Why Calvinists Should Believe In Divine Middle Knowledge, Although They Reject Molinism

By Terrance L. Tiessen

[Terrance L. Tiessen is Professor Emeritus of Systematic Theology and Ethics at Providence Theological Seminary, Otterburne, Manitoba, Canada.]

I. Introduction

In Providence and Prayer: How Does God Work in the World? I proposed a model of divine providence that I dubbed “middle knowledge Calvinism.”[1] More recently, Bruce Ware has argued for the same position, which he identifies as “compatibilist middle knowledge.”[2] As William Lane Craig has observed,

Christian theologians have typically affirmed that in virtue of his omniscience, God possesses counterfactual knowledge.. .. Not until Friedrich Schleiermacher and the advent of modern theology did theologians think to deny God knowledge of true counterfactuals. Everyone who had considered the issue agreed that God has such knowledge.[3]

What theologians disputed was “so to speak, when God has such counterfactual knowledge.”[4]

It is possible that some of the confusion about “when” God has the knowledge of counterfactuals derives from a definitional ambiguity. I am using the term as Craig does when he says that “counterfactuals are conditional statements in the subjunctive mood.”[5] One might assume that the term refers only to events which never happen but which would have happened if circumstances had been different, that is, that no counterfactual ever actually occurs. By this very strict definition of a “counterfactual,” both Molinists and Calvinists would have to assert that God knows counterfactuals only in his free knowledge, since they could not be known, by definition, until after God had chosen a particular world and thereby decided what would actually occur.

Response to the middle knowledge Calvinist proposal has indicated both misunderstanding and disagreement. Some have argued that God’s knowledge of counterfactuals (that is, of purely hypothetical events, the knowledge of what creatures would do in particular circumstances, which may or may not ever be realized) is part of his necessary or natural knowledge, and others have argued that it is part of God’s free knowledge. In either case, it is contended, there is no reason to posit divine middle knowledge. In this article, I will examine particularly the classic Calvinist objections to the concept of divine middle knowledge. Thereby, I hope to clear up some misunderstandings concerning the proposal and to demonstrate that the affirmation of divine middle knowledge is both consistent with and beneficial to the traditional Calvinist belief in meticulous divine sovereignty.

Robert Dabney asserted that the “history of the controversy on scientia media” presents an instance of the rule “that usually mischievous errors have in them a certain modicum of valuable truth. Without this, they would not have strength in them to run, and do mischief.”[6] It is precisely the “modicum of valuable truth” that I am trying to extract from the Molinist proposal, correcting it with soft-determinist rather than libertarian freedom.[7] It is more important to me to reach agreement amongst Calvinists that God makes significant use of his knowledge of counterfactuals (or true hypothetical events) than it is to reach agreement about when, logically, God has this knowledge. Nevertheless, I believe that this knowledge is sufficiently different in kind to be designated separately as middle knowledge. For this to work, however, it is essential that middle knowledge not be confused with Molinism in spite of its history within that framework.

II. The Value Of Divine Middle Knowledge And The Shortcomings Of The Molinist Proposal

The middle knowledge Calvinist position affirms that God has three kinds or logical moments of knowledge rather than only two. The “three-stage” or “three-kind” construct has usually been identified with Molinism, but the two-stage understanding was taught by Thomas Aquinas and by Reformed orthodox theologians. In the two-stage or two-kind model of divine knowledge,[8] God knows all the things that are possible by virtue of their consistency with God’s own nature (natural or necessary knowledge), that is, he knows everything that could be, he knows all possible worlds. God also knows everything that will be, but God knows this as a knowledge of his own will (hence, as “free” knowledge). But proponents of the three-stage or three-kind model of divine knowledge believe that there is a significant difference between the knowledge God has of things which are possible by virtue of their consistency with God’s own nature (his natural or necessary knowledge) and his knowledge of what creatures would do in particular circumstances, which may or may not ever occur, depending on which of the many possible worlds God decides to actualize. Since God’s knowledge of what possible creatures would do is significantly or categorically different from his knowledge of what they could do (necessary knowledge) but is logically prior to God’s knowledge of what actual creatures will do, this kind of divine knowledge is rightly termed “middle knowledge.” It comes logically between God’s necessary or natural knowledge and his free knowledge.[9] The logical order is as follows:

necessary knowledge - middle knowledge - [the decree] - free knowledge 

(of what could be) — (of what would be if ...) — (of what will be)

The significant difference between the Molinist and the Calvinist proposals is that Molinists believe that moral creatures are libertarianly free (that is, that they have the power of contrary choice) whereas Calvinists assert that moral creatures have the liberty of spontaneity. This difference in the definition of creaturely freedom is important because it enables Calvinists to insist that creaturely freedom of a morally responsible kind is compatible with divine determination, whereas Molinists believe these two to be incompatible. For this reason, compatibilist freedom is also labelled “soft-determinist” freedom as opposed to the indeterminist kind of freedom which Molinists believe in.

Where Molinists and middle knowledge Calvinists are agreed, however, is that God’s knowledge of what particular creatures would do in all possible circumstances is immensely useful to God. It enables him to choose the world history best suited to his nature and purposes for the world without having to force his will upon his creatures. In knowing how a particular creature would act in a given set of circumstances, God is able to choose the world in which the combination of the actions of God and his creatures would eventuate in exactly what God desired, but this can be brought about with minimal direct intervention. Thus, God need not force or coerce his creatures in order to have things turn out as he wished.[10] This has valuable implications for understanding how God is genuinely responsive to human prayer[11] and how God meticulously governs human history in a manner that preserves his absolute sovereignty without taking away the moral responsibility of the human beings who bring about that history. It helps us to understand why God is not morally responsible for the evil acts of his creatures,[12] even though these are all part of his eternal purpose and are, therefore, done according to God’s will.[13] God’s use of middle knowledge is part of his being the “all wise God” who established his eternal purpose wisely, drawing upon all that he knows, both by necessity and through contemplation of the possible worlds that he might choose to actualize if he wished.

William Lane Craig has helpfully stated the way in which Molinism gives us an understanding of the providence of God in Acts 2:23 and 4:27–28:

For via his middle knowledge, God knew exactly which persons, if members of the Sanhedrin, would freely vote for Jesus’ condemnation; which persons, if in Jerusalem, would freely demand Christ’s death, favoring the release of Barabbas; what Herod, if king, would freely do in reaction to Jesus and to Pilate’s plea to judge him on his own; and what Pilate himself, if holding the prefecture of Palestine in A.D. 27, would freely do under pressure from the Jewish leaders and the crowd. Knowing all the possible circumstances, persons, and permutations of these circumstances and persons, God decreed to create just those circumstances and just those people who would freely do what God willed to happen. Thus, the whole scenario, as Luke insists, unfolds according to God’s plan. This is truly mind-boggling. When one reflects that the existence of the various circumstances and persons involved was itself the result of a myriad of prior free choices on the part of these and other agents, and these in turn of yet other prior contingencies, and so on, then we see that only an omniscient mind could providentially direct a world of free creatures toward his sovereignly established ends. In fact, Paul reflects that “none of the rulers of this age understood this; for if they had, they would not have crucified the Lord of glory” (1 Cor 2:8). Once one grasps the doctrine of divine middle knowledge, one is led to adoration and praise of God for his breathtaking sovereignty.[14]

