Monday, December 10, 2007

Weakening transworld depravity

Assume Molinism. Plantinga has shown that if transworld depravity holds, then God could be justified in creating a world that would contain evil, since any world containing a significantly free creature is a world that would contain an evil, and it is worthwhile for God to create a world that contains a significantly free creature. Transworld depravity is the thesis that, given what the conditionals of free will in fact are, in any world in which there is a significantly free creature, that significantly free creature sins. This is intuitively a highly improbable thesis, as has been pointed out by more than one author. (Quick argument: Suppose that Jones faces only one choice in his life: a choice between doing an evil he enjoys only slightly and a great good he enjoys greatly. Suppose Jones has no bad habits and is clearheaded. Then the probability of his choosing the good is fairly high. But there are infinitely many possible creatures like Jones. The probability that all of them in a situation like that would choose evil is low.)

But there are weaker theses than transworld depravity that would get Plantinga the claim that God is justified creating a world that would contain evil. Here are some:

  • Given the conditionals of free will, any world containing a significantly free creature contains at least one free creature (perhaps a different one) who sins.
  • Given the conditionals of free will, any world containing at least a billion significantly free creatures contains at least one free creature who sins. But God would have good reason to create a billion significantly free creatures, especially if the majority of their choices were good.
  • Given the conditionals of free will, any world containing infinitely many significantly free creatures contains at least one free creature who sins. But God would have good reason to create infinitely many significantly free creatures, especially if the majority of their choices were good. (And, yes, for aught that we know our world could be like that. We do not know that there isn't an infinite number of significantly free creatures in our world.)
  • Given the conditionals of free will, any world containing infinitely many "strongly significantly free" creatures contains at least one free creature who sins. A creature is strongly significantly free if it is significantly free and the structure of incentives for one of its significantly free acts is such that neither choosing good is overwhelmingly probable nor is choosing bad overwhelmingly probable. There seems to be a value in strong significant freedom.
  • Given the conditionals of free will, any world containing at least aleph10000000 strongly significantly free creatures contains at least one free creature who sins. But God can have very good reason to create a world that contains at least aleph10000000 strongly significantly free creatures.
And while we likely are in a position to say that the thesis of transworld depravity is improbable, it is not clear that we are in a position to say that every thesis like one of the above is improbable.

44 comments:

  1. Thanks, Alex!

    Q. Smith's logical argument from evil is directly relevant here.

    Here's the nub:

    G. God exists and is omnipotent, omniscient, and wholly good.

    E. There is evil.

    A person is externally free with respect to an action A if and only if nothing other than (external to) herself determines either that she perform A or refrain from performing A.

    A person is internally free with respect to an action A if and only if it is false that his past physical and psychological states, in conjunction with causal laws, determine either that he perform A or refrain from performing A.

    A person is logically free with respect to a wholly good life (a life in which every morally relevant action performed by the person is a good action) if and only if there is some possible world in which he lives this life and another possible world in which he does not.

    It is possible to be internally-externally free but logically determined with respect to being morally good. This is the case with God, who is both internally and externally free but who does only good actions in each possible world in which he exists.

    Now the first argument.

    1. God possesses the maximally valuable consistent conjunction of great­ making properties.

    2. If it were intrinsically better to be logically free with respect to a morally good life than logically determined, and this logical freedom were consistent with God's omnipotence and omniscience, then God would possess this logical freedom.

    3. Logical freedom with respect to a morally good life is consistent with omnipotence and omniscience.

    4. God is logically determined with respect to a morally good life.

    Therefore

    5. It is false that it is intrinsically better to be logically free with respect to a morally good life than logically determined.

    Now, there can be a finite disem­bodied mind, for example, an angel, who is not causally influenced by its prior psychological states or anything else. More directly,

    6. It is possible that there is a nonomniscient mind x such that: for each possible world W in which x exists, and for each circumstance in which x is faced with a moral choice, x knows all the factual and moral truths he needs to know to make a correct choice.

    7. This mind x is neither causally determined nor causally influenced by any external or internal factors.

    8. Necessarily, if a perfectly free mind knows all the moral and factual truths needed to make the morally correct choice in any morally significant circum­stance in which he finds himself, then this mind will make the correct choice.

    9. If the individual essences of some necessarily good creatures were to be instantiated, the instantiations of these essences would always do what is right.

    10. If God is all-good and the proposition "God creates free humans and the free humans He creates always do what is right" is consistent, then any free humans created by God always do what is right.

    11. It is consistent that God creates free humans and the free humans he creates always do what is right.

    12. It is possible that: free humans who always do what is right exist without there being any natural evil, and if God creates these humans, he will not create natural evil.

    Now, the conjunction of (10), (11),(12), and (G) entails

    -E. There is no evil.

    Another argument is the following.

    A being properly eliminates an evil state of affairs if it eliminates that evil without either eliminating an outweighing good or bringing about a greater evil. A good state of affairs g outweighs an evil state of affairs e if the conjunctive state of affairs g and e is a good state of affairs.

    13. An omniscient and omnipotent [and wholly] good being eliminates every evil that it can properly eliminate.

    (13) is a necessary truth. If a state of affairs is eliminated by its actualization being prevented, and if a possible world is a state of affairs (a maximal state of affairs), then (13) entails

    14. God prevents from being actual any world W1 that contains evil if there is another creatable world W2 containing at least as much good as W1 and no evil.

