Evil and eternal life
December 14, 2009 — 11:54

Author: Alexander Pruss  Category: Problem of Evil  Comments: 50

Let’s say I climb Mt. Everest, and then enjoy a delightful view from the top. But as I climb the mountain, I undergo various horrendous sufferings. And after I get back down, I have to undergo extremely painful surgery. Suppose that, so far, the overall value assessment is negative. If that’s all that is involved, then climb wasn’t worth it. The view was nice, and the good of achievement was nice, but, by far, it just wasn’t worth it.

But let me add a little more to the story. I did this when I was 20. I am not permanently traumatized by the suffering, and indeed by the time I am 30, my memories of the hideous pains are no longer unpleasant. But I continue to have memories of the beauty of the climb and of the camaraderie, memories of the grandeur of the epic struggle, and these memories continue to be fairly pleasant. Moreover, the feeling of accomplishment, of having overcome the pains, is nice to have. I then live on for fifty more years, continuing to have pleasant memories of that climb.

While the goods achieved at the time of the climb were not worth the suffering, when combined with the value of half a century’s worth of memories, even when these memories are not particularly intensely pleasant, they may be worth it. Suppose you say the contrary. Well, then, replace the fifty years with five hundred or five million. Eventually, the cumulative value of enjoying these memories will overshadow the bads which were confined to one decade of one’s life (the climb, plus about ten years during which the memories of pain were painful). (This of course reminds one of Parfit’s Repugnant Conclusion. But I think there is nothing repugnant here.)

What this shows is that given a long enough life-span, our evaluations of whether some activity was worth doing can change as a result of the values of memories and of retrospective awareness of achievement. Of course, in principle, they can change in either direction. A pain might have been such that it by itself would have been worth it for the sake of an achievement, but if in fact the memories of the pain continue to be painful while the memories of the achievement fade, then it’s no longer worth it.

Observe also that while it would be possible to have the pleasure of the memories of having climbed Mt. Everest without having ever had the pains, that pleasure would then be an empty pleasure, and hence either devoid of value or of much lesser value.

Suppose that in fact we live forever. There is no good argument to the contrary that doesn’t presuppose the non-existence of God (here we can insert a discussion of arguments for materialism and of arguments against resurrection based on the need for causal continuity), so someone who offers an argument from evil against the existence of God cannot rely on the denial of the claim that we live forever. The above remarks show that remembering over a significant length of time can act as a value-multiplier, and when the length of time is long enough (and in particular when it is infinite), this can completely swamp the original assessment.

Moreover, memory is not the only such effect over an infinite life-span. A small change in character when prolonged over a very long time can make an enormous difference. Suppose that climbing Mt Everest, in my story, made me slightly more considerate. We might well question whether this was worth all the suffering. If I were to live in this more considerate way for five minutes, it perhaps wouldn’t be worth it (Socrates will disagree). But again we have value-multiplication–if the difference remains over a sufficiently long interval, the increase in value, summed (actually, integrated) over that time, will swamp the pains of gaining it. In fact, that small difference, over an infinite amount of time, will make for an infinite good. Moreover, if virtue leads to virtue, we might here have compound interest at work.

Of course, God could produce the better character directly and maybe he could induce in us false memories. But the value of that infinite good having been produced in the right way may very well be quite high. (One thinks that the value of a good G’s having been produced in the right way is proportionate to the value of G.)

These value-multiplication processes can be selective. Thus, God might very well ensure that our memories of pains not be painful to have (to do that God would need to heal traumas, etc.), while ensuring that our memories of goods be pleasant. Nor would there be anything dishonest in God’s doing this. In fact, I think most of us have plenty of memories of pains where the memories aren’t themselves painful, and this is no defect in this.

If this is right, then selective value-multiplication processes working in an infinite afterlife might very well, and in quite understandable ways, swamp the kinds of value assessments we get from this-worldly considerations. This possibility, which indeed is not merely a possibility but a fairly high likelihood if God exists, might not entirely undercut inductive arguments from evil, but I think it blunts them quite significantly.

And if one is worried that this undercuts our own reasons to prevent evils, all I can do is point one to this older post of mine.

Comments:
  • christian

    Alex,
    That seems plausible enough for pain endured on a climb. But does it seem plausible if we are considering other cases, e.g. a child’s being raped an bludgeoned to death? Where do we find fond memories here? We can then run the inductive argument on these kinds of evils.

    December 14, 2009 — 14:04
  • Observe also that while it would be possible to have the pleasure of the memories of having climbed Mt. Everest without having ever had the pains, that pleasure would then be an empty pleasure, and hence either devoid of value or of much lesser value.
    Why grant you this assumption? It seems groundless to me. If God were to decide to give me only the pleasant memories of successfully climbing Mt. Everest, and none of the typically accompanying painful memories, why would that necessarily less valuable than having both the pleasant and the painful memories?
    Suppose that I climb Mt. Everest, and subsequently suffer a traumatic head injury, the nature of the injury is such that it causes me to be unable to recall/form new memories of pain episodes, so that I that all that I can subsequently recall of my heroic ascent are the pleasant memories, are my memories somehow then less valuable then they were before the trauma? I for one think not.
    How can the value of a given memory be effected one way or another by the existence of other memories?
    I value the memory of my long dead grandfather’s face (who died when I was 6) even though I have no other memories of him. In fact I probably value it more because it is my only memory of him.

    December 14, 2009 — 14:23
  • John A.

