In Kantian Subjects, Karl Ameriks adds new dimensions to his rich account of the complex relation between Kant and his German successors. Building on his earlier work, Ameriks solidifies and expands (in the book’s first part) his highly influential reading of Kant by engaging with some recent scholarship, and he further develops (in the second part) his forward-looking account of what contemporary philosophy can take away from the early Jena Romantics, by ‘pointing out underappreciated ways in which they carry forward Kant’s spirit in a manner most appropriate for our own times’ (p. 4).

The first part of the book covers a wide array of central topics in Kant. Perhaps the most central claim here, which Ameriks defends from a variety of angles, is that Kant’s notion of autonomy must not be understood along subjectivist lines, as a capacity to do what one wants or to ‘create’ values ex nihilo without any antecedent normative constraints. Rather, as Ameriks persuasively shows, in the Groundwork Kant expounds the principle of autonomy based on his prior conviction that this principle has a supreme normative authority and necessity which cannot be grounded in merely contingent features such as our arbitrary preferences (pp. 19–21, 24–5, 58–9). The idea of moral autonomy, for Kant, entails that the will is governed by a normative law that, qua law, possesses a strict universality or intersubjectivity that can only derive from reason, where the term ‘reason’ refers, from the start, to a shared faculty that contains objective standards for right or good agency (pp. 23, 99–101). Closely connected to this central issue are two further important themes: first, Ameriks’s insightful proposal to interpret Kant’s talk about ‘making’, ‘giving’ or ‘prescribing’ laws of reason in a way that avoids the absurd implication that these phrases refer to contingent empirical processes of literal creation (pp. 24–8, 100–2, 113–17); second, his compelling response to the common worry that Kant’s identification of freedom with moral autonomy implies that evil choices cannot be regarded as free and imputable (pp. 26–34).

Ameriks also, with characteristic subtlety, preserves the important connection between these normative issues and core metaphysical features of Kant’s doctrine. What struck me as especially insightful here is his very helpful discussion of how we might make some literal sense of Kant’s obscure but indispensable notion of ‘noumenal duration’ (pp. 80–5), and his suggestion that in Kant metaphysical characterizations of the human mind rely upon a prior grasp of our normative competences and rational standards (p. 109).

I want to discuss in some detail two further topics from the first part of Kantian Subjects. In chapter 3, Ameriks engages with Eckart Förster’s account of why the Kantian philosophy remained unsatisfying for post-Kantian idealists such as ‘Förster’s heroes, Goethe and Hegel’ (p. 44). Ameriks argues that this dissatisfaction arises from a very ambitious, ultra-Cartesian conception of philosophy as strict science that rests on absolutely certain foundations (p. 37). It need not arise if one favours ‘a more modest conception’ (p. 41) of what Kant’s critical philosophy can and should aspire to. On this modest conception, Kant’s doctrine begins from given objective facts (pp. 38, 41) and argues regressively from these facts to their synthetic a priori presuppositions. In this context, Ameriks seeks to downplay the impact that Garve’s criticism of Kant’s moral theory in the Canon had on the development of Kant’s practical philosophy after 1781 (p. 45). Without taking a stand on the issue of what influenced Kant’s post-1781 thinking, I agree with Ameriks that (pace Garve) the Canon does not commit a petitio principii by first arguing for the existence of God from the validity of the moral law and then making the validity of the moral law depend on God (pp. 46–7). Kant assumes, apart from any theological commitments, that there are valid a priori moral laws. He justifies this assumption via deference to ‘the moral judgment of every human being’ who clearly reflects on the idea of a moral law (A807/B835). This starting point is not outrageously dogmatic in light of the modest conception of Kant’s project.

