Although God knows whatsoever may or can come to pass upon all supposed conditions, yet hath He not decreed anything because He foresaw it as future, or as that which would come to pass upon such conditions. (Westminster Confession of Faith, 3.2)
I’ve heard it said that many Calvinist writers currently favour some form of the doctrine of middle knowledge. I’ve also heard that among the roll call are the names of John Frame, and John Feinberg, but I have not checked this. I hope not. Terrance Tiessen and Bruce Ware have openly avowed their commitment to Calvinist middle knowledge. Terry and I have been involved in a conversation on the matter, arising from his article 'Why Calvinists Should Believe in Divine Middle Knowledge Although They Reject Molinism' (Westminster Jnl., 2007), which is to be published in due course in that journal . So this leaves Bruce, whose presentation differs somewhat from Terry Tiessen’s. I devote this short Analysis to his views as they are to be found in his God’s Greater Glory, (Crossway, 2004)
Two preliminary points
Bruce Ware offers his thoughts about middle knowledge as a contribution to establishing and further understanding God’s asymmetrical attitude to good and evil. Although God works all things after the counsel of his own will, the way in which he works his will regarding evil is different from the way in which he works good, since being immaculately holy, he cannot be the author of sin, that is, he is incapable of having evil imaginations and wicked desires that immediately bring forth evil.
There is a long tradition in Reformed Theology of addressing this question, if only because the Reformed among all theological outlooks are the most frequently taxed with the objection that their views entail that God is the author of sin. Various responses have been offered; for example, that evil is a privation or loss and therefore that as a matter of metaphysical necessity God, who has fullness of being, cannot author evil; that God is a distinct type of causal agency, he is the primary cause, not to be confused with secondary, creaturely causes; and finally that God brings about evil by willingly permitting it. Each of these positions is repeated ad nauseam in the literature; they are not exclusive of each other, but can be combined to make a cumulative case against the charge of God’s authorship of evil. There remains, of course, an element of mystery, of how it is that the divine ordination meshes with human sin and evil. Augustine and Calvin, for example, routinely refer to the matter as ‘ineffable’. Bruce Ware, and others, offer 'Compatibilist Middle Knowledge' as one way in which such ineffability can be mitigated.
What is middle knowledge? This is the doctrine that between God’s natural knowledge, his knowledge of all necessities and possibilities, and his free knowledge, what he has freely planned to bring to pass, there is a middle knowledge, his knowledge of what his free creatures would do in a vast variety of different circumstances. This third type of knowledge is an invention of Fonseca and Molina (two Jesuit theologians) tailored specifically to harmonise God’s sovereignty and libertarian free choice and divine grace and evil. Jacobus Arminius borrowed or adopted this device in his Protestant account of divine foreknowledge and human free choice. (The details can be found in Alfred Freddoso’s Introduction to his translation of Molina, On Divine Foreknowledge, (Cornell University Press, 1988) and in Richard Muller’s book on Arminius, respectively.)
To begin with, it should be noted that the very notion of middle knowledge collapses if its underlying rationale and necessity are changed. That rationale and necessity rest upon the Jesuit and Arminian commitment to the freedom of indifference. As it is impossible to have a soufflé without eggs, so it is impossible to have middle knowledge without the freedom of indifference. Indeed, I think it is true that it was Molinism and the debate that it raised, and Arminianism and the debate that it raised, that made it clear to the Reformed, clearer than it had been before, that they would be better, much better, to be committed to a compatibilist account of human freedom than to some form of incompatibilism.
Enter Bruce Ware
Enter Bruce Ware. Ware has an explicit commitment to compatibilist human freedom; he thinks that it is revealed truth, and besides that, he argues (like Jonathan Edwards) that forms of incompatibilism are logically defective. (GGG 78) So what can possibly be the attraction of middle knowledge?