To this description of the usefulness of God’s knowledge of hypothetical events (counterfactuals), I give a hearty “Amen,” but then I move quickly to assert that the situation that Craig has so nicely described is possible only if creatures have compatibilistic freedom. One of the attractions of Molinism is that it attributes libertarian freedom to creatures, since most people in our culture, and even within the Christian church worldwide, accept only that kind of freedom as authentic or genuine.[15] If Molinism were coherent, it would therefore be a very powerful option for Christians who believe that God is strongly sovereign and that creatures are genuinely free. Calvinists, on the other hand, have rejected the common notion that creaturely freedom is libertarian because it “reduces human choosing to arbitrariness,” and because it is not compatible with the strong view of divine sovereignty that is taught in Scripture.[16]

Calvinists can appreciate the intention of Molinism to maintain strong divine sovereignty even though moral creatures are deemed libertarianly free,[17] but we believe that the “grounding objection” to Molinism is valid.[18] We agree with other critics of Molinism (both synergists and monergists) that it is not possible for anyone, including God, to know what a libertarianly free creature would do in a hypothetical situation. If people had the power of contrary choice, when they decided to act in a particular way they would have been able to choose otherwise, even though everything else in the situation remained the same. Consequently, it is by definition impossible to know what a libertarianly free person will decide in a situation, no matter how completely one knows the person and all of the circumstances in which the decision is made, until the occasion actually occurs and the person makes a decision. Of course, God can know the odds that a person will act in a particular way, but if people are libertarianly free, there is always the possibility that they will choose against the odds.[19]

Craig has argued that “what makes it true that ‘If I were rich, I would buy a Mercedes’ is the fact that if I were rich I would buy a Mercedes.”[20] Unfortunately, the best that an omniscient God could do in the case of a libertarianly free creature is to predict the odds that he would or would not buy a Mercedes. By definition, if Craig is libertarianly free and if he becomes rich, he could still, in that particular instance, do the least likely probable thing! The most God could know ahead of time is that, “If Craig were rich, he would be 99 percent likely to buy a Mercedes.” The problem is that, given the huge number of decisions that creatures make in the history of a world like ours, the accumulating effect of the decisions made improbably would ultimately make prediction, even by an omniscient God, very imprecise. Thus, God’s knowledge of the “might-counterfactuals” of which Gregory Boyd speaks[21] does not end up being much use to God in his providential planning.

Whether or not one accepts the validity of the grounding objection to Molinism is very important. If it is true that God knows counterfactuals of creaturely freedom as Molinists argue, but if God could not know these counterfactuals if creatures are libertarianly free, then creatures do not have libertarian freedom.[22] Since I believe that God does know counterfactuals and that he could not do so if creatures were libertarianly free, I affirm the compatibilist proposal that creatures have the freedom of spontaneity. This makes it possible for God to know what they, being who they are, would do in any possible situation.[23] Whether God has this knowledge as part of his necessary knowledge or as middle knowledge, it is extremely useful to God in the logical moment at which he decides which of the possible worlds he will actualize.

III. God’s Knowledge Of Future Counterfactuals Is Not Part Of His Free Knowledge

It is important to reiterate the sense in which I am speaking of “counterfac-tuals,” namely, to refer to “conditional statements in the subjunctive mood.” As I noted earlier, if counterfactuals are construed as statements about possible events that are never actualized, then God would know these statements only after his decree to create a particular world. But it is clear that some theologians who argue that God knows counterfactuals as part of his free knowledge are not restricting their statement to this stringent definition of the term. Their position reveals an important misunderstanding of the nature of things in middle knowledge proposals, whether Molinist or Calvinist.

1. The Dominicans

Craig notes that

Catholic theologians of the Dominican order held that God’s counterfactual knowledge is logically subsequent to his decree to create a certain world. They maintained that in decreeing that a particular world exist, God also decreed which counterfactual statements are true. Logically prior to the divine decree, there are no counterfactual truths to be known. All God knows at that logical moment are the necessary truths, including all the various possibilities.[24]

As a result, when God decreed that a particular world would be real, God also decreed which counterfactuals are true. “Thus he decrees, for example, that if Peter had instead been in such-and-such circumstances, he would have denied Christ two times. God’s counterfactual knowledge, like his foreknowledge, is logically posterior to the divine creative decree.”[25]

Craig reports that “the Molinists charged that the Dominicans had obliterated human freedom by making counterfactual truths a consequence of God’s decree, for on the Dominican account it is God who determines what each person will do in whatever circumstances he finds himself.”[26] The Molinists, on the other hand, were trying to make room for creaturely freedom by arguing that what particular creatures would do in given situations is determined by those creatures, not by God. In principle, I agree with the Molinists against the Dominicans on this point, but I can only see how the construct works if creatures are soft-deterministically free, for reasons cited above.

2. Walls And Dongell

Jerry Walls and Joseph Dongell have assessed the situation in the same way that the Dominicans did. Commenting on my middle knowledge Calvinist proposal, they state that 

the difficulty with this suggestion is that the only way to make sense of God’s knowing counterfactuals of compatibilist freedom is to assume that those choices are causally determined by God. That is, God’s knowledge of such counterfactuals is his knowledge of what choices he would have causally determined in different circumstances.[27]

Accordingly, they deny that compatibilists are able to speak of divine “permission” in “any meaningful sense.”

Walls and Dongell badly misconstrue my proposal because, in regard to counterfactuals, what God knows is what creatures would choose to do in particular circumstances, apart from any intervention on his part, and what they would do if God changed the situation in some way by actions of his own. In God’s decree (which provides the content of his free knowledge), God is then able to decide whether he will simply permit people (or angels) to do what they would naturally do in those circumstances, or whether he needs to introduce activity of his own Spirit, either by way of dissuasion (even allowing an evil spirit to deceive, as with Ahab’s prophets), or by promptings of his own. It is only in God’s decree that he decides which of the huge number of counterfactuals will be actualized, and how they will come to be, whether they are actively permitted (as in the case of evil) or brought about through more specific divine activity (as in the case of all the good which redounds to God’s glory). At the “middle” moment, when God knows all the hypotheticals of creaturely action, what God knows is what we would do in such situations, not what he will do and not even what he would do. God deliberately permitted Joseph’s brothers to sell him to traders headed for Egypt and he permitted Satan to put ideas into Judas’s heart, but he knew all of the individuals involved in these situations so completely that he knew exactly how they would respond, if left to themselves, in each case.

Thus, in my middle knowledge Calvinist model, before God decides what he will do by way of creating a world and ordering its history, God knows how particular creatures would act if they found themselves in particular sets of circumstances. This knowledge enables God to choose a world whose history is exactly the history that he wisely purposes, but to bring about that history through a combination of the morally responsible actions of rational creatures and God’s own actions. God does this in such a way that all glory for everything good must be attributed to God, but all culpability for evil done by creatures justly lies in the will of those creatures. God is able to choose a great number of people to salvation, but then to bring about their salvation in such a way that all who are saved by grace contribute nothing to their salvation which would give them ground for boasting, and no one is condemned who does not freely choose to reject God’s grace.[28]

3. John Laing

John Laing makes the same error that Walls and Dongell make. He concludes his analysis of my middle knowledge Calvinist proposal with the statement: “The soft determinist may claim that God possesses knowledge of counterfactuals of compatibilistic freedom, but she cannot claim that such knowledge is prevolitional; it must be part of God’s free knowledge.”[29] Let me explain why this assertion is not the case. Laing cites the statement: “(1) If John were offered an extra night’s stay for free, he would (freely) decline.” Laing then argues that “the proponent of middle knowledge must make the further claim that the truth of (1) is in no way dependent upon the will of God,” and Laing opines that moderate Calvinists would be reluctant to make such a claim.[30] But Laing’s conclusion is wrong because he fails to take into account the important difference between middle knowledge and free knowledge which is that the former, by definition, is not dependent upon the will of God but is prior to it. I am, therefore, not reluctant to assert that God knew that Laing would decline an extra night’s stay for free if offered it, in those particular circumstances. As a counterfactual, this is not dependent upon God’s will; it is the way things would be if God chose to actualize the world in which Laing, being exactly who he is now, got into that particular situation. If God had reasons for wanting things to turn out differently, he would not allow the particular circumstances to obtain, or he would have brought into Laing’s personal formation influences that would make him a person who would choose differently than this particular John Laing would do.

