Let the title be a warning: it really does read like that.
* * *
In Part Three, §51 of the Prolegomena to any Future Metaphysics Kant presents four transcendent ideas, corresponding to his four categories from the tables of §21. These seem to be the great insoluble problems of metaphysics, and Kant is attempting to give an account of how that insolubility is rooted in the way in which the world must be divided into that of appearances, which can be known by observation, and that of things in themselves, which cannot. The first two problems (antinomies) he categorizes as mathematical because they rely upon reason applied to transcendent questions of space and contradiction. The latter are “dynamical” because of their correspondence to causality in time. How does Kant propose to use his understanding of the division between the world as it is experienced by us and as it is in itself to solve these latter, dynamical antinomies?
The problem presented in the third and fourth categories of reason is apparently that we can, using arguments acceptable to our reason, prove both that freedom is necessary and that everything is causally determined in nature (in the third); and that there is both some necessary first being to begin the chain of causation, and that there is nothing necessary, but everything is contingent (in the fourth). Indeed, at 340 Kant says that “both thesis and antithesis can be established by equally evident, clear, and irresistible proofs — for I pledge myself for the correctness of all these proofs — and reason thus sees itself divided against itself.” I suppose that when he speaks here of clear, irresistible proofs he means something like the arguments of Aristotle and the scholastic tradition in favor of the thesis and of Hume in favor of the antithesis.
These antinomies are presented as neatly corresponding to the tables at §21, so that the question of freedom goes with the judgments and concepts as to relation, and that of contingence to those of modality. It is not immediately apparent why this should be the case, other than because Kant likes things to neatly correspond to one another. The concepts of relation are substance, cause, and community; it seems reasonable that the question of freedom would also be one of cause, since we might attribute two kinds of causes to the effects that we observe: those of mind, to which belongs choice; and those of natural laws, to which belongs necessity. How substance and community might be related to questions of choice seems less clear. The concepts in the fourth category, of modality, are of possibility, existence, and necessity, and it seems reasonable to suppose that they would lead eventually, in considering transcendental ideas, to the question of whether all things are ultimately contingent, and therefore sometimes exist and sometimes only have the possibility of existing, or if there is something that exists of necessity.
There is an additional correspondence outlined by Kant between these categories and our a priori intuitions of space and time. The first two (infinity and simplicity), he says, correspond more to that of space (343), and the latter (freedom and necessity) to time, so that the thesis is premised upon an a priori intuition and the antithesis upon what can be known by experience. In the proceeding sections of the Third Part of the Prolegomena there’s also a discussion of how our experience of the appearances of the external world requires the pure intuition of space, and our experience of the soul through empirical psychology requires that of time, since the world is experienced more fundamentally as extended, and the soul as continuous.
There seem, then, to be two sorts of idea clusters: one with space, the world of the external senses, quantity, and quality, of which there might be either infinity or simplicity; and another with time, the internal senses of psychology, relation, and modality, which might be of freedom or causality according to law, and of possibility or necessity, respectively. As a priori intuitions, however, space and time depend upon the structure of our own minds, and therefore cannot be applied to things as they are in themselves. Thus, there’s a further division between the world of appearances, which can be described accurately and known in space and time, and the world of things in themselves, which cannot. Our judgments based upon experience can only ever apply to the world of appearances, so that everything depending upon that experience, in space and time, is only applicable within that sphere.
How might the distinction between the world of appearances and that of things in itself, then, resolve the transcendental questions? In matters of space it inserts a contradiction in the terms of the question, so as to make the questions themselves nonsensical, and in matters of time it allows for the possibility that each of two seemingly contrary positions be true in their own sphere. “I must not say, of what I think in space or in time, that in itself and without these my thoughts it is in space and time” (341), and to ask transcendental questions of the world of appearances is to do precisely that.
So, in the case of the seeming contradiction whereby we can argue both that there is such a thing as free will, and that everything must be causally related with a kind of necessity, it is possible to keep both by saying that freedom belongs to the realm of things in themselves, while causal necessity belongs to that appearances. That is because causal necessity, based upon how events unfold in time, lies not in things out in the world, or even in our souls as they are in themselves, but rather in our a priori ideas. Therefore, while we may rightly organize our experience of appearances according to constant laws of cause and effect, and judge the truth or falsehood of our perceptions as those of appearance, by their adherence to these laws, that series of causation lies not in the things in themselves, but only in our experience of it. At the same time, however, it might be possible that there is a freedom belonging to things as they are in themselves, outside of the causality of our concept of time.
Our understanding of necessity is in a similar position, being something belonging properly to the structure of our own minds, and not to things as they are in themselves. “For if only the cause in the appearance is distinguished from the cause of the appearance, insofar as the latter can be thought as thing in itself, both propositions can well stand side by side” (347). I took that to mean that all causes within appearances are and must be contingent, as causes within our experience of time; and we cannot get outside of those experiences, time being itself not outside of them. Nonetheless, there can be something that is necessary in order to cause the appearances, from outside. With things in themselves, however, there can be no necessity, that being an understanding of ourselves.
How helpful is all this, really, for responding to the problem presented by Hume? On the one hand, Kant gives a very neat and orderly account of why it is impossible to have any sensory impressions of such fundamental things as space, time, causality, or necessity, and yet we rely upon them, must do so, and are indeed justified in doing so, provided we remember that we are only giving an account of appearances, and not of things as they are in themselves. It becomes possible, then, to maintain both that even so essential a substance as one’s own soul as a thing in itself should be entirely hidden to experience, and yet might well exist; that there might be things necessarily causing our experience of things and yet that necessity itself should be both imperceptible (because it is intuited a priori), and be lacking in things as they are in themselves.
It’s a persuasive solution, yet as I attempt to understand and re-phrase it, the language Kant uses upon it takes away somewhat from its force. While I’m reading or writing on Kant, for a brief period of time, I seem to understand what he’s getting at, and am convinced that while he might not be giving as exhaustive an account of our categories of experience as he hopes to, yet it is sensible and unites many perplexing difficulties. Then I go and have a bite to eat, take a nap, and try to read what I’ve written: just as happens when I first try to read something from the Prolegomena, I see words on the page, and they evidently mean something — they meant something to me when I wrote them — but i no longer have a direct intuition of their meaning, and it just looks like so much jargon. It seems, somehow, like a fault that is not trivial to be forced to account for our actual experience of the world and of reason by means of a system of language neatly placed atop a mountain of technical distinctions and divisions into two of this kind depending upon three of that kind and two more of the other organized by way of four of something else. It might remain balanced, but precariously so.