Kant & Russel's Objections to Ontological Argument

I am going to continue discussion of the Ontological Argument that I spoke of in my last two posts, both with Anselem’s Original Argument and Descartes' variation on it. Aquinas did make an objection to Anselem's argument and offered in its stead a different set of arguments to prove God's existence, but I will instead focus on Kant's objection since it is more to be noted.

Kant, in the Critique of Pure Reason makes the argument that to speak of existence as a perfection of something is absurd, because existence can't really be a characteristic of something. If a thing exists then it necessarily has the quality of existence. If a thing doesn't exist then it cannot have any characteristics. Thus, to be speak of a thing as perhaps having or not having the quality of existence doesn't make sense. Personally, I don't find this argument, since I think you might be able to object to it. But it does suggest something which Kant himself wasn't aware of, an important characteristic of the type of logic that Anselem is using, naming Aristotelian logic. Aristotelian logic has the limitation that it can' speak of things that don't exist. In other words, if you try to use as a step in Aristotelian logic "all griffin's have golden wings" then the argument can lead to untrue conclusions because the statement implicitly assumes griffins exist. Kant himself asserted that Aristotelian logic was complete and that nothing needed to be added to it. He was proven wrong as a number of advances in logic beginning in the nineteenth century expanded logic's possibility. In particular, one could begin to speak of things that don't exist or that may or may not exist using if/then statements. For example, "if griffins exist, then all griffins have golden wings," can be used because it doesn't assume that griffins exist.

This is the reason why I find Russel's refutation of the Ontological argument definitive, because he shows that the existence of God is implicitly assumed as an initial premise of the argument. In other words, the argument is trying to prove that God exists, but already secretly assumes that God exists. Russel's argument appears in his famous 1905 essay "On Denoting." Russel rephrases the Ontological argument in a more logically straightforward way that is consistent with the original: "The most perfect Being has all perfections; existence is a perfection; therefore the most perfect Being exists." Russel shows that the the first step in the argument, "the most perfect Being has all perfections," must be rephrased in order to be meaningful. And that the only way to make it meaningful and to have a meaningful argument is to begin with, "There is one and only one entity x which is most perfect; that one has all perfections." But this is exactly what we are trying to prove and which the argument doesn't prove. Thus, the argument does nothing.

Similarly, we might also rephrase the Ontological argument so that it explicitly doesn't assume that God exists, by using the if/then form. Namely, "If the most perfect Being exists then that Being has all perfections; existence is a perfection; therefore..." From here the only conclusion that we can draw, since we started with an if/then statement and the second step only qualifies the "then" part of that statement, is "...therefore, if the most perfect Being exists, then the most perfect Being exists."

I find Russel's refutation more definitive, because he shows that the argument doesn't actually prove anything at all. It's just logical sleight of hand used to secretly import as an assumption what is ostensibly being proved, namely that God exists.

Hume has an interest response, about whether we can prove the existence of anything prior to experience, which I will discuss next post.