While the law of non-contradiction says that no proposition can be both true and false in the same instance and in the same sense, the law of excluded middle says something slightly different. The law of excluded middle (LEM) says that all propositions are either true or false. In other words, no proposition has the quality of being somewhere between (in the middle of) true & false – propositions are strictly TvF; True or False. They cannot be neither true nor false.
Now, take the subjunctive counterfactual conditionals of creaturely freedom (the molinist type counterfactual conditional), which is the object of God’s middle knowledge according to the Molinist view. That kind of conditional will say that if some subject S were to be presented to situational circumstance A, in which S was presented with a choice between x and y, S would freely choose x. Now, consider that this subjunctive conditional is an intelligible proposition. We can articulate it as a proposition in the following way: “[it is true that] if S were presented to A, S would freely choose x.” Now, this sentence is intelligible, and therefore is a proposition. We now have to ask whether this proposition has a truth value, and what that might be. If the LEM is applied, then the proposition is at least either true or false.
- Any intelligible proposition has a truth value.
- Molinist subjunctive counterfactual conditionals are intelligible propositions.
- Therefore, Molinist subjunctive counterfactual conditionals have a truth value.
- God knows all and only true propositions.
- There are some true propositions of the Molinist-type (or middle-knowledge type).
- Therefore, God knows some true propositions of the Molinist-type to be true.
Against this we could press a strong version of the Principle of Sufficient Reason (PSR), according to which there are no contingently true propositions which do not have some sufficient reason for why they are true. Since there is no sufficient reason for why Molinist-type counterfactuals are true, there simply are no ‘truths’ of the Molinist-type.
In response, I have come to think we can, maybe (I’m still not sure) avoid the power of this objection in the following way. First, we can say that the Principle of Sufficient Reason applies not first of all to propositions, but first of all to contingent substances. Certainly this becomes attractive if we adopt a deflationary theory of semantics, according to which propositions do not robustly ‘exist’ and have properties (like being true), but rather just describe things which do exist (true-ly). Thus, the Principle of Sufficient Reason, for the purposes of articulating a persuasive cosmological argument, would look something like this: “There are no beings which do not contain in themselves the reason of their existence, and which do not have any reason of their existence.”
Then, for the purposes of dialectical reasoning, we can adopt a weaker Principle: namely, the WPSR which says that for any contingently true proposition there is possibly some sufficient reason for it’s being true. For more on the WPSR, I would check out Pruss here and here, and Rasmussen here. This weaker principle could simply say that we are always and everywhere justified in looking for a sufficient reason, and that if one is even possible we ought to prefer accepting it to not accepting any sufficient reason (notice this is an epistemic rule).
Then, the only objection left against Molinism would be the following: maybe premise 2 in the first argument above is wrong. Maybe Molinist-type subjunctive counterfactual conditionals simply are not intelligible – how could there be a fact of the matter about what somebody would freely do in some circumstance which is never actualized, even while it remains logically possible that they not do the very thing they would do? Here I find myself at a stalemate. I’m just not sure. Maybe I could argue my way to the truth of Molinist-type counterfactuals by doing a conceptual analysis on what is entailed by maximal greatness, aside from omniscience. For example if omnibenevolence is entailed, and if it in turn entails that there be no gratuitous evil, and whether there is gratuitous evil depends on whether God can orchestrate history well, which requires middle knowledge, then it follows from omnibenevolence that God has middle knowledge.