Wednesday, April 24, 2013

Molinism is not Determinism

A frequent criticism of Molinism among laypeople is that it equates to causal determinism. “God chose to actualize the circumstances in which you do X, so that you could not have done any differently than you did. This means middle knowledge equals determinism.” My contention is that there are several misunderstandings within this type of critique.

First, and perhaps most fundamentally, middle knowledge cannot possibly entail determinism, by its very definition. This is because the knowledge is in between two other knowledge-moments. God’s free knowledge is everything known to God because he decreed it, including the actual world and its contents. God’s natural knowledge is his knowledge of all necessities and possibilities, including things that never come to pass. Middle knowledge, however, is distinct from free knowledge, because its content does not depend on the will of God (God’s will is logically posterior to the contents of middle knowledge); further, it is distinct from natural knowledge, because its contents are not necessary.[1] Thus, by definition, middle knowledge counterfactuals are not deterministic.

However, let us suppose that the objector really means something like, “There cannot be any such thing as middle knowledge, since God’s actualization of a world means an agent cannot act any differently than he does act.” This is also problematic. What has happened is that the objector has failed to tease out precise meanings for these terms. What happens in God’s decree (on Molinism)? God decrees that the world be actual. This decree is based on a) necessary truths (which must be a part of any world God creates, by definition), b) truths of how God will act (which are entirely and completely up to God, encompassing the physical truths of nature and any actions of God with and for his creatures), and c) counterfactuals of creaturely freedom (which are, again by definition, not up to God [though whether or not God created a world with free creatures is, of course, entirely up to God—free will is God’s gift]). Since the truth or falsehood of the counterfactual is so logically prior to God’s actualization of a world, God’s decree is only that the counterfactual be actual (rather than merely abstract).

What is impossible, therefore, is not that the subject acts any differently. What is impossible is the composite state of affairs of “In C, X would freely do S at T; C &T are actual; X would not freely do S [or X would freely not-S].” It is the all-too-common modal fallacy to distribute the necessity over the entire composite to its individual conjuncts. Middle knowledge, and Molinism, are very difficult concepts. I am hoping that by this article some confusion may be avoided.

[1] Although the contents of middle knowledge appear in natural knowledge, they are only necessarily possible, not necessary in themselves.


  1. Thanx for article

    Firstly: it is at question if something as God´s non-essential knowledge (i.e. scientia media and free knowledge) is logically and metaphysically possible regarding God´s essence. This - that it is possible - is asserted in this article: But I am really not sure about it.

    Secondly, there are two more objections against molinism: that it a) violates the law of universal God´s causality, and b) entails a passivity in God (because God passively learns which decision will free creature make), which would destroy God as actus purus, as pure perfection. What do you think about these objections and what would you reply to them?

    And one more comment about free will and determinism problem generally, to the end. Free will is a theoretical entity that requires that there be a sufficient reason to assert it. Do we really have it? Isn´t molinistic "circumstancial determinism" something that makes free will concept unnecesarry, because these completely determined circumstances (which include the current finite knowledge of the "decision-making" subject) are sufficient for explainig of why the subject chosen A and not non-A? Isn´t from this point of view free will something as "by Occams razor cuttable invisible dwarf" that we assert next to the gravity force when offering the explaination of why the material object on earth fall down when dropped?

    1. Hi Jakub, thanks for the questions! Most theists will not want to say that God only possesses essential (or "necessary" or "natural" knowledge), because then either one of two things follow: either everything done, including all acts by God, are logically necessary, or else God doesn't know anything outside of possible/necessary truths (which, presumably, even open theists would deny, since they still affirm God knows what *he* does). Perfect Being Theology, based loosely on Anselm, can lead us to conclude that God's essence will include omniscience.

      As to the other two objections, I assume that (a) means that God should be the direct cause of everything. But most theists will not accept this, and even the ones that will must overcome the counter-objection that this appears to violate both free will and God's perfection (since God would be the direct cause of sin). As to (b), Molinism doesn't mean that God must learn what the creatures choice will be, or, depending on how technical you want to get, it does, but that learning isn't passive. God knows the truths of how free creatures *would* act given varying circumstances, and he knows that passively in the sense that he did not causally determine the actions themselves. However, he doesn't learn them, because that implies there is a time where he lacks this knowledge and then gains it. The logical moments in this discussion are only showing logical relationships. Now, if one means, literally, that God gains knowledge about what free creatures *will* do, this is not passive, but active. Why? Because God actively chooses which world is actual. This means that how I would freely act is up to me, but how I will act in reality is up to God, given that God decides which of those worlds (made up of choices I would freely make that are up to me). So, I still freely choose what I do in this actual world, but God chooses that it is actual, and not counterfactual.

