Books Providence

Responses to Boyd’s open theist model of providence

 The questions that expose the incoherence of the neo-Molinist account of divine providence . . . establish that the God of open theism is an ambivalent and arbitrary warrior who cannot be trusted to rule in every situation in a way that minimizes evil and maximizes good for his creatures. (Helseth, 222)

Molinism [handles the problem of evil better than open theism] for God permits horrible evils only in view of morally sufficient reasons, whereas the God of open theism, once he sees disaster looming on the horizon (Hitler’s coming to power, for example), does not interfere to prevent terrible evils that anyone could have seen coming. (Craig, 230)

Compared to the traditional universe I inhabit, the open theist world is rotated ninety degrees to the left. . . . To change the metaphor, every word in the open theist dictionary possesses a meaning different than the words in the old one I still use. (Highfield, 231)

As with the other models in Four Views on Divine Providence, the other three contributors get a chance to respond to Gregory Boyd’s open theist proposal.

Response by P. K. Helseth

In addition to concerns arising from Boyd’s “neo-Molinism,” which Paul Helseth raised in response to W. L. Craig’s essay, Helseth identifies 3 other lines of critique:

  • Boyd’s account is “descriptively inadequate.” His description of his own view is “excessively optimistic,” but his description of the Reformed view is “woefully deficient.” Boyd “ignores the nuances and flattens out the distinctions that rescue the Reformed understanding of providence from sounding silly,” comparing his own view to a straw man of his own making (209). For example, his analogy of God with a demented scientist “fails to take seriously the Reference insistence that secondary causes are real rather than just imaginary causes” (210).
  • Second, Boyd makes human experience, rather than Scripture, paramount “as the standard for determining what is and is not plausible.” This is seen, for instance, in Boyd’s complaint that the compatibilism affirmed by Reformed theologians lacks any analogy to his own or “any conceivable experience” (210). Like the Trinity, Helseth posits, the compatibility of God’s sovereignty and human freedom and responsibility is mysterious, but it is “received on the authority of what God has revealed in his Word” (210).
  • Third, Boyd’s model is “vulnerable because it is marred by an incoherence that threatens” its integrity (210). An instance of this is found in Boyd’s treatment of Peter’s denial of Jesus as an example of the compatibilism that results from solidification of a person’s character. But “Peter had not yet irreversibly become the decisions that he had previously made,” so that Boyd’s explanation of Jesus’ foreknowledge doesn’t work, though a deterministic model does.

The remainder of Helseth’s critique focuses on the third item, and he draws on Boyd’s other writings to make his point. He carefully describes Boyd’s account of human freedom and moral responsibility and then evaluates Boyd’s description of when God can intervene and work in a coercive way. Helseth argues that Boyd’s treatment of Peter’s denial of Jesus “presents insurmountable difficulties for the neo-Molinist account of divine providence,” because it presents God as “willing to violate the self-determining freedom of responsible moral agents in order to bring about states of affairs that he really wants to bring about” (217). The problem Helseth sees in Boyd’s account is “that God intervened coercively with Peter before Peter’s character became ‘crystallized in the form of an irreversible character.’” In fact, “Boyd contends that God’s compatibilistic intervention  with Peter was the ‘loving but necessarily harsh’ means by which Peter’s character was ‘permanently changed,’ thereby conceding that Peter’s character was never irreversibly established or ‘eternalized’ in the first place” (218).

In Helseth’s estimation, what Boyd has done demonstrates “both the coherence of a more full-bodied form of compatibilism and the utter unworthiness of the God of open theism” (218). Boyd is seen to have put himself on the horns of a dilemma, either he “must concede that the ‘God of the possible’ can know what openness theologians insist it is logically impossible for him to know . . . or he must acknowledge that God knew what Peter would do because God knew that he would orchestrate circumstances that would make Peter betray Jesus, in which case God forced Peter to sin” (219). Either way, Boyd has made it much more difficult for himself “to rescue God from being tarnished by the problem of evil” (220), because he “cannot say that God is always doing ‘all he could do’ to minimize evil and maximize good when what he is doing falls short of violating the covenant that he has established with free agents” (222).

