My compatibilist model: a response to some questions

John Johnson wrote a lengthy comment on my post responding to Jerry Walls and my compatibilist proposal. He raises some substantive questions and I think it better to deal with them in another post rather than to reply in a lengthy comment or a number of smaller comments. Because John’s questions are of a sort often raised to positions like mine, I think they deserve careful consideration.

1.  If God is meticulously in control, why is he not culpable for all the evil done in the world?

Johnson would like to see a case for compatibilism that makes sense to him, but he does not think that my proposal successfully exonerates God from culpability for the evil that takes place in our world. That objection to compatibilism has often been raised in my conversations with Arminians (such as Roger Olson and Jerry Walls), and I consider it very serious. I don’t really expect to succeed where better Calvinist theologians than I have failed, but my approach differs in some important ways, and I’d like it to be on the table when synergists study this topic. I am not attracted to debate and apologetics, but I am keen to contribute to  clarification of the issues involved in this discussion. I became a Calvinist kicking and struggling, as many do, and it was biblical exegesis not philosophical argument that brought me to that position. I would rather plead mystery than allow rationalization to silence Scripture, but I do not want to appeal to mystery too quickly, lest I settle for nonsense instead.

The nature of human agency must be sufficiently robust to ground human moral responsibility

First, I concur with synergists that authentic human agency is essential to theodicy. Hard determinism of the mechanistic or materialistic variety is clearly an incompatibilism, and it can lead only to fatalism and to acknowledgment that God is the only genuine actor in the world and hence is morally responsible for all that occurs. Given all the discussion and debate that I have read and heard, I think that we are at an impasse in regard to this particular issue. Soft compatibilists are convinced that the sort of voluntarist freedom which Jonathan Edwards described suffices to give humans genuine moral agency. Incompatibilists are convinced that nothing short of libertarian freedom will suffice, but even that is deemed insufficient if it is part of a hard compatibilist model such as Thomas Aquinas classically propounded.

The nature of human freedom is not explicitly described in Scripture but we must deduce it from what Scripture does clearly teach

I concur with Paul Helm that Scripture does not define the nature of the freedom which God has given to moral creatures, so this is not a dispute which can be resolved exegetically. Nonetheless, I believe that what Scripture tells us about God’s control and what it tells us about human agency are best held together within a soft compatibilist framework.

I am not aware of any place in the writings of J. I. Packer at which he has defined the nature of human freedom, but I have a hunch that he holds to a libertarian view, while not adopting the Thomist account. He is clearly a compatibilist, but I would call his model a “mysterian model” (or perhaps an “antinomist model”) rather than the soft compatibilist model that I have appropriated from Edwards and many Calvinist theologians at work today. I derive the term “antinomism” from Evangelism and the Sovereignty of God, where Packer asserts that Scripture teaches two inescapable truths: God is meticulously sovereign and humans are morally responsible. He grants that this seems contradictory or paradoxical, but he argues that it is not a contradiction but is an “antinomy.” This he defines as “an appearance of contradiction” (18).

“An antinomy exists when a pair of principles stand side by side, seemingly irreconcilable, yet both undeniable. There are cogent reasons for believing each of them; each rests on clear and solid evidence; but it is a mystery to you how they can be squared with each other. You see that each must be true on its own, but you do not see how they can both be true together.” (18-19)

Packer posits that, in biblical theology,  a paradox is always dispensable and comprehensible, but that “an antinomy is neither” of these. “It is not deliberately manufactured; it is forced upon us by the facts themselves. It is unavoidable, and it is insoluble.” So, when we encounter an antinomy such as the relationship between divine sovereignty and human responsibility, we must “accept it for what it is, and learn to live with it” (21). We must regard them as complementary rather than contradictory and teach ourselves “to think of reality in a way that provides for their peaceful coexistence, remembering that reality itself has proved actually to contain them both” (21).

