A Theistic Argument for Compatibilism
January 29, 2013 — 17:29

Author: Kenny Pearce  Category: Concept of God Divine Providence Free Will  Tags: , , , ,   Comments: 42

One often hears it asserted that most theists are metaphysical libertarians. This seems to be supported, at least in the case of theistic philosophers, by the PhilPapers survey where target faculty specializing in philosophy of religion, who were overwhelmingly more likely to be theists than their peers in other specializations (72.3% for religion specialists vs. 14.6% overall), were also overwhelmingly more likely to be libertarians (57.4% vs. 13.7%). (Unfortunately, there doesn’t seem to be a way to compare theists to non-theists across the board, so we just have this correlation among religion specialists.)
Now, I suppose there are some reasons for this. One is that the free-will defense is widely thought to be the best response to the problem of evil, and is widely thought to require libertarianism. Another is that theists are often committed to some notion of punitive justice which is also widely thought to require free will. A third reason is that if God is a necessary being, and libertarianism is not true at least about divine freedom, then we inherit all of Leibniz’s difficulties in trying to carve out any contingency in the world at all. Now, it is perfectly possible to be a libertarian about God’s freedom and a compatibilist about human freedom. According to some (most?) interpreters, Aquinas adopts this view. On the kind of view in question, libertarian freedom would be the most perfect sort of freedom, and perhaps one might even concede that it’s a sort of freedom that we (prideful) human beings often think we have, but one would say that compatibilist freedom is sufficient for moral responsibility and is all that we actually have. Nevertheless, if the theist is committed to saying that God has libertarian freedom, then the theist is committed to saying that libertarian freedom is at least a coherent notion, and it’s easy to see why that would be at least correlated with claiming that we actually have libertarian freedom.
On the other hand, I think there are special philosophical reasons for theists to accept compatibilism. Reasons, that is, which are specially philosophical (as opposed to theological or scientific) and also specially applicable to theists. Consider the following argument:

  1. There are possible worlds in which God creates free beings other than those free beings (if any) which exist in the actual world.
  2. For any possible free being and any possible maximally specific circumstance in which that free being chooses freely, God knows what that being would choose. That is, every (maximally precise) counterfactual of creaturely freedom has a truth value known by God.
  3. Therefore,

  4. There are true counterfactuals of creaturely freedom concerning creatures who do not exist.
  5. If incompatibilism is true and any counterfactual of creaturely freedom concerning a creature who does not exist is true, then that counterfactual is a brute contingency.
  6. There are no brute contingencies.
  7. Therefore,

  8. Incompatibilism is false.

The base premises of this argument are (1), (2), (4), and (5). Plausibly, God’s omnipotence requires (1) and his omniscience requires (2). (4) Seems clearly true, and (5) is a premise in cosmological arguments, hence many theists are committed to it. (A word about brute contingencies: if one defines ‘brute contingencies’ in such a way that facts which are made true by exercises of agent-causal power are not brute, (4) is still true, since agents who don’t exist can’t exercise agent-causal power, and the argument still goes through. Because of the possibility of this construal of ‘brute contingency’, the argument does not beg the question against theistic libertarians.) So with the exception of (4), every premise is one the theist has some special reason for believing.
All the premises are, I take it, open to question (though I will comment that (1) and (4) seem clearly least questionable to me), and I am myself uncertain whether the argument is sound, and uncertain whether the conclusion is true. A great many (perhaps most) of the philosophers who endorse both theism and libertarianism see the issues clearly enough to reject at least one of the premises (typically either (2) or (5)), so I suppose I’m not saying anything terribly shocking. However, it seems to me that this argument is strong enough, from the theistic perspective, that we ought not to regard the combination of theism and compatibilism (at least about human freedom) as odd or unusual, except perhaps in a purely sociological sense. There are strong reasons, within a theistic metaphysics, for endorsing compatibilism.
(cross-posted at blog.kennypearce.net)

Comments:
  • Hi Kenny,
    Thank you for this interesting post! I look forward to the discussion it generates.
    I think you are right when you suppose that many theists believe in (human) libertarian free will because they advocate the free-will defense in response to the problem of evil. It is worth noting, I think, that if you accept Plantinga’s distinction between a defense and a theodicy, then you have available to you the option of wielding the free-will defense as a response to the logical problem of evil without necessarily being committed to belief in libertarian free will (even if you think that the ‘free will’ in the free-will defense is a libertarian account of free will). This is only a minor point, though, since many theists aim to respond to the problem of evil with a theodicy rather than a defense.
    You also mention that another possible reason that many theists believe in libertarian free will is that “theists are often committed to some notion of punitive justice which is also widely thought to require free will.” If ‘free will’ in that sentence is read as ‘libertarian free will’, I think you are right about the possibility of this being a reason that theists are inclined toward libertarianism about free will. Since many compatibilist believers in free will give up retributive punishment, this tendency by those who are committed to punitive justice toward libertarianism about free will is understandable. Perhaps theists need not be committed to punitive justice (between human creatures, at least), or perhaps theists can develop an account of punitive justice that is not ruled out on a compatibilist account of free will. More work could be done on these fronts.

    January 29, 2013 — 21:53
  • Kenny Pearce

    Hi Taylor,
    You actually can’t use Plantinga’s Free Will Defense as a defense in his sense unless you hold that human libertarian free will is metaphysically possible, but the argument I present would show that there are not even any merely possible libertarian-free creatures. Of course, if libertarian freedom is merely epistemically possible, that will be enough to show that the logical problem of evil is inconclusive. But Plantinga wants to do more than that: he wants to prove that there is no inconsistency between God and evil, and he can’t do that unless the story involved in his defense is metaphysically possible.
    I did mean ‘libertarian free will’ in the sentence you quote in your second paragraph.

    January 29, 2013 — 22:24
  • Would the Molinist not deny (4), saying that the counterfactuals are necessarily true?

    January 30, 2013 — 1:49
  • Kenny,
    You are right: if your argument is sound, the free-will defense won’t work. If a theist had other reasons for being a compatibilist, though, and was not yet convinced by your argument, s/he might think that (human) libertarian free will is metaphysically possible but not actual.

    January 30, 2013 — 5:37
  • Matt

    Kenny,
    Several libertarian-leaning philosopher-theists I know reject 2). They hold that the ability to do otherwise is required for free will. If “every (maximally precise) counterfactual of creaturely freedom has a truth value known by God,” then creatures cannot do other than what God knows that they will in fact do. But if the future is open, then there are no truth values to be known by God. This might pose problems for divine omniscience though. We’d have to cash out omniscience in terms of ‘knowing everything that it is possible to know’ and then claim that since the future is open it is impossible to know what a creature with libertarian free will will do.