Response by W. L. Craig

William Lane Craig opens his critique of Boyd’s presentation with an evaluation of his christocentric criteria for an adequate theory of providence, and Craig posits that, in this approach, “what one sees in Jesus as determinative tends to be rather subjective,” and the “criteria state necessary, but not sufficient, conditions for an adequate doctrine” (224). In place of Boyd’s criteria, therefore, Craig offers two: (1) consistency with the whole of Scripture (not just christologically relevant passages),  and (2) consistency with perfect being theology, understanding God’s attributes, such as omnipotence, omniscience, and goodness, “in the greatest terms possible” (224).

Craig affirms Boyd’s conviction that libertarian freedom is necessitated for humans by God’s other-oriented love, but he considers Boyd’s “analysis of libertarian freedom as involving the power of alternative possibilities to be defective” (225). Rather, says Craig, what is necessary is “the absence of causal constraints outside oneself that determines how one chooses,” that is, “that we have genuine ‘say-so’ about our choices” (225).

The second problem Craig finds in Boyd’s essay “is his uncritically equating control with causal determinism” (225). Boyd has thereby joined hands with Helseth here, but he then finds the compatibilism of the Westminster Confession incomprehensible. By contrast, Craig asserts, the statement by the Westminster divines “is precisely what the Molinist believes,” which makes it tragic “that in rejecting middle knowledge, Reformed divines have cut themselves off from the most perspicuous explanation of the coherence of this wonderful confession” (226).

Boyd’s “discussion of the nature of time and of God’s foreknowledge of future contingents” is Craig’s third area of concern, because Boyd runs together a “tensed” theory of time (in which the future is pure potentiality) and causal indeterminism. Craig argues that these two claims are distinct, since one could embrace a tensed theory of time and be a causal determinist, so the two should be kept distinct, even though Craig himself accepts them both (229).

Craig proposes that Boyd could avoid the charge that he denies divine omniscience if he denied that there are any future contingent truths (228). Better yet, however, Boyd’s complicated maneuver is “unnecessary, since one can embrace a tensed theory of time and causal indeterminism without sacrificing the bivalence of future contingent statements” (229). As to Boyd’s rejection of the Molinist contention that God knows the truth of counterfactuals, Craig proposes that “if knowing what he would freely do in any set of circumstances is consistent with God’s freedom, it is hard to see why his knowing what we would freely do in any circumstances is inconsistent with our freedom” (229).

Molinism handles the problem of evil better than Boyd’s open theism, in Craig’s opinion. It posits that

“God permits horrible evils only in view of morally sufficient reasons, whereas the God of open theism, once he sees disaster looming on the horizon (Hitler’s coming to power, for example), does not interfere to prevent terrible evils that anyone could have seen coming. Any reason he might have for not intervening can be adopted with equal justification by the Molinist” (230).

Response by R. Highfield

Ron Highfield finds Boyd’s view nearly incommensurable to his own, but he chooses a few decisive differences to illustrate its defects. First, with regard to the open theist vision of God, Highfield charges that it consistently ignores the doctrine of divine simplicity, and that it “reads the limitations of creatures into God,” by applying to God “the concepts of love, power, and relation . . . in the same sense that they apply to creatures” (232). The result is a portrayal of God as temporal, changeable, passible, and limited in his knowledge of the future.

Although Boyd argues that open theism best satisfies the four Christological criteria which he believes are critical for an adequate doctrine of providence, Highfield dnies that Boyd’s doctrine rises “to the genuinely christocentric,” as a consequence of focusing on principles that “by no means articulate the cardinal truths of the revelation of God in Christ” (235).

