Previous StudyNext StudyUp to Studies in Other Areas Contents Page
Return to Pastor's Home Page
Flaws in Kant’s Supposed Refutation of the Ontological Argument for the Existence of God

Many modern philosophers have assumed that Kant refuted the Ontological Argument for the existence of God. I will examine his reasoning and point out some glaring flaws. The Ontological Argument was the creation of Anselm in the 11th Century. I will use his original argument as the basis for our discussion. Through pure reason, Anselm proposed, God’s existence can be proven. He defined God as "a being than which a greater cannot be conceived." This definition seems like a reasonable one, and no one in philosophy has ever challenged its accuracy as a definition for God. Anselm argued that if it were possible to conceive of God not existing, it would not be "a being than which a greater cannot be conceived." So by the very definition of the concept of God it must also include existence, therefore pure reason tells us that God exists.

Kant attempted to discredit the Ontological Argument in two ways: first by rejecting the subject "God" along with all its predicates, including "existence," and second by arguing that "existence" is not a property like omnipotence and therefore does not qualify as a true predicate.

I will examine each of these points in that order.

Kant concedes that if we assume the subject "God," we must also accept the predicate "God’s existence." His solution is to reject both the subject and predicate and therefore avoid a contradiction. The problem with this is that by denying the subject "God," what we are really doing is denying the existence of God. We cannot deny the concept of God because we all agree that everyone has a concept of God. So we must actually deny God which means we are in truth denying the reality of God or the existence of God. So what Kant is really doing is denying the existence of God (the subject) and then denying the existence of God’s existence (the predicate).

But by doing this, he is not denying both the subject and predicate. Instead he has made both of them identical. In other words, we have 2 predicates – existence of God and God’s existence – which mean the same thing. But the subject "God" has been left intact. In order for a predicate to be a predicate, it must be different from the subject. The only way to make the predicate different from the subject is to make the predicate "simple existence" or "existing" as opposed to "God’s existence." There is no contradiction if we deny God (subject) and existing (predicate) except that we have denied the existence of existing. Since we know we exist, it seems unreasonable to deny existing. In other words we cannot deny existing of itself, and therefore we cannot deny God’s existing as a predicate unless we’re willing to deny existing in general.

And this brings us to Anselm’s point which Kant has failed to address. Anselm’s predicate "a being than which a greater cannot be conceived" must include existence. As he argued, "For the non-existence of what does not exist is possible, and that whose non-existence is possible can be conceived not to exist. But whatever can be conceived not to exist, if it exists, is not a being than which a greater cannot be conceived; but if it does not exist, it would not, even if it existed, be a being than which a greater cannot be conceived."

Because we exist we can conceive of existence. Because we can conceive of existence we can conceive of a supreme being existing. By denying the existence of the supreme being (denying Kant’s subject) we have changed the subject. The subject is no longer Anselm’s "a being than which a greater cannot be conceived" but a lesser being. Kant can never deny the existence of Anselm’s supreme being, since as soon as he denies its existence, he has changed the subject to a different being. As long as man has any concept of existence, he must be allowed to tentatively posit existence to a supreme being even if he considers the odds remote. As soon as he does that, he has proven the existence of Anselm’s being through pure reason. The only way to deny Anselm’s supreme being existence is to deny all existence. If we cannot deny all existence, we leave Anselm’s being unassailable.

Some may argue that Kant is not denying the existence of God, but the concept of God, and that this is his subject. But how can Kant logically deny the concept of God? He can examine it, explore the possibility of it, question it, but he can’t deny it unless he is capable of logically refuting it. The fact is most of mankind conceives of a God – theologian, atheist, and agnostic alike. They have a concept of God. Some embrace it, others discredit it, but the concept is still there. The only way to deny the concept of God is to logically prove the existence of God or the concept of God is impossible. Of course, this cannot be done. So the concept of God as a subject cannot be denied as long as men are able to conceive of God.

By showing that Kant cannot deny both the subject "God" and the predicate "existence" without contradiction, there is no necessity of disproving Kant’s second proposition, that existence is not a property, for the Ontological Argument to stand. However, there are serious flaws in Kant’s second argument that should be addressed as well.

Kant argued that existence should not be considered a predicate on the grounds that existence means the same thing as "being." By substituting his own definition for Anselm’s, he supposes that he has refuted the Ontological Argument. Philosophers have long argued whether existence can be considered a property (or a perfection), and it can hardly be stated that the matter has been satisfactorily settled in Kant’s favor.