    January 30, 2013 — 6:01
  • Kenny Pearce

    Roland – If Molinism is to be a type of libertarianism, then the Molinist cannot hold that the counterfactuals are necessarily true. If it is necessarily true that if I were in C I would do A, then there is no possible world in which I am in C but don’t do A. (Subjunctive conditionals with true antecedents can be true only if their consequents are also true; they behave differently from material conditionals only when the antecedent is false.) Leibniz has a view like this, but of course he’s a compatibilist.
    Taylor – Yes, I agree. I’ve sometimes wondered if this is actually Plantinga’s view, since he’s supposed to be a Calvinist. (Does anyone know?) On the other hand, if you are worried about sovereignty (in, say, a Calvinist kind of way), you’ve got a problem even if the libertarian-free people are merely possible, since there are still all these counterfactuals over which God has no control – in fact, maybe that’s why he decided not to create libertarian-free people, because the counterfactuals were bad.
    Matt – You don’t need open future to reject (2), because (2) involves counterfactual statements, not future-tensed statements. Also, if the statements in question have no truth values, then you can just say that God knows everything true. Some open theists hold that future statements do have truth values, but God nevertheless does not know them, since God knows only everything knowable, not everything true, but I think the open future view (that there are no truth values) is more prevalent among open theists. At any rate, I definitely reject the claim that if God knows my future actions I am not free to do otherwise, and I’m inclined also to reject the corresponding claim about counterfactuals.

    January 30, 2013 — 11:44
  • Dianelos Georgoudis

    I think the main reason most theists believe in libertarian free will has nothing to do with the problem of evil or with the traditional view of punitive justice. Rather libertarian free will is a fundamental aspect of our experience of life, and since it is very implausible that God would choose to create the human condition in a way that fundamentally fools us, one would need a very good reason indeed before believing that God is fooling us about free will. Such a reason does not exist, and it’s even hard to imagine how such a reason might be like. (Please observe that God’s hiddenness is not similarly fooling us into disbelieving in God, but only avoids forcing us into believing in God.)
    Further, I think the whole discussion of libertarian versus compatibilist free will is misplaced, because the very concept of compatibilist free appears to be based on a double equivocation:
    First, when theists (or the folk in general) speak of freedom of will they mean freedom of choice, whereas compatibilist free will refers to freedom of action. It’s on this equivocation that Frankfurt cases rest. So, for example, one can physically interfere with the freedom of a free agent to act on his choice, even down at the level of the firing of the first neuron in her brain – but one can’t physically interfere with her actual freedom of choice. Indeed the Principle of Alternative Possibilities that Frankfurt specifies, namely “a person is morally responsible for what he has done only if he could have done otherwise”, is false. A better rendering of the principle would be something like “a person is responsible for what he has done as a result of having chosen to do so, only if he could have chosen differently”. But this is still not good enough, because, like freedom, responsibility is a property of choices and not of actions. So, for example, imagine somebody holding a gun and making the choice to shoot somebody. Should the gun malfunction keeping him from shooting his intended victim, or should a meteorite smash his arm an instant before his finger pressed the trigger, or should he suffer a stroke which would scramble the neural signals to his hand holding the gun – any of these events would hinder the action of shooting, but would change nothing in his responsibility for having made the choice to shoot. So, finally, here is what I hold to be the right PAP: “a person is morally responsible for a choice he has made only if he could have chosen differently”. So, if a stroke keeps a person from even making the choice, then trivially that person is not morally responsible. If being drugged or ill interferes with one’s natural disposition to choose, then, as we all recognize, that person is less morally responsible for his choices.
    The second equivocation concerns the definition of compatibilist free will, namely that a person possesses compatibilist free will if that person is not physically restrained from acting in accord with his choice (itself informed by his desires and wishes and whatever) – since that definition only makes sense in an indeterministic world. A deterministic world is just a big machine. It makes no sense to speak of one part of the machine being “restrained” by another part. Rather each part of the machine is caused to move or to stop by other parts in accordance to the initial state of the entire machine and its deterministic laws of motion. Or, to put it differently, the very concept of “being restrained” presupposes the possibility of not being restrained, and thus makes no sense in a deterministic world.
    After this rather long introduction, let me offer some comments on your argument.
    I wonder about the justification of premise 5. I understand the cosmological argument entails that the causal event that creates the world is not a brute contingency, but this does not commit the theist (who embraces the cosmological argument) to the view that nothing is a brute contingency. Surely if God wants to create a world where brute contingencies obtain, then that will be so. Indeed, I happen to believe that God has reason to create the world in such a way that natural evils qua evils are mostly brute contingencies.
    As for premise 4. One will rather naturally hold that exercises of agent-causal choice are not brute contingencies, precisely because a contemplating agent is choosing. Therefore, the fact that the respective agents don’t exist in the actual world but only in a possible world being visualized by God, does not make their choices brute contingencies. If there were brute contingencies then true counterfactuals of creaturely freedom concerning creatures who do not exist would be impossible, no? Thus, it seems to me, if one accepts premise 3 one will probably reject premise 4.
    And finally a rather deep matter with premises 2 and 3. Being a theist I feel committed to a theory of truth grounded in God. Thus I believe that there are no truths independently of God, and indeed of God’s will. For this and other reasons I do not understand omniscience as God knowing all truths there are, but as God knowing all truths that God wants to know. And since I see no reason why God would want to know the truths about counterfactuals of creaturely freedom in possible words (never mind of creatures who don’t even exist), I also reject these premises. In general it seems that your argument can only be accepted by molinists.

    January 30, 2013 — 12:04
  • ryanb

    Kenny,
    I’m also wondering about (4). Is there something wrong with saying that the counterfactual claims of premise 2 are true, but not brute, since their truth is explained by Lewisian closeness relations between the counterfactual worlds containing these free creatures and the actual world?

    January 30, 2013 — 12:14
  • Kenny Pearce

    Danielos – The considerations about libertarian free will you give do not explain why theists should be any more likely to be libertarians than non-theists. They are all reasons available to the non-theist. Also, sophisticated compatibilists are able to make a distinction between freedom of will and freedom of action. For instance, Gary Watson has a paper called Free Action and Free Will (also reprinted in his Agency and Answerability). Most compatiblists today do not accept the definition of compatibilist free will you gave, and your definition is too simplistic even for many historical compatibilists.
    My premise 5 is a premise in some popular versions of the cosmological argument, namely (most of) those versions known as the ‘argument from contingency.’ It might also be thought to be entailed by certain versions of the first cause argument, or by theism more broadly, but I needn’t endorse anything like that. You say “if God wants to create a world where brute contingencies obtain, then that will be so.” But this is surely false, for in this case there would be an explanation of the contingent facts in question, namely, they obtain because God wished them to.
    I don’t understand what you are saying about premise 4. The counterfactuals of creaturely freedom pertaining to merely possible creatures are not made true by any exercise of agent-causal power, because the agent doesn’t exist to exercise that power.
    Your last comment is also puzzling. My argument cannot be accepted by Molinists, because Molinists reject its conclusion. In general, non-Molinist libertarian theists will reject (2). If that’s all you mean, then I agree with you.
    ryanb – Many people have the intuition that some kind of grounding in actual existents is needed, but modal realists, such as Lewis, of course reject that, since they think that merely possible things are just as ‘real’ as actual things. However, even if one says this, many libertarians want to say that, at least some of the time, there will be equally close worlds where opposite choices are made. This does perhaps take some of the bite out of (2), since we wouldn’t have metaphysical vagueness: the facts about the worlds would be perfectly determinate, but would underdetermine the counterfactuals, because of facts about language. Still I find it odd to suppose (indulging for a moment in a literalistic fiction) that God doesn’t know whether Adam would still have sinned if the Tree of Knowledge had been placed two feet to Adam’s left of its actual location.
    In any event, I don’t see how to get from this to the claim that there is a determinate fact of the matter but it’s not brute, but maybe I’m missing something. (Of course, if modal realism is true, then there is a really existing agent who actually made the choice, so I suppose that will help.)