  • Concerning spiritual warfare, Highfield wonders why God should not “be infinitely better at persuasion than Satan, so that God wins every time’ (236).
  • Boyd contends that God wisely uses persuasion rather than power, in order not to be coercive but, since Highfield finds no reason to assert that it is logically impossible that God could persuade all free agents to choose the good freely on every occasion, “either he does not really want to do so or he does not know how to do it.” Consequently, “either God sometimes intentionally allows evil to win or his wisdom is limited” (237).
  • Although Boyd’s appeal to divine love is sentimentally appealing, it fails fundamentally because it does not begin “with the eternal love among the Trinity, in which there is no suffering, no change, and no possibility of refusal to love in return” (237). Highfield offers as a better demonstration of God’s love for sinners, not that God allows us to say yes or no, but that he refuses “to take our idiotic no seriously” (237).
  • Regarding Boyd’s attempt to demonstrate that, in his model, “God wins by bringing good out of evil,” Highfield identifies three problems: (1) “even with a very generally defined view of God’s plan, . . . the God of open theism cannot know that this goal will be realized” (238). Boyd has actually anticipated this objection, but Highfield is stunned by Boyd’s statement that God has excluded from the adventure “possible story lines that could not result in God’s bringing good out of evil” (238). He asserts that this perspective “should result in a massive revision of open theism” (238). (2) “Even if an infinite number of ‘might be’ worlds fit the general criteria of divine success and God manages to have one of them realized, . . . there can be no assurance, much less certainty, that God will be able to realize anything close to the best ‘might be’ world” (239). (3) Boyd’s model of God’s bringing good out of evil is very impersonal, giving us “no assurance that God knows how to make my salvation sure or that the evil that befalls me tomorrow cannot thwart God’s plan for me” (239).

Highfield deems libertarian freedom and free agency to be “highly abstract and logically polished concepts,” which are very doubtfully able to “conceptualize the depth of the human self” or “God’s being and his way of relating to creation” (240). More seriously, Highfield is concerned that libertarian theory ignores the effects of the fall on our wills and reason, which makes us much less free to respond to God than Boyd’s construction assumes. (241). Highfield agrees with Boyd that we were “created for freedom and love,” but he insists that it is “only in Christ and by the grace of the Holy Spirit” that “we are gifted with the power freely to love him with our whole being” (242).

Personal reflections on these responses

With regard to Helseth’s response

Helseth’s decision to focus on Craig’s use of Peter’s denial as an example of God’s intervention took me a bit by surprise, but I have learned the background to this focus, from Helseth’s comments at a panel devoted to this book, at ETS in 2011. He has made this point in a number of previously published venues and had hoped for a response from Boyd, but he has not received one, so one more enunciation of the concern seemed necessary. I find Helseth’s complaint of incoherence within Boyd’s description of human moral responsibility and justified divine intervention to be quite powerful, so I hope that it will elicit engagement eventually. (I notice that Craig also considered Boyd’s attempt to “explain away” the Peter incident “as an instance of character determinism,” to be “desperate” [228].)

With regard to Craig’s response

Where Craig took me by surprise was in his contention that libertarian freedom is not the power of alternative possibilities but the “absence of causal constraints outside oneself” (225). Since that is precisely how many compatibilists define morally responsible freedom, this creates very significant confusion in the debate between incompatibilists and compatibilists. If Craig is representative of the classic Molinist understanding of libertarian freedom on this point, I have completely misconstrued Molinism, but so has everyone else I have read on the subject. What has brought this about? Could it be that Craig is feeling the pressure of the grounding objection? But, if so, has his position migrated out of Molinism into the sort of account of compatibilism that Bruce Ware and I have been proffering for some years? I am thoroughly intrigued.

Craig cites as an advantage of this definition of libertarian freedom that it “enables us to ascribe libertarian freedom to God himself and to Christ in resisting temptation” (225), but this is surely problematic. If God had only the compatibilistic freedom of spontaneity, then his decision to create was determined by his nature and God could not have done otherwise. Craig is right to identify a problem in Boyd’s “conclusion that the perfection of libertarian free will in heaven is compatibilistic free will,” but Craig’s own move in defining libertarian freedom as compatibilistic freedom is also highly problematic. They are both correct, I believe, to understand our freedom in the glorified state as compatibilistic, and Craig is correct to point out that Boyd faces a problem when he construes compatibilist freedom as better than libertarian, but Craig’s proposal that libertarian freedom is, by definition, the freedom of non-coercion, is remarkably confusing.