Laing puts forward these two conditionals:

(5) If given the choice to eat the forbidden fruit, Adam will (compatibilistically freely) eat. 

(6) If given the choice to eat the forbidden fruit, Adam will (compatibilistically freely) fast.

Laing then asserts:

According to Tiessen, God can actualize a world where (6) is true and (5) is false. But how can he, apart from Adam’s choice being in some way dependent upon the will of God? If it is dependent upon the will of God, then it seems that God is just as constrained in terms of the kind of world he can actualize in the Calvinist Molinist model as he is in the traditional (libertarian) Molinist model because Adam may never compatibilistically choose to abstain from eating the fruit.[31]

There is a grain of truth here but it doesn’t take us where Laing thinks it does. It is true that, in any world God chooses to actualize, Adam’s choice will be dependent upon the will of God in some way. Adam will only resist the temptation to eat the forbidden fruit if God gives him efficient grace to that end. But this puts no constraint upon God. God can do that if he chooses. The point is that, at the middle knowledge stage, God knows what sort of grace would be efficient if he were to create a man like Adam and allow him to face that situation. Similarly (cf. 1 Cor 10:13), God knows concerning all of us believers what grace would be necessary to effectually keep us from yielding to temptation. He knows what we could bear and what we could not bear. He is not left to wonder whether or not we might succumb to a particular temptation, waiting to discover this when we exercise our will.

In short, God’s knowledge of how particular creatures would act in particular situations is prior to his decree to create this particular world and it is, therefore, not part of his free knowledge, which is his knowledge of his own decree. If it is not part of God’s necessary knowledge, it must therefore be part of a middle knowledge. But traditionally, Calvinists have asserted that God’s knowledge of hypotheticals of possible creaturely action, or counterfactuals, is indeed prior to God’s decree and to his free knowledge; they have included it in his necessary or natural knowledge.[32] I suggest that we do better to distinguish between two kinds of prevolitional knowledge and hence that we should accept the existence of divine middle knowledge.

IV. God’s Knowledge Of Future Counterfactuals Is Not Part Of His Necessary Knowledge

In a moment, I will examine the objections to middle knowledge raised by Reformed theologians at the time when Molinism was initially proposed. But first, I want to examine a recent analysis of middle knowledge Calvinism.

1. David Werther’s Case for Including Counterfactuals of Compatibilist Freedom in God’s Necessary Knowledge.

David Werther states the following conclusion: “Given a compatibilist account of free will, an essential feature of Calvinism, there is no room for the threefold distinction between God’s natural, middle, and free knowledge: God’s knowledge of counterfactuals must be entirely free knowledge or entirely natural knowledge.”[33] Werther goes on to argue that God’s knowledge of counterfactuals must be part of God’s necessary knowledge. He asks us to assume that

J. I. Packer is asked to join a fusion jazz band and accepts. Call the state of affairs containing all that determines Packer’s acceptance A, and the state of affairs constituting his acceptance J. On a Calvinist account of meticulous providence and freedom, I take it that A will include, among other things, God’s decree and some desires and/or reasons of Packer’s. If so, then so long as God is an immaterial agent and Packer is not identical to his body, it follows that the determination of Packer’s acceptance cannot be merely a matter of causal, that is, physical necessity. Physical states of affairs may be included in A and J but do not exhaust A and J.

Since Packer’s acceptance, J, is a determined choice, Packer could not have failed to accept given A, the state of affairs constituting all that determines his acceptance. To maintain otherwise would be to affirm a libertarian liberty of indifference. Expressed more precisely, every possible world W including a state of affairs A, the state of affairs containing all that determines Packer’s decision to join a fusion jazz band, also includes J, the state of affairs constituting Packer’s acceptance of the invitation to join the band. And, if this is so, then among Packer’s essential properties is the property of responding affirmatively if A obtains.[34]

The statement regarding Packer’s essential properties certainly seems to set up a situation in which my model will have to concede that God’s knowledge of what Packer would do in state of affairs A is part of his necessary knowledge. But is the statement true? Does God’s knowledge of all the possible worlds consistent with his own nature—that is, of all the worlds he could possibly create— include the item that “every possible world including a state of affairs A also includes J”? I think not, because of the following distinction.

At the necessary stage, I think of God knowing, as part of his knowledge of himself, the principles of causation that underlie the statement. But I propose that God’s contemplation of worlds containing moral agents which he could self-consistently create, and of how the history of each of such worlds would unfold if left to itself, and how it would unfold if he acted personally within the world, indicate that a further stage, beyond God’s necessary knowledge, has been reached. Here God would not simply be intuiting, he would be deliberating, analysing, calculating “what if” scenarios that include both the creatures’ actions and his own. If he gave creatures compatibilistic freedom, he would be able to predict how they would act if left to themselves but he would also be able to predict how influences of his Spirit or changes in the history of the world prior to a time would bring about different effects. I judge that this deliberative process of assessment, and of playing out scenarios, moves beyond what God knows simply because he knows himself and all that he could do consistently with his own nature. This deliberative process makes use of his natural knowledge, but it moves beyond it and yet is prior to his decision to choose a particular world, including all its details, a world whose history comes to be as God has purposed, partly through the spontaneously free agency of moral creatures. If what I am describing need not be viewed as in any way moving beyond what God knows necessarily, then Werther and the Reformed tradition are correct, and God has only two kinds of knowledge. At this point, I see a difference that warrants speaking of the deliberative process as “middle,” since it utilizes the natural knowledge while moving beyond it and yet precedes the free knowledge, which is God’s knowledge of what he decides shall be the reality of his creation.

Werther assumes that God’s knowledge that Packer would join the band, if invited to do so, indicates a necessary property in Packer, which would make this knowledge part of God’s necessary knowledge. But this encounters a serious difficulty in regard to God’s free knowledge. Let’s assume that in the actual world, Packer is invited to join the band and accepts the invitation. Classically Reformed theologians have insisted that God has only two kinds of knowledge and that human freedom is compatibilist, so that what Packer actually will freely do is soft-determined. So, if we took Werther’s approach, Packer’s acceptance of the invitation would once again be a necessary property of Packer. But, in that case, what has traditionally been deemed God’s free knowledge would also be necessary and so all of God’s knowledge, including his knowledge of the actual future, would be necessary knowledge. Thus, Werther’s line of argument challenges not just divine middle knowledge; it also challenges a free knowledge distinct from God’s necessary knowledge. This indicates that it is very important to identify how and “when” properties become essential.

