In his new book, Kantian scholar Henry Allison aims to “provide a comprehensive analysis of the central tenets and arguments of . . . A Treatise of Human Nature” and a “critical analysis of Hume’s views from an avowedly Kantian perspective” (p. 1). This is a book of scholarship, not one for the lay reader, but, as such, it is recommended for anyone whose philosophy engages with Humean or Kantian themes. It should thus be widely read since there’s barely a day in the life of the professional philosopher where someone does not identify themselves as, or assert that someone else is, a Humean or a Kantian with respect to one issue or another.

Chapter 1 considers Hume’s distinction between ideas (thoughts) and impressions (experience), between memory and imagination, and his account of abstract ideas. In Chapter 2 it is argued that there is partial agreement between Hume and Kant on space and time. For both, space and time are not objects, but rather, in different ways, the manner in which things appear. In Chapter 3, Hume’s fork is considered: that is, the distinction between relations of ideas (conceptual truths) and matters of fact (empirical truths), and it is shown that this is distinct from Kant’s semantic distinction between analytic and synthetic statements.

Chapters 4 through 7 turn to the central Humean topics of causation and induction. Causation for Hume is epistemologically central since for him all empirical inferences are based on relations of cause and effect. Causal relations themselves, though, cannot be perceived: we see the fire and feel the heat, and merely infer that they are causally connected. This is what Allison calls a “Hume-judgment”. For Kant, though, our very experience of the world depends on the possession of certain pure concepts of the understanding–causation being one such concept. It may be said, then, that the Humean mind enters the story one step too late” (p. 111). For Kant, one sees one thing as succeeding another: “Kant-judgments” are cognitions of objective succession that essentially involve the concept of causation.

Chapters 8 and 9 contribute to the ongoing debate concerning the relation between Hume’s skepticism and naturalism. Hume provides skeptical arguments against the existence of the external world, induction, and the self, yet also provides a naturalistic explanation of the psychological origin of these beliefs. Is Hume a radical skeptic, merely aspiring to be a cognitive scientist, or is he also interested in normative questions? These are issues that continue to be explored, particularly in the work of Don Garrett, David Owen, and Peter Millican. Allison helps the reader through these debates and argues that Hume is only skeptical towards second order philosophical attempts to ground reasoning–his target, those with an “unrealizable drive for some form of explanatory closure” (p. 4); he is not skeptical with respect to first order empirical reasoning itself. This “reflexive turn“, he claims, is “one of the high points of Hume’s theoretical philosophy” (p. 5).

In Chapter 10 we turn to Humean and Kantian forms of philosophical therapy. Hume is a skeptic, but what he himself calls a “moderate” one–a “True philosopher”: one who, as a result of skeptical argument, has the epiphany that common life is all. Hume is only temporarily wracked with “philosophical melancholy and delirium”, famously instead enjoying dinner with friends and a game of backgammon. This epistemic and psychological arc is returned to in chapter 12 on ‘Hume’s Insouciance’. Hume finds tension between the skeptical conclusions of reason and our natural human tendencies to have certain kinds of belief, whereas Kant highlights tensions in the results of reason–contradictions (“antinomies”) where it can be argued, for example, that the world must have a beginning in time, and that it must be eternal. Allison sees Kant’s transcendental idealism, and the view that everyday objects are not things in themselves, as his version of philosophical therapy.

Chapter 11 covers the immateriality of the soul and personal identity, including Hume’s “second thoughts” about the latter, and the Kantian worries that Hume does not account for the unity of thought and the first person perspective. 

One of Allison’s key claims, one that figures throughout the book, is that Hume is committed to a perceptual model of cognition, one inherited from Descartes and Locke and that this is the source of his problems. Not only does he see thoughts as images–and thus as objects that are perceived–but inference itself involves the direct and immediate apprehension of content. In contrast to such a picture of cognition, Kant’s key innovation is to see cognition in terms of judgments involving concepts. He has a discursive rather than a perceptual model of thought and it is this that allows him to offer an alternative to Humean empiricism. Allison, though, is perhaps too ready to accept this view of Hume. There would seem to be more to Hume’s account than visual comparisons between images; there are also mechanistically construed habits and principles of association. And, as Allison notes, Millican has recently argued that Hume’s skepticism is intended to attack just this perceptual model of cognition. (Although, of course, whether such associations can account for thought is something that a Kantian would question.)

That there will be objections to some of Allison’s claims and interpretations is of course inevitable, but what is also true is that readers of this book will undoubtedly come away with a richer sense of how Kantian thoughts engage with those of Hume and of–by anyone’s lights–one of the crucial transitions in the history of philosophy.

© 2013 Dan O’Brien

Dan O’Brien, Senior Lecturer in Philosophy, Oxford Brookes University