    January 30, 2013 — 13:14
  • Dianelos Georgoudis

    Kenny,
    I think the non-theist finds herself in a very different position. Whereas the theist is committed to the view that God has given her sufficiently reliable cognitive faculties and is certainly not fooling her in matters as fundamental as her sense that she has the power to chose among alternatives (never mind all the issues relating to moral responsibility), the non-theist can and typically does point out how often our intuitions have been proven wrong. Indeed, contrary to the general impression, I find that when doing metaphysics the typical non-theist feels much less restricted by her sense of plausibility than the theist.
    I am curious to see how sophisticated compatibilists deal with freedom of choice in a deterministic world. I guess they construe “choice” as being a special kind of action deep within the human brain. But I will read Gary Watson’s paper.
    You write: “[It is surely false that if God wants to create a world where brute contingencies obtain then this will be so], for in this case there would be an explanation of the contingent facts in question, namely, they obtain because God wished them to.”
    Suppose God wants to create and flip a random coin. Surely there is no logical impossibility in God realizing that. That the coin is created and flipped has an explanation and is not the brute contingency. The brute contingency is that it came up heads.
    I have a special interest in this point, because on the best theodicy I can think of that’s exactly how most natural evils obtain. Namely randomly and without an individual moral justification. Just as it prima facie seems to be the case. So if you see any intrinsic problem with that idea please let me know.
    “The counterfactuals of creaturely freedom pertaining to merely possible creatures are not made true by any exercise of agent-causal power, because the agent doesn’t exist to exercise that power.”
    Right, I agree with that. Rather they are made true by God knowing how the agent would exercise that power if the agent existed and found herself in a particular state of affairs. Actually I fail to see why the existence or not of the agent in the actual world matters. God, on this line of thought, is supposed to know all true counterfactuals of creaturely freedom without having actualized any one of the possible worlds considered, and thus without the benefit of any existing agents.
    “My argument cannot be accepted by Molinists, because Molinists reject its conclusion.”
    On the other hand I thought only a Molinist would accept premise 2.

    January 30, 2013 — 14:54
  • Kevin

    I haven’t read the comments above (so maybe this is already addressed), but I wanted to ask, Kenny, but does 2 presuppose Molinism? If so, isn’t this more a reductio of Molinism than an argument against libertarianism?

    January 30, 2013 — 14:58
  • Kenny Pearce

    Danielos – If we accept that facts made true by exercises of agent-causal power are not brute, then we should also accept that facts made true by indeterministic physical causation are not brute. The argument will still go through, because the counterfactuals about merely possible people can’t be made true in this way either. (For a version of the PSR consistent with agent-causation and physical indeterminism, see Pruss’s book on the subject.)
    If God makes the counterfactuals true by knowing them, doesn’t that violate libertarianism? Isn’t this a form of theistic determinism?
    Kevin (and Danielos, some more) – I don’t think (2) presupposes Molinism. What I do think is that most philosophers who accept theism and libertarianism while rejecting Molinism see the issues clearly enough to reject (2), as they are required to if they accept the other premises. But I regard (2) as having independent support, hence I regard its rejection as a cost incurred by these views. Indeed, I think the intuitive support enjoyed by (2) is among the main attractions of Molinism. (I also think that the requirement that we reject (5) is the biggest cost of Molinism, hence my interest in this argument.) If (2) is false, then God may not know with certainty what would have happened if he were to act differently in very minor ways, and this strikes me as quite problematic.
    Here’s another way of putting the matter. There are three views on the subject:
    (a) Theological Compatibilism
    (b) Molinism
    (c) Non-Molinist Libertarianism
    According to (a), God’s knowledge of CCFs is ‘free’ knowledge (i.e. knowledge of God’s own creative decisions). According to (b) it belongs to the special category, middle knowledge. According to (c) God has no such knowledge. If you reject compatibilism but also don’t accept (or can’t make sense of) this special category of middle knowledge, then you are stuck with (c). I regard it as being ‘stuck’, because it involves denying that God has a sort of knowledge which, it seems to me, he ought to have. For myself, any inclination I have to accept Molinism is due to a prior inclination to accept (2). I don’t imagine that I’m the only one for whom the order of support goes in this direction.

    January 30, 2013 — 15:24
  • I tried to post this earlier, but I guess it didn’t go through. 2 doesn’t follow from divine omniscience. Omniscience plus their being facts about what possible free creatures would do in any given circumstance does give you 2. But it’s controversial at best whether there are any such facts.

    January 30, 2013 — 19:52
  • Kenny Pearce

    Justin – I agree that one plausible definition (perhaps the most plausible definition) of omnipotence is knowledge of all truths, and that it is controversial whether such counterfactuals of freedom could be truths. I said it was plausible that omniscience requires knowledge of such things, and I stand by this (relatively weak) claim, for several reasons: first, it’s plausible (though controversial) that there are such truths, in which case a certain plausible definition of omniscience will entail that God has such knowledge. Second, it seems to me that many religious believers find it intuitive that God has such knowledge, and one way for something to be plausible is for it to capture intuitions. Third, it seems to me that such knowledge might be important to guiding divine actions in ways that are important to an adequate doctrine of providence. (How can God make good decisions if he doesn’t know what the outcomes of alternative courses of action would have been?) I don’t take any of these to be anything like a proof that omniscience requires (2), but I do say that any one of these three reasons, individually, would be enough to render it plausible. All three of them together would render it extremely plausible, were it not for certain counterarguments.