This antinomist or mysterian approach is attractive, and I think that it is much better than denying either that God is meticulously sovereign or that humans are morally responsible. But because I am reluctant to appeal to mystery too quickly, I continue to work at the development and enunciation of a soft compatibilist model which provides more explanation.

I do not think that any of the synergist models of divine knowledge adequately account for the degree of control attributed to God by Scripture, both in his general providence and in his work of salvation. Cutting to the bottom line, here are my perceptions of the shortcomings of each of the three main synergist models.

Simple divine foreknowledge

I have frequently suggested that the future of open theism within evangelicalism depends upon the response of classic Arminians who affirm simple divine foreknowledge (SFK). It is not at all surprising that monergists have often expressed horror at open theism’s contention that God does not know the future libertarianly free acts of creatures. But, as a Calvinist, I think that the strength of the objections voiced by classic Arminians is rather surprising, and I think it follows from their failure to realize how little benefit God would gain from SFK. As open theist philosopher William Hasker put it, the problem with SFK is that by the time God knows the future it is too late for him to do anything about it. If God does not choose the world’s future (as monergists assert), but simply knows it, his comprehensive prescience of the future acts of moral creatures necessarily includes his knowledge of his own actions in every moment of the future. Consequently, he is no more than a passive observer of the future and does not decide upon his own action in the future. This does not do justice to the biblical portrayal of God’s work in the world, as I (and other monergists) understand it, and I actually find open theism a more coherent and dynamic model than classical Arminianism, even though open theism is even further from my own model of God as meticulously in control in his creation.

The most helpful modification of SFK that I have encountered is the proposal that God knows the future incrementally. It is as though he watches the movie for a short time and then stops it and decides how he will act. Then he plays the movie again, which now includes responses to his own action. In essence, this sounds to me rather like a somewhat less dynamic version of open theism. In open theism, God knows the present comprehensively, and the future is still open or undecided. So God is continually assessing the global situation and deciding how he will act, although he can not know with certainty how creatures will then act in the situation which is different from what it would have been if God had not contributed his own action. In the incremental foreknowledge (IFK) proposal, God makes these decisions before the history of the world begins so that the future he knows comprehensively is a future in which his own action contributes meaningfully to the outcome. But that outcome is still not exactly what God had hoped for, because he is unable to predict with certainty what libertarianly free creatures will do until they act (since he can only know the probability of counterfactuals) and so, by the time the creature has decided how to act, even in the IFK model it is too late for God to act differently than he did. The critical problem for IFK is, therefore, that until a libertarianly free creature has chosen what s/he will do, even God can only predict probabilities. Although God has more control in the IFK situation than in SFK, he still has much less control than in the meticulous providence that speaks loudly in my reading of the biblical narrative.

Open theism

I have already summarized the open theist proposal, in my discussion of IFK. Their contention that the future is open for God as well as for us, because even God cannot know the future libertarianly free acts of moral agents, is the distinctive of this synergistic model. In terms of coherence, this model has much to commend it, and I have already stated why I don’t think their proposal reduces God’s control as drastically as many theologians, both Calvinist and classic Arminian, claim. Unlike many fellow Calvinists, I disagree with open theism’s explanation of why God cannot know the future in the way that classic Arminians say he does. This is not because I affirm absolute divine timelessness. On that point, I concur with open theists that, at least since creation, God experiences events in sequence. (Here I have good Calvinist company.) But I have accepted W. L. Craig’s argument that God’s foreknowledge is propositional and should not be construed by analogy with God’s having seen the whole movie already. Rather, God knows all true propositions, and these are essentially tenseless. Any statement made in the present or present tense, that is true, can also be made as a true statement regarding the future.

