Reflection and Clarity on the Molinism Debate
Wednesday, I participated in a debate on Molinism v Calvinism. I tag teamed with Richard Bushey against TurretinFan and Josh Sommer. It seems inappropriate for me to evaluate if our side “won” or not – that should be left to the audience. Nevertheless, there are a few issues I think deserve comment. In general, most of the debate hinged on definitional issues rather than true disagreements. There seemed to be a significant amount of talking past each other. In addition, I did not make my points with the greatest clarity. I would like to elucidate more clearly so future discussions can be more profitable.
Use of Scripture
As a preface, I want to address why I didn’t go straight to Scripture and argue whether libertarian free will is taught. I don’t think that it would be an appropriate thing to try to find in Scripture. The Biblical authors did not think or write in the categories of compatibilism or libertarianism, thus, I think it would be inappropriate to try to proof-text. Additionally, the case that I was arguing was for sort of a minimalist Molinism. The single criterion for this is a metaphysical commitment to the reality of middle knowledge. Even if God chose to create a world where causal determinism is true, He still has middle knowledge and He could have chosen to create a world with libertarian creatures. Lastly, I don’t think that anything new is going to come up in the interpretive debate. We can exegete Joshua 24 and Romans 9 all day long but both sides are so entrenched, I don’t think the battle lines are going to move anywhere. Again, I don’t think the Bible speaks in these categories. What can be clearly known from Scripture is that God has complete foreknowledge and foreordination, God exhaustively sovereign, man is morally responsible, and man makes choices.
The Possibility Argument
There has been some kerfuffle over this argument already. In the second rebuttal, Josh summarized my points as an argument for middle knowledge based on the possibility of middle knowledge. Whether that’s what he intended to say or if he meant “possibility of libertarian free will” both would be technically correct (the best kind of correct). The gist of the argument is that God has His properties necessarily . That is to say, if something is true of God’s nature in one possible world, it is true in every possible world. So, the reasoning goes something like
(1) If it is possible for God to have middle knowledge (MK), then God has MK in some possible world.
(2) If God has MK in some possible world, then, God has MK in all possible worlds.
(3) If God has MK in all possible worlds, then God has MK in the actual world.
(4) If God has MK in the actual world, then God has MK.
Josh is correct in saying that this is an argument from the possibility of MK; what he didn’t mention were the other two steps that I used in the argument.
(5) If libertarian free will (LFW) is possible, then it is possible for God to have MK.
(6) LFW is possible.
From there, it’s modus ponens all the way down, so, we can use hypothetical syllogism to condense it down to
(7) If libertarian free will is possible, then God has MK.
This is logically equivalent to
(8) If God does not have MK, then LFW is not possible.
What about the arguments in favor of the possibility of LFW?
First, I argued that the burden is on the one making an impossibility claim. Now, I also used this in my defense against the grounding objection, so, a detour is warranted here. In logic, saying something is possible is simply to say that the referent does not entail a contradiction. In other words, if some proposition p is possible, then p does not entail a contradictory statement. Conversely, if p is impossible, then, it does entail a contradiction. But, the one who claims that p entails a contradiction must prove it. Another way to put this is that possibility claims are the null hypothesis and must be demonstrated as false. Our Calvinist opponents did attempt to do this. As best as I understand it, their counterargument tried to show that LFW prevents God from being sovereign.
(9) God is sovereign over all things.
(10) If LFW is possible, then, there are some things over which God is not sovereign
(11) LFW is possible (Assumption)
(12) Therefore, God is not sovereign over all things.
(13) Therefore, God is sovereign over all things and God is not sovereign over all things (9&12)
Upon further questioning, it seemed to me that they were defining “sovereignty” as entailing causal determinism. This would render the counterargument trivial and equivalent to “God cannot causally determine an indeterminate event”. Since this issue is deserving of further discussion, I won’t push it further here. I will simply say that I don’t think causal determinism is required for God to be truly sovereign.
Second, I pursued a less abstract route and cited pre-Fall Adam as a potential candidate for genuinely having LFW. There were a few reasons for this. First, the possibility argument is constructed so as to do an end-run around the never ending exegetical battle over whether the Bible teaches compatibilism or libertarianism. Still, it’s understandable that we get out of the realm of pure theoretical and back to Scripture. Second, invoking pre-Fall Adam allows us to do another end-run around a different exegetical slug fest. See, if pre-Fall Adam does have LFW, then that seals the deal for the possibility argument – Adam has LFW, God knows what Adam would do, therefore, God has middle knowledge. On the other hand, if pre-Fall Adam does not have LFW, then this eviscerates a large chunk of the exegetical case against LFW. Verses like Romans 8:7, for example, don’t carry any weight establishing compatibilism. If a sinless Adam can’t be libertarian, then post-Fall evidence doesn’t add anything to the equation.
To re-emphasize, the purpose possibility argument is just to break down the metaphysical divide between Calvinists and Molinists. The argument claims that God has middle knowledge irrespective of whether or not He created humans with LFW. There is a big difference between a Calvinist who thinks God doesn’t have middle knowledge and a Calvinist who thinks God has middle knowledge but chose to create a deterministic world.
 Two notes: First, I know strong adherents to divine simplicity are cringing at “God has properties”, but, you know what I mean. Second, this is not to say that God’s accidental properties are necessary. There are some possible worlds, for example, where God did not create anything, so, “being Creator” is not true in that possible world. This argument, however, is focused on God’s pre-volitional omniscience which is essential and necessary.
