The Ontological Argument

Of all the philosophical arguments that claim to prove the existence of God, perhaps the one that has given rise to the greatest amount of debate and controversy is that of St. Anselm of Canterbury (1033-1109). Anselm's basic argument can be restated as follows:

We begin by defining God as "that than which nothing greater can be conceived". This means that nothing can be conceived that is greater than God. Now suppose God does not exist. In other words, suppose that God does not have the attribute of existence. In that case, something with all of God's attributes plus the attribute of existence could be conceived. But since this something would have all of God's attributes, plus the attribute of existence, it would have one additional good attribute — and therefore, it would be greater than God. So in this case something greater than God would be conceived. But that's a contradiction, since by definition nothing can be conceived that is greater than God. Therefore, God must exist.

The ontological argument is one of those arguments that right away appear to have something wrong with it, although explaining exactly what is wrong is not so easy. It took almost seven hundred years until what is generally regarded as the definitive refutation of the argument was presented by Immanuel Kant (1724-1804) — though to be fair, David Hume (1711-1776) had the same basic idea. But even today Kant's objection remains somewhat controversial.

Hume pointed out that existence does not add anything to a concept: "...to reflect on anything simply, and to reflect on it as existent, are nothing different from each other. That idea, when conjoin'd with the idea of any object, makes no addition to it." [A Treatise of Human Nature, Bk. I, Pt. II, sec. 6.] This idea was taken further by Kant, who famously claimed that existence is not a predicate. Anselm's error was to treat existence as an attribute or property, like being circular or being green. But according to Kant, when we say of something that it exists, we are not ascribing a property to it, as we are when we say of something that it walks or thinks or is green. As he put it: "By whatever...predicates we may conceive of a thing...we do not make the least addition to the thing when we further declare that this thing is. Otherwise, it would not be exactly the same thing that exists, but something more than we had thought in the concept; and we could not, therefore, say that the exact object of my concept exists." [Critique of Pure Reason, translated by N. K. Smith, B 627.]

Others before Kant and Hume had already hinted at this same insight. Pierre Gassendi, for example (a contemporary of Descartes), maintained that existence cannot be an attribute in the same sense in which other qualities are, because existence is that by which something either has or does not have attributes. This is not exactly Kant's point, but it is at least on the right track. And, in fact, this same basic point goes back at least as far as Aristotle.

As already stated, Kant's objection is not accepted by everyone. Consider Kant's argument again. He points out that if we have a concept of x and we then say that x exists, we do not add a new attribute to x. For if by saying that it exists we were adding something to the concept of x, then what exists would be something slightly different from the original concept: it would be something just like x, but with the additional property of existence.

This sounds right, but all a critic of Kant needs to say is that, if existence is a property, then it follows that the concept of a real object is incomplete until it contains existence as one of its attributes. Then claiming that x exists would not add some extra idea to the concept of x; it would instead complete the concept. An example might make this clearer.

Kant believes that a concept that describes a real object x is a complete concept even if it does not have existence among its attributes. This is certainly the usual way of treating concepts: if someone completely describes x by its size, shape, color, and so on, and then an object answering that description is brought forward, we would call that object "x", whether or not the description included "x exists". However, this should not convince someone who believes that the concept is not satisfied by an existing object unless existence is part of the concept. One might claim that the concept one has in mind is of a non-existing thing with such and such a size, shape, color, etc. After all, we do conceive of non-existing things, such as unicorns and mermaids.

On the other hand, suppose that we define a unicorn as an animal exactly like a horse in every respect except that it has a single horn in its forehead, and suppose further that an animal answering to this description were discovered. Would it be reasonable to object that the real animal is not a unicorn because it contains one additional property that the unicorn lacks, namely, existence? It would seem absurd to claim such a thing. Kant therefore certainly appears to have been right.

Another way to support Kant's view is this. To say that x exists is to say that the concept x is instantiated — that there is an instance of x in the world. It makes sense to say of a concept either that it has an instance or that it does not. But it makes no sense to say the same of an object. Existence, then, is not a property of objects; rather, if it can be said to be a property at all, it is a property of concepts.