Many recent commentators have noticed how Adorno, in his late works, borrows Kant’s definition of enlightenment to define key areas of his own critical practice. These discussions, however, have failed to notice how these late borrowings present an image of Kant’s enlightenment which is diametrically opposed to his previous discussions. By tracing the development of Adorno’s engagement with Kant’s essay, I discover Adorno deliberately sublating Kant’s definition as to enable its incorporation into his own works. Further, the article will examine some problems which appear to arise for Adorno when borrowing Kant’s definition of enlightenment in his late works, which coalesce around the topics of negativism and the prospects for societal change.

In §3 of the Critique of Judgement Kant argues that if the feeling of pleasure were a sensation distinct from whatever representation gives rise to the feeling, then we would be – in the terminology of the Metaphysics of Morals – rational beings (vernünftige Wesen) but not moral beings (Vernunftwesen); we would inescapably (and blamelessly) be hedonists. I reconstruct this at first glance strange argument and suggest, first, that Kant’s actual view of pleasure is an attitudinal theory that avoids the problem of hedonism. Second, the argument of §3 is to be understood in the context of Kant’s emphasis on moral feeling and its cultivation in his writings since the Critique of Practical Reason.

After providing a brief overview of Marcus Willaschek's Kant on the Sources of Metaphysics, I critically reconstruct his account of ‘transcendental realism’ and the role that it plays in the dramatic narrative of the Critique of Pure Reason. I then lay out in detail how Willaschek generates and evaluates various versions of transcendental realism and raise some concerns about each. Next, I look at precisely how Willaschek's Kant thinks we can avoid applying the ‘supreme’ dialectical principle (for every conditioned there is a totality of conditions which is unconditioned) to the domain of appearances. Finally, I call into question Willaschek's efforts to appropriate the lessons of the Transcendental Dialectic without following Kant into transcendental idealism.

According to Kant’s Critique of the Power of Judgement, in the end all estimation of magnitude is sensible, or ‘aesthetic’, and the absolutely great in aesthetic estimation is called ‘the mathematical sublime’. This article identifies the relevant sensible element with an inner sensation of a temporal tension: in aesthetic comprehension, the imagination encounters an inevitable tension between the successive reproduction of a magnitude’s individual parts and the simultaneous unification of these parts. The sensation of this tension varies in degree and facilitates aesthetic estimation in general. But in the special case where it comes to involve a sense of exceeding our imagination’s limit, we judge a magnitude to be mathematically sublime. Pace Kant, I argue that this limit is private and contingent rather than transcendental, such that the judgement of the mathematical sublime is neither universal a priori nor necessary.

Neo-Kantian philosophers see accountability as a key property of autonomy, or of social freedom more broadly. Autonomy, among those theorists, is, I contend, implicitly co-conceived with responsibility, producing a quasi-juridical conception of autonomy and a limiting notion of freedom. This article criticizes the connecting of freedom with accountability on a number of grounds. First, various conceptions of autonomy not only operate without a notion of accountability, but, in fact, would be impaired by an accountability requirement. Second, the neo-Kantians are unable to defend the freedom enhancing properties that are supposedly brought about by the giving of reasons for one’s beliefs and actions. Third, the project of accountability is indifferent to personal outlooks, not because it takes a holistic perspective, but because of its interest in social convergence.

Jürgen Habermas’ discourse theory of morality should be understood, in metaethical terms, as a constructivist theory. All constructivist theories face a Euthyphro-like dilemma arising from how they classify the constraints on their metaethical construction procedures: are they moral or non-moral? Many varieties of Kantian constructivism, such as Christine Korsgaard’s, classify the constraints as moral, albeit constitutive of human reason and agency in general. However, this constitutivist strategy is vulnerable to David Enoch’s ‘shmagency’ objection. The discourse theory of morality, by classifying the constraints on the metaethical construction procedure (principles (D) and (U)) as non-moral, can avoid this problem.