2. The Classic Calvinist Objections To Divine Middle Knowledge

1. Libertarian Freedom and Synergism. Richard Muller observes that the problem of middle knowledge, in the minds of seventeenth-century theologians, was

not merely a speculative problem concerning the way in which God knows future contingents and conditionals—it was a broader theological problem concerning the underlying intention of the theory of scientia media, namely, the affirmation of a synergistic soteriology, and the use to which the doctrine was put by various groups viewed as heterodox by the Reformed.[35]

In this vein, Robert Dabney speaks of middle knowledge as “first invented by the Jesuit Molina, in order to sustain their semi-Pelagian doctrine of self-determining will, and of conditional election.”[36]

2. The Assertion of Divine Contingency upon the Creature. Given the Reformed emphasis on the sovereign God’s independence of his creatures, it is not surprising that theologians should be nervous about a proposal which makes God dependent upon his creatures, even if only in his knowledge of them. A. A. Hodge, for instance, argues that middle knowledge “makes the knowledge of God to depend upon the acts of his creatures exterior to himself. This is both absurd and impious, if God is infinite, eternal and absolute.”[37] Thus, Richard Muller speaks of middle knowledge as “a foreknowledge of future conditionals or conditional future contingencies arising from the free choice of creatures prior to the divine willing.”[38] From Muller’s perspective, the problem is that middle knowledge “understands God neither as simply willing a particular possible world rather than another nor as having willed a particular (actual) world, but as foreknowing and reacting to the result of a finite contingency or conditionality as prior to and apart from his willing”[39] This concern is aroused particularly because of the libertarian freedom that Molinism attributes to created moral beings.

What troubled Reformed theologians like Cocceius was that, in the middle knowledge scheme, “God knows what will occur contingently upon certain conditions lying outside of his will: these conditions are not mere possibility nor divinely willed actuality, but foreknown conditions, foreknown as actual apart from the decree, at least for the sake of stating the contingency”[40] Cocceius argued that “there can be no scientia media. .. because there can be no being independent of the divine will.”[41] An echo of this concern is heard in Stephen Charnocke’s statement that God

hath not one knowledge of himself, and another knowledge of the creatures; but by knowing himself, as the original and exemplary cause of all things, he cannot be ignorant of any creature which he is the cause of; so that he knows all things, not by an understanding of them, but by an understanding of himself.[42]

Voetius complained that middle knowledge

makes God’s understanding depend upon an external object, makes His idea come to Him from another source and so makes the eternal reasons of events in the priority of nature to be in the creature before they are in God and creatures here not to depend on God, but God to depend on the creatures as His pattern.[43]

More recently Paul Helm has complained that Bruce Ware’s compatibilist middle knowledge perspective “combines two contradictory positions.” Ware “avows meticulous, exhaustive divine control,” but he also “affirms a difference in principle between God’s relation to persons and his relation to their circumstances.” Helm suggests that “this can only be because of the autonomous powers that Ware attributes to people.”[44]

Robert Dabney describes well the manner in which God foreknows the volitions of agents who are free in the sense which compatibilists assert, namely, “through His infinite insight into the way their dispositions will naturally act under given circumstances, placed around them by His intentional providence.” With these words Dabney has described the key idea in my proposal concerning God’s knowledge of counterfactuals, and Dabney says: “The Calvinist should by no means flout it; but accept, under proper limitations.” But Dabney goes on to argue that “the term mediate is not accurate, to express this orthodox sense; because it seems to imply derivation subsequent, in the part of God’s cognition said to be mediated, from the independent will of the creatures.”[45]

3. The Uncertainty Entailed in God’s Knowledge of Future Conditionals. Muller posits that

both the Reformed and the Socinians recognized,. .. albeit from different perspectives, that the notion of a certain divine foreknowledge of future conditionals is a rather unstable concept: in order for God to know the conditional conditionally, God would have to be ignorant of its resolution in actuality. In other words, the “if “ of the conditional would have to represent an indeterminacy and uncertainty in God himself.[46]

4. The Unknowability of Conditional Future Things Apart from the Divine Will. Dabney posits that God “cannot certainly foreknow an act, unless its futurition is certain.”[47] Similarly, Francis Turretin observes that “things not true cannot be foreknown as true.” He then argues that “conditional future things are not true apart from the determination of the divine will.”[48] Whether or not the Sidonians would have repented, for instance, depends upon the divine will. Apart from God’s will “no cause of this thing can be imagined.”[49] Thus, Turretin argues that “what is conceived to be determinately from God can also be pronounced to be determinately; but what is conceived only to be possibly can be pronounced to be only possibly.” Turretin denies that “the coexistence of a free act on hypothesis can be conceived to be determinately antecedently to the decree.” He grants, rather, “that it may be possibly.” Thus he suggests,

It is true that Peter would possibly sin if placed in a given order of things antecedently to the decree; but not determinately so as to make it true that Peter would actually and in fact sin if placed in such an order of things. This could not be certain unless from a permissive decree of God.[50]

5. The Attribution of Discursive Reasoning to God. Muller cites a conviction of the Reformed orthodox that “the process of discursive reasoning itself cannot be attributed to God,”51 and he argues this point from “the Reformed assumption of an eternal or successionless duration of God.”52 We must not speak of

a discursive or successive process of knowing and willing in the divine mind, according to which God weighs contingent possibilities in his scientia necessaria and then chooses which one to actualize based on his desire to actualize one particular result rather than another. (And even this rejected position would not need a middle knowledge between God’s necessary and free knowledge.)[53]

Similarly, Robert Dabney contends that “we must also believe that God knows all things intuitively and not deductively. A deduction is a discovery. To discover something implies previous imperfection of knowledge.”[54]

3. A Response to These Calvinist Objections to Middle Knowledge

I noted at the outset of this article that I am much more interested in gaining acknowledgement from fellow Calvinists that God knows contingent hypotheticals (or counterfactuals) and makes use of this knowledge in the formation of his wise plan than I am in getting agreement that God knows these truths in a middle “moment.” Muller observes that none of the seventeenth-century Reformed theologians “denied the divine foreknowledge of future contingency What they denied was that future contingencies could be construed as sets of foreknown conditions known by God as other than mere possibilities and yet also known by God prior to God’s willing them.”[55] Nevertheless, I find it both valid and useful to distinguish categorically between the propositions God knows to be true because of his self-knowledge (which includes all possibilities) and the propositions that God knows would be true if he were to will that creatures of a particular kind existed in particular situations. In other words, I find it helpful to speak of a category of things that are not just possible but that would be so, if the particular circumstances were actualized. Since only a few of these will be decreed by God to be actual, most of them become counterfactuals, in the strict sense. In response to the key Calvinist objections to middle knowledge, I offer the following response.