    Alexander
    Interesting argument. You wrote:”Of course, God could produce the better character directly and maybe he could induce in us false memories. But the value of that infinite good having been produced in the right way may very well be quite high. (One thinks that the value of a good G’s having been produced in the right way is proportionate to the value of G.)”
    If God could produce in us the same results – planting within us false memories of painful experiences – without us having to experience the pain would that not be a better alternative given that the outcome is the same? It seems that if we can achieve the same outcome with less pain then we ought to. It seems to me that a ‘infinate good having been produced in the right way’ needs to be explained and I do not see how it can be without assuming what we are trying to prove. In fact, I am not sure what it means to be ‘produced in the right ways’ especially if the memories would be the have the same epistemic value. I am assuming that I would not know that God has given me false memories. (Kind of reminds me of Russell’s five minute argument.)
    This seems to open up a line of attach for those who think the existence of horrendous evil counts as evidence against there being a God. Assume that universalism is true and that all of us will be eventually saved (does not your argument presuppose this) – then if God could have created us in a state of being saved with false memories of horrendous evils occurring to us that are no longer hurtful because we are saved then He should have done so. What value is added by having us actually suffer if the results can be obtained without us actually suffering, only believing (falsely) that we did.

    December 14, 2009 — 14:43
  • Luke Gelinas

    Neat! It seems crucial that what’s doing the justifying work is some organic whole–not just pleasure, or pleasure in the memory of climbing, but pleasure in the memory of having actually climbed that particular mountain. This might also fit pretty well with our intuitions about what’s needed for the long-term defeat of evils in an individual’s life–the horrendous suffering isn’t just needed for any old pleasure, but for pleasure in the very activity which produced the suffering.
    About the process being selective, there does seem something slightly prima facie morally objectionable about someone tweaking with our memories like that. Maybe God could, at some point, ask for consent–or maybe there are facts about what we would consent to in this sphere.

    December 14, 2009 — 15:53
  • Thanks for all these helpful comments.
    APT:
    I am afraid I phrased the sentence you criticize poorly. The thought was this: It might be possible for God to induce false memories of having climbed Mt Everest and of having overcome much adversity in so doing, even though in fact one did not do that. And then I was simply thinking that the pleasure of a non-veridical memory is either worthless or of significantly less worth than the pleasure of a veridical memory.
    John:
    It seems to me that what is particularly valuable is a pleasant veridical memory of achievement. A pleasant false memory of achievement is of little if any value. The value in the memory lies not just in the pleasure associated with it, but in the fact that what one takes pleasure in is real.
    Luke:
    I think that whether a memory of a pain is itself painful is not a function of the content of the memory. Thus, two people could remember exactly the same kind of experience, and do so equally well, but one will be pained by the memory and the other not. (In fact, it is even possible to have painful memories of pleasures.) So the selectiveness doesn’t require God to tweak the contents of the memories, but only whether the remembering is painful or pleasant. I think there is a natural healing process in humans whereby memories of painful experiences cease to be painful. This process does not always operate, but God’s making the process operate is a form of healing. And I don’t think God needs to ask our permission to heal us.
    Christian:
    Your objection worries me the most.
    Obviously, this works better for some evils than for others. We have no reason to think a single theodicy would work for all evils. After all, we would expect that if God exists and permits evils, he might have different reasons for permitting different evils.
    However, the theodicy does provide a tool for multiplying goods which may be useful in a wide variety of cases. Suppose, for instance, that suffering makes some good possible, such as the good of forgiving the attacker (perhaps posthumously). Then the selective value-multiplication process may increase the value of that good.
    Perhaps there is some mysterious way in which in an afterlife we might rejoice forever in part in the value of the freedom which people misused to hurt us.
    Everybody:
    Here’s another analogy. Consider a very short but extremely intense pain. It may be quite bad. But if it’s very short, then it may not be that hard to outweigh given enough time. For instance, it might be worth having a very short but extremely intense pain in order to relieve a very minor chronic pain that otherwise one would have for fifty years. (To be honest, I am so afraid of pain that I probably wouldn’t go for the deal, no matter how brief the intense pain is. But that might be a form of prudential akrasia on my part.) But compared to eternity, our earthly lives are an infinitesimal blip.

    December 14, 2009 — 16:17
  • Let me add that the primary interpretation of the Mt Everest analogy had the climbing of Mt Everest be our entry into heaven. Thus, there need not be an achievement in this life that is facilitated by the pains and sufferings and which is later remembered. Nonetheless, where there is such an achievement, I am happy to embrace that case, too.

    December 14, 2009 — 16:28
  • christian

    Alex,
    Your response seems right. Here, though, I think there may be a problem:
    Perhaps there is some mysterious way in which in an afterlife we might rejoice forever in part in the value of the freedom which people misused to hurt us.
    Perhaps quite the opposite happens. Perhaps we get a selective form of disvalue-multiplication. Instead of rejoicing, that is, we spend a rather long time languishing.
    At any rate, I doubt that there are people who rejoice in the freedoms of their rapists or in the freedoms of their children’s murderers. If we don’t do it here and now–and given that we do find the opposite reaction, anger and sadness, to be fitting–then it’s very hard to see why things would be any different after death.

    December 14, 2009 — 17:00
  • Well, just about anything of objective value can, I think, be rejoiced in under some circumstances. (It may, of course, take a really long time before one becomes capable of it.)
    I do not know how unpleasantly sad one would feel about having been raped or murdered a million years earlier. Suppose, for instance, that all the trauma is gone (if earthly therapy can remove some of the trauma, heavenly therapy can surely do a lot better) and 900 thousand years ago one has become reconciled with the person who did it, indeed they are now a close friend–reconciliations of that sort do, indeed, happen. In that case, I don’t think one would necessarily feel unpleasantly sad. One might even feel joy that love was able to overcome such a gap. (What if one is in heaven and the perpetrator is in hell? I don’t know what to say then. Maybe one runs a Tertullian line about joy at divine justice, but that line makes me uncomfortable.)