      Now, to the last question: this above article is meant to show that Molinism does not have circumstantial causation (determinism). One has to ask which way the arrow runs, as it were: Does God know it because it is what we would do, or would we do it because he knows? Molinism asserts the former, and we have no good reason to assert the latter. I hope that helped! :)

    2. Hello Randy
      Thanks for reply. Here are my next counter-objections or arguments:

      "either everything done, including all acts by God, are logically necessary"
      Well, where exactly do you see the problem in this? :-)

      "This means that how I would freely act is up to me, but how I will act in reality is up to God, given that God decides which of those worlds (made up of choices I would freely make that are up to me). "

      Well, don´t you think there is a contradiction? On one hand, God chooses which of the possible worlds will be actual. But that possible world is completely determined, its possible-ness is completely concrete, it is not only partly determined, or indetermined, inexplicit. But regarding this, it must consist also of my "so-called" free acts, but how can they be free then? Well, maybe you define the freedom of choosing in some different meaning (for example as thomists do) but I would doubt it ...

      "I still freely choose what I do in this actual world, but God chooses that it is actual"

      Again, in other world - actualised possible world includes both its actuality and alll its fully determined components (up to the smallest details) that necesarilly include my so-called free acts.If the possible world wouldn´t include it, it would be some other possible world - not the one God actualises.

      "Does God know it because it is what we would do, or would we do it because he knows? "
      Well, I understand, but as you surely know, God´s science is equal, identical with his creative power. God creates through his knowledge. I still think that molinism is circumstancial determinism, because, as I have written in my previous post, the circumstances seems to be sufficient condition for making particular choice by "free" agent (and molinism says it: God precisely knows, which alternative would the agent choose in what circumstances - the set of the circumstances could be almost infinite). If the circumstances are sufficient for explanation, why would we have to add the free will (as the theoretical entity)? It makes it redundant. Or, do you think that the "circumstancial explanation" still has some explanation gap? What?

      By the way, pity you can´t speak czech :-D I have my own philosohphical blog ( and I am sure you would be able to effectively comment my posts :-)

    3. Helo Jakub, I wish I did speak Czech! You speak English very well. :)

      As a Christian, I affirm that God can choose to do certain things, and that he also is the foundation for logic and reason itself. As such, it follows that God's actions are not causally determined by something outside of himself.

      Regarding the next objection, I do not think such a possible world that is made actual is fully determined. It occurs to me that perhaps the word "determined" is being used in two different ways. The first way is causal determinism, which indeed is contradictory to freedom. The second way is an epistemological "determined," which means, roughly, "to find out and know" or something. So I think the actual world is determined by God in that he knows exactly what the contents of the world are, and God chooses which world is actual, but the only actually causally determined fact about some world is that it is actual. :)

      As to the circumstances objection, Molinists think there are possible worlds, both describing an agent freely doing X and freely doing not-X; if this is correct, then circumstances are not sufficient for explaining the choice. It is the will of the person! Now, perhaps someone can make the argument that it is circumstances alone which determine these choices, but that would be a different argument.

      Thanks again for commenting! :)

    4. Hi Randy, thanx for reply again
      (BTW I would appreciate if you had last-comments add-in at your blog ... :-) )

      In what sense you regard God as foundation of logic and reason itself?

      "As such, it follows that God's actions are not causally determined by something outside of himself."
      I understand, but - to what do you reply by saying this?

      To the next paragraph. Maybe I wrongly used the word "determined" instead of word "specified". What I silently had on mind when I was writing that paragraph was the thesis by Aristotle from ninth book of Metaphysics, that each possibility is fully specified and particular, it is not some "unspecified possibility of anything". And in this sense I thought that God has to actualise fully specified possible world. But it is true, and I haven´t twigged this previously, that if some possibility is being actualised, then it isn´t necessary that it has to by fully actualised by only one and the same agent. There could be more "actualisers", like when two fitters construct one fully specified device. Similarly, these two "fitters" of possible world are God together with free agents.
      But what I still have problem with is that although free agent could be regarded as "actualiser", still completely all his being has been given to him by God, and the particular choice, that co-creates possible world, is also being. So in this sense the created actualiser is like a puppet in God´s hands. Surely, there is a problem which you mentioned above, that God cannot be author of evil decisions. Well, there is a well-known solution of this: evil is not being and God can cause only being, so the responsibility of evil decision - which is ontologically non-being - is fully put on created agent. Although I am really not happy with this solution, there is a logic in it. But I still see one problem here: if this solution is true, then it really seems that only evil acts are authentically free acts, only it expresses true freedom of decision. Because from the free-agent´s point of view, they are completely "a se", and in no way "ab-alio"

      "Now, perhaps someone can make the argument that it is circumstances alone which determine these choices, but that would be a different argument."

      In what sense that would be a different argument? I think this is a core of the problem ...

      BTW, I am a christian too, although in "permanent spiritual crisis" and other problem ...