    15. There is some possible creatable world W2 containing only God and an infinite number of necessarily good free rational creatures who perform an infinite number of good acts. (See the talk about angels above.) In other words, for any possible creatable world W1 containing evil and an infinite number of free rational creatures who perform an infinite number of good acts, there is another possible creatable world W2 containing no evil and an infinite number of necessarily good free rational creatures who perform an infinite number of good acts.

    -E. There is no evil. From G, (14), and (15).

    Adapted from http://qsmithwmu.com/
    a_sound_logical_argument_from_evil.htm

    Which points would you counter? I think an informed scholastic or Catholic philosopher would deny (3), (8), (10), (14), and (15), though I don't have enough DETAILED idea what would his reasons amount to. So, I would be very grateful for your help.

    Thanks again!

    ReplyDelete
  2. Yes, I think responding to Quentin Smith's argument would shed a good deal of light on the issues Alex has shared with us.

    Last year QS was saying to me that his argument that Vlastmil is calling our attention to has never had a good response to it. I also asked Al Plantinga then if he'd respond to QS’ argument we're talking about here and AP did not have time then to do a serious response to this argument, which he called "a formidable" challenge though.

    I do not remember the details now but when I pointed out to QS that if one holds the doctrine of divine simplicity his argument would fail. He responded by agreeing with me but then saying there is no plausible doctrine of divine simplicity at any rate. Then I called his attention to some literature like Barry Smith's A Most Unlikely God and its review (?) by Bill Vallicella. Now QS is reading some recent works on this doctrine by Brian Leftow and Jeff Brower of Purdue who has two new papers on the doctrine of divine simplicity.

    Hope some of you could see the connection between the doctrine of divine simplicity and QS’s argument that I fail to recall at the moment and that might be one way of handling QS's "formidable" challenge. But this route is not available for AP since he rejects the doctrine of divine simplicity.

    Cheers,

    Tedla

    ReplyDelete
  3. Smith just gives a more careful version of the argument that if being able to do otherwise is not needed for valuable freedom in the case of God, then it's not needed for valuable freedom in our own case.

    Anyway, the argument does not strike me as at all difficult to counter.

    Smith claims that it is possible to have creatures that are externally-internally free but not logically free. But this is logically impossible for a simple reason.

    Take Smith's definition: "A person is externally free with respect to an action A if and only if nothing other than (external to) herself determines either that she perform A or refrain from performing A." Well, suppose that x is not logically free with respect to an action A in some situation S, so that he logically has to do A. Suppose, too, that God has created x and put x in S. That x is in S logically entails that x does A, since x is not logically free. Hence, by causing x to be in S, God has determined that x do A, contrary to the definition of external freedom.

    So, lack of logical freedom entails lack of external freedom in the case of a creature.

    It also seems to me that logical freedom may be incompatible with internal freedom for beings that live in time. Smith defines internal freedom thus: "A person is internally free with respect to an action A if and only if it is false that his past physical and psychological states, in conjunction with causal laws, determine either that he perform A or refrain from performing A." But if the person is not logically free in respect of A, but logically has to do A in circumstances S, then his past physical and psychological states, in conjunction with causal laws, entail that he does A in S. I assume here that for a being that lives in time, a choice is not performed at the first moment of existence, but takes some time. This can be taken as a stipulative definition of "lives in time".

    I am also assuming throughout that an earlier state S together with a set of laws L "determines" a later state T iff S plus L entails T and S is causally prior to T.

    What have I shown?

    1. External freedom entails logical freedom in the case of creatures.

    2. Internal freedom entails logical freedom in the case of beings that live in time.

    God neither lives in time nor is a creature, hence neither argument applies to him. Thus God can have internal and external freedom without having logical freedom. But only uncreated beings that do not live in time can have this privilege.

    It might be that Smith can try to damage the argument by bringing in the notion of relevant entailment. I doubt that would help, though. For the entailments might be quite relevant. Presumably the way that a being would be logically unfree would be by having a nature that required good action. But then anybody who caused a being to exist with that nature would be relevantly determining the action, as long as the cause acted intentionally for that end, which presumably would be the case.

    ReplyDelete
  4. Dear Alex, thanks for these insights and your patience. Instructive, as always.

    Dear Tedla, I also think the argument by QS is important. In case you remember the connection with divine simplicity, let us know.

    ReplyDelete
  5. Hi Alex and Vlastmil:

    Yes, I discovered a note which was the subject matter of our conversation with him from last year, the one that I failed to recall yesterday. I post it entirely as I wrote it down to discuss with QS:

