Molinism and Other Determinisms

In which it is argued that Molinists are determinists, but this is not to their shame.

Robert Kane is one of the world’s leading experts on the philosophy of free will. He’s the editor of The Oxford Handbook of Free Will and one of the contributors to Four Views on Free Will (Blackwell, 2007). He’s written dozens of articles on the subject of free will. So it’s safe to say he knows whereof he speaks when it comes to debates over free will.

Kane is an incompatibilist, which is to say, he believes that determinism is incompatible with free will (at least, the kind of free will needed for moral agency). But what is determinism? Here’s how Kane explains the term in his book A Contemporary Introduction to Free Will:¹

An event (such as a choice or action) is determined when there are conditions obtaining earlier (such as the decrees of fate or the foreordaining acts of God or antecedent causes plus laws of nature) whose occurrence is a sufficient condition for the occurrence of the event. In other words, it must be the case that, if these earlier determining conditions obtain, then the determined event will occur. (pp. 5-6)

In more familiar terms, we say that a determined event is inevitable or necessary (it cannot but occur), given the determining conditions. If fate decreed or God foreordained (or the laws of nature and antecedent causes determined) that John would choose at a certain time to go to Samarra, then John will choose at that time to go to Samarra. Determinism is thus a kind of necessity, but it is a conditional necessity. A determined event does not have to occur, no matter what else happens (it need not be absolutely necessary). But it must occur when the determining conditions have occurred. If the decrees of fate had been different or the past had been different in some way, John may have been determined to go to Damascus rather than to Samarra. Historical doctrines of determinism refer to different determining conditions. But all doctrines of determinism imply that every event, or at least every human choice and action, is determined by some determining conditions in this sense. (p. 6)

Now here’s an interesting (to me) and perhaps surprising (to you) observation: According to Kane’s understanding of determinism, Molinism is clearly a species of determinism. (To use Kane’s phrase, it is a “doctrine of determinism.”) For according to Molinism, God has an infallible decree; God foreordains all things, including human free choices. As the Molinist will be quick to insist, God foreordains on the basis of his middle knowledge, that is, his knowledge of the counterfactuals of creaturely (libertarian) freedom. God “weakly actualizes” a possible world by creating agents with libertarian freedom and arranging their circumstances such that they freely choose what he has planned (on the basis of his middle knowledge) for them to choose. But the fact remains that on the Molinist scheme, despite its commitment to libertarian free will, God has an infallible decree and foreordains whatsoever comes to pass. As one prominent Molinist explains:

Not only does this view make room for human freedom, but it affords God a means of choosing which world of free creatures to create. For by knowing how persons would freely choose in whatever circumstances they might be, God can, by decreeing to place just those persons in just those circumstances, bring about his ultimate purposes through free creaturely actions. Thus, by employing his hypothetical knowledge, God can plan a world down to the last detail and yet do so without annihilating creaturely freedom, since God has already factored into the equation what people would do freely under various circumstances.²

Thus, according to Molinism, if God has foreordained that Sam mows the lawn next Saturday, then Sam will mow the lawn next Saturday. God’s act of foreordination is a sufficient condition for Sam’s action (P is a sufficient condition for Q if Q necessarily follows from P) and therefore, according to Kane, Sam’s action is determined by prior conditions, namely, God’s act of foreordination. (Notice that Kane explicitly includes “the foreordaining acts of God” among his examples of determining conditions.)

As I indicated, this seems a surprising conclusion. Don’t Molinists explicitly reject determinism? Isn’t the whole point of Molinism to reconcile a doctrine of meticulous divine providence with an non-deterministic account of creaturely freedom? There are a few things to say here. First of all, observe that in the quotations above, Kane makes reference (twice) to nomological determinism, the view that all events are determined by “antecedent causes plus laws of nature.” This is the kind of determinism (often misleadingly labeled “causal determinism”) that is most commonly in view when philosophers debate whether free will is compatible with determinism. Molinists uniformly reject nomological determinism. But as Greg Welty has persuasively argued, there are striking parallels between (1) contingent laws of nature (such as the law of gravity) and (2) the Molinist notion of contingent counterfactuals of creaturely freedom (the objects of God’s middle knowledge).³ Just as we are able to exploit our knowledge of the laws of nature in order to bring about certain outcomes (e.g., I know that my pulling the trigger of a gun, in conjunction with the laws of nature, will result in a bullet being expelled from the barrel toward a target) so God is able — so the Molinist claims — to exploit his knowledge of the CCFs in order to bring about certain outcomes (e.g., God knows that his placing me in circumstances C, in conjunction with the CCFs, will result in my freely choosing A). Thus, there’s a parallel between nomological determinism (where outcomes are secured by the laws of nature) and the Molinist account of divine foreordination (where outcomes are secured by the CCFs), a parallel that warrants describing Molinism as a species of determinism.