My problem with open theism is that, although fundamentally consistent internally as a synergistic model, its understanding of God’s limited control in the world takes us even further from the biblical depiction than SFK does. In an effort to explain how it is that God’s knowledge still enables him to act effectively within history, Gregory Boyd has introduced the  idea of “might counterfactuals.” He accepts the grounding objection which both Calvinists and open theists level against Molinism. So he does not assert that God knows with certainty the counterfactuals of libertarianly free decisions, but God’s comprehensive knowledge of the past and present does enable God to predict the probability or likelihood that a person will do a certain thing in particular circumstances. This is coherent and it does give God significant effectiveness, but I suggest that even this knowledge gives God less ability to affect the future in the ways he wants to than Boyd claims it does. The history of the world comes about through an absolutely immense number of decisions by morally responsible agents, and even if God knows that there is a 99.99% probability that X will do Y in a given set of circumstances, that tiny percentage of uncertainty becomes very significant when multiplied by the immense number of decisions involved.


Readers familiar with my thought will know that I have gained a great deal from the work of Molinist theologians. But the grounding objection prevents me from becoming a Molinist. Counterfactual propositions concerning future libertarianly free acts do not have truth value because no decision exists to be known unless it actually made by a particular agent in particular circumstances. If Molinism were coherent, I believe that it would provide a very plausible model to account for God’s strong omnipotence. As a whole, this actual world would be the world of God’s choice before creation, even though its history is brought about largely through creaturely decisions, and God himself would be able to act powerfully without usually restricting the libertarian freedom he gives to moral creatures

2. How does God’s comprehensive knowledge, including his natural knowledge of counterfactuals, contribute to a compatibilist theodicy?

God’s comprehensive knowledge, including his natural knowledge of counterfactuals, is an extremely important factor in my compatibilist understanding of God’s providence. So I want to respond to some of John Johnson’s questions about this.

Johnson claimed that my “idea that God can ‘predict’ what we do because he knows us so well is compatible with a libertarian view of free will but is incompatible with [my] determinism.” He asks: “Why does God need to predict something that he already knows will occur—knowledge grounded in his prior decree and not in simple foreknowledge?”

He has missed the significant difference between God’s knowledge of the actual future and his knowledge of possible futures. It is true that, in a monergist model of providence God can predict what we will do, because he knows what he has planned for the future, in his eternal purpose, not “because he knows us so well.” The important thing in my model, which has not traditionally been stressed by Calvinists, is that when God was deciding upon the history of the world he had decided to create, he knew naturally or necessarily how every “sort” of moral creature would act in all the situations which could possibly occur. This gave God knowledge of all the worlds which he might choose to actualize, worlds in which his own action (or deliberate inaction, apart from sustaining the creature in existence) was an essential factor of every hypothetical complex of circumstances. This also enabled God to give creatures maximal responsible agency, without surrendering his own overall control of the outcome.

Johnson deems my idea of counterfactuals “very problematic,” because it seems to him to undermine “the whole idea of meticulous determination.” He wonders “what would constitute a counterfactual” in my model, rather than it being “a counter-determination?”

That is an extremely important question, and its answer lies in the critical difference between God’s knowledge of counterfactuals and his knowledge of future facts. A huge number of situations which could possibly have arisen within the life of created beings never do arise, and they do not arise because God determines in his meticulous providence that they will not. Johnson thinks that “examples of counterfactuals as we see in Scripture actually seem to assume libertarian freedom (in combination with God’s exhaustive knowledge of individuals).” What he hears in my proposal is the Molinism expertly propounded by William Lane Craig. This is not surprising, because I have learned a great deal from Craig and from Molinism, so that my speaking of counterfactuals and of possible worlds naturally reminds readers more of Molina than it does of Calvin.

Though I have indeed appropriated quite a bit from Molinism, there is a critical difference between its model and mine. Molinists posit God’s knowledge of counterfactuals to be dependent upon his knowing what any  libertarianly free creature would do in any possible situation. The problem with that Molinist construct has been identified as the grounding objection. Molinism is incoherent because beings who act libertarianly, by definition, could have acted otherwise than they did in any given situation where they actually make a decision. But this makes it impossible to predict with certainty what a creature with such freedom would do in a hypothetical situation, because knowledge of their decision is missing.

