Problems from Kant (review)

The author acknowledges his debt to the "great Kant books of the 1960s, Jonathan Bennett's Kant's Analytic, and P. F. Strawson's The Bounds of Sense."Their analytical spirit lives on in this book, but the analyses are fresh, take surprising turns, and make unexpected connections to current concerns. In seventeen chapters Van Cleve deals with necessity and the a priori, space and geometry, incongruent counterparts, time, the ideality of matter, experience and objects, substance in the First Analogy, the Second Analogy, noumena, problems of the self, rational theology, and Kant and contemporary irrealism. To sustain the flow of argument, the discussion of certain details, often important, is postponed and taken up in fifteen appendices, and there are nearly seventy pages of careful notes.

Chapter 10, "Noumena and Things in Themselves," (134 ff.) criticizes in detail the prevailing "one world" or "double aspect" view of objects, namely that Kant allows just one set of objects, but two ways of "considering" them. Things in themselves are the very same objects as appearances, but considered as they are independent of our experience of them, a view developed in Henry Allison's Kant's Transcendental Idealism. Against this Van Cleve maintains (and I concur) that "appearances are intentionalia, and that intentionalia are constructions out of states of perceivers" (150). The distinction is between two separate universes of discourse, "not between two ways of discoursing about the same objects" (ibid.). There is no space to report the sharply argued detail that led him to hold the double aspect view to be unfathomable and incoherent. He takes Kant to be a phenomenalist but holding to a version far removed from the reductionist phenomenalism of Russell, Broad, Price, and others. Here the basic elements are not made out to be indubitable givens, but are, rather, sensations. The structures built on them are, according to Van Cleve, "virtual objects" in the sense in which Quine speaks about "virtual classes." To say that something is an element of a class {F:Fx} is only to say that it is F. So, likewise, to say that, e.g., a patch of red exists is short for saying that a certain representation occurs. Carnap's Kantian masterpiece, TheLogical Structure of the World (The Aufbau) could have been consulted as a comprehensive guide to the kind of virtuality here envisaged.

It is Van Cleve's stated aim never to discuss Kantian arguments without making their premise and conclusion structure clear (viii). Quite a few of them he finds persuasive; in other cases the arguments are valid, but one or the other premise doubtful. Sometimes he sides with Kant against his critics, sometimes the opposite, and occasionally, as in the case of the Second Analogy, it remains unclear what Kant's argument was. He has shown in this and earlier publications that all reconstructions are flawed, so that either Kant had no good argument, or his argument has not yet been discovered.

The title Problems from Kant must be taken seriously. The book rarely deals with textual problems, historical context, or translation difficulties. Rather, the focus is on perennial and still compelling problems. Themes from the First Critique are tracked to and connected with current concerns and sometimes recently ventilated issues are traced back to Kant, as in the last chapter, where the question is raised if Kant was an internal realist in the style of Putnam, or perhaps an anti-realist like Dummett. The contrast with Kant shows that like all complex positions, these are shopping baskets with several goods in them, some in Kant's inventory, others not.

Logical notation is used sparingly, and only where it clarifies—no attempt here to shoehorn arguments into unneeded symbolism. Deft examples, Kant's weakness, are a strength of this book. To explain the logic of considering Van Cleve asks, "Consider me apart from my shoes, am I barefoot...