This is a synopsis on Peter van Inwagen’s ‘Ontological Arguments.‘ Van Inwagen’s thesis is that the ontological argument as articulated by Plantinga is defective, because it fails to give a good reason to believe that the properties of existing in all possible worlds (N) and being concrete (C) are compatible properties. This is not a metaphysical claim; it is an epistemological one: we cannot find out whether N and C are compatible. This is because the tools of conceptual analysis are inadequate for the job; it is like trying to determine whether four sevens appears in the variegated decimals of π just by analyzing the concept of the number. Nor does it appear that we have any other method for determining the compatibility of N and C. Without justifying the conjunction of N and C, Plantinga’s maximally great being cannot be justifiably conceived as existing in all possible worlds. (And even if it could be shown that N and C are compatible, says van Inwagen, it does not follow that the properties of Plantinga’s maximally great being are essential; for example, the being could be morally good in some possible worlds and evil in others).
For van Inwagen, the ontological argument is true only if it is ontic. That is, it must be the sort of argument that proceeds from a premise that claims that a set of properties is such that there exists something that exemplifies that set of properties. Thus we can formalize van Inwagen’s argument like this:
- Either N and C are compatible or they are not.
- If they are not, then we can deduce a contradiction from an argument with the premise ‘Something has both N and C’ and show that the conjunction of the other premises in the argument are necessarily true.
- But we can’t do that.
- If they are, then we can form an argument with the premise ‘Something has both N and C’ and show that its conjunction with the other premises in the argument is possibly true.
- But we can’t do that either.
- Therefore, we cannot show N and C are compatible or incompatible (1, 2, 3, 4, 5).
- If we cannot show N and C are compatible or incompatible, then we are not in a position to know whether premise 1 is true.
- Therefore, we are not in a position to know whether premise 1 is true (6, 7).
- If we are not in an epistemic position to know premise 1 is true, then we cannot determine that the ontological argument is ontic and it fails as a piece of natural theology.
- Therefore, we cannot determine that the ontological argument is ontic and it fails as a piece of natural theology (8, 9).
UPDATE: For what it’s worth, I think this is a powerful criticism. It largely depends on whether you think conceivability is a reliable guide to possibility; I am predisposed by other works by van Inwagen to think it is not.