Middle Knowledge and Arminianism Are Compatible: A Response to Roger Olson

What exactly is the problem that Roger Olson has with Molinism? Answer: it collapses into determinism. But it isn’t clear what he means by “determinism.” His concept is ambiguous, and he seems to acknowledge this when he says, “if middle knowledge does not imply determinism, it does convey that [our] lives are predetermined.” So there seems to be two senses of what he means for something to be (pre)determined: one is with respect to being causally necessitated to act; the second is with respect to being fated to act according to some preordained plan. In Olson’s mind, the distinction makes no difference, because both senses are sufficient for what he finds problematic with middle knowledge: God is able to use it to render our actions certain. Once God does that, he says, “then determinism is at the door if not in the living room and that is inconsistent with Arminianism’s basic impulses.”

How should Molinists reply? First, they should deny that middle knowledge entails both causal determinism and theological fatalism. Second, they should argue that the property of being rendered certain is not problematic if the objects of God’s knowledge, that is, the propositions about what free creatures would do in various worlds, are grounded by what free creatures would do. Third, they should maintain that God’s use of middle knowledge is benevolent, because God is benevolent. Let us turn to the first matter.

Contrary to what notable theologians like Bruce Ware believe, it is not the case that middle knowledge is consistent with the standard view of freedom assumed by most Calvinists: compatibilism. Compatibilism is the idea that human freedom is compatible with determinism. That is to say, one’s actions can be free even if they are the outcome of a causal process that began before one was born. What is determinism? The Stanford Encyclopedia of Philosophy has a memorable way of defining it: it is “the metaphysical thesis that the facts of the past, in conjunction with the laws of nature, entail every truth about the future.” A Calvinist might modify this to say that God’s pre-creation decree, in conjunction with the laws of nature God ordains, entail every truth about the future. That is just what Calvinists mean by their appeal to God’s “absolute sovereignty” over all things. And this is not compatible with middle knowledge.

Why not? The answer lies in what “middle knowledge” is about, what sort of knowledge it is “in the middle” of: God’s natural knowledge and his free knowledge. God’s free knowledge consists of his knowledge of all truths about what will happen, and the logical “moment” when God has it comes “after” he decides to create the world. God’s natural knowledge consists of all possibilities, that is, all the truths about what could happen, and the logical “moment” when God has it comes “before” he decides to create. Middle knowledge lies between God’s natural and free knowledge: it is knowledge of all the truths about what would happen if God were to create the world Therefore, it is posterior to natural knowledge, but prior to free knowledge. That’s why it’s called “middle knowledge,” and this is what it is about: God’s knowledge of what people would freely do if God were to create a world that is feasible for him to create. We say “feasible” because there are possible worlds that God cannot actualize such as the ones where the Adam doesn’t fall or Peter doesn’t deny Christ; these are possible worlds, to be sure, but given the facts of the matter, Adam and Peter choose otherwise and so determine what sorts of worlds are available for creation. Bill Craig (2011) illustrates how it all looks this way:

The main point to remember is that the kind of freedom people have in this view is libertarian freedom—it is not compatibilistic. Truths about what people could do with compatibilistic freedom are part of God’s natural knowledge, because it is about what is possible within the constraints of God’s decree and the laws of nature he ordains. Middle knowledge presupposes contingent truths about what people would do with libertarian freedom; therefore, there is no middle knowledge if compatibilism is true. Theologians like Ware who argue otherwise are just confused about what middle knowledge entails.

But doesn’t middle knowledge entail theological fatalism? No. Fatalism is true only if one cannot do otherwise, but there is nothing in middle knowledge that implies that we cannot. I have written about the fatalistic argument elsewhere, but it will suffice to say that any argument for the incompatibility of God’s foreknowledge and libertarian freedom is such that, if one of them is true, then Arminianism is false. Interestingly enough, both Calvinists and Open Theists agree that God’s foreknowledge and libertarian freedom are incompatible. Since, this is contrary to the assumptions of Arminianism, Olson should agree that this line of attack is unjustified.

