant in determining which course of action is right. Thus the FH also provides a theory of moral relevance that is further specified in the Doctrine of Virtue, as we need to determine what counts as ‘valuing our humanity’, and in particular our autonomy, given the ‘specific nature of the type (of rational) agent’ (p. 68) we are; that is, rational human agents. Timmons concludes that the system of duties presented in the Doctrine of Virtue can serve also as a test of moral relevance. Chapter 3, ‘The Categorical Imperative and Universalizability’, takes this issue further asking whether a solution to the problem of relevant description is enough to show that FLN can provide an adequate decision procedure. Timmons concludes that a solution to this problem can potentially block many of the ‘false positives’ and ‘false negatives’, but not all of them. A sufficiently morally complex action will generate a maxim that is detailed enough so that both it and its contrary can pass the FLN test. But Timmons claims that ‘it is (not) plausible to suppose that in all such cases the correct moral verdict is that both the action and its omission are morally permissible’ (106).

While these initial chapters focus on the role of FUL and FLN as decision procedures, Chapter 4, ‘The Philosophical and Practical Significance of Kant's Universality Formulations of the Categorical Imperative’, aims to explain what further roles these formulas play in Kant's ethical theory. Timmons claims that ‘the universalizability formulation is … best understood as … a set of formal constraints on what can count as a fundamental moral criterion’ (p. 121). Timmons argues that Kant identifies three such constraints: (1) lawlike character; (2) supremacy of moral reasons; (3) respect as a proper response. Together they ground the choice not only of humanity as the moral criterion, but of what is a fitting response to humanity. As for the further practical significance, the universalization tests ‘make clear a mode of practical thinking that brings out an agent's recognition of moral reasons vis-à-vis prudential reasons’ (p. 134); thus it ‘reveal(s) a kind of duplicity in an agent's attempt to justify immoral action’ (p. 135), as she is committed to the supremacy and thus universality of the moral law, but makes an exception for herself. Timmons’ treatment of these issues is exceptionally insightful.

Chapter 5, ‘Motive and Rightness in Kant's Ethical System’, examines whether Kant accepts that the rightness of an action is independent of the motive of the action, or as Timmons renders the Independence Thesis (IT):

Timmons thinks that although Kant is not committed to IT, Kant does reject what Timmons calls the Motive Content Thesis (MCT); namely, the view that ‘certain duties involve performing actions from certain motives’ (p. 150). He defends this interpretation against the challenge that one can only conform with duties of virtues by acting from the motive of duty. It seems that what is distinctive of the duties of virtue, unlike the duties of right, is that conformity to them requires that the agent acts from the motive of duty. Timmons argues persuasively that we can preserve the character of duties of virtues, as duties that cannot be coerced, without accepting MCT. In general Timmons makes a compelling case that a distinction between acting in accordance with duty and acting from duty applies to all duties, including the duties of virtue.

However, I worry that Timmons wins the battle here but loses the war. In my view, he is right in thinking that Kant's distinction between legality and morality is pervasive. For many philosophers like Ross, for whom IT was an important issue, the notion of motive, as Timmons points out, roughly corresponds to an ultimate or final end. Timmons basically preserves this understanding of motive as ‘ultimate ends of action’ (p. 149) when discussing Kant's basic motives of duty and self-love. According to Timmons, a motive for Kant should be understood as an instance of a ‘maxim of ends’ (p. 150), and adopting a ‘maxim of ends’ is understood as ‘setting oneself to bring about some state of affairs that is the end’ (p. 149). But this seems to imply that actions from duty are performed for a further end; namely, humanity or autonomy. But this not only seems in tension with Kant's formalism (that is, the idea that in acting from duty an agent is moved by the form, rather than the matter, of the maxim), but it seems also to contradict the nature of the duties of virtue. For Kant says that the happiness of others and self-perfection are ends that are duties. There is nothing that clearly corresponds to the distinction between the end of an action and an ultimate end in Kant's terminology, but the argument of the Doctrine of Virtue seems to commit Kant to the latter. Here is the crucial passage:

Here the ends that are duties are being contrasted with those ends that are means to further ends. Moreover, if the motive of duty were already understood to include an end, there would be no reason to postulate other ends that are also duties. But it is exactly because the motive of duty (or self-love for this matter) is not a further end for Kant, we can distinguish between legality and morality even with respect to the ends that are duties: the sympathetic person, for instance, has the same ultimate end as the virtuous, benevolent person. But the sympathetic person adopts this end from self-love, while the virtuous person adopts it from duty. This means that Kant's views are essentially different from other views, such as Ross’, that accept a version of IT. For Kant, but not for Ross, whether one's action conforms to a duty of virtue depends on the agent's ultimate end. Finally, it is worth noting that a version of MCT does seem to hold for the agent's choice of the highest-order maxim, for her choice of character. In the Religion the choice of the order of the incentive, the choice of whether we subordinate the incentive of self-love to the moral incentive or vice versa seems to be our most basic choice, and choosing the latter seems to a basic form of wrongdoing.

