Paul Helm Misses The Mark on Molinism: A Response to’s “Molinism 101”

By Jonathan Thompson


July 3, 2017

Summary: Paul Helm has contributed an article to in which he presents two noteworthy objections to Molinism. First, Helm presents an internal critique of Molinism purporting to show that Libertarian Free Will (LFW) undermines God’s providential control over the world. Second, Helm argues that Molinism is incompatible with the doctrine of irresistible grace. I demonstrate that Helm begs the question with his first objection and relies on a model of LFW which Molinists aren’t necessarily committed to. I then respond to Helm’s second objection by presenting two ways Molinism could be construed so as to be compatible with irresistible grace. Along the way I also discuss some minor errors diffused through out Helm’s essay. Such errors would include comprehensional errors, dictional ambiguity, inter alia. I then conclude that LFW presents no threat to God’s providential control over the world and neither is Molinism at odds with irresistible grace.


Issue 1: Helm Underdescribes LFW
The first issue involves Helm’s explanation of LFW. He writes,

We need to emphasize that the view of free will held by Molinists both ancient and modern is what is often called ‘libertarianism’ or ‘indeterminism.’

This is fine. But Helm also writes this:

When we speak of indeterministic freedom, we mean that any human being, in a given set of circumstances, has the power to choose A or to choose not-A.

Helm’s definition misrepresents certain models of LFW and so the definition fails to be comprehensive. What needs to be appreciated here is that while it’s true that some Libertarians believe that libertarian decisions necessitate the ability to choose between alternatives, nevertheless, not every Libertarian is committed to this model of LFW. For example, Eleanore Stump[1], David Hunt[2], and Linda Zagzebski[3], who are all non-Molinist Libertarians, have argued against the relevant principle in so far as they do not conceive of it as being a necessary condition for LFW. Further, there are even prominent Molinists who repudiate this principle as being a necessary condition for LFW as well. William Lane Craig[4] and Kenneth Keathley[5] are two such examples. Thus, if the Libertarian does not accept the model of LFW that Helm purports to describe, then Helm will be misrepresenting her preferred view on the matter.

This raises the question as to what a model of LFW that does not commit itself to the necessity of alternative possibilities might look like. One such model is the one which asserts that a choice is libertarianly free just in case the agent performing the choice is not caused to do it by causes other than his/herself as an agent. This type of causation is commonly referred to by philosophers as agent causation. In short, Helm’s definition of LFW only describes certain models of it and so is inchoate.

Issue 2: Helm Is Ambiguous Concerning The Sense That Molinists Believe Certain Verses Support God’s Having Middle Knowledge
Next, Helm sketches a way in which Molinists might argue in favor of God’s having middle knowledge on the basis of 1 Samuel 23:11-13. After offering a brief explanation of the passage Helm writes this:

To the minds of the Molinists, this incident showed middle knowledge at work, for it showed that the Lord knew what would happen if a certain free action occurred (they assumed that David and the other participants were acting with free will in the libertarian sense). God knew that if David freely stayed at Keilah, then the Keilahites would freely surrender him. So David freely took evasive action, and Saul freely gave up the expedition against David when he learned of what David had done. God knew all of this (and much more besides) by His foreknowledge.

While I won’t dispute the fact that these verses have been used by Molinists to support the idea that God possesses middle knowledge, Helm, nevertheless, fails to disclose the sense in which the verses can be used by Molinists to support the doctrine and so his statement is ambiguous. The sense in which Molinists believe that these verses and verses relevantly like this support the doctrine of middle knowledge (here after DMK) is important because there is no universal agreement amongst Molinists as to whether the DMK is actually taught in Scripture. Moreover, if it turns out that the DMK is not taught in Scripture, this, as we shall see, does not necessarily imply that it is false or is unsupported by Scripture in some other sense.

At this point I will describe the two standard methods used by theologians to infer the truth of a doctrine. The second method I will describe will be especially relevant to this discussion since the method provides us a way to forestall those objections that conclude that the DMK is false simply because it (allegedly) fails to be taught in Scripture.

