Aquinas famously rejected Anselm’s argument, and Thomists have tended to follow him in doing so. It might seem that they are thereby committed to rejecting the possibility of reasoning to God’s existence a priori. This conclusion might seem to be reinforced by the fact that Thomists follow Aristotle in holding that no knowledge is possible for human beings without sensory experience, and reject the standard alternative views (Platonic theories of knowledge as recollection, Augustinian illumination theories, and rationalist theories of innate ideas).
But actually, the conclusion doesn’t follow. To see why, note first that we have to distinguish questions about the origin of the concepts we deploy when making knowledge claims from questions about the justification of knowledge claims. What Thomists hold is that we could not have any concepts at all without sensory experience. But it doesn’t follow that, even after we have the relevant concepts, we cannot ever deploy them in a priori reasoning. Thomists would deny that we could have mathematical concepts without sensory experience, because we cannot have any concepts without sensory experience. But they need not deny that, once we have the concepts, we can come to know all sorts of mathematical truths via a priori reasoning.
Second, we have to keep in mind the specific reasonwhy Thomists reject Anselm’s argument. If Anselm is right, then the proposition that God exists is self-evident, insofar as God’s existence would follow with necessity from his essence. Hence, if we had a sufficient grasp of God’s essence, we would know from that alone that he exists. Now, Aquinas agrees with that much. The problem, in his view, is that we lacka grasp of God’s essence that is sufficiently firm for us to have this knowledge. Hence, though the proposition that God exists would be self-evident to a sufficiently powerful mind, it is not self-evident to our minds.
Notice that this does not entail that we cannot reason a priori to God’s existence, in the relevant sense of that term. It entails only that we cannot reason a priori to God’s existence in the specific way that Anselm tries to. This would leave it open that we might yet reason to God’s existence in some other fashion that is plausibly regarded as a priori– for example, via an Augustinian argument from eternal truths (a version of which I defend in chapter 3 of Five Proofs of the Existence of God).
But one might do so even in a more or less distinctively Thomistic way. Hence, consider the argument for God’s existence that Aquinas develops in De Ente et Essentia, which I call “the Thomistic proof” and defend in chapter 4 of Five Proofs. According to this line of argument, God’s essence and existence are identical, and there can in principle be only one thing of which this is true. So, for anything other than God, its essence and existence are not identical. But for anything whose essence and existence are not identical, it must be caused by something whose essence and existence areidentical. So, anything other than God must be caused by God. (Obviously, this is extremely compressed. See chapter 4 of Five Proofs for a detailed spelling out of the argument and defense of its premises.)
Now, the standard way that a fully spelled out version of this sort of argument begins is by appealing to something we know through sensory experience. For example, it might start by considering the example of a dog; argue that, for any dog, its essence and existence are distinct; argue next that anything whose essence and existence are distinct requires a cause; and so on.
But there is no reason why such an argument needs to begin with something we know in that way. Suppose instead you were entertaining the following proposition: There is at least one proposition. You can know that that is true just by thinking about it. For what you are entertaining is itself a proposition. Notice that you are not relying on any evidence from the senses in judging that this proposition is true. Hence you know it a priori. Or, if you don’t like that example, take any other a priori proposition, such as a claim of mathematics or logic.
Now, what is a proposition? Is it a kind of substance, perhaps an entity existing in a Platonic third realm? Is it a kind of attribute (for example, a thought that inheres in an intellect), and thus dependent for its existence on the substance in which it inheres? Is it a material entity, such as a neural firing pattern? However we answer these questions, the Thomistic proof entails that, like everything else that exists other than God, its existence ultimately depends on God. So, unless this proposition is itself somehow identical to God (which would entail that God exists), it will depend for its existence on God (which also entails that God exists).
That gives us the following sketch of a theistic argument:
1. There is at least one proposition.
2. If this proposition is identical to God, then God exists.
3. If it is not identical to God, then either it is a substance or an attribute.
4. If it is an attribute, then it depends for its existence on a substance.
5. So, either it is itself a substance or depends for its existence on a substance.
6. But for any substance other than God, its essence and existence are distinct.
7. And anything whose essence and existence are distinct can exist only if caused by God.
8. So, this proposition is either itself God, or it depends either directly or indirectly on God for its existence.
9. So, God exists.
I call this a “sketch” of an argument because I hardly expect it to be convincing to anyone not already persuaded by a fully developed Thomistic proof (for that, see, again, chapter 4 of Five Proofs). The point is that if we fleshed out this sketch by incorporating all the relevant details of a fully developed Thomistic proof, then it seems we would have an a priori version of that proof. For the starting point is a priori, and the remaining premises are known by way of a priori metaphysical analysis (of the natures of substances and attributes, the essence/existence distinction, and so on).
(Again, in this context I mean a priori in the sense of knowable apart from sensory experience. There are other senses attached to the term “a priori” in Scholastic philosophy, such as “reasoning from cause to effect.” I’m not talking about a priori reasoning in that sense.)
Related posts:
Plantinga’s ontological argument
Theology and the analytic a posteriori
Frege on what mathematics isn’t
Augustine on divine illumination