A popular philosophical viewpoint regarding the relationship between God’s knowledge and human free actions is Molinism. I will attempt to keep this simple and non-technical, so Molinism can be roughly defined as the view that God not only knows what people will do in the future but what they would or would not do in any given circumstance. In fact, the former is more or less based on the latter; God knows what every single person would or would not do in any world he could create, and then he decides to create a particular world, of which he then knows all future actions and events (“world” here does not mean the earth but rather a possible “universe” or state of affairs).
The reason this view is popular among Christian philosophers is that it promises to harmonize two common aspects of Christian theology: Human free will and God’s sovereignty/providence. More specifically, it seeks to harmonize human libertarian free will and a meticulous view of God’s providence. Arminians and other non-Calvinists tend to espouse human libertarian freedom, the idea that external causes (such as God) do not determine a human choice because otherwise it would not be free. Many therefore often conceive of sovereignty in more general terms: God is in control, but that does not mean that God decides every little detail of creation. A plethora of Bible verses can be mustered for their position, though of course that does not necessarily mean they are interpreted rightly.
Reformed theologians, on the other hand, have built an entire system based upon an all-causing view of God’s sovereignty; R.C. Sproul famously argued that if God did not directly control every last molecule of the universe, that one molecule could lay waste to creation. They therefore seek to change the typical understanding of free will by advocating compatibilism, the belief that determinism and free will can be compatible if free will is understood a certain way. God ultimately determines human choices, but human beings are still morally responsible for their actions, not God. Needless to say, such a view of freedom gives them decided amounts of trouble in explaining evil, but they too are armed with a host of Bible references (though again, it is up for debate if Reformed people interpret these correctly).
Molinism: Middle knowledge and the middle ground
Molinism is often viewed as the middle ground between Calvinism and Arminianism. To see why Molinism is viewed this way, it will be helpful to further explain what it is saying about God’s knowledge and his decision to create. There are at least two types of knowledge that God has: Natural knowledge and free knowledge. God’s natural knowledge contains his knowledge of all necessary truths, such as the truths of logic and mathematics. Because these truths are necessary, they are not up to God to change; in fact, they cannot be other than what they are. God’s free knowledge is his knowledge of all contingent propositions of the actual world. “Contingent” means that something could be different than what it is; for example, the proposition “Michael Jordan wore #23 for the Chicago Bulls” is a contingent proposition because, while true, he obviously could have worn a different number (which he in fact did for a little while).
A question arises: How does God know future contingents? One could simply retort that if future contingent propositions have truth value, then God will know them by nature because he is omniscient. While this is true, Molinists tend to find such an assertion to be a little lacking, so they model how God could know contingent propositions by proposing that God also has something called middle knowledge, which is basically the type of knowledge I talked about above: The knowledge of what someone or something would or would not do in any situation (this type of proposition is called a counterfactual of creaturely freedom).
The reason it is called middle knowledge is somewhat straightforward; it lies between God’s natural and free knowledge and shares characteristics of each. Like natural knowledge, it is not up to God to decide because it is up to the free wills of creatures. However, like free knowledge, it is contingent because those creatures could choose differently. Let’s say, for example, that if Bob were placed in X situation, he would freely rob a convenient store. It is logically possible that he not rob the convenient store, but unfortunately he freely would if he were in that circumstance. God cannot simply determine his choice to be different because then that would violate his freedom. Thus, any world where Bob is in X but freely chooses to NOT rob the store is an “unfeasible” world for even God to create.
This is conceived as a series of logical moments that looks like this:
Natural knowledge (all logically possible worlds) –> Middle knowledge (feasible worlds) –> Free knowledge (actual world that is created)
Basically, God chooses which world among the feasible options he has and creates it. This theoretically gives God meticulous providence because God fully knows that world down to the last detail, including the actions of free creatures, and decides to create it and not different world. It also theoretically protects freedom because what humans and angels do is ultimately up to them. Though the very fact that it attempts to strike a somewhat middle ground makes it a target for criticism by both Calvinists and Arminians alike, Molinism’s clever take on divine foreknowledge and human freedom is provocative.
