One of the beautiful aspects of Molinism is its commitment to maintaining the sovereignty of God in the actions of free agents. But this idea, like many ideas worth having, is not without its fair share of criticism. Since Molinism is an attempt to explain the coexistence of the sovereignty of God and the freedom of the will, these objections will often focus on why Molinism does not accomplish preserving one or the other. Some say that by having free creatures at all, the sovereignty of God is diminished in some way. This objection targets the ability of Molinism to preserve divine sovereignty. Since this is not the point that I am focusing on in this article, I will simply say that this criticism has been addressed many times by many individuals; and that the creation of free creatures is seen by Molinists as an act of divine sovereignty rather than a limitation of that sovereignty. The objection I want to focus on instead is one that targets the preservation of libertarian freedom on Molinism. This argument contends that because God foreknows what an agent would do in a given situation, that ultimately that means there is only one possible thing that the agent could do in that situation. In other words, the argument contends that because you would not do differently the logical implication is that ultimately you could not do differently.
The Scottish philosopher David Hume famously highlighted the error in trying to derive an “ought” from an “is”. I will attempt to channel my inner Hume, and highlight the error in trying to derive a “could not” from a “would not.” I will first explain this error in its practical application to Molinism, and then I will quickly highlight the logical problems with such a derivation.
Molinists often use the rhetorical device of explaining possibility in terms of possible world semantics. This is not to say that there are actually other worlds out there somewhere, it is just a way of organizing our thoughts about potential realities as opposed to the actual reality. When God chooses between possible worlds, and he selects one to actualize, he is selecting from a pool of feasible worlds populated and differentiated by free choices. But from the moment he actualizes the world it is known what will happen, what free choices we will make. This is where the criticism comes in. If it is known what you would do, or what you will do, then in what sense could you do otherwise? But this objection is simply the result of a perfect storm of error, where the equivocation of the word “could” meets the misunderstanding of possible world semantics in just the right way. The word “could” is used to convey two similar yet distinct meanings here. In one instance we are discussing the broad possibility or ability to do otherwise, while in the other the word is used specifically in regards to ability to do otherwise within one particular possible world (often the actual world). And here is where the equivocation meets misunderstanding. If you know what it is to speak in the sense of possible world semantics, you would know that the possible worlds are divided on the basis of such decisions. That means that it makes no sense to ask whether or not we can do otherwise within a particular possible world, since if we did otherwise it would no longer be the same world, but a different possible world! So when speaking in terms of what we would do in a particular situation, does the fact that we would not do otherwise imply that we could not do otherwise? On the former definition of “could” our ability to do otherwise is not violated and consists of all possible worlds containing that particular situation, and the latter definition of “could” is nonsensical in terms of possible world semantics and demonstrates a critical misunderstanding on behalf of the questioner. So we can conclude that there is no sense in which the counterfactuals of creaturely freedom on Molinism, and the “would” that they entail, implies that one “could not” do otherwise.
In Logical Terms:
In logical terms this distinction is made much clearer. In logic, there is a relationship between what are known as necessary and sufficient conditions. In this case the necessary condition is “would not,” while the sufficient condition is “could not.” The importance of this classification is that it comes with a system of rules. You can imply a necessary condition from a sufficient condition, but not the other way around. This means that it would be correct to say that “could not” logically entails “would not,” but it would be fallacious to flip it around and say that “would not” therefore logically entails “could not.” If we put this relationship into a simple counterfactual it would look something like this: If the agent “could not,” then the agent “would not.” There is a lot that can be derived from this counterfactual from a logical perspective. This means that “would” logically implies “could” (via modus tollens), and that “could not” logically implies “would not” (via modus ponens). But more important to this discussion, this also can be used to demonstrate that attempting to derive a “could not” from a “would not” is tantamount to committing the formal logical fallacy of affirming the consequent as demonstrated below:
1) If the agent “could not,” then the agent “would not”
2) The agent “would not”
3) Therefore the agent “could not”
If the relationship between “could not” and “would not” is as the Molinist believes it to be, then to try to equate counterfactuals of creaturely freedom to an inability to do otherwise is not just wrong in practice, it’s logically fallacious in principle.