The purpose of A Guide to Kant’s Psychologism is to provide an accessible overview of Waxman’s account of Kant, Kant’s precursors, and his continuing relevance as laid out in a series of earlier books, Kant’s Model of the Mind (1991), Kant and the Empiricists: Understanding Understanding (2005) and Kant’s Anatomy of the Intelligent Mind (2014). Waxman’s central claim is that Kant’s picture of knowledge, mind and world is a kind of psychologism, inspired by his empiricist predecessors, Locke, Berkeley and Hume, but developed in a particular way that Waxman calls a priori psychologism.

The book thus combines a reading of Kant that is different from pretty much everything in the literature with considerable historical sensitivity to the empiricist roots of Kant’s philosophy. (He has little to say about Kant’s roots in rationalism, in particular Leibniz and Wolff; he mentions Wolff only once.) This many-sided book does more: it argues, contrary to many, that Kant’s work is compatible with most contemporary science including, interestingly, the theory of evolution, and it attempts to locate Kant’s view of the mind in the spectrum of contemporary naturalist, functionalist, materialist and neurophilosophical alternatives.

The book has two parts, one on the empiricists and the later Wittgenstein, one on Kant’s a priori psychologism as Waxman sees it (the second part also contains the discussions of Kant’s relationship to contemporary science). After an introductory overview of Waxman’s radical departure from what he says is an anti-psychologistic consensus in recent work on Kant (he should have added ‘in English’), chapters 2 to 4 discuss the three most-read empiricists, Locke, Berkeley and Hume, and lay out their contributions to psychologism. The target is Platonism, which is roughly the view that the necessity involved in the necessary connections found in mathematics, semantic entailment and the like is independent of the mind. Chapter 5 is about Wittgenstein’s alleged conventionalism, which for Waxman follows Hume’s and which Kant’s psychologism would also reject. The second part, comprising chapters 6–10, lays out Kant’s a priori psychologism. The sequence of chapters, ‘The Kantian Cogito’ (6), ‘The Logical I’ (7), ‘The Aesthetic I’ (8), ‘The Objective I’ (9) and ‘The I of Nature’ (10) does not reflect any sequence in the Critique of Pure Reason, which should cause an eyebrow or two to rise (see below). A concluding chapter urges that Platonism and conventionalism are the only alternatives to psychologism and assesses them, as well as laying out what Kant can still contribute to the contemporary study of the mind.

Every one of the four sides of this book would repay close examination, but readers will probably be most interested in Waxman’s psychologistic readings of Kant. What is it and is it Kant, the Kant of the Critique of Pure Reason, in particular?

Psychologism simpliciter is the view that logic, mathematics, semantic entailments, etc., are as they are because the mind is structured as it is or functions as it does. Crudely, not: minds take 2 + 2 to equal 4 necessarily because 2 + 2 equals 4, but: 2 + 2 = 4 has whatever necessity it has because minds – properly functioning minds anyway – take 2 + 2 to equal 4. The deepest and most obvious problem with psychologism is that there is no apparent way to get necessity out of contingent facts about minds. This was one of the reasons for Frege’s well-known condemnation of psychologism. Kant clearly believed that mathematics, some propositions and the like are ‘necessarily and universally’ true. We should, and in due course will, ask: can a priori psychologism do better with necessity than psychologism simpliciter?

By ‘a priori psychologism’. Waxman does not mean simply that the mind provides some of the structures of its own experience. He does mean that but there is nothing revolutionary in this as a reading of Kant. Waxman means something much stronger: that a certain kind of consciousness of self, namely unified apperception as captured in the phrase ‘I think’, structures one’s concepts and experience. (Since only consciousness is in play here, not the mind as a whole, ‘consciousness-logism’ would be a better name than ‘psychologism’ – uglier but better.) Structure how? Well, here Waxman says many things. ‘Essential ingredients of [a concept’s] content [Waxman also says ‘meaning’ – AB] [are] derived from consciousness’ (pp. 10, 11). A priori consciousness ‘implies or entails’ certain representations (p. 145). Everything universal (for example, the pure concepts of the understanding) ‘incorporates pure … self-consciousness into its representation’ (p. 147; see also pp. 169 and 182). He also speaks of ‘“I think”-generated concepts’ (p. 146; see also p. 178), and he speaks of explicating or explaining concepts by laying out the role that unified consciousness of self plays in them (p. 146 – note that the last four quotations occur within a single page).

This is a bewildering array of relationships: ‘entailment’ (which implies two things in a semantic relationship), constituent (‘incorporates’) (which implies one), causal (‘generated’) (which again implies two but in a causal relationship), and ‘explication’ (which implies two things but not in any other specific relationship). Waxman never ties the different descriptions together but the general idea is that, for Kant, unified consciousness of self as the single common subject of one’s experience, which is a priori, is not only closely linked to one’s experience, the standard Kantian view, but is somehow incorporated into and/or generates experience or central elements of experience.

Waxman offers few examples of Kant saying any of these things and I do not myself know of many. Instead, he argues that Kant is committed to a priori psychologism by other things he says. A leading argument: ‘Pure universal self-consciousness is … indistinguishable from logic’s most fundamental representation’ (p. 178). This for Waxman entails that ‘the logical universe is … part of a priori psychology’ (p. 178), specifically, the unified self-consciousness expressed by ‘I think’. Given the quotation above about incorporation, I think that he should have said that a priori psychology is part of the logical universe, but whichever way around, no texts are cited. Here at least the book is not interpretative. Moreover, the argument is invalid. It slides from qualitative indistinguishability (two things being exactly similar in the target respect) to numerical identity (there being only one thing).

