View on God

There are only three possible ways of proving the existence of God by means of speculative reason. All the paths leading to -this goal begin either from determinate experience... or they start from experience which is purely indeterminate ... or finally they abstract from all experience, and argue completely a priori, from mere concepts, to the existence of a supreme cause. Kant, Critique of Pure Reason, A59O/B618

Kant tells us that there are exactly three ways of proving the existence of God by speculative reason. In the first, we begin from "determinate experience and the specific constitution of the world" and ascend from there to a supreme cause. "The world presents to us so immeasurable a stage of variety, order, purposiveness, and beauty" (A622/B650) that we may infer a sublime and wise cause (A625/B654). This is the physico-theological proof or argument from design. In the second, we begin from indeterminate experience or "experience of existence in general" and proceed once again to a cause. Here it does not matter what the world is like, but only that it exists; if the cosmos consisted of nothing but a speck of dust, we would still need to posit a cause for it. This is the cosmological proof. Finally, we may bypass experience altogether and argue "completely a priori, from mere concepts." This is the ontological proof, most audacious of all, as it premises nothing about what exists. In this chapter I examine what Kant has to say about the cosmological and ontological proofs. I consider them (as Kant does) as attempts to prove the existence not of the God of Abraham, Isaac, and Jacob, but of a primordial being, whose identity with the God of religion must be a matter of further argument or faith.

A. The Ontological Argument

The version of the ontological argument Kant considers is that of Descartes, not Anselm. 1 It may be set forth as follows:

1. The ens realissimum (i.e., God) is, by definition, the being who possesses all perfections.

. Since (a) existence is a perfection, (b) any being that possesses all perfections must exist. 3. Therefore, the ens realissimum exists.

Kant is generally credited with originating what has become the standard criticism of the ontological argument--that existence is not a predicate. His critique contains in addition two other objections that he and his commentators do not always keep separate from the first: in a predicative proposition you may always "reject the subject," and there is something logically defective in the concept of a necessary being. I argue that one of these criticisms is cogent while the other two--including the famous one--are not.

B. Real Predicates

Kant never enunciates the slogan so often attributed to him, that existence is not a predicate. What he says instead is that existence is not a real or determining predicate, that is, "a predicate which is added to the concept of the subject and enlarges it" (A598/B626). As always, by a 'predicate' he does not mean a linguistic item but a property or a constituent of a concept. His contention may be understood in accordance with the following definitions:

A predicate P enlarges a concept C =Df ◇ ∃x(Cx & -Px). (Note that "enlarge" may be a misleading term, insofar as enlarging a predicate typi cally results in narrowing its extension.)

A predicate P is a real predicate =Df P enlarges at least one concept. 2

It follows from these definitions that a predicate P is nonreal iff for any concept C, □(x)(Cx & Px iff Cx). This makes clear the sense in which a nonreal predicate "makes no addition" to any concept: if P is nonreal, then saying that something is both C and P says nothing not already implied by simply saying that it is C. 3

Is Kant correct in claiming that existence is not, in the sense just defined, a real predicate? Yes, indeed: there is no concept C such that ∃x(Cx & -Ex). This, at any rate, is a consequence of letting the existential quantifier express existence. 4 To suppose there is something (∃x ...) that does not exist (... -Ex) is to suppose there is something that there is not.

Relative to widely accepted assumptions, then, Kant's dictum is true. The next question is, how does it show that Descartes's argument is wrong? How does the fact that existence is not a real predicate invalidate the ontological argument or make it unsound?

One common suggestion is that only real predicates may be used in definitions, in which case it would be illegitimate for Descartes to define God as a being who, among other things, exists. 5 But this suggestion is off the mark on two counts. First, Descartes is not guilty as charged. Look at his first premise; it says that God has all perfections but makes no mention of existence. Of course, in the next premise, Descartes says that existence is one of the perfections, so one may wish to say that he is implicitly if not explicitly defining God as a being who exists. But that brings us to the second point: what Descartes is charged with is no crime. There is nothing wrong with using nonreal predicates in definitions. Any tautological predicate (e.g., being red or nonred) is as much a nonreal predicate as existence, but there is nothing logically vicious about the definition 'x is square =Df x is an equilateral rectangle & x is red or nonred'. The second conjunct in the definiens is idle but harmless.

Perhaps it will be suggested that the premise that runs afoul of Kant's dictum is not the first but the second, for if existence "makes no difference" to any concept, how can it be a perfection? A perfection might be thought of as a property that contributes to the greatness of a thing, or makes an already good thing better than it would be without it. But if existence "makes no difference" to any concept, how can it be a perfection in this sense? How can an existent thing be better or more perfect than a nonexistent thing ? 6

But this objection is readily sidestepped. As I have formulated the second premise above, it consists of a premise proper (whatever has all perfections exists) and a reason for it (existence is a perfection). Perhaps Kant's dictum undermines or refutes the reason offered for the premise, but it does not refute the premise itself. Quite the contrary: it entails the premise! If existence is implied by any concept whatsoever, then in particular it is implied by the concept possesses all perfections, and that makes the second premise true.

Our verdict