  2. Hi Randy, did you get my last comment? A have sent it two days ago and it is still not here ... now I´m not sure if it was really posted ...

  3. "Molinists think there are possible worlds, both describing an agent freely doing X and freely doing not-X; if this is correct, then circumstances are not sufficient for explaining the choice. It is the will of the person!"

    I am quite sympathetic to molinism. However, no matter who I read after (Craig, Flint, Freddoso, etc.) a particular question seems to come to mind--not sufficiently addressed by any of them. It seems quite clear that Molinism amounts to circumstantial determinism (something bordering on a compatibilist description of free will). The only answer given by these writers is that the circumstances must be fully specified, but that doesn't change the appearance of circumstantial determinism (CD), does it? You claim that it isn't CD because there are some possible worlds in which the agent does X and some in which he does non-X in the same fully specified circumstances. (I'm not sure that the primary Molinist writers would agree with you here.) That's fine. That would certainly preserve LFW, at least as far as that description goes. However, you now have another problem. If that description is true, and the only difference between the two worlds at that point is the agent's free choice (not the circumstances), then how can you ensure that the world that God actualizes turns out exactly the way He knew via middle knowledge? On your account, the particular world He knew via MK--with all of the circumstances fully specified--could yet have been different simply by a single agent choosing differently (for, on your account, the circumstances actually weren't a sufficient explanation for the action--they weren't determinate in any way). Did God somehow "freeze" the world, so that they couldn't have chosen otherwise after actualization?

    I hope that makes sense, but I believe it is a significant problem.

    1. Hey Derrick, thanks for your questions! I want to tease out a few things as I go. First, it does not appear to me to be circumstantial determinism; that is, an important part of this critique is that it seems that the circumstances cause the agent to do as he does. But why think that? I think later you provide some hint (at least if I am understanding you). You ask how God can ensure that the world God actualizes turns out exactly he way he knew via MK.

      To answer, MK contains *counterfactual* truths about how agents *would* act. Possible worlds containing agents who do x can include these counterfactuals (and indeed often do), but need not. Here's all I mean by that very imprecise language: Suppose this counterfactual is true of me: "If I were in C, then I would freely marry Jodi." We can represent this counterfactual as a basic conditional as well: "If Randy is in C, then he does x" (just a descriptive conditional of how I act in a particular possible world; I don't mean material implication necessarily). But there is another world in which I do not do this; but this only describes ability, or *could* truths. I am not saying that in one and the same circumstances I both would and would-not do x; this is not only impossible for God to know which I would do, but contradictory! Instead, God would know the truths "If Randy were in C, he would freely marry Jodi," and "If Randy were in C, he could freely not marry Jodi;" these truths are not only compatible (since not everything one could do one would do), but in virtue of knowing which one I would do, God would be guaranteed to get the world he selected.

      Now it occurs to me you may be asking the mode of God's knowledge, and that I would count as conceptual, and not perceptual. But this is my best stab at it; if I didn't address your concern, I apologize--I wasn't really sure why CD was being postulated, so I tried to respond to what I thought was being asked. I hope it helped!

  4. Randy, thanks for the response. My knowledge in this field is quite limited, but growing. So, I very well may not be clearly asking the questions that are somewhere in my head.

    Considering your response, I think my question would relate to this statement: "But there is another world in which I do not do this; but this only describes ability, or *could* truths. I am not saying that in one and the same circumstances I both would and would-not do x . . ." This is a little confusing to me. Are we to assume that in both worlds--at least up to the point in question--all of the fully specified circumstances are identical? If they are, then it seems that God could not know with certainty which choice would be made in the actualized world (assuming it was one of the two) without practically destroying LFW. If it is *impossible* that there are two worlds with identical fully specified circumstances up to that point in which I choose in two different ways (respectively), then it seems that in some way (directly or indirectly) the circumstances are determining my decisions. If we are to take LFW seriously, mustn't it be possible that, given identical fully specified circumstances up to that point, there are two PWs in which two different decisions are made, at least regarding some choices?

    Sorry for being a bit unclear. It is all still somewhat fuzzy in my head, and I cannot seem to get a molinist to understand what I am getting at.

    1. Thanks for writing back! Yes, both worlds are identical in circumstances. Let's call the world where I would freely do x "W," and the world where I freely do not-x "C." W and C are obviously two distinct worlds, but they both contain the same antecedent circumstances up to the point of the choice (by stipulation). Now "x" is defined as a contingent act, meaning there are possible worlds in which I x (of which W is one) and possible worlds in which I do not x (of which C is one). However, if the counterfactual is *true*, then C is, while possible, not a feasible world for God to actualize, because it's not true that I *would* freely not-x (and even if God forces me to not-x, it won't be C, but a different world, called F, that is identical to C up to the point of the choice, except I wouldn't be freely refraining from x, I'd be forced to refrain from x, so F would be the world we had). Now all this is to illustrate that just because you would do some contingent act, it doesn't follow that you couldn't refrain from doing it (as we just showed with what appears to me to be a coherent account of worlds W and C, with the possibilities inherent to both, even if I actually take different routes in W and C).