    December 14, 2009 — 17:19
  • John A.

    Alexander
    How would we be able to tell the difference? If God did a good job they would seem veridical to us, would they not.

    December 14, 2009 — 17:25
  • M.

    1. The memory of your climb extended over a sufficiently long lifetime may outweigh the suffering induced by the climb. And a pleasant, veridical memory of an at-the-time painful event may be more valuable than a pleasant, falsidical memory of a non-existent, and hence painless, event. But why should God have allowed the experience to have been painful at all? Why not arrange it so that I get to have pleasant, veridical memories of at-the-time pleasant event, instead?
    Maybe in the Mt. Everest case, we could say that knowledge of the adversity I faced during the climb enhances the memory. But this is rarely going to be true for most serious cases of suffering. Nor is it going to be true for most trivial cases of suffering, where the pain is simply too minor, too casual, too random or too ineffectual to count as a challenge I relish having faced. (I don’t think the fact that I’ve stubbed my toes X times during my life will make me remember my achievements any more fondly.)
    2. For a non-theist without positive reason to believe in an eternal afterlife, the fact that we could explain away divine permission of certain evils by positing that an eternal afterlife exists (indeed, with our psychologies and memories wholly intact throughout) is of little help. So I don’t think this response will be persuasive to non-theists. What about to Christian theists? They’ll have to wrestle with the reverse phenomenon in hell, where plausibly memories of one’s Earthly existence make things horribly, horribly worse. Perhaps this enlarges the problem of evil by the same factor that your response (applied to the heaven-bound) shrinks it.
    3. Regarding pleasant memories of painful experiences, I’m not so sure these aren’t falsidical somehow. If an experience is overall negative and doesn’t facilitate more positive future experiences (in a way that couldn’t have been facilitated by my having had overall positive experiences, instead), then I’m not sure it’s reasonable or even psychologically possible to truly cherish the experience. At least, not unless I’m remembering it through rose-tinted glasses.

    December 14, 2009 — 17:30
  • It seems to me that what is particularly valuable is a pleasant veridical memory of achievement. A pleasant false memory of achievement is of little if any value. The value in the memory lies not just in the pleasure associated with it, but in the fact that what one takes pleasure in is real.
    This is a curious assertion. For all I know the past is unreal, and if so none of my memories of the past is veridical. Does that epistemic possibility leave me devaluing my pleasant memories of the past? Not a tick! (Of course you might argue that this just shows that I just don’t believe that the past could be unreal, but you’d be wrong on that score).
    Furthermore, how do we go about checking whether our a given memory is or is not veridical? I suppose we check it against other memories of our own and the memories of others. Where we spot discrepancy between our putative memory, and other memories of ours, or the memories of others, we have defeaters for the veridicality claim for the given memory.
    But suppose God wanted me to remember only the the pleasant things about my heroic ascent of Mt. Everest. An ascent that I never in fact made. Suppose that in order to be sure that I would believe the memories given to me were authentic, God crafted them to be consistent with not only my own non-Everest memories, but also the memories of everyone else. How then should I ever discover that they were not veridical? And if I cannot discover this, how could the memory be any less valuable than a genuine veridical memory (to me)? I certainly will never discover the memory is not veridical, and neither will anyone else.
    If this is correct then God could create pleasant memories of my climbing Everest, for which there would exist no defeaters, and which would be just as valuable as any veridical memory of my having done the same.

    December 14, 2009 — 17:41
  • Luke Gelinas

    So the selectiveness doesn’t require God to tweak the contents of the memories, but only whether the remembering is painful or pleasant.
    I think my worry is that there is a bit of tension between this thought and certain views about virtue–those that make virtue consist in certain intentional or attitudinal relations to intrinsic goods and evils. On views like this I’m not sure it will ever be better to respond to an intrinsic evil neutrally rather than with some (perhaps very very very slight, the further into the future we go) con-attitude. I’m not sure what you describe would be ‘healing’ on these views, since according to them it would involve attitudes that are disproportionate in some way.

    December 14, 2009 — 17:48
  • Gordon Knight

    I have always liked the sort of argument. It does have problems, as have already been mentioned.
    But suppose we augment the afterlife thesis in the following way. Its not just that the afterlife is more life, it is more fully realized life. That is vague so let me give an analogy.
    I dream bad dreams sometimes. But after I wake up the badness of the dream changes its significance. It’s still a bad dream, but its not nearly as bad as I took it while I was dreaming.
    This does not remove the problem of evil, but it may help us think of it in a new light. Our perspective now is as of one in a dream. In the afterlife, if there be such, our perspective will be much different.

