At the ETS meeting in Baltimore in November/13, John Laing read a paper entitled “Middle knowledge and the Assumption of Libertarian Freedom: A Response to Ware.” Though Bruce Ware and I have never collaborated, we reached similar conclusions about the usefulness of God’s knowledge of counterfactuals in his deciding what world he would create, and I appreciate the work he has done. In Providence and Prayer, I had called my model of providence “middle knowledge Calvinism,” and I explained further in a later article in WTJ (69 [2007]: 245-66) why I believed Calvinists should affirm that God knows counterfactuals in a middle “moment” (logically), even though they reject Molinism. That prompted a letter from Paul Helm which led to a very profitable discussion, at the end of which I came to agreement with him that there is no need to postulate that God knows counterfactuals in a middle “moment,” if creatures are not libertarianly free. Within a soft-compatibilist understanding of creaturely freedom, God knows counterfactuals as part of his natural/necessary knowledge. Our conversation was published in an article in WTJ (71 [2009]: 437-54). This change has had no substantive impact on my model of divine providence, but I now speak of it as “hypothetical knowledge Calvinism,” to continue my emphasis on the usefulness of God’s knowledge of counterfactuals to his plan or decree.
Laing’s objective
In his recent paper, Laing states:
the Calvinist Middle Knowledge proposal fails on several counts. Specifically, I will argue that it encounters many of the same supposed problems as Molinism, from the grounding objection to the divine weakness objection, and that it has some problems of its own, from unusual notions of ontology and necessity, to the potential for flirting with fatalism. In the end, I suggest that proponents of Calvinist Middle Knowledge reevaluate their criticisms of libertarian freedom, and adopt full-blown Molinism or (probably more realistically) move to the Thomistic position of seeing God’s knowledge of counterfactuals of creaturely freedom as part of His free knowledge (pp. 1-2).
I am always grateful for the opportunity to revisit my theological understandings through conversation with others who are interested in the subjects about which I have written. Laing’s critique is lengthy, and I want to respond to it in detail, so I’m going to interact with it in a series of posts rather than one long essay. I’ll start with his proposal that the grounding objection which both Calvinists and Open Theists level against Molinism also applies to hypothetical knowledge Calvinism. The grounding objection against Molinism is that, if creatures are libertarianly free, it is impossible for anyone (including God) to predict accurately what they would do in situations which never occur. Hence knowledge of the counterfactuals of future libertarianly free acts is impossible, and Molinism is incoherent. Since the incoherence lies in creatures having the power of contrary choice, why would Laing think that Calvinists, who generally reject that understanding of freedom in favor of the freedom of spontaneity, would be vulnerable to the same grounding objection?
Laing admits to being puzzled “by the Reformed tradition’s insistence [with which I now concur] that God’s knowledge of subjunctive conditionals is part of His natural knowledge rather than his free knowledge” (p. 3). He writes:
I assumed that the grounding of the truth of the proposition that if I were in S, I would (compatibilistically) freely eat a piece of chocolate cheesecake would have to be located in the way God created me, or something similar. Part of the problem is that I assumed the proponent of Calvinist-Middle Knowledge would agree that both
If John were in situation, S, he would freely eat a piece of chocolate cheesecake tonight
and
If John were in situation, S, he would freely refrain from eating a piece of chocolate cheesecake tonight
are possible. Apparently, my assumption was in error, and it is this that I find troubling, not so much that I was in error, though I certainly do not relish being wrong, but because of what seems to follow. According to the position advocated by Ware and Tiessen, following Paul Helm and a host of Reformed luminaries (e.g., Turretin, etc.), if the first is possible, the second is not. Since we have been working on the assumption that it is true that if I were in S, I would eat the cheesecake, we will assume that the counterfactual,
If John were in situation, S, he would freely refrain from eating a piece of chocolate cheesecake tonight
is not only false, but necessarily false; it is not possibly true. Similarly, the counterfactual,
If John were in situation, S, he would freely eat a piece of chocolate cheesecake tonight
is not only true, but necessarily true; it is not possibly false. It is this assertion that I find perplexing and disturbing” (pp. 3-5).
What grounds God’s knowledge of counterfactuals in hypothetical knowledge Calvinism?