    "Response to Smith’s Response to Plantinga’s FWD

    1) I’m wondering if a theist couldn’t plausibly reject Smith’s claim that God is logically determined to necessarily do good [besides his distinctions between external and internal freedom.] with the following response:
    a. God is not logically determined to necessarily do good, rather God’s metaphysically determined to necessarily do good which means God cannot do anything contrary to his divine nature in such a way that being essentially good entails that such a being essentially performs only good actions.
    b.Metaphysically determined does not mean logically determined. A theist can grant that logically God can perform evil acts--though that sounds an unorthodox view-- but such acts remain in the realm of logical possibility without ever getting (actually) instantiated because God metaphysically cannot perform evil acts in virtue of his divine nature.
    c. So, a being can be metaphysically determined to essentially perform good acts without logically being determined to essentially perform good acts; God can be such a being.
    2) As for whether God could have created creatures that are necessarily good and hence necessarily perform good acts, always:
    a) Contingent free creatures/persons do not essentially/necessarily possess their nature of being essentially good --Quentin’s claim-- the way a divine being would because their nature is contingently possessed and such contingent properties cannot be logically necessary properties for such beings. Such free creatures could have been created possessing different properties such as being contingently free in such a way that at least one of their free choices could go wrong. This is a logically possible property for a contingent free creature to possess. Something along the preceding reasoning could be found in Ruth Marcus’ work according to Smith p. 103 in his Ethical and Religious Thought in Analytic Philosophy of Language. I discovered Marcus’s similar reasoning above after writing down my initial response as above. Smith prefers to work with Plantinga’s views on property exemplifications in his response to Plantinga’s FWD.
    b) Plantinga’s FWD seems to also assume some such distinctions between necessarily/essentially possessed properties such as God’s and those that are contingently possessed properties for free contingent creatures.
    c) These distinctions under (2) could reflect the distinction between being logically determined and metaphysically determined mentioned under (1) as follows: It’s both logically and metaphysically possible for contingent free creatures to go wrong, at least, once in their actions; whereas it’s logically possible for God to go wrong in his actions but then it’s metaphysically impossible for him to go wrong in any of his actions.
    3) An entirely different way of responding to Smith’s response to Plantinga’s FWD is open for a theist who is also committed to the doctrine of divine simplicity. Given a coherent doctrine of divinity simplicity, Smith’s argument fails to establish that there could be beings other than God who could be essentially good. Such essentially good creatures being non-identical to God, given that he’s essentially good and necessarily existing in whom his existence is identical to his other divine attributes, whatever that means, they cannot possess divine attributes such as essential goodness. If their existence is contingent and they possess whatever properties they possess contingently and in conjunction with the doctrine of divine simplicity, we can imagine that the beings could possess goodness contingently and hence accidentally. That sets them apart from God and such a possible way of conceiving about such creaturely beings enables a theist to respond to Smith hence undermining his crucial claim that such beings could exist. One can develop such a theistic response following Barry Miller’s work on divine simplicity that could easily be among the most sophisticated defense of divine simplicity in the last 10 years. The book I’ve in mind is: A Most Unlikely God by Barry Miller (1996).
    The above seem to be open ways for a theist to take in her consideration for responses to Smith’s response to Plantinga’s Free Will Defense."

    Something that QS also admitted in conversation about which he changed his mind later on was about avoiding saying that God's logically determined and instead adopting the idea that God's metaphysically determined. When I made these distinctions he immediately responsed by saying "I should have said God's metaphysically determined and avoided saying God's logically determined. This is just a post script that did not show up in the above post.

    Hope the above makes some sense. This is Alex's area of specialty and it's not to him I want to listen which of the above thoughts are open for a theist.

    Cheers,

    Tedla

    ReplyDelete
  6. P.S. A typo: In the last paragraph I was trying to say that "...it's now to him that..." not "it's not."

    Tedla

    ReplyDelete
  7. Another way to make use of divine simplicity here is this. If divine simplicity holds, then God is God's nature, and God is God's will. Hence, constraint by God's nature is in no sense external constraint. Nor is it constraint by anything other than God's own will. On the other hand, in the case of a creature, the nature is distinct from the creature, and so constraint by the nature is constraint by something other than the will. Hence God can be valuably free though constrained by his nature, but man cannot.

    ReplyDelete
  8. On further thought, at some point I've responded in print to arguments like Smith's (somewhat less sophisticated than Smith's, but what I say there is fully relevant to Smith's):
    http://AlexanderPruss.com/papers/NewFWD.html

    ReplyDelete
  9. Hi Alex:

    It's good to hear your more condensed and to the point remarks on the relevance of divine simplicity to QS's argument. If a theist can avail herself of divine simplicity, it seems that Quentin's argument fails. However, a theist who rejects divine simplicity, like AP, has to undermine Quentin's argument on other grounds like Alex's in his post above.

    Many thanks, Alex, for adding your thoughts to the divine simplicity resource for a theist and for pointing to your article on the very issues we're talking about here.

    Many thanks again,

    Cheers,

    Tedla

    ReplyDelete
  10. I've written up (and submitted for publication) my objection to Smith: http://alexanderpruss.com/papers

    ReplyDelete
  11. I think that there are some additional problems with Smith's argument. By "perfectly free mind" in (8), Smith appears to mean a mind that is neither externally nor internally determined. (It doesn't include not being logically determined,since the point of this part of the argument is that perfectly free minds are logically determined with respect to a good life.) If so, then God is a perfectly free mind, since he is neither internally nor externally determined. But since God is omniscient, he knows all the moral and factual truths needed to make the morally correct choice in any circumstance in which he finds himself. Then (8) requires that,necessarily, God always makes the morally correct choice in every circumstance in which he finds himself, i.e., God is not logically free with respect to a wholly good life. This directly contradicts (3).

    A further difficulty involves the interpretation of(2). If God is logically determined to do good and possesses the maximal consistent conjunction of great-making properties, then it isn't intrinsically better for him to be logically free than determined. But this leaves open the possibility that for finite creatures, it is better to be logically free than determined. (2) needs to be read "If it were intrinsically better for God to be logically free. . ." Then,(5)doesn't follow, for any being other than God.

    ReplyDelete
  12. Alex,

    Some minor questions concerning the end of the part 3 of your new paper (for which I am very grateful!).