    January 30, 2013 — 20:00
  • Dianelos Georgoudis

    Kenny,
    “If we accept that facts made true by exercises of agent-causal power are not brute, then we should also accept that facts made true by indeterministic physical causation are not brute.”
    I disagee. Facts made true by exercises of agent-causal power are not brute because they are caused by the agent’s personal will. And an agent’s personal will is self-caused and self-explanatory. It’s sovereign and creative, adding something genuinely new to reality. I think Hume is right to observe that through freedom of will humans are small gods.
    On the other hand, facts made true by indeterministic physical causation are strictly speaking not caused by anything and admit of no explanation. At some point they are indeed brute contingencies. Now it is true that the physically observed effects of indeterministic physical causation are indistinguishable from the effects of agent causality – and that’s exactly why agent causality does not conflict with physical law. But metaphysically they couldn’t be more different; in fact they are opposite.
    A final thought. Since we (and thus science) can only observe the physical effects of physical indeterminism, and since such effects are indistinguishable from the effects of agent causality, we don’t know to which degree natural phenomena are caused by physical indeterminism rather than by God’s will. Given God’s special providence the theist naturally assumes that at least in part what causes natural phenomena is God’s will. Thus the structure of modern physics very nicely comports with both methodological naturalism and with theism’s fundamental understanding of the special providence by which God guides the evolution of the world.
    (Of course in everything I write above I assume that both indeterministic physical causation as well as libertarian agent causation exist. My purpose is to show that their existence is consistent with our experience of life and beautifully fits with both classical theism and with modern scientific knowledge about the order present in physical phenomena.)
    “If God makes the counterfactuals true by knowing them, doesn’t that violate libertarianism? Isn’t this a form of theistic determinism?”
    I don’t think so. Given my understanding of what “molinism” means I think that molinism is false but I don’t see any incoherence in it. Here I take it is molinism’s idea: “There is nothing logically impossible in God having as clear an apprehension of possible worlds as would be the case were they actual. Thus God’s omnipotence entails that God can know how a free agent would freely choose in any possible state of affairs – without the benefit of that agent or state of affairs actually existing. And God’s omniscience entails that God does have this knowledge about how all agents would freely choose in all possible worlds. The world we find ourselves in is just one of these possible worlds made actual by God. The fact that God knows how each one of us will freely choose does not determine how we shall choose, for we ourselves determine that. Simply God already knows what we shall freely determine, for there is no logical contradiction in God knowing that.”
    The reason I believe molinism is false is both theological and metaphysical. On the metaphysical side, unless I have a good defeater I am committed to believing that free will is exactly as it seems to me, and therefore sovereign and having the power to create something genuinely new in the world, something that would surprise even God. But if that’s the nature of human free will then middle knowledge is metaphysically impossible. On the theological side I can’t see why God would want to have complete middle knowledge (for the greatest being I can conceive is not a control freak), whereas I can see why God would not (namely for making human free will a greater thing, but also out of respect for human suffering).
    I’d like to close by discussing the relationship between theistic determinism and physical determinism. On theism, physical determinism implies theistic determinism, but not the other way around. On physical determinism all physical facts obtain by the mathematical evolution of an initial physical state. If in such a world God supernaturally interferes and changes some fact then physical determinism in that world is destroyed but theistic determinism remains true. Now I happen to believe that neither physical nor theistic determinism exist in the actual world, and that physical facts are partly caused by God’s will, and thus determined by God, and partly the effect of physical indeterminacy, and thus caused by nothing. Suppose that I am wrong and all physical facts are determined by God, thus making theistic determinism true. This won’t make physical determinism true, except in the trivial case that God chooses to determine physical facts by following mathematical rules.

    January 31, 2013 — 2:17
  • Kenny: Thanks for the response and the interesting argument. A few thoughts about your reply:
    Is it plausible that there are true counterfactuals of creaturely freedom (CCFs)? Perhaps for some. I used to think there were CCFs, but that was before I gave it any thought. Now it’s not at all obvious that there could be such truths. You might be right that many believers find it intuitive that there are CCFs. But then again I doubt most believers have even thought about it. Once they do, they might, like me, find the idea that there are such truths more puzzling than plausible. Finally, while I certainly agree that middle knowledge would be providentially beneficial, I don’t see that that’s a reason to believe in it. If an adequate account of divine providence requires something impossible, then perhaps we should lower our standards of adequacy.

    January 31, 2013 — 6:06
  • Kenny:
    i. It’s a moral failing in x to have a character C such that there are circumstances D that are not themselves evil and yet x is determined by C and D to sin.
    ii. If the first sinner’s sin was determined by character and circumstances, then either the first sinner was in evil circumstances or had a moral failing.
    iii. The first sinner wasn’t in bad circumstances when sinning.
    iv. The first sinner didn’t have a moral failing prior to sinning.
    v. So, the first sinner’s sin wasn’t determined by character and circumstances.
    I am inclined to think that the restriction to “circumstances that are not themselves evil” in (i) might need to be removed. But I could imagine someone thinking that even a person without moral failings might be determined to sin under horrendous torture. (I don’t buy that. If a torture or other temptation is sufficient to make a person without moral failings to sin, it is the sort of circumstance that takes away responsibility.)
    But now typical Western theists will want to say that there is nothing evil prior to the first sin. So the first sin must have been not determined by character or circumstances.
    Now if you were only arguing for compatibilism, this wouldn’t be a problem for your argument. But your argument, were it sound, would establish that creatures cannot be free and undetermined.

    January 31, 2013 — 8:50
  • Dan Johnson

    Kenny, in response to your question about Plantinga’s view:
    Plantinga told me one time that he is “more of an Arminian Calvinist.” I was rather surprised, because that sure sounds like a contradiction to me. But he really does think of the Arminian movement, historically speaking, as a kind of outgrowth of or sub-tradition within the Calvinist movement; and that is a fairly plausible take on the actual history of the movement, it seems (though my knowledge is limited).
    With respect to the free will defense being available to Calvinists who are compatibilists: Plantinga thinks that Calvinists can think that the sort of ability that libertarians think is necessary for free will is metaphysically possible, and that even if it isn’t necessary for moral responsibility, it may still be valuable by virtue of being a kind of imitation of God. So he thinks the Calvinist can still use his defense.
    I initially thought as you do (that a Calvinist can’t use the free will defense), but a conversation with him has made me less sure. I guess I still think that the sort of ability that libertarians think necessary for free will is not metaphysically possible, but whether you are committed to thinking that will depend on your motivation for being a Calvinist.

    January 31, 2013 — 9:24
  • Kenny Pearce

    Justin – I’m not saying there are no arguments against it. As you say, many people believe it initially. I would add to this that none of the arguments against it are decisive. That’s enough for me to call it ‘plausible’.
    Alex – I agree that this line of thought creates severe problems of the sort you point out, especially for the specifically Christian theist who wants some sort of doctrine of the Fall. On the other hand, given that a literal historical reading of the early chapters of Genesis is untenable, I’m not totally sure what to make of the theology here, and I’m also not sure that, even on a literal historical reading, the story would really be anything like a solution to the problem of the origin of evil in a world created by a good God. But I will say this: the theist who thinks that creaturely free action of the undetermined sort is at least possible (even if not required for moral responsibility, and even if not had by us) has resources to say some helpful things on this topic which are denied to someone who accepts my argument.
    Dan – It’s true, as a matter of historical fact, that Arminians in the strict historical sense (as opposed, e.g., to Wesleyans) are part of the Reformed tradition, and there’s a lot more to Reformed theology than just the so-called ‘doctrines of grace’ (i.e. the five points) on which Calvinists and Arminians disagree. It’s also true that Arminianism is radically misunderstood (relative to the correct historical usage) by most American Protestants today. (In this connection, let me recommend the blog and writings of theologian Roger Olson.) In any event, I think you are right that what a Calvinist says about this depends on what the motivation for being a Calvinist is. I think having those brute contingencies, undetermined by God, floating around, would bother most, but perhaps there is no logical inconsistency. A theology student once told me that some scholars attribute to Calvin himself the view that Adam and Eve were libertarian-free before the Fall; I don’t know about the interpretive claim (I’ve never read the Institutes), but I can see how that view could fit pretty well with the rest of Calvinist theology, and it would also allow the Calvinist to avail herself of the FWD.