    December 14, 2009 — 19:20
  • John and APT:
    It doesn’t matter that we can’t tell the memory is non-veridical. The mere fact that it is non-veridical, whether we can tell or not, makes it less valuable. Compare this: Which would one rather have? A spouse who is always faithful or a spouse who always cheats on one but who is able to engage in a perfect simulation of faithfulness? Nozick’s Experience Machine is relevant here.
    M.:
    “For a non-theist without positive reason to believe in an eternal afterlife, the fact that we could explain away divine permission of certain evils by positing that an eternal afterlife exists (indeed, with our psychologies and memories wholly intact throughout) is of little help.”
    Let J be the proposition that the evils of this world are ones such that if God exists, God would be justified in permitting them. Let A be the proposition that there is an afterlife of such a sort as to make my argument work. Let G be the proposition that God exists.
    Suppose the theist and the non-theist can come to an agreement that:
    1. If A, then J. (Because of my argument.)
    2. If not-A, then not-J. (Because of the inductive evidence that the non-theist points out.)
    3. If not-J, then not-G. (This is an uncontroversial necessary truth, at least if we phrase J correctly.)
    Now, the non-theist wants to press the problem of evil as evidence against the existence of God. But in light of (1), all that the non-theist can establish is (2). And that’s not enough to provide evidence against the existence of God, unless there is evidence against J. In other words, the onus is on the non-theist here, as she is the one trying to show that the evils of this world make it plausible that not-J.
    “But why should God have allowed the experience to have been painful at all? Why not arrange it so that I get to have pleasant, veridical memories of at-the-time pleasant event, instead?”
    So, as you note, in the case I gave, the pain in the first place was a crucial part of the story–it accentuated the achievement. I think this may be more common than you think, especially when one considers that the achievement need not come to fruition in this life, but perhaps only in the afterlife.
    The minor evils are perhaps more of a problem. However, there is a way in which sometimes minor discomforts are transformed in memory, and not in a false way. Perhaps you were camping as a child, and while camping you slept on rather hard ground. A very minor discomfort, but it annoyed you at the time, and you griped about the whole trip. But now in hindsight you might see it as a significant part of a camping experience which was, in fact, happy. Such transformations do in fact happen to us, and in the case of minor evils, we are not justified in thinking of any particular evil that it will not be thus transformed.
    “What about to Christian theists? They’ll have to wrestle with the reverse phenomenon in hell, where plausibly memories of one’s Earthly existence make things horribly, horribly worse. Perhaps this enlarges the problem of evil by the same factor that your response (applied to the heaven-bound) shrinks it.”
    Well, we don’t actually know whether the memories of the earthly experience in hell make things worse or not. Nor do we even know that the total suffering in hell is infinite. It could be that hell involves asymptotic healing, with the total suffering (the integral of suffering over time) being finite. We can say that hell is no greater than what the folks in it deserve. If we think they deserve infinite suffering, infinitely multiplied by memories, then it will not trouble us to suppose that they get just that; if, on the other hand, we think that they deserve only a finite amount of suffering, then we should think that hell only contains a finite amount of suffering.
    Luke:
    I think one can have a con-attitude to evils without suffering at the thought or memory of these evils. For instance, I definitely have a con-attitude towards the event of my breaking my toe when I jumped over a stream and hit my foot against a rock on the other side. But my memory of the experience is neither pleasant not unpleasant, though the experience was decidedly unpleasant. At the same time, I have a pleasant memory of a walk with family, pretty sunny day, etc.
    I am not proposing that we will take pleasure at the evils themselves, but at the associated goods.

    December 14, 2009 — 19:35
  • John A.

    Alexander
    “Compare this: Which would one rather have? A spouse who is always faithful or a spouse who always cheats on one but who is able to engage in a perfect simulation of faithfulness? Nozick’s Experience Machine is relevant here.”
    You are giving us a false choice and I do not think that Nozick is relevant here. I might not choose to go into Nozick’s machine, but if I were in it unknowingly, how would I know? I am in agreement with APT. Between having veridical or non-veridical memories, if we cannot tell which is which it makes no epistemic difference which they are. If I in fact think my spouse is faithful and my experiences confirm that she is, then the fact that she is not will not matter to me because I will not know it and, more importantly, I will have no reason to doubt her faithfulness. My point is that God, if he is completely good, could, and should, create us as saved beings with memories that we think are veridical of experiences of horrendous evils that we have overcome. (Of course, paradoxically, he will be a deceiver, but…) From an epistemic point of view, false memories will serve the same function as your thought experiment of climbing Mt Everest and the painful surgery but without the actual (horrendous) experience of evil. I presume that, from a moral point of view, if one can obtain the same end without actual suffering then that is the route that one ought to employ.

    December 14, 2009 — 20:19
  • It doesn’t matter that we can’t tell the memory is non-veridical. The mere fact that it is non-veridical, whether we can tell or not, makes it less valuable.
    Less valuable for whom? Who is it that the non-veridical memory is accruing less value to in this case? Is it me? I don’t see any difference, in terms of value-to-me, between a veridical memory and a non-veridical memory with zero defeaters (for all I know all of my memories just are non-veridical memories with zero defeaters). In any case, both sorts of memories are pleasant, both seem equally veridical to me, and neither can be proven to be non-veridical by any means whatsoever. From where I sit, it looks like it is not the supposed veridicality of the memory, but rather the absence of possible defeaters for the memory, that gives the memory greater value. And God could certainly give me a pleasant non-veridical memory without any possible defeaters.
    Though I must plead ignorance of Nozick’s case.

    December 14, 2009 — 20:27
  • Gordon:
    The dream case is helpful. It’s also helpful to think of a certain kind of novel where, suddenly, at the end “everything fits”, and in a surprising way.

    December 14, 2009 — 20:53
  • APT:
    The Nozick case is this: You plug yourself into a machine which will then give you a bunch of non-veridical experiences simulating whatever it is that you think is valuable in life. Surely that’s not what we really want.
    APT and John:
    That epistemically I am in the same position in situations A and B does not entail that I am equally well off in A and B. One way to see this is in terms of preference. We desire that our spouse be faithful, rather than that our spouse appear faithful. And the same is true for most other desires. We desire to be respected by our colleagues and friends, rather than to think we’re respected by them. We desire to know the answers to philosophical questions, rather than to think we know the answers. (If we just desired to think we know the answers, we could brainwash ourselves into that state.) That’s the phenomenology of our desires, and I think it is reasonable to suppose that it reflects the value structures involved. We do not, I think, desire to be the hero of a super-happy version of the Truman show where we never find out what’s going on.
    Consider the following scenario. George is an adult who has total amnesia. It is known that he will never recover his memories. We could convince him that he was a great hero, someone who secretly saved many lives. He would never find out he wasn’t, and would be very proud of himself. Should we? Surely not. Why not? Because it’s not true.