    1. Do you think, like Swinburne, that

    if a mind x exists, and S is a circumstance in which x is faced with a moral choice, and x in S knows all the factual and moral truths he needs to know to make a correct choice in S, and x is not in S influenced by external or internal factors other than the mentioned knowledge, then x makes a correct choice in S?

    2. If yes, is it a strict entailment?

    3. Do you think, like Swinburne, that temptation is a necessary condition of a (strongly) significant free choice of a creature?

    4. What about the fall of Satan? Did God tempt him? Is not every tempting sinful? Does not Christ say in the gospel that temptations are not from God?

    ReplyDelete
  13. Vlastimil:

    Ad 1: No.

    Ad 2: Not applicable.

    Ad 3: No. What is required is incommensurability between some value one takes to inhere in the bad choice, insofar as one understands this value, and some value one takes to inhere in the good choice, insofar as one one understands this value.

    Ad 4: The best story I know is that given by St. Thomas.

    Best wishes,
    Alex

    ReplyDelete
  14. Alex, you write,

    (Quick argument: Suppose that Jones faces only one choice in his life: a choice between doing an evil he enjoys only slightly and a great good he enjoys greatly. Suppose Jones has no bad habits and is clearheaded. Then the probability of his choosing the good is fairly high. But there are infinitely many possible creatures like Jones. The probability that all of them in a situation like that would choose evil is low.)

    I don't off hand see the relevance of this for Plantinga. I'm not even sure I see the improbable event you're refering to. Look, all of the agents you invoke in the counterexample are, by hypothesis, TWD. If they are TWD, then the chances that all of them go wrong is not low--it is rather 1. That is, if the CFF's are true of each agent--and by hypothesis they are true--then it is certain that each will freely do something wrong, no matter what world God creates them in.
    The way you describe the relevant CFF's, they sound like some kind of probabilistic subjunctive. But for Plantinga they are definitely not probabilistic. So there is no problem here, I think.

    ReplyDelete
  15. Mike:

    I am arguing that the TWD hypothesis itself has low probability.

    ReplyDelete
  16. I am arguing that the TWD hypothesis itself has low probability.

    But who denies that? Plantinga is not arguing that it is probable, but only that it's possible. As you no doubt know, he is resonding to the logical problem of evil there. So all he needs is that such a world be possible. We can quibble about Plantinga's liberal modal epistemology, but given such an epistemology, we can reasonably assert that there is such a world.

    ReplyDelete
  17. Nobody is claiming that TWD has high probability, and certainly Plantinga is using TWD simply to counter the logical problem of evil. However, Plantinga's argument does little to counter the following claim: It is extraordinarily improbable that God and evil co-exist. No surprise--it's not intended to counter that claim. What I am trying to do is to find a way to counter that claim.

    ReplyDelete
  18. I see. This post reads like a criticism of Plantinga on TWD on the grounds that the hypothesis is improbable. But it is irrelevant to Plantinga's goals that it is improbable. So the criticism must be that the improbability of TWD shows that Plantinga's argument would not be of much help with probabilistic arguments from evil.
    I take it you're arguing that some form of TWD might be advanced in response to such a probabilistic argument. One of the things you say is this,

    Given the conditionals of free will, any world containing infinitely many significantly free creatures contains at least one free creature who sins. But God would have good reason to create infinitely many significantly free creatures, especially if the majority of their choices were good. (And, yes, for aught that we know our world could be like that. We do not know that there isn't an infinite number of significantly free creatures in our world.)

    Tell me how this makes it more probable that God exists along with evil. According to Plantinga, individual essences exist at every possible world--all of them do. The only question is which individual essences get instantiated at each world. So, there is nearly no doubt that, for any given world w, there is some essence E such that if God instantiates E in w, then E goes wrong in w. But then, why should God instantiate such an essence? He can instead instantiate infinitely many others--and have infinitely many left over. Indeed, since there are infinitely many individual essences in each world, it is very likely true that, for any essence E that God might instantiate and that would do something wrong, there is another essence E' that God might instantiate instead that would not go wrong (or would go less wrong). So there is never any loss in value in not instantiating an essence that would do something wrong.

    ReplyDelete
  19. Mike:

    Well, in a way, it is a criticism of Plantinga. For I think we want a defense to be at least contain the kernel of a theodicy. A defense that posited some circumstances that we are all but certain do not obtain is not entirely satisfactory. And such defenses do not seem hard to give. Take a world which contains an infinite number of people, all of whom except one are morally perfect. The imperfect one once commits a very minor sin--he causes a minor pain to afflict someone else for a few seconds. The other person then forgives him, and this forgiveness brings the malefactor to repentance and to a realization of the importance of virtue and the badness of vice. The malefactor then overcomes his vice and never sins again. It seems pretty plausible that God would be justified in creating such a world. But this defense, even if it works, is not very satisfactory, because we know for sure we're not in a world like that.

    Anyway, back to the issue at hand. The problem is that the values of the conditionals of free will depend on the antecedents, and the antecedents will include a statement of what other persons exist in the world. So it is not at all clear that it is probable that God would be able to create infinitely many people none of whom sin.

    Let's take your substitution of essences idea. First, choose an infinite set of essences of persons. Some of these persons would sin and some wouldn't. Take one who would, say Curley. We could replace Curley's essence with the essence of somebody who wouldn't sin, say Margaret. Fine. Now take another sinner, say Adolf. Replace his essence with the essence of somebody who wouldn't sin, say Patrick. But by replacing Adolf with Patrick you've changed the truth values of the antecedents of some of the conditionals of free will for Margaret (assuming Adolf or Patrick existed before the end of Margaret's agential life). When we picked out Margaret, we picked her out in such a way that she wouldn't sin in the original context, the one that had Adolf in the place of Patrick. But Margaret might well sin if Adolf is replaced with Patrick.