Chapter 6, ‘Kant's Grounding Project in the Doctrine of Virtue’ (co-written with Houston Smit), is an immensely rich chapter that aims to show how the duties of virtues are derived from the categorical imperative. This is a particularly difficult challenge, given that, despite appeals to something like the value of humanity throughout the Doctrine of Virtue, there is nothing like a systematic derivation from the formulas that a reader of the Groundwork might have expected. Overall, this chapter is a major achievement, one of the most important papers on the Doctrine of Virtue published in the recent past. Based on Timmons’ view that FH functions as a moral criterion, Timmons and Smit make a compelling case that the particular duties can be derived from certain mid-level principles that specify the normative significance of the value of humanity. One example of such mid-level principles is ‘the moral harm principle’, which says that ‘one can fail to respect one's own humanity (by performing) actions … that destroy or damage these end-setting powers that constitute one's humanity’ (p. 192). The principle is a clearly plausible specification of FH and serves to derive the ethical duties to oneself qua animal. It does obviously well in explaining the prohibition on suicide, but less well, as they acknowledge, in explaining the prohibition on masturbation or selling one's tooth. The arguments for the derivation of these duties from the moral harm principle is pursued in greater detail in Chapter 7, ‘Perfect Duties to Oneself as an Animal Being’. Timmons tends to see the defence of the duties as stemming from limitations on Kant's application of his own principles, or from his reliance on dubious empirical premises. Decoupling Kant's bizarre or outdated views from his ethical system has obvious advantages, but I am left with the suspicion that something important here was not captured: Kant does seem to think that there are fundamental duties that are grounded on the teleological organization of a human being, rather than on the more abstract nature of our rational powers.

Chapter 8, ‘The Moral Significance of Gratitude in Kant's Ethics’ (co-written with Houston Smit), gives a detailed account of the grounds (that is, the actions from an agent that can generate duties of gratitude), the nature and the justification of duties of gratitude in Kant's ethics, as well as an understanding of the vice of ingratitude as resting on ‘a failure to properly appreciate and respond to the ground of “real” self-esteem’ (p. 263). One of the many highlights of this insightful piece is the discussion of Kant's baffling claim that a debt of gratitude cannot be fully discharged even if ‘I return to my benefactor fifty times what he gave me’ (27: 442–3; cited at p. 253). In particular, it shows why such a non-dischargeable duty does not conflict with the ‘ought implies can’ principle. The idea is not that an act of beneficence requires me to keep showering my benefactor with gifts and services, but that I must preserve an ‘ongoing esteem for the benefactor’ (p. 255).

Chapters 9 and 10 discuss Kant's account of moral evil and vice. Chapter 9, ‘Love of Honor, Emulation, and the Psychology of the Devilish Vices’ (also co-written with Houston Smit), discusses the origin of the devilish vices, envy, ingratitude and Schadenfreude. Although the basis of our self-esteem is our dignity (recognition self-esteem) and a comparison of our actions and character with what the moral law requires (appraisal self-esteem), a natural tendency to compare ourselves as others is essential to our pursuit of self-perfection. Striving to emulate those who ‘make manifest by example the kinds of behavior and attitudes that one should strive to acquire’ is an essential element of our moral development. However, in doing so we often fall into the mistake of taking self-esteem to be comparative: I take my worth to be measured in comparison with the worth of another agent. This error gives rise to a tendency of wishing (and acting) against the well-being and moral development of others. Timmons is careful to point out that not all instances of envy, ingratitude and Schadenfreude are devilish and he provides a compelling account of when they are devilish. Chapter 10, ‘The Good, the Bad, and the Badass’, provides a systematic account of Kant's doctrine of radical evil and the degrees of evil in the famous discussion of these topics in the Religion. By insisting that the rigorist views apply primarily to transcendental, rather than empirical, psychology, the chapter admirably tries to show how Kant's rigorism can be made compatible with various aspects of our common experience of evil, such as (1) that evil comes in degrees; (2) that it is not necessarily the expression of a stable disposition and; (3) that it is often found in agents who are also partly good. The chapter shows how Kant's denial of the possibility of a diabolical will can be made compatible with cases in which agents seem to be pursuing evil for its own sake.

In sum, this is an outstanding collection that will be immensely valuable to anyone interested in Kant, Kant's ethics or any aspect of practical philosophy discussed in the book.