First, let me say that I am happy that Laing understands my position correctly. Having grasped it accurately, however, he argues that neither Ware nor I “has really offered a clear explanation for how counterfactuals of compatibilist freedom may be grounded” (p. 5). I will not attempt to respond for Ware, but I will try to explain for the benefit of Laing and others who share his puzzlement what grounds God’s knowledge of counterfactuals.
On Laing’s original assumption that I believed God’s knowledge of counterfactuals to be grounded in his free will, he assumed that God “could create [people] so that their characters lead to different action,” though he does not find that stated clearly in the model (p. 5). That assumption was seriously wrong, because the situation is completely otherwise when God knows counterfactuals naturally or necessarily. In that case, God’s will has nothing to do with this hypothetical knowledge, rather, it rests on the nature of things, on the fact that creatures of a particular sort (including the complex of their inheritance, their nurture, their culture, their habits, their values, their motives, their “affections” [as Jonathan Edwards termed them], and their inclinations etc.) would act in a particular way in a particular set of circumstances. God knows this truth about personal causation as he knows the laws of mathematics, and so he can predict exactly how a particular person would act in hypothetical situations, most of which will never occur.
To use Laing’s own example, God knows Laing so completely that he knows whether or not he would voluntarily eat a piece of cheesecake offered to him in a given set of circumstances. God’s knowledge of a truth about that hypothetical instance is not grounded in his knowledge of Laing’s will (as per Molinism) but of his nature. God simply knows what any person exactly like Laing would choose in that specific situation. This is enormously useful to God when he decides which of many possible worlds he will actualize. It leaves creatures voluntarily free much of the time in those worlds, while giving God the ability to know what they would do. Hypothetical knowledge Calvinism thus provides grounds for God’s knowledge (soft-compatibilistically free agents and the principles of agent causation) which Molinism does not provide when it posits that Laing could eat or not eat the cheesecake in exactly the same circumstances.
I grant that God can know the probabilities (as Open Theist Greg Boyd emphasizes) but he can not know the certainties and, given the immense number of decisions made by moral creatures in a world populated by as many people as ours, over so long a time, even if the probability that Laing would eat (or refrain from eating) were extremely small, the sum total of those uncertainties would make God’s choice of any world very risky. This would put God in precisely the position of having to act responsively in the moment which Open Theists have so well described.
Molinism, hypothetical knowledge Calvinism, and the role of the human will
I concur with Laing’s statement of the grounding objection as it is leveled against Molinism but these objections do not apply to hypothetical knowledge Calvinism.
First, Laing observes that, “since counterfactuals refer to non-actual persons and/or non-actual states of affairs, their truths cannot be grounded in the activity of the agent spoken of in the counterfactual (never mind questions of backward causation” (p. 6). This is true in regard to Molinism, because it ascribes libertarian freedom to persons. If, in exactly the same circumstances, Laing (being who he is in the hypothetical situation) could eat the cheesecake or refrain from eating it, since God does not know which choice he makes in an instance that is still only hypothetical, God cannot predict what Laing would do in that hypothetical situation. Without an act of will on Laing’s part, God could at best only say how probable it is that Laing would or would not eat the proffered cheesecake.
But this problem does not exist for the soft-compatibilist (i.e., the person who affirms that God’s meticulous sovereignty is compatible with the moral responsibility of creatures who act voluntarily) because people’s choice would be predictable with certainty if one knew both them and their circumstances completely, and knew the causative principles governing the choices that such people would make in such circumstances. So, on this point Molinism fails the grounding objection but hypothetical knowledge Calvinism does not.
Molinism, hypothetical knowledge Calvinism and the role of God’s will
Laing then presents the second key feature of the grounding objection to Molinism: counterfactuals “cannot be grounded in God or God’s will because that would make them either necessarily true or at least not independent of His will” (p. 6).