    December 14, 2009 — 21:09
  • John A.

    Alexander
    “Consider the following scenario. George is an adult who has total amnesia. It is known that he will never recover his memories. We could convince him that he was a great hero, someone who secretly saved many lives. He would never find out he wasn’t, and would be very proud of himself. Should we? Surely not. Why not? Because it’s not true.”
    If you knew the truth regarding the life George lived would you tell him if it was a life full of personal suffering and lose? I think that it is possible to argue that it is morally permissible to lie in certain situations.
    Regarding what we desire. I can agree with you, but my point is that it makes no difference. I can desire x and believe that I have x, but this does not change the epistemic fact that if I think I have x, and do not, but there are no defeaters for me holding this belief, then it makes no practical difference whether my memories are veridical or not. What messed Truman up was that there were defeaters.
    Let us assume that universalism is true and that God decides to short circuit the process and create us a saved beings with memories that we take to accurately reflect the life we have lived. These will include the memories of overcoming horrendous evils. But the events never happened. What epistemic difference does it make to how I view my life that these events did not happen if I happen to posses memories which I believe accurately reflect my life? It is not that I would choose this, but if I did not know, how am I being harmed especially when I would be acting the same way had the events actually happened.

    December 14, 2009 — 22:47
  • M.

    Now, the non-theist wants to press the problem of evil as evidence against the existence of God. But in light of (1), all that the non-theist can establish is (2). And that’s not enough to provide evidence against the existence of God, unless there is evidence against J. In other words, the onus is on the non-theist here, as she is the one trying to show that the evils of this world make it plausible that not-J.
    Right. I think most non-theists will simply assign a low prior credence in the existence of an eternal afterlife that preserves our psychology and memories, which doesn’t require assigning a low prior credence in the existence of God. This is similar to how theodicies which explain natural evils as the effects of freely-exercised demonic agency aren’t very convincing to non-believers.
    Here’s another way of thinking about it. Take an afterlife A in which my memory/psychology is unaltered, and an afterlife B in which my psychology is such that memories of painful events remain painful for me. Assume (harmlessly) there are no other kinds of afterlife. As a non-theist, I don’t find the existence of B much less probable than the existence of A; nor do I find C, the non-existence of an afterlife, implausible. Let E be the statement that a particular prima facie negative experience I’ve had is indeed bad. Then P(E) = P(E|A)P(A) + P(E|B)P(B) + P(E|C)P(C). Running induction against other hidden goods tells us that P(E|B) and P(E|C) are extremely high. Since additionally P(A) = P(B) and P(A) + P(B) + P(C) = 1, it follows that P(E) will be relatively high. So I’ve still gotten to the probable existence of gratuitous evils.

    December 15, 2009 — 0:16
  • christian

    I don’t know Alex. There are a number of different issues here. Suppose John freely does the wrong thing, all of the time. There’s certainly a sense in which nobody should rejoice in his freedom. But this is hard. Maybe we can distinguish someone’s freedom from their performing actions freely, and then rejoice in one and not the other. I find this hard to do, myself.
    But suppose I’m strange in this regard. My point can be put in a way that doesn’t take a stand on this issue. When someone looks back and remembers being raped, even if they come to forgive their rapist, and even if they come to appreciate the value of the freedom of their rapist, there is still the memory of being raped. Perhaps the memory will fade in time, as well as the intensity of one’s mental anguish, but I suspect there will always remain a deserved sense of being done an injustice. The disvalue multiplies here, as the value of reaching the summit would multiply, through recollection, in the case of reaching a summit after a painful hike.
    Of course, one might simply remember the pain of the hike too and forget the sense of accomplishment. So, I suppose, even in your original case it could be that disvalue gets multiplied. I think another premise is needed. It seems that you’re picturing the afterlife in rose-tinted glasses, but it could just as likely go in a much darker direction. If things go as some Christians think they will, perhaps there is a way to defend this premise. But then we have a cost. Some versions of the argument from evil might suppose there is no eternal life, a thesis which it’s objectors will deny. It seems the same thing can be said in the other direction. The kinds of justifying goods you’re appealing to would only be available if a certain conception of the afterlife were true. This is a thesis that proponents of the argument from evil will deny.
    The issue then turns to assessing prior probabilities, I think.

    December 15, 2009 — 1:50
  • John A.

    Alexander
    “Eventually, the cumulative value of enjoying these memories will overshadow the bads which were confined to one decade of one’s life (the climb, plus about ten years during which the memories of pain were painful). (This of course reminds one of Parfit’s Repugnant Conclusion. But I think there is nothing repugnant here.)”
    To change our focus a little – would it ever be possible for a person to say “I have overcome horrendous evils but I wish that I did not have to go thru the experiences I did”? It seems that your position requires that a person who overcomes horrendous evil be in a position where they understand that evil to be a necessary step in their personal development such that they could say “I am who I am today as a result of going thru the horrendous evils that I went thru and I am glad that I went thru and overcame these experiences. I am a better person for it.” Even if we grant that someone could say this and believe it to be true can that person then say, “but I wish I did not have to go thru that experience. I was not that bad a person before these experiences and I would give up being better if it meant not experiencing what I did.” If that person can say and mean this, then it seems that the problem of evil has not been resolved even if there are no longer have any painful memories. Besides, what about those who never overcome the horrendous experiences they have?
    I have a question for you (and anyone else): Does God violate our free-will by making us experience things we would not knowingly and freely choose to experience had we been given the chance?