But didn’t God render the Fall certain by the use of his middle knowledge? If we mean by “rendering something certain” to establish the fact that some event will be the case (though not necessarily the case), and by “using middle knowledge to do so” we mean appealing to foreknowledge about which contingent events would occur if God were create our world, then the answer is ‘yes.’ What is supposed to follow from this? Olson makes his main point:

That use of middle knowledge, providentially to render the fall certain, necessarily implies a plan in the mind of God that makes the fall not only part of God’s consequent will but also part of his antecedent will. And, as everyone knows and agrees, the distinction between God’s consequent will and God’s antecedent will is crucial to Arminianism’s argument that God is not the author of sin and evil.

Here is his argument stated more formally:

  1. God uses middle knowledge to render the Fall certain only if God has a plan that makes the Fall a part of his antecedent will (premise).

  2. God uses middle knowledge to render the Fall certain (from Molinism).

  3. Therefore, God has a plan that makes the Fall a part of his antecedent will (MP, 1, 2).

  4. If God has a plan that makes the Fall a part of his antecedent will, then God is the author of evil (premise).

  5. Therefore, God is the author of evil (MP 3, 4).

Of course [5] is an unacceptable result contrary to the assumptions of Arminianism. What to make of this argument? Certainly, it is deductively valid, but are all of its premises true? I think not. There is no good reason to believe premise [1] is true, and one good reason to believe it is false. With respect to the antecedent/consequent distinction, Molinists can affirm, along with other Arminians, that God desires unbroken fellowship with humans only if humans obey God. That is to say, there is no reason why a Molinist could not affirm that God’s antecedent will is for unbroken fellowship with us, but that he consequently wills our obedience to be a condition of that fellowship. Why couldn’t then God ordain a world where everyone obeys? Well, it is possible that there is no such world feasible for God to create, that is, human beings go wrong in every world where God desires a meaningful relationship with them. This is the heart of Alvin Plantinga’s famous Free-Will Defense, and even if we reject the particulars of Plantinga’s famous argument (the possibility of transworld depravity), it does not necessarily follow that the worlds in which God seeks to accomplish his goals, like building lasting relationships with his creatures, are all without sin. Therefore, Olson’s argument fails.

A concern: what grounds the fact of the matter concerning Adam and Eve’s decision to disobey? This was forcefully brought up by one of Olson’s readers who noted that the truth value of counterfactuals of creaturely freedoms must be explanatorily posterior to the creature’s free choices. As he remarks “this is a huge problem for Molinism because the providential usefulness of middle knowledge is predicated on its being explanatorily *prior* to actual creaturely choices.” What this means is that God knows a libertarian free choice of a creature only if the choice is determined by the creature; it cannot be the case that God knows it without the creature’s existence in an a priori fashion. Thus Molinism is internally inconsistent, and Olson couldn’t agree more.

I think this worry is unfounded. Molinists like Alfred Freddoso and Tom Flint have developed a response that preserves the a posteriority of God’s knowledge concerning the choices of free creatures that do not exist. Here is how it goes.

1. The statement “It will be the case that Adam and Eve decide to disobey God” is now grounded if and only if the statement “‘Adam and Eve decide to disobey God’ is now grounded” will be the case.

From (1) we can deduce that there was a time when God knew what Adam and Eve would do, and if God knows our days before we exist (Psalm 139:16), then God knows what Adam and Eve would do before they exist. Both Molinists and most anti-Molinists agree (save Open Theists). Now call the world we inhabit Alpha:

2. The statement “It would be the case that ‘Adam and Eve decide to disobey God (if God were to create Alpha)’” is now grounded if and only if the statement “‘Adam and Eve decide to disobey God’ is now grounded” would be the case (if God were to create Alpha).