That noted, the first method for arriving at the truth of a doctrine simply involves validly inferring the doctrine by employing the appropriate hermeneutical and exegetical principles on to the text from which the doctrine is derived. The utilization of this method presupposes that Scripture does, in fact, teach—either implicitly or explicitly—the relevant doctrine.

In contrast, the second method used by theologians involves arriving at a theological truth via philosophical argumentation. This method is distinct from the first in that its utilizer assumes that a particular teaching or philosophical account of a teaching is underdetermined by Scripture. For example, the Bible teaches that God is eternal, but it is plausibly underdeterminative with respect to whether God is timeless or everlasting throughout time.[6] Scripture, taken by itself, is therefore consistent with both views on the matter and so we must rely upon further philosophical argumentation to adjudicate between these competing theories.

So, with respect to the DMK, if it turns out that the doctrine is not, in fact, taught in Scripture, notice that it will be of no avail for the non-Molinist to conclude that the DMK is false because of this. This is because the doctrine may still be consistent with what Scripture teaches and also be the best model purporting to explain the underdeterminative data relevant to the doctrine. This means that if the non-Molinist wants to demonstrate the DMK is false they must either raise a philosophical objection against the view or else show that the doctrine is inconsistent with some aspect of Scripture.

That said, the point of highlighting these distinctions is to give non-Molinists pause. More specifically, before the non-Molinist raises the objection that 1 Samuel 23:11-13 doesn’t demonstrate that God possesses middle knowledge, they should first consider the sense in which the Molinist believes this verse (and verses relevantly like this) to support their view in order to avoid discussions in which disagreeing parties talk past one another.

Issue 3: Helm Presents a Trivial Reason Why The Reformers Rejected Middle Knowledge
Following this, Helm explains why certain reformed Christians have and should reject positing middle knowledge. Helm writes this:

Since the Reformed held that all that occurs is unconditionally decreed by God and that men and women are responsible for their actions, they saw no need for a third kind of divine knowledge, a middle knowledge, which depended upon God foreseeing what possible people would freely do in certain circumstances. The Reformed interpreted the Keilah incident differently. God did not simply see what Saul would do; He ordained that Saul would come down if David remained. He ordained that David would depart from Keilah upon hearing what Saul would do. And He ordained that Saul would change his mind.

Helm does not define what is meant by an “unconditional decree” so I can only speculate with respect to what he means by it. Perhaps then, being a Calvinist, Helm means to use the word ‘unconditional’ in the same sense that Calvinists use the word ‘unconditional’ when speaking of unconditional election. For those unfamiliar with this doctrine, unconditional election is the Calvinist doctrine that asserts that God elects who will be saved according to His own purposes, without regard to His foresight of the faith or works (in the meritorious sense) in man. Ex hypothesi, an unconditional decree would then be God’s decision to create a world according to His own purposes, without regard to His foresight of the faith or works in man. Hereafter, we shall understand the phrase ‘unconditional decree’ in this way.

Now return to Helm’s statement. Helm states that middle knowledge is unnecessary because the Reformers “held that all that occurs is unconditionally decreed by God. . .”. Let us unpack Helm’s use of the term ‘occurs’. If, by ‘occurs’ Helm just means to say that God unconditionally decrees all that has/will happen in the actual world, then Helm’s statement, taken by itself, presents the Reformed person no good reason to prefer the determinist-Calvinist model (hereafter D-C model) of the decree over the Molinist one. Why? Because Molinism is compatible with there being an unconditional decree. This is because Molinism does not assert nor does it imply (so far as I can tell) any position regarding whether or not the divine creative decree is conditional. So, if Helm is purporting to present a reason that Reformed persons should prefer the D-C view over the Molinist one, Helm must provide some argument for why the D-C model is preferable.

Suppose however, that by “occurs” Helm means to denote God’s decreeing the truth value of counterfactuals in addition to his decreeing of future contingents. This interpretation would further explain why Helm found it appropriate to note that, on his view, God decreed the counterfactual statements relevant to the Keilah incident. This interpretation, while presenting a true reason that the Reformers rejected middle knowledge, nevertheless turns out to be trivial. Why? In order to understand why this is the case, we must first understand a crucial difference between the Molinist and D-C model of the divine creative decree. On the Molinist view, God’s knowledge of counterfactuals is logically prior to the divine creative decree and so the truth value of these counterfactuals aren’t up to God to decide, but rather the creature. In contrast, the D-C model affirms that God’s knowledge of counterfactuals is logically posterior to the divine creative decree. Further, their truth value is decreed by God.