I take this to be a very important concern, because it expresses an issue that also creates nervousness about hypothetical knowledge Calvinism among many other Calvinists. They fear that if God makes use of his knowledge of counterfactuals concerning moral creatures, in a deliberative manner, when choosing a world (that is, determining his decree), he is making himself dependent upon the creature. There is no validity to this concern, however, precisely because no critical divine dependence occurs, given the situation I have outlined. It would be a problem if God were dependent on the actual choices of libertarianly free creatures, as in Molinism, but it is not a problem when the choices of voluntarily free creatures are predictable provided one knows the principles of causation which apply when particular creatures are in particular circumstances. What God knows counterfactually is not dependent on real creatures, it is dependent on principles which God knows naturally, like he knows that 1+1=2. These hypothetical facts are necessary truths, but their necessity, the principles of causation, is somehow grounded in God’s own being rather than his will. They are like rules of logic. God did not have to create this or any world, but if he were to create a world it would have to be coherent with the being of God himself. God is not dependent upon the will of creatures in this world because it was he who chose this world, with all its peculiarities. He chose this particular world, not on the basis of his knowing what the creatures in it would libertarianly freely choose to do (as Molinism asserts), but because he knows the causative principles about the action of soft-compatibilistically free agents. This is a very important plank in my understanding of compatibilism.
Hypothetical knowledge Calvinism with respect to Molinist responses to the grounding objection
Laing describes some of the ways in which Molinists have attempted to answer the grounding objection and he explains why these fail. I concur with his analysis, but I definitely do not concur with his statement that “It is hard to see how proponents of counterfactuals of compatibilistic freedom can do any better with grounding” (p. 6). I think that I have done a great deal better, but I must leave it to my readers to judge how well I have succeeded.
Laing goes on to restate Bruce Ware’s explanation of how the truth value of counterfactuals concerning the future acts of morally responsible creatures is grounded, but he objects that Ware has only explained how God knows counterfactuals, not what grounds their truth. Laing thinks this to be the case because Ware asserts that God knows how the person will choose in accordance with his desires (God’s Greater Glory, p115, n. 10; cited on p. 7). But Laing wants to know: “What grounds the truth of counterfactuals regarding the development of individual desires?” He speculates about ways in which Ware could address this question, and he deems “most promising” a suggestion made by Paul Helm, in his dialogue with me, namely that they may be grounded in the divine mind (WTJ 71 (2009): 445). Of this most promising option, however, Laing asserts that it is “no different” from his own libertarian position, so “Calvinist Middle Knowledge” has the same grounding problem as Molinism.
Once again, I think that Laing is incorrect. To say that the principles of causation relative to persons with compatibilistic freedom are known by God naturally is to assert that God knows them because he knows himself perfectly. Here I refer back to my earlier comment about the necessary coherence of any world God were to create with God’s own being. I speculate that Laing’s assertion that Helm and he are asserting the same thing derives from Laing’s assuming that what Helm means by “God’s mind” is the same thing as Laing means by God’s will. That cannot be the case since Helm believes God to know the truth of counterfactuals necessarily. I can’t promise that Helm would accept my own words concerning the necessary coherence of any world he would create with God’s own being as synonymous with his own statement. What I can say is that I agree with Helm’s assertion that God “would know, intuitively and immediately, what A in C would do” (p. 445). Helm unpacks that further:
For as part of his natural knowledge God has the idea of A in C as a possibility, along with his knowledge of A possessing innumerable different beliefs and desires in innumerable different sets of circumstances, and (given such knowledge), in his wisdom God creates A in C, creates him down to the last atom and molecule, evil apart, and immediately sustains his life nanosecond by nanosecond, even as, while being sustained by his Creator, A in C perpetrates evil. Such a state of affairs does not necessarily involve divine coercion or compulsion, not in the usual senses of these words, though it will if God in his wisdom decrees to create A as being in some circumstances not responsible for his actions. It is not that God in his wisdom permits possible persons such as A to exist, rather he brings it about that they exist by decreeing that they do, and (in his wisdom) he permits their perpetration of evil.
I completely agree with Helm at this point, and I propose that what he and I are asserting with regard to God’s natural knowledge of how a person A would act in circumstances C, where A is soft-compatibilistically free, is definitely not vulnerable to the grounding objection that is so frequently brought against Molinism. It is the soft-compatibilistic nature of human freedom (rather than an incompatibilist construction of libertarian freedom) which grounds the truth value of the counterfactuals which God knows necessarily. By contrast, Molinists assert humans to be libertarianly free and this is why they are unable to explain what grounds the truth value of the counterfactuals concerning the acts of those free creatures, which they propose God knows in his middle knowledge.
In short, Laing’s first objection to hypothetical knowledge Calvinism fails. The grounding objection to Molinism does not apply to hypothetical knowledge Calvinism.
In future posts, we’ll look at other criticisms Laing has made of hypothetical knowledge Calvinism, but that is enough for this post.