    January 31, 2013 — 10:39
  • Kenny:
    I don’t need any literal reading of Genesis here. All I need is the broad idea that all evil other than sin is posterior to sin and that there was a first sin. (The last follows from the plausible claim that there have only been finitely many sins.)
    Actually, I think I only need a weaker claim than that all evil other than sin is posterior to sin: That all moral failings are posterior to the first sin. For I think that any circumstances (e.g., severe torture) that would determine to wrongdoing a person without moral failings are going to be exculpatory. But sin is, by definition, not exculpated.

    January 31, 2013 — 12:34
  • Kenny,
    You ask: “How can God make good decisions if he doesn’t know what the outcomes of alternative courses of action would have been?”
    I can make good decisions without knowing what the outcomes of alternative courses of action would have been. All I need for good decisions is conditional (causal) probabilities. 🙂
    I take it you want to pack a lot into “good”, maybe “the best”.

    January 31, 2013 — 12:47
  • “Now, I suppose there are some reasons for this.”
    Another reason might be that there is a large overlap between non-naturalism and theism, and one motive for being a compatibilist is that it’s hard to fit libertarianism with naturalism (pace Kane).

    January 31, 2013 — 12:59
  • Kenny Pearce

    Alex – I agree that your point does not require a literal reading. Here’s how I see it: your view would need to be supported by either reason or revelation (or, preferably, both). Not every non-literal reading of Genesis (together with the other Biblical texts that refer to the Genesis story) supports your view. I reject the literal reading of the Genesis story, but I don’t presently have anything to replace it with, so I don’t know whether revelation ultimately supports your view. A reading of Genesis that did support your view would be easier to square with St. Paul’s theology (and the broader Christian tradition) than one that didn’t, so I concede that your view enjoys some prima facie revelational support.
    As far as reason, I don’t think anyone has a philosophically adequate account of why God allows evil at all. That’s not to say that there hasn’t been any progress, but there hasn’t been anything like a solution. Furthermore, I think Robert Adams’ arguments that God might not create the BPW can be used to make plausible the claim that God would create beings who are essentially such that they start out life in a morally defective state of character. (This could be combined with soul-making theodicy in interesting ways.) I think I might very well be such a creature. Since this view is on the table for me, I’m not in a position confidently to endorse your argument.
    In short, the denials of your iii and iv are both epistemically possible, from my perspective.
    On ‘good decisions,’ I mean the sort of advanced planning, control, etc. required by an adequate doctrine of providence. Although it might have looked like it, my question was not purely rhetorical; I’m open to the possibility that it might have an answer. I just find it puzzling.
    I completely agree with your last point. (Although I’m not sure Kane disagrees with the claim that it’s hard – after all, it takes him a whole book to work it out! But he certainly thinks it’s possible, and that’s a claim most other philosophers reject.)

    January 31, 2013 — 13:20
  • Dianelos Georgoudis

    “How can God make good decisions if he doesn’t know what the outcomes of alternative courses of action would have been?”
    Detectives trying to solve a case, scientists searching for mathematical patterns within physical phenomena, mathematicians trying to find laws within abstract rules, philosophers trying to find the deepest orders in the human condition, the common person trying to make ends meet – they all try to discover a good solution, and they do this basically by trying out hypotheses and choosing the one that works best. But this intellectual process does not exhaust the power of free will, nor does it express its essence. Indeed such quests for a solution are mechanizable. In principle one could built a machine, feed it the relevant data and criteria of success, and have it mechanically search for the best solution. For all the above purposes then free will (by which I always mean libertarian free will) is not required.
    I’d like to suggest that apart from the issue of moral responsibility, what free will gives us is the power to create, that is to bring into reality genuinely new things that weren’t there before, things that delight all including God.[1] A good case in point is the work of the artist. Contra Plato the artist does not copy from nature, nor does the artist have access to some ideal realm of artistic forms, searches there, and makes a physical representation in paint or sounds of what she has discovered. Rather what is significant in the artist’s work is creation itself. What gives value to the work of the great artist is not copying or reflecting a pre-existing value, but that the great artist had the power to *create* that value. And even though the great artist follows rules, by her work she also modifies and builds on them. If there is an ideal world of artistic forms the great artist expands it.
    Incidentally, I am claiming the above not in the abstract, but testifying from some personal experience with painting. At those magic moments where everything falls together it seems one forces the paint to be right, by one’s own will one shifts the whole of reality in a way that makes the colors work and shine. The value of the artistic work is not grounded in something pre-existing but in one’s will. By the power of one’s will one infuses power to the work.
    Coming now to the initial matter Kenny asks about, I think God’s condition in the creation of the world is much closer to the artist’s creative condition than to the detective’s , or the scientist’s, or the philosopher’s condition of discovery. The goodness of the world is not grounded in that God contemplated all possible worlds and actualized one which God valued to be second to none. Rather the goodness of the world is grounded in that God created it.[2] And, given free will, it is also grounded in our own creative participation in the world. (I think it is mainly in this sense that creaturely free will is fundamental in creation.)
    Which brings me to the issue of the place of free will in our own condition (as well as to my beloved soul-building theodicy): I order to create one need not be the rare great artist. Rather in the defining moral dimension of the human condition we are all called to be creators of the good life. The universe and our good choices are like the painter’s canvas and colors. And as the great artist is recognized in her great work, thus the great person is recognized in her great life.
    Since the start of the powerful discoveries of the physical sciences in the age of enlightenment I find that mechanical customs of thought have invaded our cognitive processes in many fields where they don’t belong. An important case in point is ethics, where for no clear reason moral philosophers try to discover some rule-based system by which to measure ethical value. This would be akin to an artist or a critic trying to discover a rule-based system by which to do measure the value of artistic work (and this is not only a waste, but keeps them from understanding art). I find that’s not at all how it is, and that when one contemplates one’s own everyday moral life one apprehends the following: The good choices illuminate one’s being and bring it closer to God (the “treasure in heaven”), but exactly which choice will thus shine is not fundamentally a matter of conformance with a given set of rules, but a matter of one’s interior self-transcendent loving. Thus, to mention an evocative example, what grounds the goodness of Mary Magdalene’s choice to use the expensive oil to wash Jesus’s feet is not conformance with an ethical rule, but her loving Jesus. In ontological language this becomes trivial: One becomes like the God of love by letting love shape one.
    As for the visible world: What one chooses and what effect one’s choice ends up having in the world is not important, only that one’s choice is moved by Christ-like love is. On the other hand one can safely trust that God orders the visible world in such a way that one’s choices for love will ultimately have a good effect, that out of a good spiritual tree good physical fruit will issue. Even when this would seem not to be the case, as when one for love does not resist evil. The world is God’s garden and is not of our concern; nor do we have any power over the world. No matter how splendid it is, we are not of the world. The world is there to give us freedom and to be enjoyed, and not to be worried about. This is why our Lord and also all the great moral teachers teach non-attachment, that we should work for secret love and not for discernable results.
    [1] A sense by which there is no evil in God’s creation, is that only good can be created. Only good can enlarge reality and become a part of it; all evil is transient.
    [2] The theodicist does not try to see the value of the world and thus understand why God has created it, but to see why God has created the world and thus understand its value.

