Responding to James White’s Anti-Molinist Critiques: The Truthmaker-Style Grounding Objection

By Jonathan Thompson


August 1, 2018

Summary: In The Dividing Line episode Behold the Secular Woman & WLC on Molinism (Once Again) James White argues that Molinism lacks the explanatory resources to sufficiently answer what determines the truth value of counterfactuals of creaturely freedom (hereafter, CCFs). In this essay I highlight an ambiguity in White’s use of the word ‘determine’. I then address his objection, refuting it under a variety of interpretations. I also address potential counterresponses White might provide. As we shall see, one noteworthy interpretation of White’s argument turns out to be the perennial truthmaker-style grounding objection to Molinism.

Video Source:

Disambiguating White’s Objection
To begin, White plays a segment from the Reasonable Faith podcast where William Lane Craig states that, on Molinism, God does not determine the truth value of CCFs.1 White responds with the following:

So God does not determine [the truth value of counterfactuals of creaturely freedom]. Who [or what] did? I guess, for the Molinist, you can’t ask that question. 2

First, a preliminary note: When White states “. . . for the Molinist, you can’t ask that question” I take White to be speaking idiomatically so as to convey his belief that Molinism is explanatorily deficient in so far as it is unable to provide an explanation with respect to who or what determines the truth-value of CCFs.

That said, what does White mean by his use of the word ‘determine’? He doesn’t say. However, what needs to be appreciated here is that although the word ‘determine’ can be used in a causal sense,—such as it is used in denoting causal determinism—it can nevertheless be properly employed in non-causal senses as well. To give an example, it is true that the combined quantity of the individual members of a given set of numbers will determine its sum. Here, the sense in which the combined quantity of numbers determines the sum of the set is non-causal because numbers, sets, and mathematical operations are abstract objects and are therefore causally impotent. Another way of expressing this relation between these numbers is to say that the set of numbers non-causally makes its sum such in virtue of its members. That is to say it is by the numbers merely existing within the set by which the set obtains its sum. A more colloquial way of expressing this relation would be to say that the joint quantity of numbers of which the set is composed serve to explain the sum of the set. That distinction in mind, White fails to clarify for his listeners if he’s using the word ‘determine’ in a causal or non-causal sense.3 White’s use of the word ‘determine’ is therefore ambiguous. In light of this ambiguity, in the forthcoming sections I shall consider both causal and non-causal interpretations of White’s objection and demonstrate that regardless of which interpretation White prefers, his worry can be sufficiently answered.

On The Causal Sense of ‘Determine’
Moving forward, if by his use of the word ‘determine’ White means to argue that Molinism is explanatorily deficient because it fails to provide an explanation that sufficiently causes a CCF to obtain its truth value, then White will be begging the question against those Molinists who are agent causal libertarians. This is because White will be assuming without argument that there must be a sufficient reason for why a CCF has a particular truth-value. What needs to be understood here is that on agent causal libertarianism, nothing can sufficiently explain why an agent performs some action over another.4 To illustrate the point, we can conceive of two otherwise identical logically possible worlds,  W1 and W2. In W1 Nappa is given the free choice between A and B, and chooses A rather than B. Moreover, Nappa has reason R1 explaining why he chose A. Now let’s suppose that in W2 Nappa is presented with the same choice, but instead chooses B rather than A. Further, once again Nappa, has reason R1 explaining why he made this choice. In this case, R1 would explain why Nappa chose A in W1. But notice that R1 would also explain why he chose B in W2. However, even if we concede that R1 partially explains why Nappa made these choices in each respective world, notice that R1 does not sufficiently explain why he chose A rather than B in W(or why he chose B rather than A in W2). This is because given agent causation, there just isn’t any sufficient explanation for why Nappa chose one way over another. At most, the agent causal libertarian can say that Nappa, as an agent, caused the choice even though they can’t sufficiently explain why he made it. Thus, it will be impossible for the agent causal libertarian to provide a sufficient reason for why Nappa chose A rather than B. In sum, given agent causal libertarianism, the fact that a CCF is true rather than false (or vice versa) is a brute (i.e. inexplicable) fact, something which the relevant objection we’re considering excludes as an option from the outset.

But let us now further suppose that White has non question begging reasons for thinking that brute facts aren’t ontologically acceptable. For the sake of argument, let us stipulate that his reason for disbelieving in brute facts stems from his commitment to a certain version of the principle of sufficient reason (hereafter, I will simply refer to this version of the principle of sufficient reason as “the PSR”) according to which every fact (i.e. true proposition) has an explanation for its being true. After all, Determinist-Calvinism neatly conforms to this principle in a way that Molinism does not in so far as the former view posits God’s decretive will as the sufficient explanation for the truth of some contrastive counterfactual proposition. But why should anyone accept the PSR? After all, White presents no argument on behalf of it and so begs the question in favor of the principle. Perhaps though we can extend charity to White and stipulate that his justification for his belief in the PSR is that he simply intuits its truth similarly to how we intuit other non-inferentially justified metaphysical truths about the world such as, No object is larger than itself, Everything that has a shape has a size, Nothing can be red all over and green all over. The question before us then is “Is the PSR plausible?”