    OK, so maybe we need to now replace the ordered pair (Margaret,Adolf) with some pair that won't sin. And then continue--replace an ordered triple, and so on. However, I don't know of a good reason to think that the process would actually converge to a stable collection of individuals.

    Maybe, though, we could do this. Say that x has transworld perfect virtue if x leads a morally perfect life in every possible world in which x has significant freedom. Isn't it likely that there are infinitely many persons that have transworld perfect virtue, and God could put them together in a world where they each have significant freedom, without introducing any other persons?

    Well, no, that is not obvious at all. First of all, it is not clear that all possible combinations of persons are compossible. If essentiality of origins is true, then it is not possible for me to exist in a world where my parents do not. And there might even be contradictions between essences. It could perhaps (I am not sure about this--in fact, I kind of doubt it, but someone might think this) be that certain persons have the essential property of living under one set of universal laws of nature and thus cannot coexist with persons who have the essential property of living under another set of universal laws.

    Second, it is not obvious that it is at all probable that there are any non-divine persons with transworld perfect virtue. Consider. Any possible person could, probably, find herself in an infinitude of possible circumstances. It seems plausible that the probability that all of the relevant conditionals of free will are in her favor--that there is no conditional of free will specifying a circumstances under which she would sin--is actually zero, just as the probability of rolling heads an infinite number of times in a row is zero. (That something has zero probability does not entail it is impossible.) So for any particular individual, the probability is zero. Can we claim that, nonetheless, there is a high probability that there is some possible non-divine person who has transworld perfect virtue? I am not sure.

    Maybe there are other ways of trying to put together a world with infinitely many morally perfect significantly free persons. But the task is far from easy.

    ReplyDelete
  20. Alex,

    I'll have to respond piecenmeal, since this is a pretty long reply. But quickly you say,

    The malefactor then overcomes his vice and never sins again. It seems pretty plausible that God would be justified in creating such a world.

    I agree. But such a world would not serve Plantinga well. Recall, he is trying to show it might be impossible that God actualizes a world in which there is no evil at all.

    ReplyDelete
  21. The problem is that the values of the conditionals of free will depend on the antecedents, and the antecedents will include a statement of what other persons exist in the world. So it is not at all clear that it is probable that God would be able to create infinitely many people none of whom sin.

    So what if subjunctive conditionals do not validate strenghtening/weakening antecedents? We have infinitely many individual essences from which to choose. This is not going to significantly affect our chances of actualizing only those who do no wrong.

    ReplyDelete
  22. Well, no, that is not obvious at all. First of all, it is not clear that all possible combinations of persons are compossible. If essentiality of origins is true, then it is not possible for me to exist in a world where my parents do not. And there might even be contradictions between essences.

    This is not consistent with Plantingan metaphysics in which the individual essences of every possible creature exist in every world. There could be no contradiction. Further, God's creating you is a matter of his instantiating you in a world. He could instantiate you alone, if he willed, on this model. So there is no room for essentiality of origins. We are mixing two distinct metaphysical views here.

    ReplyDelete
  23. Mike:

    That we have infinitely many essences to choose from does not guarantee that it is likely that a subset could be chosen with the appropriate properties... At least, an argument is needed.

    As for compossibility, while all the essences coexist, it does not follow that the entities exemplifying them can coexist.

    ReplyDelete
  24. Let's take your substitution of essences idea. First, choose an infinite set of essences of persons. Some of these persons would sin and some wouldn't. Take one who would, say Curley. We could replace Curley's essence with the essence of somebody who wouldn't sin, say Margaret. Fine. Now take another sinner, say Adolf. Replace his essence with the essence of somebody who wouldn't sin, say Patrick.

    But this seems beside the point. The conclusion I'm after follows from the possibility of a world in which no one goes wrong. I suggest that it is likely that such a world is possible. There is no reason to proceed by way of substitution of created essences.

    ReplyDelete
  25. But what reason do we have to think it is probable that such a world is feasible?

    ReplyDelete
  26. Limits to the worlds that God can actualize depend on certain contingent properties of individual essences: in particular the property of being TWD. Other contingent properties God can modify, since he can choose a world in which most of your contingent properties are different. For a world taken at random (such as ours), I don't see the basis for worrying that a best world could not have been actualized. Some think it was actualized (though I concede, I don't). There is no reason to believe, for instance, that there is a large distribution of TWD's across possible worlds. I have no reason to believe that there are thousands actual TWD's, but suppose we say there are likely 10 thousand. It would not matter to actualizing a best world. We have infinitely many other individual essences.

    ReplyDelete
  27. We can set up the problem as follows. For each possible person x, let G(x) be the set of all feasible worlds w such that:
    (i) x is significantly free in w; and
    (ii) nobody sins in w.
    Note that x has TWD iff G(x) is empty.

    Now, for any two possible persons x and y that lack TWD, it can still be the case that G(x) and G(y) have an empty intersection--they could have no worlds in common. This could happen for several reasons. First, it might be that there is no possible world that contains both x and y. Second, it might be that while there is a possible world that contains both x and y, there is no feasible world that contains both x and y. Third, it might be that although there is a feasible world that contains both x and y, in every such world, x sins. Fourth, it might be that although there is a feasiable world that contains both x and y, and in which x does not sin, in any such world y sins. Only if we can eliminate all four of these cases will G(x) and G(y) have non-empty intersection.

