I TURN NOW from pure exposition of Kant’s project to some critical assessment of it. I shall begin in this chapter with a consideration of two apparent problems which in fact have fairly straightforward and satisfactory solutions, before proceeding in subsequent chapters to further problems which may be more troublesome. It is my hope that this exercise will shed a little further exegetical light on Kant’s project, test its strengths and weaknesses, and also suggest some philosophical morals of broader significance.
A first apparent problem is this rather basic one: Why, given its striking differences from the “metaphysics” of the precritical tradition, does Kant even call the new “metaphysics of nature” which constitutes the core of his reformed metaphysics a “metaphysics” at all?
There is an answer to this question. Although Kant’s “metaphysics of nature” cannot simply be equated with either the general metaphysics or the special metaphysics of the precritical tradition, it does nonetheless bear certain important resemblances to each of them. First, as I indicated earlier, it bears a strong formal and material resemblance to precritical general metaphysics, or general ontology. That is why Kant in the Critique represents it as the successor-discipline to general ontology.1 Second, the etymologically derived force of the word “metaphysics,” namely that of a discipline dealing with matters “beyond nature” (meta ta physika),2 had been cashed out by precritical metaphysics in terms of the supersensuousness which it had understood to pertain to the concepts and principles of general metaphysics in certain areas of their application, but above all to those of special metaphysics.3 As we saw, Kant in the critical philosophy draws a distinction between such supersensuousness as a characteristic of concepts and principles and the broader characteristic of a priority, and renounces the former as a feature of the concepts and principles of his own “metaphysics of nature,” replacing it with the latter. Accordingly, for him, the notion of a priority replaces that of supersensuousness as the distinctive mark of “metaphysics” answering to the etymological force of the word: the generic idea of what is “beyond nature” now gets reconceived to mean what is “beyond experience,” not in the sense of what is supersensuous, but in the sense of what is a priori.4 This, then, constitutes a further reason for Kant’s conception of his “metaphysics of nature” as a “metaphysics”: somewhat like traditional metaphysics, and especially traditional special metaphysics, it has, if not quite the supersensuousness which they bore as the distinctive mark of their metaphysicality, then at least the closely related though broader feature of a priority, which now takes over in that role.
A second apparent problem is this: Is there not something odd about Kant’s very project of defending metaphysics, given that, as we have seen, the contents and even the general conception of the discipline are for him indeterminate and undergo transformation in the course of the defense? Does this not amount to defending a mere word?
Part of the answer to this question has in effect just been supplied: while it is true that for Kant the contents and the general conception of the discipline are to a considerable extent indeterminate and subject to transformation during its defense, there do nonetheless remain certain strong resemblances between the “metaphysics” from which he starts out and the “metaphysics” with which he ends up. A further part of the answer, though, lies in the fact that this shift in the contents and the general conception of the discipline as it undergoes development conforms to a general model to which Kant subscribes of how sciences do and should develop. According to his general model, (1) such changes in the contents and conception of a discipline during its development are quite normal, and (2) they represent the gradual emergence to explicitness of an “idea” of the discipline which was implicit in, but only imperfectly grasped by, the minds of inquirers at the start.5 Moreover, while point (2) seems quite dubious, it is really only point (1) that is essential for addressing the problem with which we are here concerned, and point (1) looks extremely plausible in light of more recent work in the history and philosophy of science.6
In short, both of the problems just raised in fact seem to have reasonably straightforward and satisfactory solutions.