1. Libertarian Freedom and Synergism. Repeatedly, in my reading of classics of Reformed orthodoxy, I have sensed that the massive resistance to the concept of middle knowledge is primarily a resistance to synergism. This seems to have prevented most Reformed theologians from considering how middle knowledge might function within a monergistic framework such as I am putting forward. Take, for instance, Leonhardus Rijssen’s statement that

all acts of a created will are subject to divine providence, so that nothing is indeterminate; this [line of argument] proves absolutely the interdependence of primary and secondary causality and of the creature and the creator—so that all creatures depend on God not only for their existence (non tatum in esse), but also in their activity (sed etiam in operari). Therefore, a middle knowledge (scientia media) that is supposed to have as its object a free determination of the will dependent upon no higher cause is impossible.[56]

Rijssen rightly criticizes Molinism because of the libertarian freedom it attributes to moral creatures; but his argument presents no difficulty for compatibilist middle knowledge, since this entails a contemplation of potential realities that would only be real if God wills them to be, in which case they would indeed be dependent upon God.[57]

I agree with Muller that the Reformed objection to the Molinist model of divine knowledge “belongs to an attempt to maintain the soteriological monergism of the Reformation over against various synergistic theologies.”[58] Francis Turretin, for instance, observes that the middle knowledge proposal was “adopted by the Socinians and the Remonstrants who courageously defend it so as to preserve free will in the citadel.”[59] I grant this to be true, but middle knowledge can be appropriated for other reasons. I too am committed to monergism, but I regret that the urgency of defending God’s meticulous sovereignty in providence and in redemption seems to have prevented Reformed theologians from considering the usefulness of incorporating compatibilist middle knowledge into their own monergistic understanding of God. As often happens at times of theological dispute, the baby got thrown out with the bathwater. Rejection of Molinism because it used middle knowledge in defense of libertarian freedom need not lead to rejection of middle knowledge within a monergistic framework.

Voetius complains that, if middle knowledge is true, “(3) From this knowledge we on our part will be able to assign the reason and cause of divine predestination.—(4) Thus, predestination as regards this knowledge will have to be called postdestination rather than predestination, as regards the temporary object.”[60] But my model is clearly not subject to this criticism. God does not choose people to salvation based upon his foreknowledge of who will believe in the particular circumstances he foresees are going to occur, but he knows what circumstances need to exist for those whom he chooses to salvation to come to him voluntarily

2. The Assertion of Divine Contingency upon the Creature. I fail to see why it should be deemed problematic that God knows what a creature would do because of (i.e., based upon) his knowledge of that possible creature. This is not a difficulty so long as the finite contingency or conditionality that God takes into account is only a possibility until God decides whether or not it will actually occur. In other words, God’s knowledge is not dependent on an act of any real creature. The point is simply that, in deciding what world he will actualize, God uses knowledge of how particular creatures would act in particular circumstances, so that he can realize his purposes with minimal direct involvement or intervention. Thus, the nature of contingency or dependence at work in this sort of knowledge does not endanger God’s essential independence in any way. It is certainly inappropriate, concerning my compatibilist understanding of divine middle knowledge, to suggest that humans have “autonomous powers,” as Paul Helm complains concerning Ware’s articulation of this model of divine providence. This is not the synergism of Molinism and humans have no autonomous powers. They actually do only and always what God eternally purposed that they should do when he chose to actualize this particular world.[61]

Cocceius is therefore unnecessarily concerned about beings “independent of the divine will” (see §4.2.2 above). No beings are independent of the divine will. At the middle stage, these are only possible beings in the mind of God, beings that he might or might not will to actualize, as he wishes. Should God will to actualize them, their existence will have been contingent upon his will, and God will know of their actuality because he knows his own will. Consequently Dabney’s concern that middle knowledge would be mediated “from the independent will of the creatures” is legitimate in regard to Molinism but not in a monergistic appropriation of divine middle knowledge.

When Turretin discusses predestination, he objects that a middle knowledge proposal would assign a reason for predestination other than God’s purpose and good pleasure “because the foreseen consent of the will of Jacob placed in such circumstances would be at least the condition without which God could not predestine to salvation Jacob rather than Esau.”[62] I share Turretin’s concern to protect God’s sovereign freedom and independence, but his objection to middle knowledge is unfounded, provided that compatibilism is maintained. God’s sovereignty in his decree is not diminished in any way by his using middle knowledge in the process of determining the actual future. It is not problematic to grant that certain circumstances are a necessary part of Jacob’s salvation, provided that God retains the decision to actualize the world in which such circumstances occur or, alternately, to actualize a world in which such circumstances do not occur, a world in which Esau would be the heir of the promise.

Muller complains that middle knowledge “confers a quasi actuality upon the possibility,” and he objects that this “cannot be, inasmuch as all actuality is conferred by God.”[63] Similarly, Turretin objects to future certainties known prior to the decree, because he thinks of them as having the sort of actuality that only the decree can produce.[64] This is simply not an issue for my proposal, since the future hypotheticals have no actuality as events, although they have truth value. They are, in the strict sense, all counterfactuals, until God decides which ones will be factuals and which will be (strictly speaking) counterfactuals. Prior to the decree these objects of God’s knowledge do not have actuality. They become actual truths only if God actualizes them, but prior to his decision they are truths of a particularly significant kind. But, and this is critical, they are more than the mere possibilities which God knows naturally; they are truths about the particular worlds between which God chooses, possible worlds which God knows prior to his decree by his rational contemplation of the way in which particular creatures would behave in particular sets of circumstances.

3. The Uncertainty Entailed in God’s Knowledge of Future Conditionals. It is not a problem that, to use Muller’s words, “in order for God to know the conditional conditionally, God would have to be ignorant of its resolution in actuality” (see §4.2.3 above). In fact, that is precisely the point of its “middleness.” Knowing this category of possibilities, God will then decide what becomes actual. Thus, this middle knowledge of God does not “represent an indeterminacy and uncertainty in God himself,” as though there were a deficiency in the being of God. The uncertainty exists precisely because, at this point, God has not decided what will be certain. He has not yet determined whether this will be a factual or a counterfactual (in the strict sense), so it is indeterminate until he does so. God’s uncertainty at this point, concerning which of these hypothetical contingents will become actual, simply derives from the fact that God has not yet decided what world he will create and govern. Since Calvinist theologians have generally granted that God has this sort of knowledge as an aspect of his necessary knowledge, the logical priority of this particular knowledge to God’s decree and hence to his free knowledge should be of no concern. This is not a peculiarity of the middle knowledge concept.

4. The Unknowability of Conditional Future Things Apart from the Divine Will. We meet a serious confusion in Dabney’s statement, “God cannot certainly foreknow an act, unless its futurition is certain” (see §4.2.4 above). What Dabney and many other Reformed theologians have failed to grasp is that these counterfactuals are concerning something that would be future and certain, if God chose to actualize the world in which they occur. God can know them, even though they never actually occur, given compatibilistic freedom. Dabney argues that if the causative connection between two events is certain, then “when God foreordained the existence of the second cause, He equally ordained that of the effect. But there are but the two sources, from which the certainty of its futurition could come,”[65] (namely, God’s effectuation without any second cause and his effectuation through natural second causes[66]). Fair enough, but only if God did foreordain the world in which this agent existed as a second cause would the effect come about. Nevertheless, the truth God knows about the possible event is grounded in the nature of the creature as that creature would exist at that time, by God’s foreordination, were the situation actual. Since none of this would come about apart from the divine decree, we need not fear that God’s independence of his creatures is compromised at all. If the situation were actually future, it would be because God had decreed it, but as an item of middle knowledge, prior to God’s decree, it is a knowledge that is grounded in the nature and disposition of the creature.[67] It is precisely for this reason that we should distinguish God’s knowledge of these true hypotheticals concerning possible futures from his knowledge of other merely possible events.