    It is harder yet for a triple of distinct possible persons x, y and z to satisfy the condition that the intersection of G(x), G(y) and G(z) is non-empty--there are yet more ways for this condition to fail. And it is harder yet for an infinite sequence x1, x2, x3, ... of distinct possible persons to be such that the intersection of G(x1), G(x2), G(x3), ... is non-empty.

    Lemma: There is a feasible world containing infinitely many significantly free persons iff there is a sequence of infinitely many significantly free persons x1, x2, x3, ... such that the intersection of G(x1), G(x2), G(x3), ... is non-empty.

    Now it seems to me far from obvious that it is likely that there is a sequence of infinitely many significantly free persons x1, x2, x3, ... such that the intersection of G(x1), G(x2), G(x3), ... is non-empty.

    Maybe there is some argument out there that it is likely. But until such an argument is constructed, "it is not clear that we are in a position to say" (to use the wording from my post) that it is likely that there is a feasible world containing infinitely many significantly free possible persons none of whom sin.

    ReplyDelete
  28. Alex and Mike,

    Thanks for this discussion.

    Alex,

    1. There seems to be a typo in your new paper on QS's atheological 1997 argument from evil. Section 4: " ... if Smith were right that creaturely logical freedom without external and internal freedom is possible ..."

    Shouldn't there be " ... if Smith were right that creaturely external and internal freedom without logical freedom is possible ..."?

    2. Section 3: "It seems that if Patricia is logically determined to lead a wholly good life, this is reflected in her dispositional psychological and physical states in such a way that anybody else who has the same psychological and physical states will share that logical determination—she has the dispositional state of being such that in any set of possible circumstances she will act well, after all. It would be really odd if logical determination to lead a wholly good life were not supervenient on the psychological and physical state of the person. Whence the source of the logical determination, then? Does the haecceity of the person in some mysterious way predetermine how the person must act? (This kind of view would seem to share the disadvantages of Molinism without Molinism’s advantages.)"

    What is the problem with the haecceities?

    3. As for the relation of "metaphysically valuable freedom" (your concept) and "logical freedom" (QS's concept):

    Was Christ logically free with respect to his wholly good life?

    Was Mary logically free with respect to her wholly good life? Did not the grace given to her made her logically unfree?

    4. Is not a human's being able to refrain from his doing A in a circumstance C a necessary condition of his freely doing A in C?

    It seems to be the case. At least, standard libertarian definitions of a human free action suggests as a necessary condition something like the fact that the human could have done other than A even if every factor prior to A had been the same.

    4. Suppose God does not permit a human to refrain from A in C. More concretely, suppose God does not permit the Pope to promulgate a non-true doctrine when he promulgates a doctrine. How, then, can it be the case that the Pope freely promulgates a true doctrine?

    Thank you.

    Best wishes,

    V.

    ReplyDelete
  29. Of course, the last question is numered as 5.

    ReplyDelete
  30. Vlastimil:

    Ad 1: Thanks!

    Ad 2: Well, it would be weird if haecceities determined how one acted. Maybe not utterly absurd, but very strange--there would be necessary truths about how somebody would have to choose, not grounded in anything other than the numerical identity of the person. (Maybe they wouldn't be much more problematic than conditionals of free will, though?)

    Ad 3: Christ wasn't logically significantly free. But that's OK, because he (qua person, and moral evaluation is of persons) was not a creature.

    I don't know whether Mary was significantly free or not. It may be that she was free to reject grace.

    Ad 4: The pope cannot definitively promulgate a false doctrine to the Church as a whole. But he has lots of other options. Instead, he might re-promulgate the doctrine that God exists, or he might dance a jig, or he might intend to promulgate a false doctrine and fail (e.g., because his voice fails and he can't find any paper, or because a heart attack takes him), or he might promulgate a false doctrine non-definitively or to only a part of the Church, vel cetera. He has lots of freedom there--just not the freedom to "successfully" promulgate a false doctrine definitively to the whole Church.

    ReplyDelete
  31. Lemma: There is a feasible world containing infinitely many significantly free persons iff there is a sequence of infinitely many significantly free persons x1, x2, x3, ... such that the intersection of G(x1), G(x2), G(x3), ... is non-empty.

    Here's a sketch of a proof. For any created agent x, it is possible that x dies before he commits a sin S. Let the life of any agent x, that is terminated prior to committing any sin be L(x). For each x, L(x) is a member of G(x). Let God simultaneously create each of infinitely many agents on isolated islands in world w, and let God allow each agent x to live exactly the duration of L(x). In w we have the intersection of G(x0), G(x1) . . . G(x00).

    ReplyDelete
  32. Mike:

    That's neat.

    One problem is that conditionals of free will might depend on other people's actions even if there are no causal connections there. Moreover, the conditionals of free will may depend on when people's lives terminate.

    After I read your comment, I was thinking of the following setup. God creates a world with infinitely many causally isolated persons. Then he lets everybody who would sin die just before they would sin. Some people will die before they exercise significant freedom because their first act would have been sinful. But, intuitively, it seems likely that infinitely many should survive this process to act rightly at least once.