That said, notice that the statement that the Reformers rejected the DMK because they affirmed that God unconditionally decrees counterfactuals implies the statement that “The Reformers rejected the Molinist model of omniscience because they affirmed the D-C model of omniscience.”  This, I confess, is platitudinous. Of course if a subject, S, already accepts some proposition, P, then S by virtue of believing P will also reject not-P (assuming that S is rational and understands what not-P is)! Instead, a more insightful approach to the topic concerning the Reformers relationship with Molinism would be to present reasons that the Reformers preferred the D-C model of the decree over the Molinists one (besides the fact that they weren’t Molinists). Unfortunately, Helm neglects this task.

Issue 4: Helm’s Use of the Word “Foresee” is Ambiguous
Of further note, Helm’s use of the word ‘foresee’, as it is used in the initial quote under Issue 3, requires disambiguation. The word ‘foresee’ just means to know something beforehand. Notice, however, that this definition implies nothing about how this knowledge is obtained. Moreover, Helm, so far as I can tell, doesn’t provide any contextual hints indicating whether or not he thinks that Molinists believe God’s knowledge to be acquired through a conceptualist or perceptualist model of divine cognition. Helm’s use of the word ‘foresee’ is therefore ambiguous, thereby leaving Molinism prone to be misunderstood.

In an effort to forestall this ambiguity, it will be helpful to explain the difference between perceptualist and conceptualist models of divine cognition. William Lane Craig explains the differences:

The perceptualist model thinks of God’s cognition on the analogy of sense perception. . . [God] somehow looks ahead in time and “sees” what is there.[7]

This is in contrast to a conceptualist model of divine cognition which “thinks of God’s knowledge more on the analogy of innate ideas”.[8] On this model “[God] didn’t get this knowledge from anywhere; He just has it innately.”[9]

That in mind, what needs to be appreciated here is that Molinism adheres to a conceptualist model of divine cognition. For this reason, when Kenneth Keathley, in his book “Salvation and Sovereignty” explains Molinism, he notes, “Since God is omniscient, He innately knows all things. . . God never ‘learns’ or has things ‘occur’ to Him. He already knows all truths.”[10] So, on Molinism, God’s foreseeing of what would happen under a particular set of circumstances should be understood as nothing more than His having an innate intuition.

Issue 5: Helm Begs The Question By Asserting That LFW Includes The Ability To Thwart God’s Decree Without Providing Justification
Next, Helm presents an argument against Molinism:

Not only is middle knowledge unnecessary to an all-knowing, all-decreeing God, but the Molinists’ conception of free will makes it impossible for God to exercise providential control over his creation. Why? Because men and women would be free to resist His decree. God can only bring to pass the actions of free agents via his middle knowledge of what they would freely do if…

As I understand the objection, Helm is arguing that Molinism is false because if it were true, then it wouldn’t be the case that whatever future contingent God has decreed will be actual. This would further seem to be the result of creatures possessing LFW (as Helm has illicitly defined it).

Perhaps the most evident problem with Helm’s objection is that he asserts that libertarian freedom includes —presumably, by implication or entailment—the ability to resist the divine creative decree, but Helm provides no justification for this assertion. Helm thus begs the question in favor of this claim. Furthermore, Helm’s argument purports to demonstrate that God’s failure to possess providential control over the world is the result of human beings possessing the ability to choose between alternatives. As we have already seen, Helm’s model of LFW is underdescriptive. Thus, if the Molinist does not accept a model of LFW that requires an agent to be able to choose between alternatives, then Helm’s critique of Molinism will simply not apply to her.[11]

Issue 6:
Helm’s Argument That Molinism is Incompatible With Irresistible Grace Relies on an Implicit Mischaracterization of LFW
Finally, Helm proceeds with his last critique of Molinism. He writes,

Further, given the Molinist view of freedom, it is impossible for God to bring about the conversion of any person by the exercise of His effective call, for in the view of the Molinists it is always possible for an individual to resist God’s grace.