I concur with Craig that an affirmation of God’s foreknowledge of counterfactuals “is the most perspicuous explanation of the coherence” between God’s “freely and unchangeably ordain[ing] whatsoever comes to pass,” on one hand, and God’s sinlessness in the occurrence of evil and the genuine liberty and contingency of secondary causes, on the other hand. I made precisely this argument in favour of what I originally called “middle knowledge Calvinism.” What Paul Helm eventually convinced me of, however, is that God’s knowledge of counterfactuals is “middle” only if creatures have libertarian freedom, defined as the power of alternative possibilities. So, nothing has changed in the fundamental shape of my model of providence, but I concur with those within the classic tradition who posit that God knows counterfactuals naturally. This is possible, however, only if creatures have the freedom of spontaneity, as Craig himself has here asserted. But, again I ask, has Craig not, thereby, abandoned libertarian freedom, and hence left Molinism behind?

To compound my confusion, Craig later asserts that he affirms both a tensed theory of time and causal indeterminism (227). I don’t understand how his belief in causal indeterminism goes together with his rejection of the principle of alternative possibilities as the essence of libertarian freedom. Given Craig’s brilliance, I am obviously missing something critical in his surprising definition of libertarian freedom, and a reader of this post may be able to help me out.

It is surprising that Craig finds fault with Boyd’s denial that God knows counterfactuals of creaturely freedom when God knows counterfactuals of his own free action. The critical difference between these two situations, in Boyd’s construct, is that, in the case of God’s knowledge of his own actions in all possible situations, what God knows is what he has decided he would do. He knows his own will, so that whatever creatures do, God’s own action is predetermined by his own free decision. The libertarianly free creature’s action, by contrast, is unknown to the creature until he decides upon it in the moment when the situation requires a decision. Unlike God’s predetermined decision, the creature is himself ignorant of what he would do, if he has libertarian freedom, in the usual sense of the principle of alternate possibilities.

Craig is correct in his assessment that Molinism handles the problem of evil better than open theism. Indeed, Molinism would be an excellent theological formulation of biblical teaching were it not for the grounding objection. Thankfully, the advantages Craig identifies for Molinism are retained in hypothetical knowledge Calvinism. I too believe that “God permits horrible evils only in view of morally sufficient reasons” (230), and I assert that God’s natural knowledge of counterfactuals enabled God to choose a good world, one in which many of his holy purposes are achieved in and through the voluntary actions of moral creatures, in such a manner that creatures are morally accountable for all the evil of their sin, and God is glorified in all the good that comes about by his own gracious work within the history of his creation. His victory over evil is certain because it has been planned from before creation was begun, and good is brought out of the evil that God deemed necessary in the big picture.

With regard to Highfield’s response

Although my model of providence reaches very similar conclusions to Highfield’s,  it is less incommensurable with Boyd’s because I approve of some of the revisions of classical theism that Boyd has made. Nonetheless, I found myself frequently concurring with Highfield’s criticisms of open theism’s portrayal of God, in regard to its adequacy as a portrait of the God revealed in Scripture. I am particularly sympathetic with Highfield’s doubt that God’s knowledge of “might” counterfactuals gives God significant ability to ensure that he can accomplish anything but an extraordinarily general plan for creation.

As I have indicated in previous posts, I acknowledge that Highfield’s caution about theological speculation may be beneficial. But I found that, in his response to Boyd, I sometimes wondered how we can know God, if he is as much beyond creaturely logic as Highfield proposes. I know that rationalism is dangerous, and I welcome warnings that the Creator-creature distinction not be muddied, but in my own theological work I find it helpful to seek formulations that illuminate God and his work through the use of widely plausible (though hardly “common sense”) logic. Consequently, I find myself less critical of Boyd’s methodology than Highfield is, even though I question that Boyd’s method leads us where he believes it does.