Middle knowledge: A misfit tool for Calvinists
Though people both sides often resist it, Molinism or something akin to it is often employed by both Calvinists and non-Calvinists. Someone who wants to be more Reformed may, for example, adopt the theory of middle knowledge but still try to uphold a strong double predestinarian stance on election: God chooses individuals logically prior to his survey of feasible worlds and only chooses those worlds where his elect choose to accept him, disregarding worlds where, say, Rob the Reprobate freely comes to faith. Someone who does not want to go that way can view election differently. For example, one could say that God chooses to set the terms of salvation (by faith in Christ), and any who come to believe in the world he creates is therefore elect. This is why one of my professors likes to view Molinism more like a tool in the theologian’s or philosopher’s toolbox rather than a full-blown alternate system.
While I largely agree with his point, I think one can still ask if the tool is better fit for one system or another, similar to how a hammer is more consistently useful for a carpenter than for a cook. In this regard, I think Reformed people are kind of barking up the wrong tree by trying to employ a theory of middle knowledge because they have to give up too much to make sense of it (and my professor also agrees with this).
First, let us say that a Calvinist takes Molinism at face value: In that case, this Calvinist would have to concede libertarian free will and the fact that God’s middle knowledge is not up to God to change. While such a view could conceivably be worked into a something that is Reformed-leaning, as I explained above, it ultimately grates against the Reformed view of God’s sovereignty because there is something contingent that really isn’t up to God but up to creatures. This limits what kind of world God can create, which is not something that a consistent Calvinist is going to want to say. It also remains debatable if such a Calvinist can really hold to a classical form of unconditional election because even if one conceives of God’s electing choice occurring logically prior to middle knowledge, God must still choose worlds where his elect freely come to faith (in a libertarian sense), which may cause problems for the Reformed ordo salutis. It ultimately still sounds like God’s electing choice is at least partially based on his knowledge of someone’s faith rather than purely by his will, which sounds more classically Arminian than Calvinist.
Most Calvinists would not take Molinism at face value for these very reasons and instead try to marry compatibilism with middle knowledge. They will even argue that compatibilism actually makes better sense of middle knowledge because there is “something” that makes it true rather than it “happening” to be true if libertarian free will was the case. I’m not going to get into truth-maker theory and the grounding objection here; I will merely point back to the logical progression of natural, middle, and free knowledge and what compatibilism is. The progression starts with God’s knowledge of broad logical possibility, which is then limited by what free creatures would or would not do in any situation, and from which he then decides which world to create. However, if theological compatibilism is true, then God can and does determine what agents do without violating freedom and moral responsibility. If this is the case, then it makes no sense to say that God’s choices of creation are limited by what free creatures would do in a given circumstance; there is nothing that limits God’s choices other than what is broadly logically possible. Depending on how you take it, middle knowledge therefore collapses into either natural or free knowledge because what a creature would or would not do is determined by God. In some sense, God still knows counterfactuals of creaturely freedom, but he would know them only as extensions of how he himself would decide to determine them.
Thus, I happen to agree with those Calvinists who are hostile to Molinism and middle knowledge in this: They rightly see that these philosophical ideas ultimately aren’t friendly to them. Reformed philosopher Paul Helm is well aware of this and critiques another Calvinist, Bruce Ware, for trying too hard to utilize middle knowledge. Arminians like Roger Olson have critiqued Molinism as well, but not only do I think many of his criticisms are confused, the important agreement on libertarian free will instantly makes Molinism more of a live option for non-Calvinists, though this does not mean that there aren’t issues that have to be smoothed over. Given the fact that Luis de Molina developed the theory of middle knowledge as an attempt to reconcile God’s foreknowledge with libertarian freedom, it seems a bit wrongheaded for Calvinists to try to employ it. They either have to concede libertarian free will with large consequences to their theology or they have to try to ram the square compatibilistic peg into the round middle knowledge hole.