But I find the arguments, which I will not discuss, for the incompatibility of omniscience
and freedom, if not indisputably correct, at least pretty convincing, and I will therefore
not reply in that way. (And I think that the attempt of Augustine and Boethius and
Aquinas to solve the problem by contending that God is outside time—that he is not
merely everlasting but altogether nontemporal—is a failure. I don't mean to say that I
reject the proposition that God is outside time; I mean that I think his being outside time
doesn't solve the problem.) I will instead reply to the argument by engaging in some
permissible tinkering with the concept of omniscience. At any rate, I believe it to be
permissible for reasons I shall try to make clear.
In what follows, I am going to suppose that God is everlasting but temporal, that he is not
“outside time.” I make this assumption because I do not know how to write coherently
and in detail about a nontemporal being's knowledge of (what is to us) the future. Now
consider these two propositions:
X will freely do A at t.
Y, a being whose beliefs cannot be mistaken, believes now that X will do A at t.
These two propositions are consistent with each other or they are not. If they are
consistent, there is no problem of omniscience and freedom. Suppose, then, that they are
inconsistent, and suppose free will is possible. (If free will isn't possible, the free-will
defense is self-contradictory for that reason alone.) Then it is impossible for a being
whose beliefs cannot be mistaken to have beliefs about what anyone will freely do in the
future. Hence, if free will exists it is impossible for any being to be omniscient. Now, if
the existence of free will implies that there cannot be an omniscient being, it might seem,
by that very fact, to imply that there cannot be an omnipotent being. For if it is
intrinsically impossible for any
end p.201
being now to know what someone will freely do tomorrow or next year, it is intrinsically
impossible for any being now to find out what someone will freely do tomorrow or next
year; and a being who can do anything can find out anything. But this inference is
invalid, for an omnipotent being is, as it were, excused from the requirement that it be
able to do the intrinsically impossible. This suggests a solution to the problem of free will
and divine omniscience: why should we not qualify the concept of omniscience in a way
similar to the way the concept of omnipotence is qualified? Why not say that even an
omniscient being is unable to know certain things—those such that its knowing them
would be an intrinsically impossible state of affairs. Or we might say this: an omnipotent
being is also omnisicent if it knows everything it is able to know. If we say, first, that the
omnipotent God is omniscient in the sense that he knows everything that, in his
omnipotence, he is able to know, and, second, that he does not know what the future free
acts of any agent will be, we do not contradict ourselves—owing to the fact that (now)
finding out what the future free acts of an agent will be is an intrinsically impossible
action.
I must admit that this solution to the problem of free will and divine foreknowledge raises
a further problem for theists: Are not most theists committed (for example, in virtue of
the stories told about God's actions in the Bible) to the proposition that God at least
sometimes foreknows the free actions of creatures? This is a very important question. In