Thus, Turretin is mistaken in his argument that “conditional future things are not true apart from the determination of the divine will” (see §4.2.4 above) and so they cannot be known by middle knowledge. God can know that, if the external revelation and the particular inner work of the Holy Spirit which God was bringing to bear upon the people of Chorazin and Bethsaida had been brought to bear upon the Sidonians, many of them would have repented (Matt 11:21). There are degrees of hardness of heart, and God would have known the relative hardness of the hearts of people in these two instances. Indeed, would Turretin not want to say that, had God chosen to bring the Sidonians to repentance, he would have known ahead of time what it would take to bring that about? Surely he does not want to suggest that God might have decided in his eternal purpose to bring the Sidonians to repentance but that he would not have been sure just what that might take when the time came! God could know what people like those in Sidon would do if instructed with a particular message and if the Spirit of God did a particular sort of work in their hearts and minds. This is the essence of compatibilist freedom, and we should not shrink away from its implications for God’s knowledge because of a fear of the synergism that was part and parcel of Molinism.

Turretin contends that we can only know for certain that Peter will in fact sin if we know that God has decreed this to be so. I grant that Peter would actually and in fact sin only if God decreed to permit him to do so. But, given God’s knowledge of connections, habitudes, and so forth, which Turretin himself grants, God would know, hypothetically, whether Peter would sin in particular situations or not. This is more than just knowing that Peter might possibly sin; it is knowing that he would sin, in that situation. The primary issue between Turretin and me, therefore, is not whether or not God has this kind of knowledge but whether it is necessary or middle.

5. The Attribution of Discursive Reasoning to God. I fail to see why it should be deemed problematic to conceive of God as reasoning discursively. The objection seems to be a perceived consequence of God’s absolute timelessness. But, given the manner in which such timelessness is generally conceived, why could God not deliberate in the way that I propose? Whatever framework is used by Reformed theologians to account for the transition from necessary to free knowledge, from knowledge of all possibilities to choice of what will be real, can easily accommodate a middle moment. The Reformed tradition already allows for a process or function or activity involved in God’s choosing, from among the immense array of possibilities, what he will bring about (directly and indirectly). This applies easily to the intermediate process of contemplating how particular creatures (if God were to create them) would act in particular situations (if God were to ordain them), if allowed to carry on without intervention from God. The proposed movement from middle knowledge to free knowledge requires no inferential or discursive process that differs significantly from the traditionally conceived transition from God’s necessary knowledge to his free knowledge. The interposition of a middle moment is not problematic, provided it does not include a change in the nature of creaturely freedom from soft-determinist to indeterminist freedom.

Dabney is concerned to avoid the implication that God learns things through deduction and discovery. But Dabney himself has spoken of God knowing the possible, including all possible universes, and then selecting the one that “pleased Him best.”[68] That entails deliberation and sounds successive, even though the succession is logical and need not be temporal, so the interposition of another “stage” should be no problem. Further, inference need not imply “previous imperfection of knowledge.” It is, rather, an inner deliberation in which God uses powers of reason and deduction, but the outcome adds nothing to God’s knowledge in any way that could be identified as an improvement, as though the state prior to this deliberation were less perfect than the one afterwards.

Even in the traditional construct, God’s eternal purpose is wise. It is the “plans” of the Lord that he realizes and these are wisely considered. And God’s knowledge of himself is at least logically prior to his decision to create something other than but dependent on himself. I am less sure that such a discursive process must be “middle.” At the very least, however, a valuable distinction should be made between the kind of knowledge I am postulating and the things God knows through perfect knowledge of himself. If two sub-categories are defined within the necessary knowledge I have no great quibble, but, since the objects of divine knowledge of which I am speaking are essentially the same as are spoken about by Molinists, it seems best to acknowledge the usefulness of their concept of middle knowledge, while defining differently the nature of the creaturely freedom entailed. This radically modifies the construct and, in fact, gives God’s knowledge the ground that is lacking in Molinism.

Consider, for instance, Molina’s proposal about God’s answer to David concerning what the men of Keilah would do if he spent the night there. I grant that God’s statement to David about what the men would do may not have been knowledge of a counterfactual truth but simply a statement concerning their ill intent.[69] Nevertheless, in my proposal, prior to God’s decree, God was able to know in which world(s) David would be betrayed and in which world(s) he would not be betrayed so that God could choose between them. If we do not grant this knowledge to God, and if we do not assume that God pondered this knowledge in establishing his decree, then God lacks something essential to the wisdom of his decree. To collapse this sort of knowledge into the category of the necessary fails to acknowledge its distinctiveness.

In God’s natural knowledge, he has the knowledge of logical relations, causal relationships, and so on, that grounds his more particular knowledge in the middle stage. But it is only when God begins to contemplate the creation of a world with various kinds of creatures that he has reason to assess how such world histories might come out, and then he considers what it would take to bring a particular man like Joseph into Egypt in a situation where God would be able to provide a context for the development of the covenant people as a nation.

V. Conclusion

By now I hope that my reasons for appropriating the concept of divine middle knowledge into a monergistic model of divine providence are clear. Molinists have correctly identified the benefits that this concept provides. The fact that their own proposal is incoherent because creatures are assumed to be libertarianly free should not prevent compatibilists from assessing the validity of the middle knowledge construct. God’s knowledge of what particular kinds of creatures would do in particular possible circumstances is not dependent upon God’s decree; nor is it something inherent in God’s knowledge of himself and of all that could possibly be consistent with himself (i.e., his necessary or natural knowledge). This is something that God would only contemplate if he were to deliberate about creating a world distinct from but dependent upon himself. Thus, we do best to distinguish this as a distinct logical moment in God’s knowing which is still prior to his deciding upon the history of the world.

Once God decided that he would create a world, the kind of knowledge which Molinists have described is essential to God’s wise planning. Since this is a knowledge that God uses in forming his eternal purpose, it logically precedes his free knowledge, which is a knowledge of his own will. Hence, it is only right that it should be dubbed “middle” knowledge. Affirming that God has knowledge of this kind does not diminish his independence in any way, but it best portrays the kind of interaction between God and his moral creatures that compatibilists believe to be true.