To reveal the flaw in Kant’s definition, I give an example from my back yard. For fifteen years I had a thriving fruitless mulberry tree. Then slowly it began losing its green leaves and finally died. For two years the dead tree stood in my yard until a heavy wind storm blew it over. It was then sawed into logs, and after awhile burned in the fireplace. The tree passed through five stages:

VIBRANT EXISTENCE – its flourishing fifteen years

LIVING EXISTENCE – when the main north branch stopped producing leaves and the south branch’s leaves became less vital

MORBID EXISTENCE – when the tree died but remained standing in my back yard

PARTIAL EXISTENCE – when the tree was sawed into firewood logs

NON-EXISTENCE – when the logs were consumed by the fire

So we see the tree made a progressive journey from vibrant existence to non-existence. This was done over a period of time in degrees. The tree never really stopped existing until it was burned, but what changed was the quality of its existence. Some may argue that its existence didn’t change but its essence changed. But in essence the tree remained a fruitless mulberry tree in my back yard. If we want to say that every time a cell died or a new leaf grew that its essence changes, then we would have to say that the old tree’s existence was constantly changing. Tree 1 ceased to exist as it was changed into tree 2 and one second later tree 2 ceased to exist as it was transformed into tree 3. If we apply this to the complexity of humans, we can see how this would be ridiculous. Either I exist as the essence of ME (even though my body is constantly in flux) or we have entirely changed the dictionary meanings of essence and existence. (But even if we do accept this ridiculous idea of constantly alternating existence with non-existence, we still have an overall change in the degree of existence as each new existence marks a progression.) But I think its only reasonable to say that what did change in the tree was the degree of its existence. It was slowly progressing from vibrant existence to non-existence. In essence, it was still the same old fruitless mulberry in my back yard, but its degree of existence was changing.

When Kant looks at God, he only considers existence and non-existence, but this is only part of the picture. He ignores the degree of existence of God. He never considered the question of the vibrant existence of God. From the analogy of the tree we can easily understand how different degrees of existence can lead to the idea of perfection of existence which would be somewhere beyond vibrant existence. Therefore, we can conclude existence is not only a property but also a perfection. It therefore must be considered a predicate.

Kant’s example to prove the unnecessariness of existence to a concept actually proves Anselm’s point. Who could agree with Kant that the concept of 100 dollars is the same as holding 100 dollars in your hand? Having 100 dollars in my hand is much greater than 100 dollars in my thoughts. Existence of 100 dollars is greater than non-existence of 100 dollars especially if I want to buy something. This example of existence improving the concept can be repeated in many areas. Here are some additional examples:

The concept of a sunset as compared with actually seeing one

The concept of a musical symphony as compared with actually listening to one

The concept of physical intimacy as compared with actually participating in it

The concept of holding your own baby as compared with actually doing it

In all these examples existence has improved upon the concept. This is not because the concepts were inadequate or incomplete. We find even after attending the symphony for many years, the concept in our mind is not equal to actually listening to the musicians. Existence makes a difference. Therefore it must be a property. To deny existence as a predicate is absurd.

I would like to conclude this discussion of existence as a predicate with a look at a typical philosophy textbook argument explaining Kant’s position. Beck and Holmes argue that existence cannot be a predicate by giving the following comparison: They start with two statements, 1. Dogs bark and 2. Dogs exist. They then compare 1a. If there are any dogs, they bark, and 1b. There are dogs, and they bark, with 2a. If there are any dogs, they exist, and 2b. There are dogs, and they exist. Here they claim that the second set functions differently since it conveys no new information about dogs whereas the first set does. They make the same mistake Kant made.

What they fail to mention is that the word "are" is a synonym for "exists." In other words 1a. is really saying If there exist any dogs, they bark. Now it is apparent the fallacy of this position. For statements 1a. and 1b. contain two predicates (exist and bark) whereas 2a. and 2b. contain only one predicate (exist) which accounts for the difference in the two statements. If we would make 1a. and 1b. with only one predicate to have a fair comparison with 2a. and 2b. we would get 1a. If there bark any dogs, they bark, and 1b. There bark dogs, and they bark. Now 1a. and 1b. are just as suspect as 2a. and 2b. Obviously all four statements need more information. Here are my proposed readings: 1a. If people hear barking, then dogs exist, and 1b. People do hear barking, therefore dogs exist, compared with 2a. If people feel a divine presence, then God exists, and 2b. People do feel a divine presence, therefore God exists. Don’t expect to find these examples in any philosophy textbook soon.


This discussion of the Ontological Argument 1994 by David Humpal. All Rights Reserved.

Previous StudyNext StudyTop of Page