Since Molinism seems to have such a strong place in the comments section, I want to try out an issue on the concept of prevolitionality. The issue is central to discussions of Molinism, because Molinists take the truth of counterfactuals of freedom (CFs) to be prevolitional for God: that is, they are true prior to, in the logical order of things, anything that is true as a result of God’s will.
Some Molinists, such as Tom Flint, also claim the prevolitionality claim is the same as claiming that there is nothing God could do to make such CFs false. I have resisted this identification in print, insisting that the presence of snow in a certain spot in northern Indiana is prevolitional with respect to me but not beyond my control: I could have put a piece of cardboard over the spot and it would not have had snow on it.
Flint replied by claiming that the case of God is different. He thinks God’s case is more like what would be the case had I thought about the spot in Indiana, and consciously refrained from interfering, claiming “Here, a full explanation of why the snow is there would have to take account of the various acts and processes going on in my mind. So the groundï¿½s being snow covered would obviously be both post-volitional and non-resilient.” (A claim is resilient for S iff S lacks counterfactual power over its truth-value.)
I think this response is inadequate. If we begin with the concept of prevolitionality, the natural understanding of it is in terms of causality. We might say that a truth is prevolitional for S iff it is not true in virtue of any act of S’s will, but the notion of “in virtue of” seems obviously causal. If we take prevolitionality in causal terms, there is a straightforward understanding of causation on which, even if I decided not to go to Indiana and interfere, no act of my will caused the presence of snow on the spot in question.
Consider an analogy. A squirrel is sitting at the edge of the green when I hit my tee shot, which happens to go in the hole for a hole-in-one (if you’ve seen me golf, you might question this possibility, but indulge my fantasy for a bit). There is a causal process involving my swing, its trajectory and speed, the density of the ball, the character of the club, the wind patterns in place, as well as the texture of the grass and amount of moisture in the soil with reference to which the causal story of the hole-in-one can be told. The squirrel plays no role in this causal process, even if the squirrel considering leaping at the ball and then declined to do so. Neither does the thought life of Vladimir Putin, who that very morning considered hiring an assassin to kill me, but thought better of it. None of these ï¿½woulda-coulda-mighta-shouldaï¿½ events (to make my post formally precise and elegant, purged of this tinge of Southern heritage, let us call the events in question ï¿½WCMS eventsï¿½) plays a causal role at all. Events that might have, or could have, or should have, or would have interrupted the causal process are not themselves part of that process.
To make sense of this ordinary causal notion and to rule out any causal roles for WCMS events, there must be a way for things to go ï¿½on their ownï¿½ independent of interference. It is worth noting that central to Molinism is the claim that, because of counterfactuals of freedom, there is a way for things to go ï¿½on their ownï¿½ independent of Godï¿½s interference. Combine the causal notion above (or some quasi-causal surrogate of it for those who think causation is a purely mundane phenomenon) with this feature of Molinism and, even for God, there is a distinction between that which is prevolitional and that which is resilient.
So on what I claim is the intuitive notion of causality, Flint is mistaken. In response, there are two paths I see. First, Flint might argue for a more Millian account of causation, on which WCMS events are causes (since they are part of the field in which the intuitive causes produce their ordinary effects). Such views of causation have serious problems, however, so Flint may incline in a different direction. He may argue that prevolitionality should not be understood in causal terms, but rather in explanatory terms. This option fits the language of his response, which clearly adverts to the theory of explanation and especially to the concept of a full explanation.
The problem I see with this second route is that I can see no reason to favor an approach in terms of explanation over an approach in terms of the ordinary notion of causation. Without an argument for preferring explanation over causation in the account of prevolitionality, Flint’s argument that prevolitionality implies lack of control fails, since it has an arbitrary premise. So he needs an argument. But what would such an argument look like?