It seems to me that the PSR is challengeable. For consider the Big Conjunctive Contingent Fact (hereafter, BCCF), which is the conjunction of all contingent facts there are. Does it have an explanation? It seems not. As Craig notes,

. . . if the explanation of the BCCF states a contingent fact, then it, too, must have a further explanation, which is impossible, since the BCCF includes all the contingent facts there are. But if the explanation states a necessary fact, then the fact explained by it must also be necessary, which is impossible, since the BCCF is contingent. Therefore, not every fact can have an explanation.5

Now if Craig is correct here, then we will have a rebutting defeater to the relevant PSR, thus rendering it harmless to the Molinist.

Alternatively, if the Molinist does not want to reject the PSR, then she could consider whether or not she genuinely shares White’s intuition about it. If she does not intuit the truth of the PSR, then it follows that while White himself has rational justification for adherence to the PSR, he has nevertheless failed to present the Molinist with any rational justification in favor of it. It would therefore be irrational for the Molinist to adhere to White’s belief in the relevant PSR since the Molinist would lack justification for adherence to it. Therefore, if White thinks that the Molinist should be committed to the PSR, then he owes them some type of argument on its behalf.

Now, has White provided any such argument? Not in the particular webcast I’m addressing, however, elsewhere White has voiced his opinion that the view that there exist brute CCFs is contra-biblical. Specifically, White has stated the following:

[On Molinism, CCFs] do not find their source and origin in God. . . That’s not the God of the Bible. . . My God is not confronted by anything that’s outside of his control that then can constrain him to do anything!6

Notice that White implies that there is nothing tout court that God is met with which can constrain His actions. But also notice that White makes no attempt to provide any exegesis in support of his claim. White instead resorts to merely voicing his own personal incredulity against the Molinist position.

Second, White’s claim that nothing can constrain God’s actions is demonstrably false. To demonstrate the point, consider the property of God’s moral goodness. Christians have traditionally conceived of God as being morally good necessarily. Furthermore, given that it is necessarily true that God is morally good, then it follows that the proposition expressing that truth, given White’s own Calvinism, will be known to God as a part of His natural knowledge and so will logically precede His decretive will. Furthermore, since the proposition concerning God’s goodness is necessarily true and precedes His decretive will, He is therefore not free to decree the truth value of the relevant proposition. Even further, since, God is morally good necessarily, then it follows that God cannot perform evil. Thus, it is in virtue of God’s own morally good nature by which His actions are constrained that He cannot perform evil. So, it turns out that we actually have a knockdown argument against White’s view that there is nothing tout court which can constrain God’s actions.

Suppose, however, that White rejects the orthodox view concerning the necessity of God’s goodness and instead affirms that there are no necessary truths. Instead, let us suppose that White believes that all truths—including one’s about God’s own nature and truths about the libertarianly choices of libertarianly free creaturesmust be unilaterally decreed by God.  This view is known as universal possibilism.7 On this view, God could have brought it about that even logical contradictions be true and tautologies be false. So, for example, on universal possibilism it would not be necessarily true that a triangle is three-sided or that God = God. Rather, on this view, God decrees the truth value of all truths simpliciter.

Is universal possibilism plausible? Not at all. For we may ask White if the proposition “There are no necessary truths” is necessarily true. If he responds ‘yes’, then universal possibilism is self-refuting since the proposition “It is necessarily true that there are no necessary truths” is itself a necessary truth. If White responds ‘no’, then that entails that it’s possible there are necessary truths, but if it’s possible that necessary truths exist, then that entails necessary truths exist in some possible world. Furthermore, if it’s possible that necessary truths exist in some possible world, then it also follows that necessary truths exist in every possible world—for that is entailed by what it means to be necessary. Even further, if necessary truths exist in every possible world, then it follows that necessary truths exist in the actual world—for the actual world just is a possible world. Therefore, necessary truths exist in the actual world.8 Universal possibilism thus implausible since it entails that the law of non-contradiction is not a necessary truth.

On The Non-Causal Sense of ‘Determine’
At this point we shall consider the non-causal interpretation of White’s use of the word ‘determine’. According to the non-causal interpretation of White’s argument, something is determined if and only if there is some existent in virtue of which some proposition is made true. In line with this definition, White’s criticism should then be understood as the claim that there are no true CCFs because there isn’t anything in virtue of which these CCFs are non-causally made true. In other words there are no truthmakers for CCFs. In academia this objection is known as the truthmaker-style grounding objection.9 Hereafter I shall refer to this objection as such.