    The problem here (and I think in your proposal as well) is with the potential interconnectedness of everything via strengthening of antecedents of conditionals of free will. If one could posit that facts about what is happening in places that are causally isolated from an agent do not affect what the agent does or would do, the above arrangement would work fine. I actually think such a posit is one that Molinists should agree with, but in fact I suspect many of them wouldn't (Tom Flint argued against a stronger version of this posit once in correspondence with me, and his arguments would apply to this version as well).

    Anyway, what you say here is good enough to convince me that if Molinism is true, probably there is a world where there are infinitely many significantly free persons none of whom sin.

    But we can make this more tricky. Presumably, God has good reason to want significantly free persons to interact with one another. There is great moral value in such interaction. So, is there a world containing infinitely many significantly free persons none of whom sin and who interact with one another, in such a way that every pair of persons is connected through some chain of interactions?

    A minor problem with your example is that death might well count as an evil, and so the world described is not a world without evil. Here is something better. God simply takes away the free will of those who are about to sin, and has them unfreely act righteously. That is not an evil.

    ReplyDelete
  33. Alex,

    Thanks, very helpful.

    Let me pose again the general question which went unnoticed:

    1. Is a human's being able to refrain from his doing A in a circumstance C a necessary condition of his freely doing A in a circumstance C?

    It seems to be the case. At least, many libertarian definitions of a human free action suggests as a necessary condition something like the fact that the human could have done other than A even if every factor prior to A had been the same.

    Let me further a general, basic, and well-known worry about the compatibility of God and freedom of creatures.

    2.

    God is the first cause of every being different from him (claimed by traditional theism), including free (?) volitions of creatures.

    Christian theism: "... it is God which worketh in you both to will and to do of his good pleasure" (Phillipians 2:13).

    Aquinas, STh. Ia, q. 83, a.1 ad 3: "God ... is the first cause, Who moves causes both natural and voluntary."

    Isn't all of this in conflict with something you embrace in your New FWD (2003, section 5)?: " ... a person is not free if all of her actions are brought about by another person ... It is logically impossible that a person x is significantly free and yet another person, y, for every choice of x brings it about that the x makes a right choice or brings it about that x makes a wrong choice."

    What would be your strategy for harmonizing all these insights?

    3. A similar tension can be found between the two following positions.

    First, see your and R. Gale's New Cosmological Argument (1999): one of the results of the paper is that God freely and intentionally brings it about that the Big Conjunctive Contingent Fact of the actual world, including propositions describing free actions and volitions of creatures, is true.

    But in your new paper The Essential Divine Perfection Objection to the Free Will Defense (2007, section 3) you indicate a different claim: " ... everything except possibly creaturely free choices is under divine control." Similarly, in A New FWD (2003, section 7) you wrote: God's "... free creative action explains all contingent true propositions other than those explained by the free actions of created persons ..."

    I do NOT want to cavil about fabled inconsistencies in your beliefs. I do not take them as clear inconsistencies; just try to understand.

    ReplyDelete
  34. 4. As you wrote, Christ qua person was not a creature; a person x is logically singnificantly free with respect to an action A only if x qua person is a creature; so, Christ was not logically singnificantly free with respect to his wholly good life.

    Couldn't, then, one infer as follows? Christ qua person was not a creature; a person x suffers only if x qua person is a creaure; so, Christ did not suffer.

    ReplyDelete
  35. Vlastimil:

    One suffers insofar only if one has a created nature. But not everybody who has a created nature is a creature. Christ is the counterexample.

    The problem about God's sovereignty and grace and free will is a really tough one. First, note that St. Paul is not saying that when you sin, it is God working in you to sin!

    What I get from Augustine is the following hard idea: If you sin, that's your own work; if you act righteously, that's God's work.

    St. Paul is probably talking about grace in Phillipians 2:13, and the state of the believer. I think it would be an over-reading to take the text to exclude our possibility of rejecting grace. If we reject grace, then we've freely rejected grace.

    My favorite model for this is one that is closely akin (or identical?) to something Eleonore Stump has been defending. Suppose that I am drowning. You jump in to pull me out. I can do nothing, in which case you pull me out. Or I can deliberately wriggle out of your grasp, in which case you don't pull me out. If I do nothing, it seems correct to say that my being pulled out is all your own work. It would be inappropriate for me to try to take credit for my not wriggling out of your grasp, and to say that thereby the achievement is mine. No, it is you who have done the work. But if I do deliberately wriggle out of your grasp, then I am responsible for my own drowning, woe is me. May God have mercy on us all.

    In Richard Gale's and my cosmological argument we oversimplify the free will issue a bit. Still, it's not completely unreasonable to say that by explaining the cause (our power of free choice, which was created by God), some explanation is given of the effect (the particular choice made), even if the connection here is indeterministic. I have a long discussion of this stuff in my PSR book.

    ReplyDelete
  36. Thanks, Alex.

    As for the issue of compatibility of God and freedom of creatures, Wm. Most, SJ has an excellent book going along the Augustinian and Stumpian lines (which I like, too), available online. Its summary: http://www.catholicculture.org/library/
    most/getchap.cfm?WorkNum=214&ChapNum=31

    ReplyDelete
  37. Hi Alex

    Congratulations for a refreshingly audacious blog! I wonder what you’d make of the following twist to your drowning story: Having saved you, the life-guard has sex with you on the beach and you subsequently file charges of rape against him. You were not unconscious, you could have wriggled out of his grasp or at least cried ‘NO’, both while you were being ‘saved’ and while you were being ‘raped’, but you did not. Is the life-guard guilty of raping you?