In fact, as a reading of the first Critique as a whole, Waxman’s account faces formidable challenges. I first consider three to do with the ‘I think’, then turn to some others.

1. 1. In the first edition of the Critique, the term ‘I think’ does not even occur until the chapter on the Paralogisms, i.e. not until halfway through the book (the chapter begins in A341). Even worse for Waxman, that chapter was apparently one of the last bits of the Critique to be written. If so, Kant had all but finished the work before he first mentioned ‘I think’. Even transcendental apperception makes its first appearance only as late as halfway through the Transcendental Deduction (A108).

2. That there is no mention of ‘I think’ until after A341 partly explains the paucity of quotations from Kant in the book – in the first edition, there are no relevant passages to quote until after A341. (All but one of Waxman’s references to the first edition in his crucial chapter 6 are to passages after A341.) It also accounts for the fact that the order of chapters 6 to 10 does not follow any order in Kant – there is no relevant order in Kant to follow. (I would argue that much the same is true of the second edition but do not have space to do so here.)

3. 2. When Kant finally gets to ‘I think’, he tells us that it is the sole text – not of his work but of rational psychology, a view that Kant is attacking (A344/B402). Further, the chapter’s goal is negative; Kant urges that the self as presented by ‘I think’ displays nothing about the nature of the self (A355). Neither seems to be consistent with Waxman’s claims.

4. 3. If the use of ‘I think’ by rational psychology is a main target of the Paralogisms chapter, one would expect Waxman to give that chapter a lot of attention, especially given the recent ground-breaking work on Wolff by Corey Dyck and others. He does not.

5. 4. In the first Critique, the derivation of the Categories starts with Aristotelian forms of judgement, not unified self-consciousness. Waxman is aware of this and urges that the forms of judgement must supplement the ‘I think’ (pp. 149 and 193). He also sees that, for Kant, mathematics has its origins in space and time (p. 164, though he goes back on this later – pp. 178, 179). Kant says that mathematics is synthetic, not analytic (‘analytic’ in Kant’s sense has to do with meaning, not Waxman’s), so grounded in the structure of experience, not self-consciousness. But then, what is left for unified consciousness of self to do? For Kant, especially in the second edition, unified consciousness of self is a datum, something whose possibility is to be explained, not something that explains other things.

6. Further: Kant allows that unified consciousness of a diversity of representationally integrated objects is necessary for unified consciousness of self. But Waxman wants unified consciousness of self to be incorporated into, a constituent of, a representationally integrated field of objects (pp. 147, 169, 182). Given what Kant says about the grounds of the categories and mathematics, why would he adopt such a view?

7. 5. Kant aims to show which propositions are ‘necessarily and universally’ true (‘synthetic a priori’) (in the Transcendental Analytic) and which cannot be warranted (through much of the Transcendental Dialectic). There is a huge problem getting general necessity out of the ‘I think’. The necessary and universal propositions that Kant thinks he can warrant are true of everything we can know. However, each instance of unified consciousness of self is at most a necessary feature of only one mind. I do not see how Waxman can make the leap from something true of me, even necessarily true of me, to anything necessarily true of everything. Yet he says that he is doing so repeatedly (e.g. on pp. 169, 190 and 193).

8. 6. Would consciousness of self being a priori help? The term ‘a priori’ in Kant is slippery. At different times, he means at least three different things by it: learned independently of experience, acquired in some way other than by upbringing and social interaction (these capture what he means when he calls something ‘pure’) and necessary and universal. Grant that some form of self-consciousness is a priori in the first two senses (roughly Waxman’s definition of ‘a priori’). That something is a priori in either or both of these senses would not entail that anything is necessary or universal, the third sense.

9. Putting (5) and (6) together and pace Waxman (cf. p. 169), his a priori psychologism seems to have as severe a problem warranting necessity as psychologism simpliciter.

10. 7. Kant has two official methods, transcendental argumentation (in particular, to the necessary conditions of experience) (A94/B126) and construction in the imagination (roughly, thought experiments about the necessary features of kinds of thing) (B14–18; A713/B741). These are not methods for studying the role of the ‘I think’ in experience.

11. 8. Finally, contrary to Waxman’s whole project, in the first Critique Kant had little to say about the origins of elements of cognition and what he did say was mostly negative.

Waxman’s view is original and challenging even if, as I think, its target is at most a minor and late topic of the Critique of Pure Reason as originally written. That said, late in the writing, Kant did begin to say striking things about ‘I think’: for example, ‘“I think” … is not an experience, but is the form of apperception. … [It] attaches to and precedes every experience’ (A354). Such murky sayings do indeed deserve detailed study. Because such remarks first appear only late in the writing, perhaps later works might be a richer hunting ground for Waxman. I am thinking in particular of Kant’s writings on the role of autonomous, self-legislating self-consciousness in morality and his (apparent) view that physical science is a self-positing of some kind in the (to be sure unfinished and highly repetitive) Opus Postumum.