From (2) we can deduce that God knows what Adam and Eve would do if he were to create Alpha (the world in which we now inhabit). But Molinists and anti-Molinists like Olson disagree! Why? If it is because knowledge of counterfactuals of creaturely freedom must be explanatorily posterior to actual creaturely free choices, then there is nothing in (2) that denies this. If Adam and Eve’s existence is required for their choices to be known, then how is it that from (1) we can deduce that there was a time when God knew what Adam and Eve would do before they were made? I suspect that we can’t, and that would help explain why Olson’s reader wrote what he wrote—he is an Open Theist afterall. But Arminius wasn’t and Arminians like Olson aren’t, so it seems that this critique from our Open Theist friend proves too much; if it is true, it undermines Arminianism altogether.

How, then, does God use middle knowledge for providential advantage? Take for example the case of the atonement. Let us cite Acts 2:23 and 4:27-28 (NIV) as the data of Scripture to be explained:

This man was handed over to you by God’s deliberate plan and foreknowledge; and you, with the help of wicked men, put him to death by nailing him to the cross. […] Indeed Herod and Pontius Pilate met together with the Gentiles and the people of Israel in this city to conspire against [Y]our holy servant Jesus, whom [Y]ou anointed. They did what [Y]our power and will had decided beforehand should happen.

The standard Calvinist interpretation of these passages is to explain away the “tension” between human freedom and God’s sovereignty with an appeal to compatibilism. On Molinism, it is by virtue of their middle knowledge that each member of Godhead knew what we would do if the Father sent the Son, and yet the Father sent him anyway so as to make it possible for us to share in the love and fellowship of the triune God. The “tension” between God’s sovereignty and our responsibility disappears in a way that it doesn’t on Calvinism, because on Calvinism it is utterly mysterious how we remain responsible (or even obstinate [Acts 7:29]), for our sins even though they are determined by events long before we were even born (for an extended argument along these lines, see McCall, 2012).

The way God uses middle knowledge for his providential advantage, then, is that in light of what he knows about what would happen with specific people in specific circumstances, God decides to create just those people and just those circumstances to freely carry out what God intends (for more see Craig’s essay in Jowers et al., 2011). This makes Olson worried. But why? He tells a story that is supposed to represent the problem:

Suppose I know one of my students so well that I know (beyond any possibility of being wrong) that if I suggest he read a certain book he will misunderstand the subject of our course and go on to fail it. Without the book, he would pass the course. I suggest he read the book. Why? Well, perhaps because I need someone to fail the course. I don’t grade on a curve and the dean is worried that I am not upholding academic standards. All my students pass with flying colors. My career is in jeopardy as is the academic credibility of the school. So I use my middle knowledge of the student’s dispositions and inclinations to bring it about infallibly that he fails the course. Nothing I did took away his free will. He read the book voluntarily (no external coercion was used, only inducement).

In the form of a rhetorical question, Olson says that it is obvious who is responsible for the student failing the course. What to make of this? The worry here is that God could do bad things with his middle knowledge. But there is an obvious response to this: God is not so cruel to be manipulative for malevolent ends, nor is he so foolish to rob us of our moral responsibility by making himself an object of moral blame. Yes, it’s true: if God abused his middle knowledge like this, he would not be good. But God is good, and the bare possibility of abuse (if that is possible for God!) does not negate the use.

To conclude, I think Olson’s objections against the compatibility of middle knowledge and Arminianism fail. But whatever we might think of his objections, we should embrace and take comfort in God’s providence over all things. While God’s omnipotence may be a cause for fear, there is a comfort to be found in a love so perfect that it drives out fear (1 John 4:18). It does not matter if God’s providence is “meticulous” (an ambiguous term to say the least); it matters whether his providence is loving. I suspect there is an aversion to “meticulous” providence because today’s Calvinists so readily underemphasize God’s universal love for all people; the idea that God might use it to give everyone “optimal grace” is foreign to the modern Christian (see Blanchette and Walls, 2013). Molinism can help subvert that. But there may be more to it than a tired debate between Calvinists and Arminians; we may not like the idea of God being in charge at all. In that case, it is worth echoing the Reformed philosopher Jamie Smith when he says, “How tragic that Christians are so captive to the spirit of the age that proclamation of God’s sovereignty is no longer heard as comfort” (from Twitter).

[This post is cross-posted at The Society of Evangelical Arminians website.]