    February 1, 2013 — 5:58
  • Mike Almeida

    Is it plausible that there are true counterfactuals of creaturely freedom (CCFs)? Perhaps for some. I used to think there were CCFs, but that was before I gave it any thought. Now it’s not at all obvious that there could be such truths.
    It would be strange, I think, if there were no such truths. We assume that there are such truths all the time, especially in moral reasoning. Since the world is very likely indeterministic, the only candidates for free action are undetermined free actions (small exceptions, perhaps, in some contexts).
    1. If I hadn’t shot Smith, he would have (freely) killed Jones.
    2. If Jones hadn’t been arrested, he would have (freely) robbed another bank.
    3. If I had flipped the coin, it would have fallen tails.
    Worries about the truth of these counterfactuals seems to come from their chancy context. But this is a mistaken reason to doubt their truth. Say the coin is fair, and consider the description ‘tossing the coin’. Under that description, there are lots of worlds in which the event results in tails and lots of worlds in which it results in heads. But that’s because the reference class of the event is very large under that description. There are many reference classes for this event. What I did was toss the coin, but I also did something much more specific: I tossed the coin at a certain velocity, under certain weather conditions, at a certain angle, etc. Described in this way, what are the chances that it falls tails? It will be nearly certain that it does (barring any weird quantum events occurring). Similarly, for free, undetermined actions of moral agents. The chances that Smith tells the truth might be 50/50. But the chances that Smith tells the truth in the specific physical, psychological and social conditions he happens to be in will go to something close to one. What has to be kept in mind is that chance is time and description dependent. As the description of the event fills out and the time of the event gets closer, the chances of some specific outcome occurring goes up close to 1. So there should not be any problem from chance on counterfactuals of freedom.

    February 1, 2013 — 9:36
  • Heath White

    Mike,
    The theory you are describing is that CCFs of the form “if p were the case, q would be the case” are true just in case the conditional probability of q on a very specific p is pretty close to 1. I do not know if your point is to defend Molinism, but that theory will not do the work the Molinist wants it to do. For it will not support modus ponens in all cases: there will be some (maybe tiny) fraction of cases where the conditional probability of q on p is high, and therefore (on this view) “if p were the case, q would be the case” is true; God makes it the case that p; but q doesn’t occur. Since God has to organize his universe taking into account a bazillion of such CCFs, some non-negligible number of his decisions will be suboptimal. These mistakes would ramify throughout the creation and essentially we would have an Open Theist view where God is a really good probability-estimator and does the best he can.
    This view is a reasonably good argument that some CCFs are highly probable. But that is not the same as being true.
    I share Kenny’s skepticism, or doubt anyway, about true CCFs.

    February 1, 2013 — 15:34
  • Kenny Pearce

    Heath – For the record, Mike was quoting the commenter Justin Capes, not me. I have a strong initial inclination to think that CCFs all have truth values, but the responses to the argument I gave which I take most seriously are (a) rejecting (2), or (b) endorsing the conclusion. It would be awesome if there was a way to deny (4), but I’m not seeing it. I am quite strongly committed to (1) and (5). So I suppose you could describe me as doubting, if all you mean by that is that denying (2) is a live option for me.
    On the other hand, I think your comment brings up a good reason for supposing that CCFs have truth values: I’m not sure I understand how a proposition with no truth value can have a probability of being true. I suppose it could be that P(~p) = .1, P(p) = .1, and P(p is indeterminate) = .8, or something like that (where we could be talking about any number of kinds of probability). But if none of them were determinately true, what would the probability mean? I can also understand, ‘if p were true, then the objective probability that q would be .9,’ but I don’t think that makes P(if p were true, q would be true) = .9. Perhaps the fact that (at least some of) the ones with true antecedents have determinate truth-values is enough to get things off the ground, since that ensure that there are some circumstances in which a given counterfactual would have had a determinate truth value. The whole thing is quite puzzling to me.
    I believe Robert Adams has a view where some CCFs have determinate truth-values and others don’t. Maybe something like that can work.

    February 1, 2013 — 16:32
  • Kenny:
    I think that when we say things like: “Were p to hold, q would likely hold”, we’re neither affirming a counterfactual whose consequent is a probability claim (A→(P(B)>0.75), say), nor are we affirming the probability of a counterfactual (P(A→B)>0.75, say). I think what we’re doing is something like affirming that there is a high conditional probability (P(B|A)>0.75, say), or, better, a high causal conditional probability.

    February 1, 2013 — 20:58
  • Dianelos Georgoudis

    Mike,
    “I tossed the coin at a certain velocity, under certain weather conditions, at a certain angle, etc. Described in this way, what are the chances that it falls tails? It will be nearly certain that it does (barring any weird quantum events occurring).”
    As it turns out this problem has been analyzed both theoretically and experimentally. The micro-turbulence of the air appears to have a negligible effect, and thus *when the coin flip is stopped when coming down for the first time* a practiced magician can control the result pretty well, and a specially built coin flipping machine virtually perfectly. On the other hand, even though I don’t have any actual results, I feel pretty confident that if the coin is allowed to bounce repeatedly on a wooden table, things change rather dramatically. The thermal vibration of the coin’s molecules when it hits the table, the catastrophic breakage of wood fibers where the coin hits – all add small indeterministic disturbances to the path of the coin, which will quickly add up. My bet is that even the most precise coin flipping machine we can build will fail to produce results which are significantly removed from the statistical average of 0.5. And my bet is that a purely theoretical analysis will produce the same result. Incidentally there is a delightful article about the issue of coin flipping written by an eminent statistician and bearing the for a philosopher ominous sounding title “The Problem of Thinking Too Much”.
    “Similarly, for free, undetermined actions of moral agents. The chances that Smith tells the truth might be 50/50. But the chances that Smith tells the truth in the specific physical, psychological and social conditions he happens to be in will go to something close to one.”
    Here I strongly disagree. The physical processes in Smith’s brain which drive her decision are many billions of times more complex than the flipping of a coin, and in an intedeterminic system the loss of determinism does not grow linearly but exponentially. (I am not here referring to the loss of predictability which is also quickly lost in chaotic systems even when they are deterministic, but to the broadening of the probabilistic function in a complex indeterminitic system.)
    Consider a person you know being hungry and opening the fridge to pick something to eat. In the fridge there is a piece of cake and a banana. Knowing that this person has a sweet tooth and is not particularly health conscious you quickly estimate that she will choose the cake with a probability of 0.8. Now, given our best understanding of matter which is quantum mechanics, and assuming (as is virtually certain) that the probability of her choice is determined by the physical state of her brain, there is a fact of the matter about the value of the probability p she will choose the cake. There is to my knowledge no scientific reason whatsoever to suspect that p lies significantly far from 0.8, the value you estimated after quickly considering the issue and using only very rudimentary data. Actually there is a lot of reason for believing that our simple estimates about the probabilities of free choices are close to the factual probabilities, namely that we are capable of operating quite effectively in a social environment, which depends on us making fairly good such estimates most of the time.