    December 15, 2009 — 6:49
  • Mike Almeida

    The above remarks show that remembering over a significant length of time can act as a value-multiplier, and when the length of time is long enough (and in particular when it is infinite), this can completely swamp the original assessment.
    Alex, I think there’s a quasi-repugnant-conclusion that might be a serious problem. Go to a world in which everyone’s life is utter torment with no relief until death. Make it as bad as you can imagine; it does not matter at all how terribly bad it is. In that world a barely perceptible pleasure over an infinite (finite, for that matter) amount of time swamps the life long torment. Nonetheless it is hard to see how God would be justified in permitting the torment. First the benefits are barely perceptible and second they are realizable with much less pain.

    December 15, 2009 — 7:21
  • Good questions, again.
    Mike:
    If the very same benefits can be realized with much less pain, then the theodicy won’t work.
    John:
    It is possible to have all sorts of wishes. It is possible to wish that Fermat’s Last Theorem were false. I do not think an unfulfilled wish is necessarily a source of unhappiness.
    I don’t see how our free will is violated when we’re made to undergo things we wouldn’t have chosen for ourselves. Maybe our autonomy is, but I don’t think we have claims of autonomy against God.
    Being deceived is in and of itself bad. That doesn’t mean we have to tell people everything. There are morally salient differences between (a) not telling someone something true; (b) telling someone something false; and (c) creating an illusion. In case (a), we need not be intending a deception.
    M.:
    “I think most non-theists will simply assign a low prior credence in the existence of an eternal afterlife that preserves our psychology and memories, which doesn’t require assigning a low prior credence in the existence of God.”
    There is a question whether this credence assignment is rational.
    Moreover, such a credence assignment directly gives one an argument against the existence of God:
    1. If God exists, we have eternal life (stipulate that this means: with psychological continuity).
    2. We do not have eternal life.
    3. Therefore, God does not exist.
    The argument for (1) is by reflection on the great non-fungible value of the life of persons, such that the permament loss of any human being’s life (or at least of a virtuous human being’s life; but I take it that the non-theist also assigns a low credence to virtuous people having eternal life) is a loss of cosmic proportions. (Why cosmic? Well, think of someone you love. Suppose that the Virgo Cluster of galaxies contains no life and its destruction would not adversely impact life. Wouldn’t be better for the Virgo Cluster to be destroyed than for that person to permanently cease to exist?)
    Now, it’s also plausibly pretty unlikely that there be eternal life if there is no God. Thus, the non-theist should set her credence for there being eternal life roughly equal to her credence for the existence of God, prior to the consideration of evils (other than death as such–there is stuff to think through here).
    The argument from evil (other than death as such) then is supposed to make the non-theist significantly further lower her credence that God exists. However if my argument shows that there is no problem of evil if there is an eternal life (it doesn’t, but I am imagining that it is fully successful), then the problem of evil (bracketing death as such) cannot lower the credence that God exists below the credence that there is eternal life. And that’s basically where it already was for the non-theist. So, yes, my argument would succeed in neutralizing the distinctive evidential contribution of the problem of evil (bracketing death as such).
    “Take an afterlife A in which my memory/psychology is unaltered, and an afterlife B in which my psychology is such that memories of painful events remain painful for me. Assume (harmlessly) there are no other kinds of afterlife. As a non-theist, I don’t find the existence of B much less probable than the existence of A; nor do I find C, the non-existence of an afterlife, implausible. Let E be the statement that a particular prima facie negative experience I’ve had is indeed bad. Then P(E) = P(E|A)P(A) + P(E|B)P(B) + P(E|C)P(C). Running induction against other hidden goods tells us that P(E|B) and P(E|C) are extremely high. Since additionally P(A) = P(B) and P(A) + P(B) + P(C) = 1, it follows that P(E) will be relatively high. So I’ve still gotten to the probable existence of gratuitous evils.”
    Given the high correlation between eternal life and the existence of God (P(eternal life|God) is very high by my argument earlier in the comment, P(eternal life|~God) is tiny, etc.), you cannot assume that A and B are equally likely, because P(A|God) and P(B|God) are not equal.
    All that said, I think, as a methodological principle, that if too much in an argument rides on priors, something has gone wrong. One is then running into the problem of neutral evidence (see John Norton’s “Cosmic Confusions”).
    Christian:
    A sense of having been done an injustice need not hurt.

    December 15, 2009 — 8:59
  • John A.

    Alexander
    “. I do not think an unfulfilled wish is necessarily a source of unhappiness.”
    I agree, but does it not have to be the case that all experiences of evil will result in some greater good coming about even if it takes a long long time? (Remember that I am arguing from a universalist perspective) This seems to be implied by your example that given enough time all memories of painful experiences will lose their sting and be seen as necessary for some good to come about. If it is not, then there are cases of evil that do not result in good and if God has foreknowledge then He ought not to allow these to happen. I am assuming that the standard theodicy that God has reasons that he knows that justifies Him in allowing evil to exist is true.
    “Being deceived is in and of itself bad. That doesn’t mean we have to tell people everything. There are morally salient differences between (a) not telling someone something true; (b) telling someone something false; and (c) creating an illusion. In case (a), we need not be intending a deception.”
    A consequentialist would not agree with your 1st sentence, but assuming it is true I think that the distinction between a, b and c is forced (especially if my intention is to deceive or mislead – it is important to remember that for the deceiver a successful deception is good. (Hence the reason why Descartes’ Evil Genius is such a problem -for him, deception is not an imperfection.) Can you give an example.
    Your right, God does not violate my free-will- He violates my autonomy. I am not sure that we cannot makes claims of autonomy against God though. I would like to see that argument if you have time or if you can direct me to a source. My intuition is that if God created us with free-will then it is not fair of Him to punish us additionally to any harm we might experience as a direct result of actions that we knowingly and freely choose to perform. In short, why Hell (or Heaven for that matter)?
    I appreciate your responses to my comments. Thanks

    December 15, 2009 — 9:57
  • M.