General comments on the book as a whole

All in all, I think that this four-views book is helpful. The authors all articulate their views clearly, and they are incisive in their critiques of one another’s perspective. It might plausibly be suggested, however,  that Helseth and Highfield’s model is so close that there are really only three models in this book, though there is a slight difference in method between the two monergist proposals. A Thomist model, with its hard compatibilism conjoining libertarian freedom with divine determinism, would have provided an interesting voice in the conversation, but since this book is clearly a conversation between evangelicals, I understand why this direction was not taken. In my own response to the book, I have had more interchange with Craig’s view than any of the others, and this makes me think that a hypothetical knowledge Calvinist contribution would have enriched the conversation. But I can see how it might have been deemed not significantly more different from Helseth’s classic compatibilist proposal than Highfield’s model is.

I think that an emphasis on the usefulness of God’s hypothetical knowledge adds to the explanatory power of the compatibilism that Helseth enunciates, but I think that it would also set up an interesting interchange with Craig’s Molinism and Boyd’s “neo-Molinist” open theism, so that the contrast between monergism and synergism might be less stark. That being said, I am grateful to Dennis Jowers for coordinating this project, and to the four contributors for their participation. Jowers adds a helpful conclusion in which he identifies agreements and disagreements between the contributed models, and assesses their strengths and weaknesses. When he was done, however, I concluded that he finds all of the models defective in significant ways, and I was eager to see his own constructive alternative. That is not a bad way to leave things, since it illustrates the complexity of this doctrine and the difficulty of restating just what Scripture tells us about God’s work in the world.

Previous posts in this series: Part 1, Part 2, Part 3, Part 4, Part 5. Part 6. Part 7, Part 8


By Terrance Tiessen

I am Professor Emeritus of Systematic Theology and Ethics at Providence Theological Seminary, Canada.

2 replies on “Responses to Boyd’s open theist model of providence”

I think I can help you understand Craig’s position a little; although I’m by no means an expert, so all statements that follow should be read implicitly starting with “I think…” 😛

To start, Craig takes “compatibilism” as the view that [external] casual determinism is compatible with human free choices. When he talks about “casual indeterminism”, then, he means the view that denies that this causal determinism is compatible with human free choices. Thus he takes causal indeterminism as the central tenet of libertarianism/incompatibilism.

He denies that the principle of alternate possibilities is a necessary or sufficient condition for free will in light of the Frankfurt counterexamples (see here: That said, he does still hold that often (if not always) we *do* have the power to choose otherwise, even though this power does nothing for us in terms of our free will. Because we still can choose otherwise, the Molinist distinction between possible and feasible worlds (and thus the distinction between necessary knowledge and middle knowledge too) is still meaningful.

Thus he takes the causal indeterminism as both a necessary and sufficient condition for free will, and since it involves a denial of compatbilism, it is safe to still call it libertarian free will (I guess you could think of it as libertarian free will that has been corrected in light of the Frankfurt counterexamples).

Now, because I’ve been reading your blog for a bit (great blog by the way, I always look forward to your posts) I suspect that you might have a slightly different understanding of the term “compatibilism”. Please correct me if I’m wrong, but as I understand it, you use the term “soft compatibilism” to refer to the kind of free will that you employ in your model. I’ve been curious for some time now (and perhaps this is an oppostunity for you to answer my curiosity) why you categorise your view as a kind of compatibilism in the first place. I know that you undestand God’s hypothetical knowledge as being part of his necessary knowledge, and I suspect that you agree with Craig that causal indeterminism (as I’ve used the term here) is a necessary and sufficient condition for free will. I further suspect, then, that what you see as being affirmed as compatible in your view is that an action being free is *compatible* with the agent not being able to choose otherwise. If I’m right, then I think what you call “soft compatibilism” is exactly what Craig calls libertarianism. What differentiates your views then is that you deny that anyone has the power of contrary choice, while Craig affirms that we infact do have such power. The confusion, I think, comes from you two focussing on different theses (he focuses on causal determinism, while you focus on alternate possibilities) when it comes to the labels compatiblism or incompatiblism.