Let us try to answer this first by seeing what according to Bruce Ware is left of middle knowledge. He notes that libertarian freedom is problematic. (GGG 112) But once we abandon libertarian freedom, according to Ware, and recognise that there is a necessary connection (that is, a causally necessary connection), ‘between knowledge of a given state of affairs and knowledge of what the agent would choose in that particular setting…’ then all is plain sailing; more light is immediately thrown on the meshing of sovereignty and responsibiity. For given this connectedness, ‘God could know what the agent would choose by knowing fully the circumstances in which the agent would make his choice’. (GGG 113-4) As Ware explains, ‘The set of factors in which the agent makes his choice constitutes a set of individually necessary and jointly necessary sufficient conditions for forming within the agent a strongest inclination or highest desire by which he then makes the one choice that is in accordance with that highest desire’. (GGG 114)
But if God knows what Jones, if placed in circumstances C, would do, then this is surely part of God’s natural knowledge, his knowledge of all necessities and possibilities. Among these possibilities are segments like those described by Ware. Middle knowledge ‘works’ for the compatibilist, Ware says (GGG 115), but it only works because it is not genuine middle knowledge, a distinct type of divine knowledge, but part of God’s natural knowledge. It works precisely because it is not middle knowledge. The examples that he cites from Scripture (GGG115-9) are of God’s knowledge of unactualised possibilities, possibilities therefore that are counterfactual, but they are not counterfactuals of human (libertarian) freedom, and so are not middle knowledge at all. Ware can only be saying what he does because of an imperfect grasp of what the medieval theologians, and following them the Reformed and the Reformed Orthodox, have meant by God’s necessary knowledge. I fear that what results is not so much middle knowledge as muddle knowledge.
Of course what particularly interests Ware is the use that this alleged middle knowledge can be put to in elucidating God’s relationship to evil. (GGG 119f.) Here we must note that he is keen to elucidate that relationship, to show more about how it can be that God ordains evil but is not its author. (This is a recurring theme, GGG, 112, 113, 122, 130.) His idea is that middle knowledge explains more of how it can be that such a relation is indirect and permissive, in that God’s decree respecting some possible acts of evil is routed through his knowledge of certain counterfactuals.
Ware represents God as deliberating as he passes from his natural knowledge to his free knowledge via his middle knowledge, his knowledge of counterfactuals. In a footnote (GGG, 120 fn. 12) he warns the reader that such language is a mere façon de parler, since all God’s deliberations are in ‘eternity past’, atemporal. But I doubt whether Ware can have it both ways. I doubt whether he has a hope of preserving a special kind of knowledge, compatibilisit middle knowledge, which is timelessly eternal. For Molina, the contents of God’s middle knowledge were timelessly eternal but they were distinct because they had a distinct content, they were (one and all) possible instances of the liberty of indifference, which immediately gave such knowledge a status distinct from God’s natural knowledge. If Warian middle knowledge is distinct from natural knowledge, what makes it distinct? The only possible candidate for an answer to that question is that it is the knowledge of what God will permissively decree that he ‘comes to’, that he deliberates over such knowledge in deciding which evils he will permissively decree. Ware’s account is, I believe, infected with temporalism. The classic grammar of God , bequeathed to us from Augustine, creaks: temporalism, mutability, ignorance - each began to show their heads
That apart, even as Ware describes this deliberation, whether it is a temporal process or not, it is not at all clear that such middle knowledge helps us with evil. This is what Ware says ‘By controlling the complex of factors prompting the natures of moral agents to develop a strongest inclination within a given situation, God could effectively redirect the choice and action that the agent could carry our’.’ (GGG 121). This cannot be God ‘occasioning’ a choice, as Ware suggests (GGG123) ). But nor can it be a description of God changing the complex of factors, as Ware supposes (GGG 121) but of him simply choosing one set of possibilities over another. God does not change possible people, like a marionette may be changed by being painted red instead of its present green before the show begins. Rather God simply selects a possible person placed in one set of circumstances (one set of possibilities) or that person placed in another set (another possibility).
The only kind of middle knowledge is that devised by Molina and his frères. There are no other kinds, and Ware’s attempt to show that there is another kind, compatibilist middle knowledge, and that it will help us deflecting from God the charge of being the author of evil, is consequently futile. There is divine counterfactual knowledge, knowledge of what might be and of what might have been, but that is part of God’s natural knowledge. And there remain the standard ways of parrying the charge of God and evil: evil as privation, primary and secondary causation, and the Augustinian idea of ‘willing permission’. But the mystery remains as well. Which is how we might expect things to be.