With respect to the relevant objection, notice that it takes for granted that there are at least some propositions (i.e. CCFs) that have truthmakers (where a truthmaker is some existent in virtue of which some proposition is non-causally made true). This view is known as truthmaker theory. Now, certain versions of truthmaker theory are suspect, but nevermind.10 Moreover, White provides no argument on behalf of the notion that CCFs require truthmakers, but let’s ignore that too. Even if we were to grant truthmaker theory, there’s simply no good reason to think that Molinism is incapable of providing sufficient truthmakers for CCFs. Jennifer Jensen, for example, suggests that, for all we know, what grounds CCFs (if they require grounds) are either counterfactual facts or else non-categorical states of affairs.11 12  So, ex hypothesi, if CCFs require grounds, then what could ground a proposition like Nappa would freely do some action, A in fully specified set of circumstances, C, is the fact that Nappa would freely do A in or else the non-categorical state of affairs in which Nappa freely does A  in C. Thus, Molinism, in principle, is capable of satisfying the truthmaker thesis.

In Defense of Counterfactual Facts/States of Affairs as Truthmakers
Let us now consider some objections to the solution I just offered. According to William Hasker, he intuits that only categorical states of affairs can ground conditional truths.13 Accordingly, the non-categorical states of affairs Molinists posit cannot ground CCFs. By way of responding to Hasker’s objection, the Molinist should first consider whether or not they intuit the truth of Hasker’s claim about what can ground conditional truths. If they do not, then it will be irrational for the Molinist to adhere to Hasker’s view. This is so because the Molinist would lack justification for adherence to the intuition. Moreover, I don’t claim to intuit that non-categorical states of affairs can ground conditional truths, only that it’s not obvious to me why one should rule them out. Therefore, if Hasker thinks that Molinists should be committed to the intuition that he appeals to, then it is incumbent upon him to first provide us with some type of argument on behalf of the intuition. Hasker provides no such argument.

Next, there is an objection that comes from Theodore Sider. In his discussion on the grounding objection to presentism, he complains that the non-categorical states of affairs presentists posit in their response to the grounding objection to presentism are “ontological cheats”. Specifically Sider states that,

What seems common to all the cheats is that irreducibly hypothetical properties are postulated, whereas a proper ontology should invoke only categorical, or occurrent properties and relations. Categorical properties involve what objects are actually like, whereas hypothetical properties ‘point beyond’ their instances.14

I take it that Jensen has successfully refuted Sider. She does this by highlighting that Sider begs the question in favor of his view since he gives no argument on behalf of his claim except to assert that he wants to call non-categorical candidates for truthmakers “cheats”. Not only that, but Jensen has also rightly pointed out that Sider is begging the question against the Molinist.15 In proving this accusation Jensen supposes for the sake of argument that all true propositions have truthmakers. Now according to Jensen, if we take it to be the case that all truths require truthmakers, then it would seem to be wrongheaded to go looking for the truthmakers of CCFs among the occurrent properties and relations Sider wants to ground them in since CCFs are themselves literally counterfactual. Hence, by restricting good candidates for truthmakers to categorical states of affairs, Sider seems to be begging the question by ruling the aforementioned candidates out at the outset.

A final objection worth considering is a variant of The Indispensability Argument for Platonism. This variant of the argument purports to demonstrate that Molinists who happen to be anti-realists about abstract objects cannot consistently ground CCFs in counterfactual facts or states of affairs because, on anti-realism, neither CCFs or counterfactual states of affairs exist in a metaphysically heavyweight sense. The brevity I seek to preserve in this article will not permit me to discuss this objection in any depth. However, I have already defended the notion that Molinist-antirealists can consistently engage in abstract discourse without being ontologically committed to the existence of abstract objects in my essay Responding to James White’s Anti-Molinist Critiques: Abstract Objects to which my readers may inquire.

In conclusion, I’ve demonstrated that White’s anti-Molinist argument suffers from an ambigity stemming from his use of the word ‘determine’. Specifically I demonstrated then the word ‘determine’, as White uses it in his argument, is understood in a causal sense, he simply begs the question against the agent-causal libertarian by assuming from the outset that CCFs must be non-brute. I also demonstrated that if White intended the word ‘determine’ to be understood in a non-causal sense, then his objection can still be sufficiently answered by the Molinist by grounding the truth of CCFs in counterfactual facts or states of affairs. So far as I can tell, the Molinist response to this perennial truthmaker-style objection is perfectly defensible.




3On pages 64-65 of Jennifer Jensen’s “The Grounding Objection to Molinism,” PhD diss. (University of Notre Dame, 2008) she makes a parallel observation and distinction regarding the use of the word ‘making’ as it is employed in grounding objections to Molinism.