Reviewing the debate, I wasn’t clear on what I was trying to accomplish in the response to the grounding objection. My response to the grounding objection happens in two steps. First, I advanced two critiques that serve to defuse the objection. Second, I advanced three potential solutions that provide an account of grounding pre-volitional CCFs. Unfortunately, I did not manage my time well and spent the majority of the time on the defusing arguments rather than a balance.
First, the purpose of the defusing arguments is to neutralize the grounding objection. This is why I said in the debate that these are kind of non-answers that will not satisfy the Calvinist.
(i) The first defuser is that the grounding of CCFs is an inscrutable question, akin to asking how God knows anything. This isn’t to downplay the importance of the question, rather, it is a preface to say that this is an extremely complicated question. Many times, I’ve seen a sort of “heads I win, tails you lose” scenario play out. The Calvinist asks “How does God know these CCFs?”; the Molinist responds she doesn’t know and the Calvinist claims victory. Alternatively, the Molinist responds with an in-depth theory of grounding and the Calvinist hand waves that away, saying it’s unbiblical or too complicated. The reality is that the question is complicated, not Molinism. Because of the nature of God and the limits of man, every attempt theological reflection must at some point appeal to mystery; while unsatisfying, I think it’s legitimate to appeal to mystery here, saying that we perceive these truths and God knows them by virtue of His omniscience.
(ii) The second defuser is to say that the grounding objector bears the burden of proof. As discussed above, the reason is because the grounding objection is an impossibility claim. It claims that middle knowledge combined with an axiomatic grounding principle leads to a contradiction. So, the burden is on the grounding objector to elucidate what grounding principle he is using and demonstrate the contradiction. This would look something like the following wherein the grounding principle is G, middle knowledge is M, and A is some derived proposition.
(14) (G⋅M) ⊃ (A⋅¬A)
Further, I made the claim that given any grounding principle either (a) CCFs could be demonstrated to be grounded on that principle or (b) the principle could be rejected for independent reasons. I would like to add on to the previous discussion that there is a difference between an impossibility claim and an uncertainty claim. For example, “pre-volitional CCFs cannot be grounded” is not the same as “I don’t see what grounding there is for CCFs”. The first statement is an argument, the second is just an uncertainty claim. Perhaps it seems unintuitive for you, but that doesn’t make it false. In the spirit of fairness, I admit that I do the same thing – there are certain claims of Calvinism that don’t seem to fit together to me; that doesn’t make them false, it means I personally don’t find them to be persuasive.
Again, just to re-emphasize, the purpose of these defusers is not to rebut the grounding objection but simply to bring it to a point of neutrality. If you are not already a Molinist, then, these will not be satisfactory. If you are already a Molinist, these can serve to remove the objection and you can still be warranted in holding your position for other reasons.
Second, I advanced three potential solutions that provide an account of the grounding of pre-volitional CCFs. These do claim to rebut the grounding objection.
(i) The first potential solution is Molina’s own view of “supercomprehension”. In this view, God knows via scientia naturalis what He is able to do. In that knowledge, He knows that He can create a libertarian free creature or even an impersonal indeterministic scenario. Molina claims that in order to truly know what He can create, God must know every aspect of that object including all possible relationships with other possible objections. In the case of indeterministic objects, that includes knowledge of what would occur. At the 1:07:15 mark in the debate, I read the relevant quote from Molina on this topic.
(ii) The second potential solution is the conventional “essence solution”. This basically posits that God contemplates uninstantiated human essences in which the CCFs are grounded. I didn’t go into any detail in the debate, so, I will save that for further discussion.
(iii) The third potential solution is what’s called “Complete Concept Molinism” which builds off of Leibniz’s thought about complete concepts. Ultimately, it grounds CCFs in what are called choice functions. I think that this is the most promising approach for solving the grounding objection. I’ve linked to the relevant paper here.
Something else that didn’t get discussed much is what Richard and I mean by “Reformed Molinism”. I don’t know which one of us (or if neither of us) started using this term first, but, it can mean different things. I remember the first time I used it, I meant that the majority of my views are traditionally associated with reformed theology – things like covenant theology, amillennialism, the 5 Solae, real presence of Christ, etc etc. When Richard started using it, he meant it as a distinction between those Molinists who lean closer to the Arminian side and those who lean to the Reformed side. For example, some Molinists hold to corporate election and others hold to individual election. Since Richard and I are in the individual camp, if we were to exegete Romans 9 or John 17, it would probably sound indistinguishable from Calvinist exegesis to most ears. Here’s a brief rundown of some of the terms we use to distinguish between the different flavors of Molinism
Mere Molinism – a commitment to middle knowledge and libertarian free will
Minimalist Molinism – a commitment to middle knowledge, but, not necessarily LFW
Arminian leaning Molinism – mere Molinism + compatible Arminian doctrines (e.g. prevenient grace, corporate election, etc)
Reformed Molinism – mere Molinism + compatible Reformed doctrines (e.g. radical depravity, unconditional individual election, limited atonement, efficacious grace, perseverance of the saints, etc), shares a close link with Congruism.
I look forward to further discussion with TurretinFan, Josh, and other Calvinists. I hope that this post helps to clarify the position that I’m coming from and ameliorate any confusion on my part in the debate. Additionally, I hope that my skeptical Calvinist brothers and sisters can see that I’m not frothing at the mouth, screaming about Servetus. I will gladly admit that should Molinism be demonstrated to fail, I would embrace the fullness of Reformed theology without any hesitation.