Out of further curiosity, do you take “hard compatibilism” to be the view Craig ascribes simply to “compatibilism”?

As for your question, “Could it be that Craig is feeling the pressure of the grounding objection?” I don’t think he sees the grounding objection as much of a threat given his response to it here: and his other responses to anti-Molinist arguments from philosphers, found here:

Thanks for your thoughts Roland.

It is helpful that you have drawn attention to Craig’s focus on causal indeterminism as the critical feature that makes human freedom libertarian, rather than the power of contrary choice (i.e. principle of alternative possibilities). Since Craig thinks that the freedom of spontaneity is sufficient, his redefining of libertarian freedom thus makes sense, in the context he had provided.

Your point about Craig continuing to affirm that creatures often do have the power of contrary choice is also very important. This explains why God can not be meticulously in control (i.e., causally determining, but I tend to dislike causative language because it easily creates misunderstanding. I prefer to use “causing” as a term to describe God’s direction action to bring about something, as opposed to the mediated action of his deliberate permission, in which cases I see the creature as the one “causing.” This is always the case when evil is done. In those cases, I do not speak of God as “causing the evil,” though I know that the Scripture does not hesitate to do so. It seems to me to contribute to the impression synergists have that [within a monergist framework] God should be held accountable for the evil of sinful creaturely acts).

The crunch, for Craig, I think, is that granting creatures frequent classically libertarian freedom also makes the grounding objection a continuing problem for him. I expect that you are right, however, that Craig does not in fact feel pressure in regard to the grounding objection. Why he doesn’t mystifies me, but I have heard him address it and we were clearly ships passing in the night. Craig remains convinced that God knows that he would buy a Mercedes under certain conditions. Somehow, the “fact” that he would do so explains God’s knowledge of counterfactuals, but I can not grasp how God knows this, so far as creatures have the power of contrary choice.

Your thoughts have been some help to me, in understanding where Craig was coming from in writing his response to Boyd. I continue to think that he has introduced considerable confusion, however, by defining freedom (at least as to its sufficiency, if not its usual actuality) in exactly the way that compatibilists frequently define it. In his response to Boyd, I see a peculiar affinity to hypothetical knowledge compatibilism at some points, while retreating to classic incompatibilism at others.

As to my own definitions, Roland, by “hard compatibilism,” I mean the Thomist affirmation that meticulous divine control (and Thomas would be comfortable with “causal determinism”) and libertarian freedom are compatible. By “soft compatibilism,” I have in mind the most common Calvinist affirmation that meticulous divine control and the freedom of spontaneity are compatible. But some Calvinists (e.g., J. I. Packer perhaps) plead mystery, and appear to affirm both meticulous providence and libertarian freedom, without appealing to Thomistic explanations.

It was interesting to hear that I have befuddled you a bit, and this may follow from a softening or tentativeness in my own position. I have been a pretty classic soft compatibilist. I remain a compatibilist, but right now I’m somewhat agnostic about what sort of freedom is necessary to ground moral responsibility. My situation is complicated. I am a compatibilist because I believe that humans are morally responsible, and I believe that God is meticulously in control. I feel the force of libertarian concerns about the adequacy of the freedom soft compatibilists have generally affirmed, a freedom of spontaneity (hence my surprise at Craig’s move in redefining libertarian freedom!). But, it is clear to me that God knows counterfactuals, and I find the grounding objection to middle knowledge persuasive, so I doubt that creatures have the power of contrary choice.

Nevertheless, perhaps I am headed towards the more agnostic position [what he calls antinomy, as opposed to paradox] of J. I. Packer. I’m still working on it. Currently, however, because of the grounding objection to knowledge of the counterfactuals of libertarian freedom (classically defined!!), I remain a traditional soft compatibilist. What is non-negotiable for me, given my reading of Scripture right now, is that God is meticulously in control (which Craig would consider “causal determinism”), and that humans and angels are morally responsible.

Leave a Reply

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

HTML tags are not allowed.

145,569 Spambots Blocked by Simple Comments