    February 2, 2013 — 2:49
  • Mike Almeida

    The theory you are describing is that CCFs of the form “if p were the case, q would be the case” are true just in case the conditional probability of q on a very specific p is pretty close to 1.
    Heath,
    I hope I’m not doing that, since I just finished arguing with Alex that the probability of the counterfactual (A N-> B) is not evinced by the fact that P(B|A) is high. All that in Alex’s last post. I actually had only one goal, and that was to point out that a common objection to CCF’s is misplaced. That objection–commonly attributed to Bergmann in his ‘might-counterfactuals’ paper, but noticed long before that–was that CCF’s are all false if they’re advanced as ‘would-counterfactuals’. And the reason for that is that the consequent of CCF’s, given the antecedent, is still chancy. So, if my point were to take any formal description, it would be closer to something like this. There is an objection to CCF’s that goes as follows: (i) it is false that were a fair coin tossed it would fall heads, since (ii) were a fair coin tossed it would not certainly fall heads. My point was (sort of) to deny (ii). More exactly, my point was that every toss of a coin is a very specific event, in very specific circumstances and while ‘tossing a fair coin’ vaguely describes the event of, say, Sue tossing a fair coin with velocity v, angle a, in conditions c, etc., the latter event does not have a .5 chance of falling tails. It has something closer to a certain chance of falling tails. And similarly for all alleged chancy events that figure in CCF’s. So, to sum up, my aim was to respond to a common objection to CCF’s that appeals to the chancy consequents of CCF’s given their antecedents. I say that the consequents are not chancy at all given their antecedents.

    February 3, 2013 — 9:08
  • Mike Almeida

    I feel pretty confident that if the coin is allowed to bounce repeatedly on a wooden table, things change rather dramatically. The thermal vibration of the coin’s molecules when it hits the table, the catastrophic breakage of wood fibers where the coin hits – all add small indeterministic disturbances to the path of the coin, which will quickly add up. My bet is that even the most precise coin flipping machine we can build will fail to produce results which are significantly removed from the statistical average of 0.5. And my bet is that a purely theoretical analysis will produce the same result
    I’m less confident, I think. Ideally, we would have to have several such tables, identical in structure and current state. We’d have to have the coin tossed in exactly the same way, hitting the table in precisely the same way. In all such events, the coin will not fall tails at .5. It will rise to close to 1 or descend to 0. But it is impossible to replicate the circumstances exactly, and this is why it will appear to have chances of .5.
    Funny that you should mention magicians tossing coins, since I’ve used the example of an expert coin tosser (which surely there can be) who tosses fair coins to fall tails at will.
    Here I strongly disagree. The physical processes in Smith’s brain which drive her decision are many billions of times more complex than the flipping of a coin, and in an intedeterminic system the loss of determinism does not grow linearly but exponentially
    I think there might be some confusion here between epistemic probability and chance. For my case to be made, you’d need to have the same person, in the very same state S in the very same circumstances C, having the very same optons O. His choice of an option too will go to 1. The complexity of the situation goes to questions of epistemic probability, not to questions of chance.

    February 3, 2013 — 11:01
  • Mike Almeida

    Dianelos, this might make my view clearer. Recall Peter van Inwagen’s ‘roll back argument’ against indeterministic free will. Van inwagen urges that if there is genuine chance of .5 that a coin will fall tails in the specific circumstances C then it must be true that, if God rolls back history from a time after the coin fall tails to time t and then lets it go forward again, the coin will fall tails half of the time (do it an infinite number of times to get the right frequency). This is precisely what I’m claiming is false. And it is false in the human choice case too. Roll back history from after the time at which Jones decided to shoot Smith, and let history roll forward. Do it over and over and you will get exactly the same result over and over: viz., Jones shooting Smith. What is true is that the general event of ‘tossing tails’ has a .5 chance, since it includes lots of people tossing coins in various circumstances and in various ways. But no particular coin has ever been generally tossed (whatever that would be). Specific coins are tossed in specific circumstances and yield specific results with near certainty. Roll back history in the case where the coin is tossed on the table and bounces, then let it go forward. The coin will bounce in the very same way and land the same way.

    February 3, 2013 — 11:18
  • Heath White

    Mike, what I’m now puzzled about is why we wouldn’t call this determinism.
    Smart was very happy to say we “could have done otherwise” given a certain general characterization of antecedent circumstances, but that if we specify the antecedent circumstances down to the detail, they were guaranteed to come out the way they did. How does your view differ from his?

    February 4, 2013 — 21:26
  • Joshua Rasmussen

    Hi Kenny,
    premise 4: If incompatibilism is true and any counterfactual of creaturely freedom concerning a creature who does not exist is true, then that counterfactual is a brute contingency.
    Objection. A contingent counterfactual, C→A, is (ultimately) explained by a necessary counterfactual, C→B. For instance, Adam would eat the fruit if Eve were to offer it to him in the garden because necessarily, Adam would have a strong desire to eat the fruit if Eve were to offer it to him in the garden. (If you don’t think this counterfactual is necessary, just fill in more detail into the antecedent.)
    Also, maybe you addressed this in the comments, but hardly any cosmological arguments these days make use of the premise that there are no brute contingencies. (I’ve come to think the premise is defensible, but my sense is that it’s not a premise that many philosophical theists accept.)

    February 5, 2013 — 6:09
  • Mike Almeida

    Mike, what I’m now puzzled about is why we wouldn’t call this determinism. Smart was very happy to say we “could have done otherwise” given a certain general characterization of antecedent circumstances, but that if we specify the antecedent circumstances down to the detail, they were guaranteed to come out the way they did. How does your view differ from his?
    It is not determinism because there is still an extremely small chance that some quantum event affects some macro events, like a walk to the university. But the extremely small chance of quantum tunneling does nothing to alter the fact that were I to walk to the university, I would arrive safely. Analyses of counterfactuals on which that comes out false are just bad analyses, that’s been the intuition anyway.

