Molinists think that if CCFs are true postvolitionally rather than prevolitionally, then human beings are not free, because facts about how we would act in whatever situation are up to God. But if God has middle knowledge, Molinists say, then facts about how we would act in whatever situation are not up to God, and hence we are free. Here I dispute the former conditional: I think that even if CCFs are true post-volitionally, their truth values are not up to God. So if all that a person who calls themselves a Molinist wants is that facts about how we would act are not up to God, then they ought to adopt my view instead of Molinism, because my view doesn’t have the rather large theoretical cost of primitively true contingent counterfactuals. In the literature, the view that counterfactuals of freedom are true postvolitionally has been called Thomism, but my view doesn’t really have much to do with what St. Thomas thought about this and I haven’t really seen anyone else say something like this (though I haven’t looked too hard), so I’m going to call it Schmomism instead.
A popular Molinist claim is that if God knows CCFs postvolitionally, then He’s responsible for their truth-values; every CCF is true or false by divine decree. I think this is false, for it doesn’t appreciate the contextual nature of counterfactuals and how the distribution of modal space is more or less relative to the antecedent of whatever counterfactual. More to the point, it seems possible that, given the Stalnaker semantics for counterfactuals, there are some CCFs with antecedents that distribute modal space such that the worlds selected by the antecedent produce true counterfactuals that are beyond God’s control; if this is the case, then there are some CCFs that are not known prevolitionally any still yet are not up to God, and thus the Molinist claim that postvolitionally known CCFs are true or false by divine fiat is false. The upshot is that God’s free knowledge and postvolitional knowledge are not coextensive.
For example, suppose God creates a world where there’s only one spot open at Oxford and two applicants, Harry and Harriet. If Harry were offered the spot, he would take it, likewise for Harriet. Suppose at this world Harriet is offered the spot; then, from this world, it is true that if Harry were offered the spot, he would take it – but this isn’t true by divine decree, it is true because the nearest antecedent world is a world where Harry is offered the spot is the world most similar to world where Harriet is offered the spot. What the Molinist claim requires is that God could make it such that the nearest world where Harry is offered the spot is also a world where Harry declines the offer; but it’s simply not up to God whether or not a given world is more similar than another world with respect to a selected antecedent – the worlds just are what they are. Now, throw in the fact that there are countless counterfactuals that God takes into account for selecting any given world and it becomes clear that the number of counterfactuals that are in both God’s free knowledge and postvolitional knowledge becomes fairly small. Of course, by creating one world instead of another world it is true that God decrees one set of counterfactuals to be true instead of another set, but what’s important here is that it is not up to God which counterfactuals are in that set or which counterfactuals in that set are true or false – the antecedents of whatever counterfactuals are responsible for that.
So, if you’re a reluctant Molinist and all you wanted was a view that said your counterfactuals of freedom were not up to God, then join me in affirming Schmomism, which does just that but without primitively true counterfactuals of freedom.