Page 8

question (given that P has libertarian freedom) admits of no answer. Simply asserting that there
is an answer does not refute the grounding objection, but only begs the question against it.

The Appeal to Possible World Semantics

Another way in which Molinists have attempted to provide a ground for CFs is to appeal to

possible world semantic regarding counterfactuals. On the standard account, a counterfactual
proposition is true if and only if, in the possible world(s) nearest the actual world in which the
antecendent is true, the consequent is also true. So, for example,

(A) If David had remained in Keilah, then Saul would have besieged the city

is true if and only if, in the possible world nearest the actual world in which David remained in
Keilah, it is also true that Saul besieged the city. So, a CF like (A) can be grounded if there is a
possible world in which both the antecedent and consequent of (A) are true that is closer to the
actual world than a possible world in which the antecedent is true and the consequent is false.

However, there is a problem facing the Molinist at this point. There are certainly possible

worlds in which David stays in Keilah and Saul besieges the city, but there are also possible
worlds in which David stays in Keilah and Saul does not besiege the city. Which of these
possible worlds are closer to the actual world? On the standard analysis, the "distances" between
possible worlds is measured by their overall similarities and differences. But, it is easy to
imagine two possible worlds, W

and W

, that are exactly the same up to a time t in which David

makes the decision to remain in Keilah, which differ subsequent to t in that in W

Saul besieges

the city, and in W

Saul does not besiege the city. Again, which one is closer to the actual world?

At first glance, it would seem that neither is closer.

Thomas Flint, however, has attempted to solve the problem in favor of MK. Describing the

grounding objection, Flint says that

a counterfactual such as (c _ z) is grounded only if z would be grounded if c were true--
in other words, only if, in the nearest world in which the agent in question is in c, she
grounds z--say, by performing action Z. But if the agent is truly free, then aren't there
worlds in which she is in c and refrains from performing Z that are just as close to our
world as any world in which she is in c and performs Z? If a Z-less world is just as close

Of course, William Lane Craig has argued the opposite point, namely, that the demand for a

ground for such propositions as (A) begs the question against libertarianism, or, to put it in his words, "The
demand for a ground for volitional counterfactual states of affairs seems misguided. It implicitly
presupposes that libertarianism and agent causation are false doctrines" (Craig, "Hasker On Divine
Knowledge," 100). But this is not the case. What the Anti-Molinist is doing is simply pointing out the
implications of libertarianism for the truth-value of counterfactuals which refer to the actions of
libertarianly free agents. It is one thing to say that "Jones freely chooses x." We can easily imagine that
such a proposition about a free agent is grounded in the categorical state of affairs of Jones freely choosing
x. But, it is another matter entirely to say that the conditional "If Jones were in c, he would freely choose
x" is grounded. Craig says it is grounded in the counterfactual state of affairs it describes. But what state
of affairs is that? It is precisely the question at issue as to whether or not there is such a state of affairs.