My concern lies with Ameriks’s claim that, despite some ‘odd formulations’, the Canon expresses ‘standard Kantian doctrine, explained at greater length in later publications’ (p. 43, cf. 45). The odd formulations include the claim that the presupposition of God and an afterlife are inseparable from the idea of moral obligation (A811/B839) because without these presuppositions the ideas of morality of God cannot be sufficient ‘incentives’ for the execution of moral commands (A813/B841). Ameriks concedes that according to the Canon, the moral law does not by itself provide ‘a clear incentive … that is, inner motivation’ but insists that Kant here nevertheless regards moral principles as ‘internally binding’ (p. 50). However, it is hard to see how moral norms can bind or obligate my will if these norms provide me with no sufficient motive or reason for action. What seems odd about the Canon is precisely its seeming proposal that moral norms are laws of reason that have necessary normative validity even though they fail to give us a clear motive for moral choice. After 1781, Kant realizes that moral principles could not claim supreme normative authority if they were unable to influence our will without assistance from extra-moral, partly non-rational (sensible) desires. Thus, he regards the idea that ‘the thought of duty’ is ‘of itself a sufficient incentive’ (MM, 6: 393) or that a priori formal laws of reason sufficiently determine the will (CPR, 5: 22) as a necessary concomitant of the idea that pure practical reason is a source of necessary practical laws (CPR, 5: 25).

In chapter 9, Ameriks seeks to defend Kant’s doctrine against the Hegelian charge of subjectivism. Ameriks shows that Kant supports his idealism by arguing ‘from (rather than to) unquestioned objective … assertions of a number of strong cognitive claims’ (p. 140). Thus, Kant’s eventual designation of space and time as (metaphysically) ‘subjective’ forms of intuition is not only compatible with but driven by the consideration that these forms are conditions of our objective empirical and a priori mathematical cognition (pp. 141–2; cf. 117–18). Ameriks then turns to the Hegelian worry about subjectivism concerning the categories. This worry concerns, chiefly, Kant’s view that the categories have no meaning or sense apart from their relation to spatiotemporal intuition. Ameriks rightly stresses that the pure (unschematized) categories do not lack all ‘meaning’ in a contemporary sense: they do have genuine representational content. He goes on to make the stronger claim that this content signifies metaphysical aspects of non-sensible, noumenal reality: it ‘corresponds to different kinds of entities’ (p. 146), for example, to substances and accidents. Apart from our sensible intuition, we cannot apply the categories to determine which things are substances or causes and accidents or effects (p. 151). Hence, the pure categories ‘lack … theoretical cognitive significance in the specific sense of “objective validity”’ but they, or the ‘looseness of “play”’ of thoughts that they enable, are ‘still within the context of a broadly objective framework’ (p. 146).

It is not so clear, however, what this broadly objective framework is meant to be. Kant frequently identifies the pure categories with logical functions for bringing given sensible manifolds to the objective unity of consciousness. In light of this identification, I find it hard to see how the categories could signify objective metaphysical determinations once they are considered in abstraction from all sensible manifolds. When Kant stresses that the pure categories are not limited to our human intuition, he adds that they are not free from their restriction to sensible intuition in general (B148, B150). Thus, they could not be applied to objects of a non-sensible intuition (B149). It seems, then, that Kant is not just careless or hyperbolic when he says that the categories, ‘without the data of sensibility’, would be ‘merely subjective forms of the unity of understanding, but without object’ (A287/B343), since they are ‘nothing but’ logical functions for combining sensible manifolds (A245; Prol, 4: 324). These phrases are clearly meant to mirror his characterization of the pure forms of intuition as ‘nothing but’ (A42/B59) ‘subjective condition(s) of sensibility’ (A26/B42).

Ameriks rightly emphasizes that we need the pure categories for practical reasoning about noumenal objects such as freedom (p. 148). But the question is whether the pure categories suffice on their own for the representation of noumenal objects, or whether they can only represent noumenal objects once their purely logical content has been supplemented (and perhaps transformed) by the objective form of practical reason. To my mind, the latter alternative is suggested by Kant’s argument, in the second Critique, that the categories acquire from the moral law a ‘practical meaning’ for their transcendent designation of noumena (CPR, 5: 49–50, 58).

This debate raises very thorny textual and philosophical issues. I suggest that one need not take a stand on these issues to rebut the Hegelian worry about the ‘subjectivity’ of the categories. Clearly, our forms of sensibility have no (positive) meaning or application in regard to things in themselves. Yet, as Ameriks very persuasively argues, this does not detract from their objective cognitive character. Thus, it should also be possible to give a sufficient defence of the objectivity of the categories via the argument that they are necessary conditions for the objective unity of our sensible experience and thereby for all our determinate theoretical cognition.