But again, someone might object, aren’t Molinists supposed to be incompatibilists about free will? How then could they affirm both determinism and free will? Is there a glaring contradiction here that Molinists haven’t noticed or conceded? No, I don’t think so. As I’ve pointed out before, while determinism can be defined in very broad terms (as Kane does), there are many different types of determinism, depending on what are considered to be the determining factors. What this means is that one could be a compatibilist with respect to one type of determinism (e.g., divine determinism) but an incompatibilist with respect to another type of determinism (e.g., physical determinism). Likewise, one could be a broad-scope compatibilist who holds that free will is compatible with determinism in the broad sense (as Kane defines it) but also a narrow-scope incompatibilist who holds that free will is incompatible with determinism in one or more narrow senses. In the case of Molinism, I think the proper conclusion is that Molinists are incompatibilists with respect to divine causal determinism, but compatibilists with respect to divine non-causal determinism (i.e., the middle-knowledge account of divine foreordination). And for that reason, Molinists should be understood — whether they like it or not — as broad-scope compatibilists (i.e., compatibilists with respect to determinism in the broad sense, as Kane defines it).

But wait a minute. According to Kane’s definition, wouldn’t a proponent of exhaustive divine foreknowledge also turn out to be a determinist? For God’s foreknowledge that Sam will mow the lawn next Saturday is a sufficient condition (logically speaking) for Sam mowing his lawn next Saturday.⁴ Recall Kane’s statement:

In more familiar terms, we say that a determined event is inevitable or necessary (it cannot but occur), given the determining conditions.

That applies as much to divine foreknowledge as to divine foreordination. If God foreknows E, then E is inevitable; it cannot not occur, given God’s foreknowledge. So there’s a sense in which God’s foreknowledge is a “determining condition.” Certainly it’s a weaker kind of divine determination than the Molinist posits, but it still fits Kane’s definition.

Yet why stop there? Even the view that there are determinate past truths about future events — for example, that it was true yesterday that Sam will mow the lawn next Saturday — seems to fall under Kane’s definition of determinism. For if it was true yesterday that E will occur tomorrow, and the past is ‘fixed’ or ‘settled’ in the way that most of us take it to be, then E is inevitable. Note: the argument here is not that E is absolutely necessary (i.e., it must occur no matter what else happens) but rather that E is conditionally necessary, where the sufficient condition is the past truth about E’s occurrence. In other words, E’s occurrence is entailed by some preceding condition (a determinate past truth). This is weak sauce, to be sure; few will be troubled by that kind of determinism. My point is merely that such a view still seems to qualify as ‘determinism’ according to Kane’s broad definition.

So, what should we conclude from all this? That we should give up talking about ‘determinism’ at all, since nearly everyone turns out to affirm ‘determinism’ of some stripe? I don’t think that’s right. But here are some more useful conclusions we might draw:

1. ‘Determinism’ is more widespread than most people assume, and it shouldn’t be treated as some kind of metaphysical bogeyman.
2. The pertinent and substantive issue is not whether a view is deterministic, but rather what kind of determinism it affirms (i.e., the ‘what’ and the ‘how’ of the determining conditions). Thomism, Calvinism, Molinism, Simple-Foreknowledge Arminianism, and even some versions of Open Theism are committed to determinism in the broadest sense. But obviously they take very different positions regarding how and in what sense events (especially creaturely free choices) are determined.
3. The only way to avoid determinism completely is to embrace some version of the “open future” view, according to which, for at least some future events, there are no determinate truths now about whether those events will occur. For theists, this will require taking what Dale Tuggy has called “the wide road to open theism”: the view that God doesn’t know the future free choices of his creatures because there are presently no truths to be known about those future free choices.⁵

I suppose Molinists might be tempted to cut off my argument at the knees, so to speak, by rejecting Kane’s definition of determinism. If they make that move, they ought to offer an alternative definition that (1) applies to all the views that are normally taken to be “doctrines of determinism,” (2) doesn’t apply to their own view, and (3) isn’t ad hoc. Perhaps they can pull off that feat, if they’re sufficiently determined.

---