    ReplyDelete
  38. I actually worry a bit that the rape idea suggests a problem for what I said. For there will be some rare circumstance in which, although the rapist is no less to blame, we would say that the victim ought to have resisted. Suppose that the victim is a 190cm tall, muscular body-builder, holding loaded Uzis in both hands, while the rapist is a 160cm tall nerd wearing too little clothing to conceal weapons. Then while the rapist is no less a rapist, we would be puzzled as to why the victim did not resist, unless the victim was under a vow of non-violence or something like that (but then why the Uzis?). And, I think, we would think that the victim should have resisted.

    Sometimes we do hold people responsible for inaction. My feeling is that we do this in two kinds of circumstances, though:
    (a) Cases where the agent should have acted but did not.
    (b) Cases where it took significant moral courage, or some other virtue, to refrain from acting.

    We would not hold someone responsible for failing to resist being pulled out of the water. Or at least we would not credit them with any positive merit. Or is that almost any? If so, this account has a problem.

    ReplyDelete
  39. Thanks, I appreciate your analysis except where I live it’s illegal to carry Uzis! Since you filed charges and the case is going to trial, I was thinking more in terms of your response to the life-guard’s line of defence if it was claimed that your behaviour while being saved and while being had sex with was taken by him as implying consent to being done to what on each occasion he admittedly took the initiative to do.

    ReplyDelete
  40. 'Assume Molinism.'

    Why?

    Molinism is trivially true.

    Molinism claims that a person will freely choose one particular way in each set of logically possible circumstances that could occur in a real world.

    Is this true?

    Let us take two sets of logically possible circumstances which could have occured in the real world, but did not.

    1) I am sitting down to breakfast in an hotel at 8:30 am on Sun 30/12/2007, and a waiter is asking 'Tea or Coffee’, and an omniscient being has infallible knowledge that I will choose tea.

    2) I am sitting down to breakfast in an hotel at 8:30 am on Sun. 30/12/2007 , and a waiter is asking me ‘Tea or Coffee’, and an omniscient being has infallible knowledge that I will choose coffee.

    Are these logically possible, and distinct, sets of circumstances?

    What are my counterfactuals of freedom in each of those 2 sets of circumstances?

    Which way would I freely choose in each of those sets of circumstances?

    If there is a definite fact of the matter about how I would choose in each of those 2 sets of logically possible circumstances, has Molinism now been proved to be true?

    These questions are trivial to answer.

    ReplyDelete
  41. "Circumstances" in the definition of Molinism explicitly excludes any facts that logically entail what action is done. That an omnisicient being knows that I will do A entails that I will do A, and hence we exclude the omniscient being's knowledge from the circumstances.

    ReplyDelete
  42. '"Circumstances" in the definition of Molinism explicitly excludes any facts that logically entail what action is done.'

    I see.

    So Molinism is the doctrine that people will always choose to punt on 4th and 14, as including any other facts about where the 4th down might be, or what the time left on the clock is, or any other relevant information, might logically entail what action is done??

    William Lane Craig has a discussion of Molinism at

    Molinism

    I quote :-

    'By means of His natural knowledge, then, God has knowledge of every contingent state of affairs which could possibly obtain and of what the exemplification of the individual essence of any free creature could freely choose to do in any such state of affairs that should be actual.'

    I see nothing in Craig's description of 'every contingent state of affairs which could possibly obtain' which excludes the possibility of a contingent omniscient being existing in that state of affairs.

    Of course, Craig would also deny that 'every contingent state of affairs which could possibly obtain' includes an omniscient being, although he also claims that his omniscient being is necessary and exists in every state of affairs.

    That is because an accurate, rather than a fudged, description of 'every contingent state of affairs which could possibly obtain', would show that Molinism is a mere tautology.


    Both of the following states of affairs are 'contingent states of affairs which could possibly obtain.

    1) I am sitting down to breakfast in an hotel at 8:30 am on Sun 30/12/2007, and a waiter is asking 'Tea or Coffee’, and an omniscient being has infallible knowledge that I will choose tea.

    2) I am sitting down to breakfast in an hotel at 8:30 am on Sun. 30/12/2007 , and a waiter is asking me ‘Tea or Coffee’, and an omniscient being has infallible knowledge that I will choose coffee.

    These 2 states of affairs fit perfectly Craig's description of a 'contingent state of affairs which could possibly obtain'


    This state of affairs cannot possibly obtain.

    '1) I am sitting down to breakfast in an hotel at 8:30 am on Sun 30/12/2007, and a waiter is asking 'Tea or Coffee’'

    That is like claiming that there could be a 4th and 14.

    True, but that is not a description of a state of affairs that could obtain.

    It is only a part of a description of a state of affairs that could obtain.

    ReplyDelete
  43. I've just realised how Craig could tighten up his argument.

    Simply change the following 'By means of His natural knowledge, then, God has knowledge of every contingent state of affairs which could possibly obtain...'

    to this :-

    'By means of His natural knowledge, then, God has knowledge of every contingent state of affairs which could possibly obtain in one or more worlds...'

    And then Craig could claim that God knows what each person would freely choose in all of those worlds.

    Mind you, he would then struggle with a proof that the person always freely chose the same way in each of those worlds.

    ReplyDelete
  44. The "natural knowledge" that Craig is talking about does not include middle knowledge. He discusses middle knowledge a bit later in his article, when he talks of knowledge of counterfactuals of free will.

    ReplyDelete