I now want to summarize Ameriks’s fascinating argument in the second part of Kantian Subjects. Ameriks here proposes that the early Romantics are especially valuable now because they have revealed to us a way in which philosophy can progress ‘toward an extensive historical, rather than pseudo-scientific, presentation of its main points’ (p. 154). This valuable view must be separated from more questionable aspects of Romanticism, such as its initial commitment to a Cartesian ideal of philosophy (as in the early Schelling, pp. 156–7), its ‘wildly ambitious teleological character’ (p. 161) or the reactionary tendencies of later Romanticism (pp. 178–80). Instead, Ameriks focuses on the forward-looking yet historically informed spirit of the Romantics as thinkers who are ‘concerned with our universal progress as historical beings’ but recognize that such progress cannot take place without taking a ‘fresh look’ at past epochs and cultures (p. 180). This implies a deeply historical model of philosophizing, according to which ‘philosophy should not present itself as a timeless axiomatic complex but must display a sequential and argumentative resolution of the specific difficulties of one’s immediate predecessors’ (p. 222; cf. p. 233).

Ameriks puts special emphasis on Hölderlin, whom he reads (especially in Hyperion) as proposing ‘the possibility of a Kantian Romanticism’ (p. 197) in response to the modern disenchantment induced by scientific and political revolutions that threaten to undermine the idea of a meaningful destination for mankind (p. 190). This response to our late modern predicament avoids what Ameriks regards as excessive aestheticism and subjectivism (in Nietzsche and others: pp. 215–17) because it remains committed to the Kantian emphasis on the equality and dignity of human beings and to the progressive social, moral and political goals of Kantian rationalism (pp. 223–25; cf. 245). Yet, the Romantic ‘non-technical style’ (p. 241) of philosophy exemplified in Hölderlin (and others) retains a strong aestheticist element: it is a universal ‘poetry’ qua imaginative rhetoric, whose appeal to feeling and imagination (p. 224) is intended ‘to have a public effect in changing human sensibility in a more progressive direction’ (p. 228). Ameriks deems this ‘poetic’ conception of philosophical expression a model worth emulating in our time because it has great potential for public appeal and ethical change (p. 250) and can thus serve as an effective way of ‘bringing about’ not just philosophical progress but also the broader socio-cultural ‘realization of enlightened aims’ (p. 243).

I find this historically informed, ‘broadly artistic’ (p. 243) model of philosophizing highly appealing. However, I must qualify my enthusiasm with two pessimistic remarks. First, it seems to me that the Romantic model of philosophy that Ameriks regards as especially valuable for our times is not just ignored but actively dismissed in current mainstream ‘Analytic’ philosophy, whose disregard for cultural-historical considerations alongside its idolatry of (on the one hand) formalism and technical rigour and (on the other hand) the natural sciences rather looks like a return to the old scientistic worldview with its (as Ameriks nicely puts it, p. 226) alienation of human subjectivity. Second, one might question whether the kind of broad public effects and changes that Ameriks hopes to result from the works of enlightened Romanticists are not rather to be expected from more dubious figures and movements. This worry is especially serious in our time, where the increasing erosion of education (not to mention Bildung) leads to ever decreasing prospects for a ‘significant literate audience that has a chance to hear, and … an uncorrupted desire to listen to’ (p. 229) Romantic voices. But already in 1915, Rilke – whom Ameriks mentions (p. 228) as one spiritual successor of the Romanticist project – laments that the First World War, ‘this nothing-but-hopelessness (Heillosigkeit) was necessary in order to exact proofs of heartiness, devotion and greatness. Whereas we, the artists … evoked nothing in those same human beings, brought nothing to ascend, could not transform anybody’ (Schnack Reference Schnack and Scharffenberg2009: 501; my translation).

These gloomy ruminations should not detract from the fact that Kantian Subjects is a magnificent book which succeeds in the unusual feat of using highly sophisticated, illuminating philosophical scholarship regarding Kant and his German successors as a way of articulating a bold and appealing vision for contemporary engaged philosophizing.