Here is an interesting article by a self-described former Molinist; he takes the tactic of granting that God has middle knowledge of libertarian actions, but as a matter of fact, there is no good reason to think we have libertarian freedom (I am summarizing quite a bit, but this appears to be the chain of reasoning). As such, he does not want to be a Molinist. I think the article is worth responding to in a few key points.
First, while he pits Molinism vs. determinism as contenders of explanation for data, this isn’t a symmetrical comparison. Interestingly, this is recognized when he locates the issue of contention as one of freedom of the will (specifically, libertarian freedom). Whether or not libertarian freedom is present (as it is in Molinism and is not on determinism) just dictates which data set is under consideration. If it is present, then determinism is manifestly false. If it is not present, then Molinism is false.
Now he recognizes that there is another option: compatibilist freedom. So it actually comes down to whether or not there are any good reasons to think that we have libertarian freedom. He says no. Why would he do that?
Well, he says, the argument for libertarian freedom is one’s intuitions, and these are properly basic. But, he claims, sometimes “our immediate desires override what we want.” While I don’t really know for sure what this means, it turns out it’s irrelevant. Libertarianism is not the thesis that “all of my decisions are libertarianly free,” but rather “at least some of my decisions are libertarianly free,” so that this consideration does not lower the plausibility of libertarianism by one iota: any plausibility that is gained from intuitive considerations should remain.
The second reason he gives is as follows: “The fact that we chose A when B was a valid option does not necessarily mean that we could have chosen B.” Once again, I frankly do not know what this means. What does he mean here, then, that B was a valid option? If he means we could have chosen B, then, yes, we necessarily could have chosen B. The world-indexed proposition “Randy could choose B in circumstances C in world W” would be true in all worlds, even deterministic ones. But regardless, he goes on to say that advocates for intuitive support for libertarianism have to argue that agents in a deterministic world would not have the intuition that they are free, which we (supposedly) do not know. I think the point here is supposed to be that this renders the intuitive support for libertarianism equivalent with support for determinism.
But I don’t see that we have to argue that, “In all deterministic worlds, no determined creature has the intuition that he is indeterministically free.” First, it may be that we do have reason to think creatures in such a world would not have such an intuition: namely, God is not a deceiver. Most people would be overwhelmingly deceived about their everyday actions. But let that pass (it is not essential to the response). Second, the mere possibility that we are mistaken in our intuitions is not enough to render these intuitions just as likely as not to be false. Why think that our intuitions need to be incorrigible in order to provide sufficient source for particular beliefs? Finally, he may be undermining his own later responses to other, relevant issues. If he thinks the possibility of intuitions going wrong is enough to nullify their use, and if he thinks that it’s just as likely (by implication) as not that we have determined, massively false beliefs about our everyday experiences and our roles in choosing them, then he has a 50-50 shot at thinking that any given belief he holds is false. But no one should hold any 50-50 belief as true. Therefore, he has acquired a defeater for literally any belief he holds, including his belief that libertarian intuitions are undermined.
Additionally, he seems unaware of the fact that there are indeed positive arguments for incompatibilism, and by a slight extension, libertarianism. This extension involves both intuition and moral reasoning. For example, certain versions of the consequence argument entail that incompatibilism is true, and that entails that one of two options is true: hard determinism and libertarianism. Most determinists do not accept hard determinism; the overwhelming number accepts soft determinism, also called compatibilism. But compatibilism is ruled out by the consequence argument. But what if he decides he will in fact be a hard determinist? Well, aside from ruling out biblical discussions of freedom, we can suggest: God would be acting against his character in forcing unfree creatures to do actions that are evil; in other words, God would be performing these evil actions alone, as the agents involved would be no more free than a child’s GI Joe action figures are. The intuition is that we are not so determined, and God is not of such poor character. The moral argument is that “if we are not free, we are not responsible; we are responsible; therefore, we are free.” If this is so, this leaves us with libertarianism. He will have to address both the consequence argument and its implications in order to move forward, unless he does embrace hard determinism.
A few other notes: first, he writes that, “The argument from the explanatory power of Molinism requires that determinism cannot account for human freedom.” Strictly speaking, this is not correct. Explanatory power is typically used to refer to the idea that the best explanations make the evidence we do see more probable than if the explanation was not correct.
Suppose determinism could account for human freedom. Why would it follow that Molinism’s explanatory power falls below that of determinism? Just run the (admittedly made up) “numbers” in your head: if Molinism were true, would we expect to see the evidence we do see or infer (human freedom, God’s omniscience, the seeming truth about counterfactuals, our intuitions that we are libertarianly free)? Sure. What about if Molinism were false? Is it just as probable? Well, pretty plainly, no. Now what about determinism? Without begging the question as to which type of freedom is true, would we expect to see the data we do see (human freedom, God’s omniscience, the seeming truth about counterfactuals, our intuitions that we are libertarianly free) on determinism? Surely not; even if possible, I doubt it’s even 50-50. But let’s say it is. In this case, then, Molinism emerges the clear winner in this category.
He offers Frankfurtian compatibilism as an alternative, but there are two issues worth addressing. First, as to this contention—“if God were to determine you to do precisely what you would have freely chosen, then perhaps freedom can be preserved”—this is what is called “overdetermination,” and counts against explanations when all else is equal (a permutation of Ockham’s razor). Second, his Frankfurtian example doesn’t actually show that freedom is preserved in this scenario. Indeed, Frankfurt examples just are examples of the lack of determination in a case where some agent chooses a particular way, and the other way is not actually open to him after all; they are not examples where a causally determined action is itself deemed free.
While there is much more that can be said, the point of my article is to suggest that the original author has not shown that there is no good reason to think we have libertarian free will, and has not shown philosophically that libertarianism and compatibilism are equally plausible.