    The argument for (1) is by reflection on the great non-fungible value of the life of persons, such that the permament loss of any human being’s life (or at least of a virtuous human being’s life; but I take it that the non-theist also assigns a low credence to virtuous people having eternal life) is a loss of cosmic proportions. (Why cosmic? Well, think of someone you love. Suppose that the Virgo Cluster of galaxies contains no life and its destruction would not adversely impact life. Wouldn’t be better for the Virgo Cluster to be destroyed than for that person to permanently cease to exist?)
    Note that this is an argument that God implies an afterlife, not that God implies an afterlife in which I do and am all the things your response to PoE requires of me. I can imagine, for instance, God arranging for us to enjoy heavenly lives so much more sublime and elevated than our terrestrial ones that we never even bother to think back to our meager Earthly pleasures – or that in our exalted states we simply cease to relish them. (You may be proud of your current car, but if you exchange it for a vastly more sleek and expensive one, you might soon find it difficult to imagine having lived with anything less.)
    Given the high correlation between eternal life and the existence of God (P(eternal life|God) is very high by my argument earlier in the comment, P(eternal life|~God) is tiny, etc.), you cannot assume that A and B are equally likely, because P(A|God) and P(B|God) are not equal.
    A and B are, strictly speaking, both cases where eternal life exists, so I assume you mean “Mt. Everest-style” eternal life here. Now, it’s true that P(A|God) != P(B|God), but I don’t think that matters. Let God* denote an omnipotent and omniscient but hugely evil entity. As a non-theist, I find God and God* equally probable (at least on my priors). So P(A) = P(A|God)P(God) + P(A|God*)P(God*) + P(A|Other)P(Other). Similarly for B. Since as far as I know, P(A|God) = P(B|God*) and P(B|God) = P(A|God*) and P(A|Other) = P(B|Other) (all by symmetry), I am licensed to conclude P(A) = P(B).

    December 15, 2009 — 10:20
  • M.:
    1. I think you shouldn’t find God and God* a priori equiprobable for much the same reason that you shouldn’t find a priori equiprobable the hypotheses (a) there is a force of nature whereby all massive objects are attracted to one another and (b) there is a force of nature whereby all massive objects other than pairs of hammers are attracted to one another. If you found (a) and (b) a priori equiprobable, you would not be justified in believing in the universal law of gravitation (OK, it’s only an approximation, but let’s bracket that), because all our observational evidence is equally supportive of (b) as of (a). (As far as I know nobody has every bothered to measure the gravitational force between a pair of hammers. If they have, then just vary the example.) There are simplicity considerations and (a) is simpler than (b), just as the hypothesis of a being that has all perfections is simpler than the hypothesis of a being that has such-and-such perfections but not others.
    If says that God and God* are a priori equiprobable, by the same token one might just as well as say (I think this is basically an argument Al Plantinga makes) that the following hypotheses are all equiprobable: There is no deity at all; there is a God; there is a God*; there is a God3 (say, he’s omniscient but not omnipotent, or whatever); there is a God4; there is a God5; …. But if they’re all equiprobable, then they each individually have zero or infinitesimal probability, and in particular the probability that there is no deity at all will have zero or infinitesimal probability. But that’s surely a pretty silly argument against atheism. Why is it silly? Maybe because the hypothesis that there is no deity has a simplicity that the hypothesis that there is a God177 does not have, and hence the two are not on par. But by the same reasoning, the hypotheses that there is a God and that there is a God* are not on par.
    2. It seems to me–and this is really hard stuff to work out–that we are running into the neutral evidence issues Norton is working on. It seems that the right thing to say is something like this. We observe this life with its evils. If there is a God, then there is an afterlife. It will do no harm to assume for the sake of the argument we know nothing about what this afterlife would be like, apart from the fact that it is good and just. In particular, we have no evidence that the afterlife is such as I describe it, but neither do we have evidence against this claim. If it is such as I describe it, then, let us suppose for the sake of the argument, the problem of evil disappears (this is a great overstatement–there are the difficulties other commenters have pointed out). But if that’s the scenario, then to argue that evil is still evidence against the existence of God, we are relying on probability assignments to something we know nothing about. And that’s problematic.
    3. I have grading to do, so I can’t try to work out the probabilistic details rigorously. I am guessing that one problem is that you may be independently assigning probabilities to eternal life hypotheses, evils and God. But these things are not independent. Consider things a priori. Let E be the existence of evils for which there is no justification in an earthly life. Let L be the hypothesis that there is eternal life, with L1 being my subhypothesis, and with there being various other subhypotheses. Let L* be the disjunction of all the afterlife subhypotheses that solve the problem of evil, including L1 if I am successful. Let G be the existence of God.
    Then, still reasoning a priori, we observe that G&E entails L* (or at least P(L*|G&E) is extremely high). Our priors, to be consistent, must take this into account when setting priors. There should be an appropriate order to the setting of priors, at least if we make use of any principle of indifference. In particular, we cannot set our priors for L* and for E before we set our prior for G. The reason for that is that the hypotheses that are explanatorily prior must have their prior probabilities set first (I can argue for this, I think; I think induction requires it). So we set a prior for G, let’s say 0.1. Now we need to set priors for E and for L*. To do that in a way that properly ensures consistency, we partition and set the priors for the conditional probabilities: P(E&L*|G), P(E&~L*|G), P(~E&L*|G), P(~E&~L*|G), and similarly conditioning on ~G. Now, P(E&~L*|G)=0. That leaves: P(E&L*|G), P(~E&L*|G) and P(~E&~L*|G). My story shows, let us suppose, that the disjunction L* is not empty, so none of these should be zero. If we’re going to be setting probabilities where we know nothing–and I think that’s foolish for Norton’s reasons–and we use some principle of indifference, we’ll give each of these probability 1/3.
    Consequently, P(E|G)=P(E&L*|G)=1/3.
    We then need to condition on ~G and consider P(E&L*|~G), P(E&~L*|~G), P(~E&L*|~G) and P(~E&~L*|~G). Since you’ve pointed out that ~G allows for some supernaturalistic hypotheses, I guess we can’t say that P(L*|~G) is zero. But it’s fairly small, still, I think, because by far the most likely thing given ~G is naturalism (being the simplest ~G-hypothesis). However, for our purposes, we won’t need to consider L* as it turns out. Let’s just set the probabilities for P(E|~G) and P(~E|~G).
    Option 1 (not the right option): Use some principle of indifference, and say P(E|~G)=1/2 and P(~E|~G)=1/2. In that case, E is indeed evidence against G, as P(E|G)=1/3<1/2=P(E|~G). But it is pretty weak evidence against G. The occurrence of some event that has probability 1/3 on a hypothesis H and has probability 1/2 on its denial is really not very impressive. Suppose we have an urn, and it either has three balls, one black and two white, or it has one black and one white ball. We draw out a black ball. Yes, that favors the second hypothesis. But nobody would say that it does so in any significant way. Even on the first hypothesis, 1/3 of the time we’d draw out the black ball.
    Option 2 (the better option): Observe that E requires the existence of fairly sophisticated beings, perhaps even of conscious beings. On naturalism, the prior probability of the existence of such beings is fairly small. There are so many possible worlds where there is just a high entropy mess. Of course there are non-naturalistic hypotheses in ~G, but the bulk of ~G is naturalism. I’ll grant, however, that if naturalism is true and if there are the sorts of beings that can have evils happen to them, then very likely E. Anyway, I think the net result is that P(E|~G) is not very large. Certainly not larger than 1/4, and quite possibly much, much smaller. So let’s say it’s 1/4. But now P(E|G)=1/3>1/4=P(E|~G). And so E is support–but very weak–for G.
    John:
    Actually, a desire-fulfillment consequentialist, like Peter Singer, is committed to the claim that being deceived is bad when we have a desire not to be deceived. And in most matters most of us do have a desire not to be deceived.
    The difference between silence and deceit/lying is easy. In silence I refuse to offer or manufacture evidence for p. I am not thereby offering or manufacturing evidence for ~p.
    The difference between deceit and lying is a bit harder, and I don’t need it for my argument. Still, roughly, it’s a difference between manufacturing evidence and breaking trust.