4To clarify, the claim here is not that there can be nothing that partially explains why an agent performs some action. For the agent may very well have reasons which serve to partially explain why they performed the action. Rather the claim here is that nothing can sufficiently or fully explain why an agent performs some libertarianly free decision over another.


6See The Dividing Line episode William Lane Craig, Molinism, The Lowered Bar, and Erasmus (47:21-48:34)

7Definition adapted from J. P. Moreland & William Lane Craig, Philosophical Foundations for a Christian Worldview (1st Edition), p. 233

8Argument is adapted from Ibid. p. 233

9The truthmaker-style grounding objection to Molinism should not be confused with the closely related causal-style grounding objection to Molinism. The former objection purports to demonstrate that Molinism is false because it is incapable of providing required truthmakers for CCFs. The latter objection purports to demonstrate that Molinism is false because CCFs lack the relationship to the subject of the counterfactual that is required in order for the action described by the counterfactual to be libertarianly free. Unlike the truthmaker-style grounding objector, the causal one does not want to know what makes CCFs true, rather she wants to know how CCFs could be true given human libertarian freedom. Answering the former question does not answer the latter question. For the the fact that x (e.g., a truthmaker) entails y (e.g.,  a true CCF) does not mean that x sufficiently explains y. For example, the fact that a piece of paper is rectangular entails that the paper is a quadrilateral. However, the fact that the paper is a quadrilateral does not sufficiently explain why the paper is rectangular. Instead, it seems that the sufficient explanation for the paper’s rectangularity must be found in either (1) the paper maker (2) the purpose that paper fulfills or (3) the circumstances under which the paper was created. For a response to causal-style grounding objections see Jennifer Jensen “The Grounding Objection to Molinism,” PhD diss. (University of Notre Dame, 2008) pp. 120-139.

10Specificially I take truthmaker maximalism, which is the view that all truths require truthmakers is dubious. In an attempt to disprove the relevant theory, Trenton Merricks has, I think, successfully argued in Truth and Ontology (Oxford: Clarendon Press, 2007), p. xvii; cf. p. 168 that if truthmaker theory obtains, then, at most, only truths about what properties are actually had by entities that exist in the actual world must have truthmakers. For negative existentials, universal generalizations, moral, modal, dispositional, tensed truths, and counterfactuals of creaturely freedom seem to furnish exceptions to the truthmaker principle. On page 181 of his book, Merricks thus concludes “Some truths are not true in virtue of how they are related to any existing entity or entities.”

11On page 115 of The Grounding Objection to Molinism, PhD diss. (University of Notre Dame, 2008) Jensen describes a categorical state of affairs as one that does not contain a conditional in its description. A non-categorical state of affairs, then, is presumably one that does contain a conditional in its description.

12See Jennifer Jensen The Grounding Objection to Molinism, PhD diss. (University of Notre Dame, 2008) pp. 112-116

13In William Hasker’s, God, Time, and Knowledge (Ithaca: Cornell University Press, 1989), p. 30. he specifically states “In order for a (contingent) conditional state of affairs to obtain, its obtaining must be grounded in some categorical state of affairs. More colloquially, truths about ‘what would be the case…if’ must be grounded in truths about what is in fact the case. However, this statement requires clarification. For as Jensen notes in The Grounding Objection to Molinism, PhD diss. (University of Notre Dame, 2008) p. 113 “In the first sentence, [Hasker] seems to identify the relata of the grounding relation as states of affairs. In the second sentence, he seems to identify the relata of the grounding relation as truths. This second statement is admittedly supposed to be colloquial but I don’t think that either statement precisely captures what he intends to say. The grounding relation is usually described as a relation between propositions and states of affairs where propositions are grounded in states of affairs. I think that the intuition that he intends to capture is something like this: conditional truths must be grounded in states of affairs that are categorical states of affairs. For example, consider the conditional truth. . .: If today is my birthday, then I will open a gift. Let’s suppose that this material conditional is true. There are several candidates for states of affairs that could serve as the ground of this material conditional: (1) its being the case that today is not my birthday, (2) its being the case that I will open a gift, or (3) its being the case that if today is my birthday, then I will open a gift. I take it that the intuition Hasker cites is supposed to rule out (3) as a good ground for the conditional because it is not a categorical state of affairs. However, (1) and (2) are categorical states of affairs that can serve as grounds for the conditional above. Thus, material conditionals can be true because they can be grounded in categorical states of affairs.”

14Theodore Sider, Four Dimensionalism: An Ontology of Persistence and Time (Oxford: Clarendon Press, 2001), pp. 40-41

15See Jennifer Jensen The Grounding Objection to Molinism, PhD diss. (University of Notre Dame, 2008) pp. 115-116

Tagged with:

About the Author

By Jonathan Thompson