For those unfamiliar with the doctrine of irresistible grace, it asserts that there is an invitation to salvation, prepared by God, uniquely for the elect which is ineluctable and so is effective in that it is met with affirmative response from the person who receives it. How might this be construed so as to be compatible with Molinism? One way this might be possible requires that we understand LFW and its implications. LFW asserts that (1) we are free and (2) that the freedom necessary for moral responsibility is incompatible with determinism. Notice, however, that this definition does not imply that we are morally responsible for every action that we commit. Thus, on Libertarianism, it is possible that there are at least some actions we perform which are causally determined. Therefore, the Molinist could maintain that our response to God’s effective call to salvation is causally determined. Moreover, this is consistent with Molinism so long as there exists the potential within the agent to perform other libertarianly free actions. Thus, this is one way Molinism could be compatible with irresistible grace.

Another way Molinism could be compatible with irresistible grace is if we were to attenuate the definition of irresistible grace. One way this could be done is by undercutting the justification the D-C adherent has for her belief that it is metaphysically impossible to resist God’s grace. For example, Romans 8:30, a common prooftext for irresistible grace, states:

And those whom he predestined he also called; and those whom he called he also justified; and those whom he justified he also glorified. (ESV)

Note that this verse asserts those who are called are justified (thereby implying that the call is met with an affirmitive response), however, this verse does not assert that those who are called must have responded affirmatively to the call. Therefore, it would be question-begging for the D-C adherent to assume that God’s effective call was metaphysically impossible to resist.

Perhaps, then, the D-C adherent believes that grace being metaphysically impossible to resist is implied by the text through the author’s use of the word ‘predestine’. On the D-C view God predestines everything that will happen through causal determinism. Further, an event is causally determined just in case there are conditions such that, given them, nothing else could have possibly occurred. Thus, given causal determinism and given that God has predestined an individual to salvation, it follows that it would be metaphysically impossible for that individual to resist being saved.

This D-C response, however, is question begging as well. In order to see why this is the case important to understand what the word ‘predestine’ means. The Greek word used for ‘predestined’ in Romans 8:30 is proōrisen (προώρισεν) which denotes that something has been predetermined or foreordained.[12] Now, because the word ‘determine’ is a constituent of the word ‘predetermine’ some D-C adherents might insist that predestination analytically entails causal determinism. This, however, fails to consider that the word ‘determine’ is a homonym, having multiple meanings, some of which are consistent with Molinism. For example, while it’s true that the word ‘determine’ can denote causal determinism it can also just mean ‘to cause something in a particular way’ or ‘to decide something’. Furthermore notice that Molinism isn’t necessarily at odds with saying that God causes our free choices in a particular way. This is because Molinism maintains that the way in which God causes our choices is by His weakly actualizing them.

All things considered, since the word ‘predetermine’ is a homonym having meanings that are consistent with Molinism, the D-C adherent would therefore be begging the question by assuming without argument that these alternative interpretations to the word are false. Of course, the failure of the D-C adherent to prove their own view obviously doesn’t imply that the Molinist understanding of the word ‘predestine’ is correct, only that the relevant use of the word is underdeterminative and so is compatible with Libertarian interpretations. In short, with respect to Romans 8:30 the Molinist could maintain that this verse (and verses relevantly like this) doesn’t imply that grace is impossible to resist, but only that it won’t be resisted by the individual who receives it.

The last way one could respond to Helm’s objection is by arguing against the doctrine of irresistible grace. Helm’s essay assumes the truth of irresistible grace without argument. This is because Helm purports to present D-C adherents (rather than Christians, in general) reasons to reject Molinism. If, however, one isn’t a D-C adherent, then Helm’s objection turns out to present no threat to Molinism.