Notes

1. Terrance Tiessen, Providence and Prayer: How Does God Work in the World? (Downers Grove: Inter-Varsity, 2000).
2. Bruce Ware, God’s Greater Glory: The Exalted God of Scripture and the Christian Faith (Wheaton, Ill.: Crossway, 2004), 110–30. John S. Feinberg has also stated: “While I doubt that an indeterminist could consistently hold that God has middle knowledge, I see no reason for a determinist to deny this” (No One Like Him: The Doctrine of God [Foundations of Evangelical Theology; Wheaton, Ill.: Crossway, 2001], 752).
3. William Lane Craig, “The Middle Knowledge View,” in Divine Foreknowledge: Four Views (ed. James K. Beilby and Paul R. Eddy; Downers Grove: InterVarsity, 2001), 120.
4. Ibid.
5. Ibid.
6. Robert L. Dabney, Systematic Theology (Carlisle, Pa.: Banner of Truth, 1985), 159.
7. Alvin Plantinga, as a Reformed philosopher, has incorporated God’s knowledge of possible worlds into his argument for the logical compatibility of the existence of a good and omnipotent God with the existence of evil (God, Freedom and Evil [Grand Rapids: Eerdmans, 1974]). Without speaking of middle knowledge, he has revived the concept in its traditional Molinist form, including the assumption that humans are libertarianly free. Consequently, my own defense of divine middle knowledge within a framework of meticulous divine sovereignty and compatibilist human freedom is traditionally Calvinistic and should not be confused with Plantinga’s proposal.
8. Richard Muller describes the way in which Reformed orthodoxy appropriated but qualified the terms used by the older scholasticism: “The ‘simple understanding’ is sometimes called scientia necessaria, since it is a knowledge that God must necessarily have, given the divine omniscience. It is ‘natural’ because God possesses it according to his nature: it is a knowledge that God has and, in fact, must have simply because God is God. It is an ‘indefinite’ knowledge with regard to things other than the divine essence itself, inasmuch as its objects are possibilities, not actualities. This ultimate divine knowing is typically juxtaposed with scientia libera. Such knowledge is ‘free’ because it rests on the divine freedom, specifically on the freedom of God to will certain possibilities and not others. The Reformed orthodox postulation that God’s free or visionary knowledge is definite and certain therefore does not mean that God’s knowledge is restricted to certain events and things. It is definite because it refers to actualities. It is certain because it is neither partial nor contingent” (The Divine Essence and Attributes [vol. 3 of Post-Reformation Reformed Dogmatics: The Rise and Development of Reformed Orthodoxy, ca. 1520 to ca. 1725; 2d ed.; Grand Rapids: Baker Academic, 2003], 411–12).
9. For purposes of this discussion, we need not decide whether God is timeless or everlasting. Most of the theologians involved in the original dispute over divine middle knowledge assumed that God is timeless, but they did not hesitate to use temporal language when speaking of the logical order entailed in God’s planning and knowledge.
10. Although he vigorously opposes Molinism, Robert M. Adams asserts the usefulness of middle knowledge, if it were possible, for precisely the reasons I have cited. See “An Anti-Molinist Argument,” in Middle Knowledge: Theory and Applications (ed. William Hasker, David Basinger, and Eef Dekker; Contributions to Philosophical Theology 4; New York: Peter Lang, 2000), 343–44.
11. See Tiessen, Providence and Prayer, 337–62; and Tiessen, “Can God Be Responsive if the Future Is Not Open?,” in Semper Reformandum: Studies in Honour of Clark H. Pinnock (ed. Stanley E. Porter and Anthony R. Cross; Carlisle, U.K.: Paternoster, 2003), 94–110.
12. See, e.g., Ware, God’s Greater Glory, 115–30. Ware states helpfully: “Perhaps we should think of God’s regulating the factors of a situation, then, as ‘occasioning’ a particular choice to be made, rather than as ‘causing’ a particular choice to be made. Because God knows the natures of each person perfectly, he knows how those natures will respond to particular sets of factors presented to them. Thus, without causing a person to do evil, he nonetheless controls the evil they do. He controls whether evil is done, what evil is done, and in any and every case he could prevent the evil from being done. But in no case does he cause the evil to be done. In this way, God maintains meticulous control over evil while his moral creatures alone are the agents who do evil, and they alone bear moral responsibility for the evil they freely do” (123–24).
13. Paul Helm is correct to point out that God’s relationship to the good acts of humans whom he graciously enables to do that good is no less puzzling than his relationship to evil human acts. In both cases, God’s knowledge of how particular creatures would respond to situations and to God’s own influence is valuable to God in his providential governance of morally responsible creatures (“The Classical Calvinist Concept of God,” calvinist-concept-of-god.html [accessed 24 January 2007]).
14. Craig, “The Middle Knowledge View,” 134–35.
15. Cf. Thomas Flint’s description of “the lure of Libertarianism” in Divine Providence: The Molinist Account (Ithaca, N.Y.: Cornell University Press, 1998), 24–31.
16. See, e.g., Ware, God’s Greater Glory, 112.
17. Thomism goes even further, since it also affirms libertarian freedom but it believes this to be compatible with God’s meticulous sovereignty.
18. E.g., Charles Hodge, Systematic Theology (New York: Charles Scribner & Co., 1878), 399–400.
19. The grounding objection has been much debated between Molinism and its detractors, and the controversy seems to have reached an impasse. The objection has been stated by numerous philosophers and theologians, synergists and monergists alike. Gregory Boyd, for instance, has identified the particular problem that Molinists face, which is not a difficulty for compatibilist middle knowledge: “How can agents be said to be self-determining when they don’t ground the definiteness of what they shall do and ever would have done in different circumstances? The definiteness unexplainably eternally precedes them, created by nothing. If agents possess self-determining or libertarian freedom, I argue, they must be the ones who ultimately resolve indefinite possibilities into settled facts” (“An Open-Theism Response [to Molinism],” in Divine Foreknowledge: Four Views, 146).
20. Craig, “The Middle Knowledge View,” 142.
21. Boyd, “An Open-Theism Response,” 146.
22. This is John M. Frame’s point: “Craig would like to believe that middle knowledge reconciles divine sovereignty with libertarian freedom. In fact, it does not. If divine creation on the basis of middle knowledge means anything, it means that libertarianism is excluded. Craig is inconsistent to affirm both libertarianism and the divine act of actualizing a complete, possible world, including all creaturely choices” (The Doctrine of God [vol. 2 in Frame’s Theology of Lordship series; Phillipsburg, N.J.: Presbyterian & Reformed, 2002], 505).
23. Fascinatingly, it would be hard to find a better description of the ground of God’s knowledge of counterfactuals of compatibilist freedom than this one offered by Gregory Boyd: “To the extent that would-counterfactuals apply to future free agents, they do so because the actions of these agents flow either from the character God has given them (habitus infusus, in classical terminology) or from the character they will acquire if they pursue a certain possible course of action (habitus acquisitus, in classical terminology). In either case the would-counterfactuals are not ungrounded, as in classical Molinism. From all eternity God knows that if he chooses to create free agent X, she will have the basic characteristics of a, b and c. And from all eternity God knows that if agent X freely follows a certain possible life trajectory, she will become the kind of person who would do x in situation z” (“An Open-Theism Response,” 148).
24. Craig, “The Middle Knowledge View,” 121.
25. Ibid., 122.
26. Ibid.
27. Jerry L. Walls and Joseph R. Dongell, Why I Am Not a Calvinist (Downers Grove: InterVarsity, 2004), 128 n. 10. Precisely the same misunderstanding is at work in David Basinger’s thought when he posits that in a theologically determinist universe, “God does in fact have a form of MK. That is, in addition to knowing what. .. will happen in the actual world, he knows exactly what would have happened in any other world he could have actualized. But his decisions are not based on what he foreknows. He does not, in other words, utilize his MK when deciding how to act. For the compatibilistic God just the opposite is true. His knowledge is based on his decisions. Since he can create any self-consistent state of affairs of which he can conceive, he knows what will or would happen in any context because he knows how he has decided or would have decided to respond to that which he encounters. Accordingly, for the compatibilist, the nature of God’s knowledge is irrelevant to his ability to control earthly affairs” (David Basinger, “Middle Knowledge and Classical Christian Thought,” in Middle Knowledge: Theory and Applications, 183).
