Consider this post a sidebar to the ongoing series on Molinism. It draws on some recent comments by William Lane Craig, arguably the leading evangelical defender of Molinism, in response to a reader’s question (bold added):
First, I don’t claim that “universal salvation is impossible because of free will.” The point here is subtle and easily misunderstood. I think that there certainly are logically possible worlds in which everyone freely places his faith in Christ and so is saved. What I’ve said is that, for all we know, such worlds may not be feasible for God to actualize (or, if some are, they may have overriding deficiencies that make them less preferable). The point here is that God’s being omnipotent does not entail that He can actualize just any logically possible world. For the persons in those worlds, were God to try to actualize them, might freely choose to reject God. We can grasp this point by realizing that which world is actual isn’t up to God alone; free creatures are co-actualizers of the world along with God by means of their free choices, which God does not determine. So it may not be feasible for God to actualize a world of free, universal salvation (without overriding deficiencies).
Craig is exactly right that on the Molinist view, “which [possible] world is actual isn’t up to God alone.” God determines some contingent truths, while his creatures determine other contingent truths by their (libertarian) free choices. God only ‘weakly’ actualizes this world. He ‘strongly’ actualizes many aspects of the world, e.g., causally determining the circumstances in which free creatures will make their choices, but God doesn’t causally determine those choices. Rather, by way of his middle knowledge, God knows infallibly what free choices his creatures would make in those circumstances, and thus by ‘strongly’ actualizing those circumstances God ‘weakly’ actualizes the world in its entirety. Even so, as Craig puts it, we are “co-actualizers” of the world, because the actuality of this world depends both on God’s free choices and on ours.
This model of divine providence has proven attractive to many Christian thinkers, partly because of its prospects for theodicy. If the actualities of this world aren’t entirely “up to God” then perhaps God can’t be held morally responsible for the fact that some aspects of this world are less than ideal (e.g., not all creatures are saved).
However, I think the way Craig puts matters in the quotation above conceals some of the oddities of the Molinist’s position. Craig makes it sound as though which possible world is actualized is “up to” both God and us, based on the actual free choices that we all make. But this is misleading for two closely related reasons.
First, on the Molinist view, which possible world will be actualized is determined, in a non-causal sense, before any creatures even exist (let alone make any free choices). The actual world may not be fully actualized until the creaturely free choices are made, but which possible world will be actualized is settled entirely by an infallible divine decree prior to God’s act of creation. In logical order: God first consults his middle knowledge of the counterfactuals of creaturely freedom (i.e., contingent conditional truths about what each possible creature would freely choose in each possible set of circumstances) to discover which possible worlds are feasible worlds (i.e., ones he can weakly actualize). God then considers which feasible world is optimal, all things considered, and decides to weakly actualize that world, which includes creating (directly or indirectly) various creatures and the circumstances in which they will make their free choices.
Secondly, on the Molinist view, we contribute to God’s decision about which possible world to actualize only as possible creatures, not as actual ones. God’s middle knowledge concerns only possible creatures, i.e., free agents that God could create. So when Molinists like Craig tell us that the actual world is “up to us” as well as “up to God” we should understand that it’s ultimately “up to us” only insofar as it depends on possible choices that we would make if we were created. When God consults his middle knowledge, logically prior to his decree to create, you and I are regarded merely as possible creatures. There can be no reference to any choices that we actually make (even future ones) because the actuality of those choices hasn’t been determined yet.
Now I’m not the first person to point out the perplexities of the idea that merely possible creatures and their merely possible choices can make a metaphysically and morally weighty contribution to the actual course of events in this world. But here’s the thing: which feasible worlds are available to God depends not only on possible creatures who will be created (such as you and me) but also on possible creatures who will never be created. If the actual world is “up to” us, in some weighty sense, then it must be equally “up to” every possible creature, whether actually created or not.
The Molinist faces something of a dilemma here. If he wants to say (as he does) that it’s partly up to us which possible world is actualized, he must concede that in precisely the same sense it’s partly up to possible creatures God never brought into existence. Conversely, if he wants to avoid saying that non-existent possible creatures have made a substantive contribution to the course of history — the course of actual history — he’ll have to deny that any of us have made such a contribution.
My guess is that most Molinists will grasp the first horn of the dilemma. But that means they’re committed to saying, for example, that which possible world is actualized is partly up to Santa Claus. After all, isn’t Santa Claus a possible creature? There doesn’t appear to be anything logically impossible about the existence of a person meeting the description of Santa Claus. Couldn’t God have freely chosen to create such a person? Is there anything in Molinism that precludes there being a possible world in which Santa Claus exists and makes libertarian free choices (e.g., about what presents to bring the good little boys and girls who also exist in that world)? The Molinist may not believe there is a Santa Claus, but it’s hard to see why he shouldn’t think there could have been a Santa Claus.
Hence it seems to me that if Craig’s Molinism is true, and the world is partly up to God and partly up to us, then it must be partly up to Santa Claus as well (not to mention a potentially infinite number of other non-existent possible creatures).
That, I think we can all agree, is more than a little strange.