      Now you say, "it seems that God could not know with certainty which choice would be made in the actualized world (assuming it was one of the two) without practically destroying LFW." But I ask, why think that? It doesn't seem that way *to me* at all, and so I am hoping you can explain why it seems that way to you (and then we'll be well on our way!). :)

  5. Thank you for continuing this conversation! All of it is helpful as I continue to re-read some of the Molinist literature. It seems that Flint begins (and only begins) to deal with my concern at the end of his chapter on the "grounding" objection (pp. 135-137 in my copy). He refers to Freddoso, which I have read, but which also leaves my problem largely unresolved. Craig doesn't go much beyond "as long as the circumstances are fully specified and we say they are freedom-preserving, then we are OK."

    Before I go too much further, let me ask a bit more about your most recent response. Are you saying that there are *no* feasible worlds in which I do x and, in the second, do not x?

    1. No problem--I love to talk about this stuff! :)

      In our example, we only considered two worlds of free action (the third world, F, was where God forced me to do x, and so F doesn't describe a world of free action): W and C (I am using these for "would" and "could" so it's easier to remember which is which). I did not attempt to answer whether or not there are any worlds feasible for God's actualization where I freely refrain from x. We're only considering two worlds that are both identical up to the point of my doing or not doing x, and since these are the only two options (the law of excluded middle), they are the only two worlds describing precisely the same fully-informed counterfactual.

      Now perhaps there are worlds where I would freely do not-x, but maybe these are worlds vastly different from W and C; maybe they are worlds where other choices would not freely be made and so are not feasible (say, a world, call it D, where Abraham Lincoln resigns from the presidency instead of fighting the Civil War--such a world, which we shall assume holds similarly except for Lincoln's resignation and my would-freely-not-do-x action, God cannot feasibly actualize, since, as a matter of fact, Lincoln would not freely resign. God could have forced Lincoln to resign, but then world we would have would not be D, but instead some other world, even if it looked quite similar on the surface--because at least one other proposition would be different. Does this help at all?

  6. Could there be two possible worlds that are exactly identical up to the point that an individual makes a decision and in each world the individual makes different decisions? I think you may be referring to this in your last paragraph but I am not sure I fully understand it. If it isn't possible for the individual to make different decisions then surely they are determined by their circumstances that were actualised by God. If it if possible for the individual to make different decisions how do you account for that when the possible worlds truly identical up to the point of the agent making the decision?

    1. Hello Pete, sorry in the delay! Yes, it definitely is possible to have two worlds, call them W and W', in which X freely does A in W and does not freely do A in W'. True counterfactuals, though, account for the difference; that is, though these worlds are possible, only one is feasible.

      But I suspect you mean what accounts for the counterfactual's being true (since your question, if taken literally, just asks for an account of the possibility of making different decisions, which God's omnipotence would seem to be quite helpful here in giving us free will; and since no direct problem is articulated, this seems sufficient)? If that is something like what you mean, it would be the agent; that is, the agent is imbued with causal powers.

      Now perhaps you mean what is the ontological ground of these choices? If this is what you mean, it seems plain to me it cannot be the agent. But I am persuaded by Merricks' work in Truth and Ontology that the principle that underlies this grounding demand would lead to a lot of bizarre consequences, such as with negative existentials (e.g., surely "there is no unicorn" is a true and meaningful statement) and things like holes. In any case, though, it may be easier to state the objection outright, if there is one (if not, that is totally OK; I appreciate the conversation!).

  7. Hi Randy,

    If I understand your response to Derrick and Pete correctly, that means that our choices are not determined by the circumstances involved. If so, then why when describing different possible worlds are the circumstances involved always brought up by Molinists? If they don’t detemine the outcome then it seems God doesn’t choose to actualize a world based on the circumstances. Rather, it seems He chooses to actualize a world based on what the free creatures will choose and how that fits in to His ultimate plan. So why when discussing this issue are the circumstance consistently referred to? I look forward to your response.


    1. Hello Jake, thanks for your thoughts! These circumstances influence, but do not cause, choices. Thus, if the circumstances were different, it may be the case that I would freely choose something different. So, suppose I were in circumstances where I was hungry. Then, if offered some food, I may freely choose to eat (suppose I would). But if circumstances were the same, except I had eaten an extra large breakfast that morning, suppose I would freely decline. It was in my power to eat or not eat, but circumstances influenced my decision. They didn't causally force my decision. That is why circumstances are relevant, but not causal. I hope this helps! :)


Please remember to see the comment guidelines if you are unfamiliar with them. God bless and thanks for dropping by!