By contrast, in my own Calvinistic model of hypothetical knowledge, God’s knowledge of counterfactuals is possible precisely creatures are not libertarianly free. Being who they are they would make a particular choice in a particular set of circumstances and they could not choose otherwise. It is for this reason that Calvinists do not affirm divine middle knowledge, but assert that God’s knowledge of future counterfactuals concerning the action of responsibly free creatures is part of his natural knowledge. As I have put it, God knows the principles of agent causation that operate in moral contexts, analogously to his knowing the principles of mathematics or logic. He does not choose these, they derive from his own nature. God knows them because he knows himself, just as he knows the actual future history of the created world because he knows his will.

Contrary to the fear of many in the Calvinist tradition, the proposal that God took into account his knowledge of the principles of creaturely agency, which I take to be the ground of God’s natural or necessary knowledge of counterfactuals, does not make God dependent upon the creature. My fellow Calvinists are very right to guard against any construction that makes God dependent. He is, by nature, completely independent of anything outside of himself. But those who criticize monergists like me, who propose that God made significant use of his hypothetical knowledge in the establishment of his eternal plan or decree, frequently have failed to grasp the essential difference between what God knows counterfactually and what he knows to be factual about the future. God’s actual foreknowledge is, as the Reformed tradition asserts, his knowledge of his own will. He knows what will happen in the future because he has chosen that future for his creation.

In the incompatibilist Molinist model, God knows counterfactuals about future free acts because he knows what the possible creatures would choose to do in those particular situations. But this is precisely what God cannot know, because those decisions are never made. Here open theists and Calvinists agree. By contrast, within a compatibilist Calvinist framework, God’s knowledge of counterfactuals is not a knowledge of the decisions a free creature would make in a hypothetical situation, even though they could have made different decisions. Rather, it is a knowledge of the principles of agent causation, knowing what a particular kind of creature would decide in particular circumstances. The difference here is immense, in that Molinism does make God dependent on the will of creatures for his middle knowledge of counterfactuals. But when God knows counterfactuals naturally or necessarily this dependence does not occur. Most of the hypothetical situations God considers in choosing which world he will actualize are never realized. Creaturely decisions in none of those hypothetical situations are being made by libertarianly free creatures. So God’s independence from his creatures is completely preserved in hypothetical knowledge Calvinism.

3. How can a soft determinist speak of God as permitting anything, since everything that occurs is as he chose it should be?

In my model of compatibilism, I had written:

Although everything that happens is part of the world God chose, and is therefore “determined” by God, his personal action is not of the same sort in every instance… The moral goodness of all moral creatures, like their life, persists only so long as God maintains it. God created everything good, but not self-sufficiently good. I take this to be the substructure of the common Calvinist distinction between those things which God deliberately effects and those which he deliberately permits. He is morally accountable for what he effects, but he only effects what is morally good. He is not morally accountable for the evil he deliberately permits creatures to do because the sinfulness of those actions derives completely from the evil disposition of the creature.

Concerning this proposal, Johnson says that he “smell[s] a bit of an equivocation here,” and so he asks: “How can a God who, ‘In his meticulous sovereignty…chose this particular world and all of its history, down to the finest detail, including the sinful acts of morally responsible creatures,’ ever be said to simply ‘permit’ something? Meticulous determinism does not allow for the language of permission, even ‘deliberate’ permission.”

The perceived problem Johnson mentions derives from his use of the word “simply,” in speaking of the way divine permission works in my model of providence. I have been reminded sometimes by Arminians, that Calvin strongly protested against those who spoke of God’s “permitting” the fall. But when Calvin does that, I hear him speaking about a mere (“simple”) permission, of the sort that synergists of Calvin’s day would regularly have affirmed. Arminians also believe that everything that occurs is permitted by God, but in their model this is true in the sense that God decided to give moral creatures libertarian freedom and thereby to permit them to do whatever they choose. It was in that sense that synergists of Calvin’s day spoke of the fall, as just one of the many things decided upon by creatures, without God’s prevention. Even open theists assert that God can and does restrain the freedom of creatures in some cases, but he does this very rarely.

