LET US, THEN, consider Kant’s defense of his reformed “metaphysics of nature” against the Pyrrhonian and the Hume-influenced skeptical problems. I shall begin with the Hume-influenced problems, those concerning the existence and reference of a priori concepts and the possibility of synthetic a priori knowledge. For, as we shall see, Kant’s solution to the Pyrrhonian problem builds on his solutions to these.
As I have already indicated, Kant’s first line of defense here is to point to apparently clear examples of a priori concepts which do exist and refer and to apparently clear examples of synthetic a priori principles which are known. Some of these examples, including some of the most compelling for Kant, in fact lie outside his reformed “metaphysics of nature” (namely, the concepts and principles of mathematics),1 but, as we saw, others lie within it (for example, the concept of cause and the causal principle). This is what I have called the common sense moment in Kant’s solution.
But Kant also develops a much more elaborate solution to the Hume-influenced problems on behalf of his reformed “metaphysics of nature.” This solution pursues a dual strategy. One side of the strategy undertakes to prove that specific metaphysical a priori concepts refer and specific metaphysical synthetic a priori principles are true. The other side undertakes to explain the possibility of their referring and being known. This dual strategy is reflected in many methodological comments that Kant makes—for example, his famous distinction between an “objective” and a “subjective” side of the Transcendental Deduction.2 Kant adheres to this dual strategy for defending the concepts and principles of his new “metaphysics of nature” against the Hume-influenced puzzles pretty consistently throughout the Transcendental Analytic of the Critique, where the concepts and principles in question are treated.3 Let us consider each side of this dual strategy in turn.
Kant’s proof that a specific metaphysical a priori concept refers or a specific metaphysical synthetic a priori principle is true always takes the form of what has come to be known in the secondary literature as a “transcendental argument,” i.e., an argument which turns on an attempt to show that the reference or truth in question is a condition of the possibility of experience (of a certain type).4 In other words (since “Y is a condition of the possibility of X” and “Necessarily if X then Y” are logically equivalent), such a proof turns on an attempt to demonstrate the truth of a conditional proposition of the general form “Necessarily if there is experience (of such-and-such a type), then a priori concept C refers / synthetic a priori principle p is true.”5 Most of the work involved in these proofs goes into demonstrating just such a conditional proposition. That done, the completion of the proof—which Kant usually treats as too obvious to require explicit development—is a simple inference by modus ponendo ponens from the truth of this conditional proposition plus the seemingly unquestionable truth of its antecedent (“There is experience [of such-and-such a type]”) to the truth of its consequent (“A priori concept C refers / synthetic a priori principle p is true”).6 Examples of such proofs are the Transcendental Deduction (which is concerned with the reference of metaphysical a priori concepts) and the Analogies of Experience (which are concerned with the truth of metaphysical synthetic a priori principles).7
One of the main attractions of this style of argument, for Kant, lay in the prospect that it seemed to offer of refuting Hume-influenced skepticism about metaphysics on behalf of specific metaphysical concepts and principles on the basis of premises which not only Kant but also the Hume-influenced skeptic about metaphysics himself would have to concede, such as that he had experience (of such-and-such a type).8
Kant’s explanations of the possibility of particular metaphysical a priori concepts existing and referring, and of particular metaphysical synthetic a priori principles constituting knowledge, are mainly in terms of his transcendental idealism. Famously, the thesis of transcendental idealism holds that the essential form of the objective world which we experience is contributed by our own minds (in contrast to its matter, which is given to us in sensation), this essential form comprising, on the one hand, the pure intuitions of space and time and the synthetic a priori principles of mathematics associated with them, and on the other hand, the a priori concepts of the understanding and the metaphysical synthetic a priori principles associated with them.9 Kant’s explanation of the possibility of metaphysical a priori concepts existing and referring is as follows. That they can exist at all actually has, in his view, a fairly straightforward explanation (one not strictly dependent on transcendental idealism): instead of being derived passively from sensation, they have their source in the active understanding, more specifically, in its logical forms of judgment.10 For example, the ultimate origin of the concept of cause, and in particular its component concept of necessity, lies in the hypothetical form of judgment “If p then q,” which expresses a robust relationship of consequence (Konsequenz).11 That such concepts are able, moreover, to refer to objects is explained mainly by the transcendental idealist thesis that they contribute essential form to objects (qua objects of possible experience, or appearances, of course, not things in themselves).12 For it will be recalled that in his letter to Herz Kant had argued that there were actually two circumstances in which the reference of a concept to an object was intelligible: either when the object caused the concept through sensation or when the concept caused the object, as notionally happens in the case of a divine archetypal intellect. Kant’s mature thesis of transcendental idealism makes the reference of metaphysical a priori concepts intelligible in a secularized version of the latter way: they in a sense cause their objects.13 A further part of the explanation is supplied by the Schematism chapter of the Critique:14 an account of how the basic logical meanings of these concepts which they derive from the logical forms of judgment get supplemented with temporal “schemata” which make their application to objects of experience possible.15 For example, in the case of the concept of cause, the basic notion of the consequence of one thing on another that is supplied by the logical form of judgment “If p then q” gets supplemented with what is basically the schema of constant conjunction (à la Hume), which renders the concept applicable to objects of experience.16
As for metaphysical synthetic a priori knowledge, transcendental idealism enables Kant to explain the possibility of this in roughly the following way: On the one hand, the fact that certain metaphysical synthetic a priori principles express aspects of the essential form of the objective realm of nature accounts for their truth. On the other hand, the idealist fact that we are responsible for the presence of those features accounts for our ability to know that they are there without prior investigation, and hence for the a priority of the knowledge in question. How can I know a priori (despite the non-analyticity of the claim) that, for example, every event has a cause? Because I constitute reality to conform with this principle.17
Such, in general terms, are the strategies by means of which Kant hopes to defend the particular a priori concepts and synthetic a priori principles of his reformed “metaphysics of nature” against the Hume-influenced skeptical puzzles concerning the ability of such concepts to exist and refer and the ability of such principles to constitute knowledge.