The fact that Helm’s essay is titled “Molinism 101” suggests that he intends to present his readers with an introduction to Molinism. Despite this fact, Helm misrepresents Molinism by incorporating in to it aspects of LFW which Molinists aren’t necessarily committed to. Moreover, Helm betrays the perceived intention of his article by failing to elucidate certain concepts essential to Molinism which are commonly misunderstood in theological circles (i.e. the sense in which Molinists believe certain prooftexts support their viewpoint, the Molinist view of divine cognition).

Apart from Helm’s explanation of Molinism, his article is unnecessarily long in that it includes a platitudinous discussion concerning why the Reformers rejected Molinism. Also, Helm’s two major objections to Molinism were wholly unimpressive. His first objection involved him merely arrogating that LFW includes the ability to thwart God’s decree. His second objection was no better in that it illicitly supposed that LFW maintains that every action an agent performs is either morally praise/blameworthy.

In short, Helm’s approach to the topic of Molinism needs improvement. It’s unfortunate that by espousing himself to these sorts of lackluster objections that Helm has, in effect, sequestered himself from what William Lane Craig has characterized as “one of the most fruitful theological concepts ever conceived”.[13,14]


[1] In Moral Responsibility and Alternative Possibilities: Essays on the Importance of Alternative Possibilites edited by David Widerker and Michael McKenna, “Moral Responsibility Without Alternatives” p.141 Eleanor Stump presents a Frankfurt scenario in which a Frankfurt victim named Jones possesses no alternative to voting for Republicans. Stump concludes from this scenario that “. . . it certainly seems as if Jones is morally responsible for his act of will to vote for Republicans, although it also seems true that it was not possible for Jones to do anything else other than willing to vote for Republicans”.

[2]  In Philosophical Studies 97 (2):195-227 (2000), “Moral Responsibility and Unavoidable Action” David Hunt argues against Frankfurt-style counterexamples to the principle of alternative possibilites in that they fail to preclude all alternatives because they rely on a counterfactual intervener to preclude alternatives to the subject’s action.

[3] In Philosophical Perspectives, 14, Action and Freedom, (2000)  p. 245 “Does Libertarian Freedom Require Alternate Possibilities?” Linda Zagzebski states that “Talk of counterfactual conditions for the attribution of responsibility. . . can help illuminate the idea of causing or making something happen, but eventually we have to get back to the idea those conditions aimed to elucidate, and that is the idea of causing an act, making it happen. Once we have identified who or what makes an act happen we have identified the potential bearer of responsibility for it. In Frankfurt cases in a non-deterministic world the agent makes her act happen in as ultimate a sense as you like; in a deterministic world it is not the agent that ultimately makes her act happen. Frankfurt cases do nothing to lead us to rescind the view that this is a significant difference. What they do show is that [the Principle of Alternative Possibilities] is a false path. The presence of alternate possibilities may be a reliable sign of the presence of the agency needed for responsibility, but it is not necessary for it.”

[4] In his article Free Will located at William Lane Craig states that he rejects the necessity of alternative possibilities. Specifically, he states “I’m persuaded by illustrations like that given by Harry Frankfurt to show that freedom does not require the ability to choose other than as one does. . . what is critical to free will is not the ability to choose differently in identical circumstances but rather not being caused to do something by causes other than oneself.”

[5] In Kenneth Keathley’s Salvation and Sovereignty (p.100). B&H Publishing. Kindle Edition. Keathley asserts in a footnote that “. . . I do not hold that an agent must have the ability to do otherwise ‘in any given situation,’ but, . . . ‘at significant will-setting moments.'”

[6] In “Time and Eternity” William Lane Craig has a chapter titled Two Views on Divine Eternity in which he dedicates a section to surveying the biblical data concerning God’s relationship to time. Craig concludes from his survey that the relevant biblical data seems to be underdeterminative.


[8] Ibid.

[9] Ibid.

[10] Keathley, Kenneth. Salvation and Sovereignty (p. 16). B&H Publishing. Kindle Edition.

[11] Special thanks to Randy Everist for helping me better understand and articulate my own objections to Helm’s view. Visit Randy’s website at



[14] Special thanks to Jesse Roach for guiding me in my essay, helping formulate my abstract as well as my conclusion.


About the Author

By Jonathan Thompson