    February 5, 2013 — 7:26
  • Dianelos Georgoudis

    Mike,
    All physical phenomena we observe have an indeterministic quantum nature. (Since we are having this discussion I assume we both assume indeterminism.) So it’s always the case that the current state of a physical system only fixes the probability distribution of the future state but not the future state itself. Now there are many cases where that probability distribution is a very narrow one around one particular state. In those cases it is often possible to find out the value of that state using the simpler math of classical mechanics. Our disagreement then is this: You suspect that the case of flipping a coin in the air and letting it freely come to rest on a wooden table is such a “classical” case where the probability distribution is very narrow and the future state of the coin is practically fixed. (This has nothing to do with epistemic probability; perhaps the future state is for all practical purposes fixed, but it is impossible to know which it will be.) I suspect that the probability distribution is not narrow, and indeed close to the assumption of randomness, namely a probability 0.5 for heads and 0.5 for tails.
    Since we can’t rollback reality and experimentally solve our disagreement, we can only study the question theoretically. Probably the critical part where significant indeterminacy enters the path of the coin is when it bounces on the table. What causes the phenomenon of elastic collisions is the repulsive force when the electrons of the coin molecules are brought close to the electrons of the table’s molecules. The loss of kinetic energy is caused by the increase of vibration on both the coin on the table, but also in the catastrophic breaking of wood fibers on the table where the coin hits – another place where minute changes can cause much larger effects. An analysis of the relevant quantum physics is probably extremely complex, even as a computer simulation, and given the absence of practical benefits I very much doubt the exercise has ever been done. What we can estimate right away is that if the quantum effects are sufficient to produce a 0.01% variation on the angle or speed of the coin after its first bounce, and given the rough microscopic form of the wood fibers as well as the effects of the thermal vibrations of the coin, the next time the coin will hit the table some 30 cm away it will encounter significantly different physical parameters. Anyway, absent a detailed computer simulation, I suppose the best way to resolve our disagreement is to ask a quantum physicist about her sense. Have you considered asking a quantum physicist colleague at the institution you work for?
    As for Peter van Inwagen’s roll back argument, the only interesting case is if the complex electrochemical machinery of the brain is such that it fixes a broad probability distribution of its future states. If it doesn’t and the brain for all practical purposes functions as a classical/deterministic machine which fixes the next choice then the whole issue is moot.
    Let me then assume that the brain’s indeterministic machinery can evolve into radically different future states of comparable probability, states which correspond to different choices. Inwagen argues that even then free will is incompatible with indeterminism. I agree with Lara Buchak’s criticism of Inwagen’s premise about the probabilistic nature of a free agent’s choice in the case of a roll back. Actually I see a general problem with using modality in the context of free agents. A human with my name (and DNA) who in a possible world makes a different choice than the one I made in the actual world is not obviously me. Personal identity – who one is – is affected by one’s choices. Going back to Inwagen’s argument, given that a free agent S makes a particular choice in the actual world, God can roll-back the world to the exact state (physical and spiritual) of the world just before that choice, but in that rolled-back world S does not exist. For the real S has made her choice. Or, to put it differently: Given that S did make a particular choice, if it were really the case that the world is rolled back with S in it, S would necessarily make the same choice again, notwithstanding the fact that the world (and her brain which actualizes her choice) is indeterministic.
    It seems that such roll back though experiments at best rest on shaky epistemic ground, and people appear to have different intuitions. I think there is a better way to think about free will: Consider the indeterministic evolution of the future probability distribution of the physical universe starting at some time at the past. This past state fixes the probability distribution of the state of the brain of any free agent S in the future, and thus fixes the probability distribution of her future choices. What would be considered to be like magic and to violate the physical closure of the universe (and is thus prohibited by God’s general providence in creating a naturalistic physical world) would be if S’s actual choices would be statistically extremely surprising (in the sense that a sequence of HHHHHHHHHH of random coin flips is more surprising than the sequence HTTHTTTHHTH even though both sequences are equally probable). All other actual choices by S are permitted. Thus physical law limits S’s freedom of choice, but on the other hand S’s freedom of choice can exist within these limits. Freedom of will gives S the power to move her brain’s state along any path which does not stray into virtual statistical impossibility. And given not only our experience of life but also what we know (or I think we know) of QM and the physiology of our brain, this is very significant freedom. For all practical purposes the typical human has the power to guide her life pretty much as she will.
    Incidentally, when I write above that physical law limits S’s choices, I mean it in the statistical sense, i.e. in the same sense that physical law allows but for all practical purposes limits a bar of gold from Fort Knox to quantum tunnel into my drawer. Should we observe people making choices which are too surprising given the physical state of their brain we would think that something supernatural is going on. But freedom of will, as we experience it and as classical theism says exists, does not require of evident supernatural effects since it is entirely compatible with the expected effects of the indeterministic structure of physical that modern science has uncovered. On the other hand there is not, and as long as the physical closure of the universe holds there can’t be, any observable evidence for free will. To use David Chalmers’s idea of a zombie world, there is a possible world in which all physical facts are identical to ours (including philosophers debating free will) and in which free will does not exist. But given our sense of free will and the absence of defeaters I think it is unreasonable to believe that our world is like that.

    February 6, 2013 — 17:18
  • WH

    3. There are true counterfactuals of creaturely freedom concerning creatures who do not exist.

    Question: Is (3) true only if there are merely possible beings?

    If yes, then (3) entails that there are merely possible beings. If there are merely possible beings, then possibilism is true (by definition). But possibilism is false.

    So, if “yes” is the answer to Question, then (3) is false.

    August 20, 2014 — 10:53
    • I agree that possibilism is false. I think the answer to your question is ‘no’. It is true, for instance, that if there had been a unicorn it would have been four-legged, but this fact is not made true by the existence of some merely possible unicorn. I don’t see why the same could not be true of facts about particular creatures which might have existed but do not.

      August 20, 2014 — 17:25
  • Michael Almeida

    Hi Kenny,

    Plantinga is no possibilist, but holds that some CCF’s are true. When you say that were the possible, non-actual being S in (maximal) state of affairs T, then S would freely do A, what you really mean is that were the creaturely essence S instantiated in state of affairs T, the instantiated essence would freely do A. All of this I’m sure you know; but it answers the question above. Creaturely essences are actual (indeed, they exist necessarily), and there is a unique essence for each possible being. These are what go proxy for those beings, for (at least some) actualists, in propositions about ‘possible beings’. Though I have no quarrel with the possibilists.

    August 21, 2014 — 15:25
  • Yes, that’s Plantinga’s theory. I didn’t figure it mattered much which theory we went for as long as we got the result that the CCFs can be true without the existence of merely possible creatures. Plantinga’s theory gets that result. Personally, I’m a little skeptical about the existence of these essences, and very skeptical about the semantic claim you make about ‘what I really mean’ by the CCF, but I don’t have settled views of my own on this, other than that any acceptable theory needs to get the result that Plantinga’s theory does get, namely, that there can be true counterfactual conditionals about things that do not exist at all.

    August 21, 2014 — 15:39
  • Looking back, I guess I did quantify over merely possible beings in premise (2) of my argument in the original post. The argument can and should be rewritten to avoid that.

    August 21, 2014 — 15:40
  • Michael Almeida

    Right, I didn’t intend to make a semantic claim; neither does Plantinga. He just wants to say that an actualist can make claims that are close enough to those that are explicitly about possible beings to substitute for them. The essences are nothing more than collections of properties so, if you have no qualms about abstract objects of this sort, then there is not much more to concern you. There is a question about Plantingan haecceities that people can legitimately doubt, I think, and that he needs for his individual essences. I’m inclined to think that there can be two distinct things that are qualitatively identical, so I’m inclined to think there are haecceities of some sort. But I could probably be talked out of it.

    August 21, 2014 — 17:15
  • Leave a Reply to Roland Elliott Cancel reply

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