Theistic Arguments Series:  Anselm's Ontological Argument

The following is an outline of Anselm's ontological argument for the existence of God:

    1. We conceive of God as a being than which no greater can be conceived.
    2. This being than which no greater can be conceived either exists in the mind alone or both in the mind and in reality.
    3. Assume that this being than which no greater can be conceived exists in the mind alone.
        a. Existing both in the mind and in reality is greater than existing solely in the mind.
        b. This being, existing in the mind alone, can also be conceived to exist in reality.
        c. This being existing in the mind alone is not therefore the being than which no greater can be conceived.
    4. Therefore, this being than which no greater can be conceived exists in reality as well as exists in the mind.

The impression I get from reading this is that it's seems too simple.  You can't just define something into existence, and prove it based on the definition.  But I think that's exactly what Anselm has done.

To demonstrate this, let me first present this argument with a different object.  I shall attempt to prove the existence of a unicorn (if we consider a unicorn to be the ideal animal) by substituting 'animal' for 'being'.

    1. We conceive of a unicorn as an animal than which no greater can be conceived.
    2. This animal than which no greater can be conceived either exists in the mind alone or both in the mind and in reality.
    3. Assume that this animal than which no greater can be conceived exists in the mind alone.
        a. Existing both in the mind and in reality is greater than existing solely in the mind.
        b. This animal, existing in the mind alone, can also be conceived to exist in reality.
        c. This animal existing in the mind alone is not therefore the animal than which no greater can be conceived.
    4. Therefore, this animal than which no greater can be conceived exists in reality as well as exists in the mind.

Obviously, the argument is false.  There are no unicorns.  Yet, the argument takes exactly the same form as Anselm's.  Clearly, there's something wrong with it.  Either it is logically flawed, or it doesn't apply to unicorns in the same way it applies to God.  My guess is that theists will believe the latter, but I think it is the former.  We can use this version to help identify exactly where the difficulty arises.

There are several avenues of attack.  I won't bother to argue that it is inappropriate to compare the greatness of things that exist to the greatness of things that are purely conceptual, although I think that is a valid objection.

But consider statement 3.  We are assuming that this animal exists only in the mind.  Statement b says that we can also conceive of this animal as existing in reality.  That's fine as far as it goes, but we must remember that we are under the assumption that it doesn't actually exist in reality.  So, although we can conceive of something that exists in reality, we are not assuming that it actually does exist in reality.  How, then can we arrive at statement c with a new presumption that such an animal does exist in reality?  For statement c clearly assumes that there is an animal that exists in reality that is greater than the one which we have assumed in statement 3.  That is a direct violation of the assumption of statement 3. 

To say that this greater animal exists is based on the definition given in statement 1.  And that points us to the real problem with the argument.  The definition in statement 1 assumes the existence of this animal.  If its existence were not assumed, then statement c would not make sense.  But statement c does assume that this animal exists in reality (the animal than which no greater can be conceived).  In other words, the argument begs the question, because it depends on the existence of the animal (by definition) to prove the existence of the animal.  And Anselm's argument begs the question, too.

If we change the definition in statement 1 to say "We conceive of a unicorn as an animal than which, if it exists, no greater can be conceived", then we could no longer use statement c to show that there is a greater animal than the one we have assumed in statement 3.  The whole argument, then, becomes moot.  And I think this objection applies just as much to Anselm's version as it does to my version.