"All-Crushing" Critic of Metaphysics 249The other two kinds of proofs, namely, the cosmological proof and the
physico-theological proof (otherwise known as the argument from de¬
sign), fare no better. They make more sense — up to a point. Still, they do
not prove the existence of God, as God is understood by theists. The cos¬
mological proof, according to Kant, runs thus:
If anything exists, an absolutely necessary being must also exist. Now I, at least, exist.
Therefore an absolutely necessary being exists. The minor premise contains an expe¬
rience, the major premiss the inference from there being any existence at all to the
existence of the necessary. The proof therefore really begins with experience, and is
not wholly a priori or ontological. For this reason, and because the object of all possible
experience is called the world, it is entitled the cosmological proof. (A6o4f=B632f)
For Kant, this proof, which he himself endorsed in his Only Possible Argu¬
ment, hides many pseudo-rational principles and a "whole nest of dialec¬
tical assumptions, which the transcendental critique can easily detect and
destroy" (A 609=6637). Indeed, given what he thinks he has proven in the
section on the antinomies, it cannot possibly work.
Kant also believes that the physico-theological proof fails, although he
has more respect for it. Indeed, he claims that this
proof always deserves to be mentioned with respect. It is the oldest, the clearest, and
the most accordant with the common reason of mankind. It enlivens the study of na¬
ture, just as it itself derives its existence and gains ever new vigour from that source.
It suggests ends and purposes, where our observation would not have detected them
by itself, and extends our knowledge of nature by means of the guiding-concept of a
special unity, the principle of which is outside nature. This knowledge again reacts on
its cause, namely, upon the idea which has led to it, and so strengthens the belief in a
supreme Author [of nature] that the belief acquires the force of an irresistible convic¬
tion. (A624f=B652f)
It is not a proof. In particular, it cannot prove that there is a perfect being,
such as the theistic God. Indeed, neither the cosmological nor the physico-
theological argument can do that, for both presuppose that the ontological
argument is valid, and both therefore fail.^16 The last argument has persua¬
sive force. It does not prove what it is intended to prove, but it helps us to
understand nature as an ordered or created whole.
The subject matters of the dialectic are, accordingly, far from being
entirely useless. They have to do with fundamental questions that are un¬
avoidable for us. Kant believed that they are expressions of deep "inter¬
ests" of reason that cannot simply be dismissed. Metaphysical speculation
is as inevitable for us as breathing. These questions concern the forms of
reason - what Kant calls the "transcendental ideas." The ideas, which are