What troubled Calvin was that some in his own church community were speaking in terms that we might call “proto-Arminian.” I do not believe that Calvin would have had any objection to my speaking of God’s permitting some things, including Adam’s fall, since I insist that Adam’s action in Eden was not permitted as just one of the multitude of decisions that Adam made in his lifetime by God’s permission. As I use the term, it refers to the critical difference between direct and indirect agency on God’s part. There are things that God brings about directly, things which would never have occurred without God’s action to move them in the direction they take. All good done by creatures is of this sort. But, not all creaturely action is God’s direct doing. For reasons usually hidden from us, God decided in many instances that he would not act efficaciously to move creatures to do good, but would allow them to do evil.

Because of what I have called the “moral entropy” in creation, analogous to the natural entropy that physical scientists observe, evil will occur if not restrained by God’s own action. When God chooses not to prevent creaturely evil-doing, we can rightly speak of God as having “permitted” it. The egregiously evil acts of humans in events like the holocaust did not occur in spite of God’s utmost effort to prevent them from occurring, as a synergist would have to posit; they occurred because God chose not to prevent them. They came about because of his “deliberate permission,” and they are part of the whole history of the world which, in its totality is good, and within which every evil act is explicable but only by God, who has not revealed to us how they contribute to his overall good purpose.

Of all the evils God deliberately permitted sinful people to commit, none was more evil than putting to death as a criminal and a blasphemer the sinless man who was uniquely God’s Son. But no greater good has ever come out of any evil than has come from the substitutionary death of Jesus for sinners. His crucifixion was not something God was trying to stop but could not, nor was it something he had foreseen but not chosen. It was the evil God deliberately sent his Son to endure, infinitely good in its results because it was God’s way of redeeming all who believe in him in the ways he makes himself known to them. God determined that Christ should be crucified, but he did not prompt to evil those who brought it about. That was the work of the arch adversary who unwittingly brought about his own destruction.


When we talk about the respective roles God has chosen to reserve to himself or to give to his creatures, we are in complex and momentous territory. This issue is large, indeed it is comprehensive, and we can only settle upon a position by grasping a sense of the entirety of God’s revelation in this regard. So this is something that we will almost certainly tweak as we go through life, re-encountering passages of Scripture often read before but suddenly heard differently, and many of us will have paradigm shifts occur in our theology, perhaps more than once. This is not a bad thing, I think. It is something to which we should always be open, lest our theological system become a strait jacket, preventing us from reading and hearing Scripture with an openness to the illumination of God’s Spirit, which can delightfully surprise us by casting upon texts often read before, a light which we have never seen on previous readings. I love the hopefulness with regard to God’s self-revelation which this fosters in us as we read God’s word and do theology.

This entry was posted in Divine Knowledge, Providence, Theology Proper and tagged , , , , , , , . Bookmark the permalink.

3 Responses to My compatibilist model: a response to some questions

  1. John Johnson says:


    Thank you for the time you have taken to answer my concerns. I especially appreciate your concluding remarks. I have but given this a quick read-through and I hope to put aside some time in the near future to think carefully about the ideas you have articulated here and then, hopefully, dialog further.

    As always, I greatly appreciate the teacher’s heart that is evident in your writing and your obvious love of the truth. Thanks again.


  2. Andrew says:

    Have you read Packer’s Concise Theology? I read him as being a soft compatibilist, to use your terminology.

    • Terrance Tiessen says:

      Thanks for that lead, Andrew. I have not read that work but I can keep my eyes open for this, if/when I do get to it. What I’d be looking for is a specific statement about the nature of the freedom God gives humans, which is compatible with God’s meticulous sovereignty.

Leave a Reply

Your email address will not be published. Required fields are marked *

HTML tags are not allowed.

123,658 Spambots Blocked by Simple Comments