28. Synergists doubt that this actually works as monergists portray the situation. I grant the weight of some of their concerns and have tried to address those concerns, without abandoning monergism, in my proposal of “universal sufficient grace,” in ch. 11 of Who Can Be Saved?: Reassessing Salvation in Christ and World Religions (Downers Grove: InterVarsity, 2004), 230–58.
29. John D. Laing, “The Compatibility of Calvinism and Middle Knowledge,” JETS 47 (2004): 467.
30. Ibid., 455.
31. Ibid., 466.
32. E.g., John Frame: “God in Scripture often speaks of what would happen in conditions other than those that actually occur.. .. From a Reformed point of view, however, it is difficult to see why this kind of divine knowledge must be isolated as a third kind of knowledge, alongside necessary and free knowledge.. .. When God knows possible worlds, does he not also, by virtue of that knowledge, also know all possible free creatures and their possible actions? So, from a Reformed point of view, there is no reason why we shouldn’t regard God’s knowledge of contingencies under the category of necessary knowledge” (Doctrine of God, 502–3; emphasis added).
33. David Werther, “Calvinism and Middle Knowledge, “ http://www.arsdisputandi.org/publish/articles/000122/article.htm (accessed 26 August 2004), abstract.
34. Ibid., §3.
35. Muller, Divine Essence and Attributes, 420. I take this to be Karl Barth’ s concern also, so that his objections to middle knowledge are potent with reference to Molinism but do not address a monergistic model like my own (cf. Karl Barth, Church Dogmatics II/1, The Doctrine of God [ed. G. W. Bromiley and T. F. Torrance; trans. T. H. L. Parker et al.; Edinburgh: T&T Clark, 1957], 575–86). See also Archibald Alexander Hodge, Outlines of Theology (rev. and enl. ed.; New York: Thomas Nelson & Sons, 1891), 43.
36. Dabney, Systematic Theology, 156.
37. Hodge, Outlines of Theology, 148.
38. Muller, Divine Essence and Attributes, 419.
39. Ibid.
40. Cocceius, Aphorismi prolixiores (vol. 7 of Opera omnia; Amsterdam, 1675), §4; cited by Muller, Divine Essence and Attributes, 421.
41. Ibid.
42. Stephen Charnocke, Discourses Upon the Existence and Attributes of God (new ed.; London: James Blackwood & Co., 1880), 292; emphasis added.
43. Gisbertus Voetius, Selectarum disputationum theologicarum pars prima (Utrecht, 1648), 320; cited in Heinrich Heppe, Reformed Dogmatics: Set Out and Illustrated from the Sources (rev. and ed. Ernst Bizer; trans. G. T. Thomson; London: George Allen & Unwin, 1950), 80.
44. Helm, “The Classical Calvinist Concept of God.”
45. Dabney, Systematic Theology, 159.
46. Muller, Divine Essence and Attributes, 421.
47. Dabney, Systematic Theology, 157.
48. Francis Turretin, Institutes of Elenctic Theology (3 vols.; ed. James T. Dennison, Jr.; trans. George Musgrave Giger; Phillipsburg, N.J.: Presbyterian & Reformed, 1992), 1:214 (3.13.10).
49. Ibid. Cf. Heppe, Reformed Dogmatics, 77.
50. Turretin, Institutes, 1:217–18 (3.13.20).
51. Muller, Divine Essence and Attributes, 424.
52. Ibid., 429.
53. Ibid.
54. Dabney, Systematic Theology, 50. See also the objections to discursive reasoning by God in Heppe, Reformed Dogmatics, 70–72.
55. Muller, Divine Essence and Attributes, 420–21. I found it intriguing that John Gill offers so fine a description of what I am affirming as God’s middle knowledge. He does it after he has described the “knowledge of simple intelligence” and before he speaks of the “knowledge of vision,” but he does not mention middle knowledge. Perhaps he avoided the term because he was aware of the negative response it usually elicits from Reformed theologians, or perhaps he simply included this knowledge in God’s necessary knowledge. Gill writes: “ [God] knows what might be, and in course, would be, should he not prevent them by the interposition of his power and providence, and which he determines to do: so he knew the wickedness and treachery of the men of Keilah to David, and that if he stayed there, they would deliver him up into the hands of Saul, and therefore gave him notice of it, that he might make his escape from them, and so prevent their giving him up, according to his determinate will, 1 Sam. 23. 11, 12. God knows the wickedness of some men’s hearts, that they would be guilty of the most shocking crimes, and that without number, if suffered to live, and therefore he takes them away by death; and that such is the temper of some, that if they had a large share of riches, they would be so haughty and overbearing there would be no living by them: and that even some good men, if they had them, would be tempted to abuse them, to their own hurt, and therefore he gives them poverty” (Body of Divinity, or A System of Evangelical Truths, Deduced from the Sacred Scriptures [London: John Gill, 1769–1770; repr., Atlanta: Turner Lassetter, 1950]), 61.
56. Leonhardus Rijssen, Summa theologiae (Bern, 1686), 3.24, controversia, arg. 3; cited by Muller, Divine Essence and Attributes, 428. Similarly, Turretin wrote: “(3) If all the acts of the created will fall under the divine providence so that none are independent and indeterminate, no middle knowledge can be granted (which is supposed to have for its object the free determination of the will, depending upon no superior cause)” (Institutes, 1:215 [3.13.11]). Note that, here again, Turretin’s objection is to libertarian freedom, not to middle knowledge per se.
57. It should be noted, however, that Rijssen did not even allow for a so-called hypothetical knowledge,” such as put forth by Gomarus and Walaeus (Muller, Divine Essence and Attributes, 429).
58. Ibid., 432.
59. Turretin, Institutes, 1:213 (3.13).
60. Voetius, Selectarum disputationum pars prima, 320; cited in Heppe, Reformed Dogmatics,80.
61. Cf. Helm, “The Classical Calvinist Concept of God.”
62. Turretin, Institutes, 1:216(3.13.14).
63. Muller, Divine Essence and Attributes, 429.
64. Turretin, Institutes, 1:217 (3.13.19).
65. Dabney Systematic Theology, 157. Ibid., 158.
66. Ibid., 158.
67. In this regard, I wonder whether John Frame has fully grasped the Molinist proposal. He writes: ‘‘Indeed, when God ‘chose to create one of these possible worlds,’ was he not foreordaining everything that would come to pass in that world? Craig may say that God’s choice was motivated by his knowledge of Peter’s libertarian free choices, but God’s choice itself necessarily limited Peter’s freedom, so that he would no longer have that kind of free will’’ (Doctrine of God, 504–5). Does this not confuse what God knows in the two distinct moments of his middle and free knowledge? We can put aside the issue of libertarian freedom since it is part of the Molinist proposal but is not essential to the concept of middle knowledge. What Frame’s statement fails to recognize, I think, is that Peter’s decision is sure in this world after God chooses this world, but that its immediate determiner in this world was Peter. It is a fact that God chose the world in which Peter denies Jesus but it is a world in which Peter freely denies (again, setting aside the question of whether that freedom is libertarian or soft-determinist). God knew that Peter, being the person he had come to be by the moment of temptation, would deny Jesus if put into exactly that situation. Knowing this, God was able to choose that this Peter would end up in this situation, and God needed to do nothing directly to bring it about that Peter denies Jesus. God permits this to happen by purposely allowing Peter to be who he was and by letting him get into that particular set of circumstances. Underlying Frame’s objection to Craig’s Molinism, I think, is the grounding objection and that objection is legitimate, but it must not prevent us from seeing the intrinsic rightness of the basic concept of middle knowledge, abstracted from the assumption that creaturely freedom is libertarian.
68. Dabney, Systematic Theology, 50: “And unless God had before Him the ideas of all possible universes, He may not have chosen the one which, had He known more, would have pleased Him best; His power was limited.”
69. Following Turretin, Institutes, 1:216 (3.13.15).