    December 15, 2009 — 11:44
  • John A.

    Alexander
    Let me concede for the sake of the argument that I agree with you regarding our desire not to be deceived. The fact that you and I exist is good evidence that we are the type of beings that do not want to be deceived. But, my original issue was whether God, at the time of creation, being a universalist, would have been better served had He created beings that were ‘saved’ with ‘memories’ of experiencing and overcoming horrendous evils. (We would not be these beings.) Would it be reasonable for such beings to ask if their memories were veridical when all the evidence suggests that they are? Because they have these memories they do not doubt that they ‘experienced’ what they did and overcame the evils they ‘faced.’ The only difference between your scenario and mind is that my ‘saved’ people did not really face any evils. But, from an epistemic point of view they react as we would if we were the climber in your thought experiment. So the question is, what does God gain by allowing evilto exist and having people experience horrendous evils when the same results can be achieved without evil?

    December 15, 2009 — 16:54
  • M.

    1. I’m not sure a hypothesis’ simplicity translates into evidence in the way required here. For example, suppose I use a computer program to generate a secret binary digit. I then use a completely different program to generate a million other binary digits, each of which turns out to be 0. The hypothesis “All of my numbers are 0” is simpler than the hypothesis “My first number is 1, and all million others are 0,” but this can’t seriously count as evidence that the first number is 0, since it was generated by a different process. For similar (if terribly undeveloped) reasons, I don’t feel simplicity should make God more subjectively probable than God*. Why should perfections X and Y bear on a completely different kind of perfection Z? Unlike your hammer example, I can’t think of some particularly obvious relationship (like a universal force of gravitation in the example) that would likely render perfectly powerful and intelligent beings perfectly good. Just a rather hazy thought.
    2. I wasn’t aware of Norton’s paper before you linked it earlier and will have to consider it further. You may very well be correct here.
    3. I think we should be conditionalizing on some background assumptions about the world, including the existence of conscious beings. The germ theory of disease doesn’t say anything about whether conscious beings exist, whereas Christian Science’s theory of disease perhaps does; but I don’t think this tells us anything useful about the relative (prior) probabilities of the two theories.

    December 15, 2009 — 17:01
  • John:
    Assuming your beings are smart enough, they might think of your argument and start worrying about the veridicality of your argument. 🙂
    M.:
    In the case where there are random binary digits, you’ve already specified how the digits are to be generated–namely, randomly and I assume independently (as otherwise I don’t think the example works). Conditioning on this specification, we must assign the same probability to each sequence.
    But it’s different if we don’t have this specification in our background. Suppose that you come across a very long binary string as you walk along the beach, stretching off towards the horizon (it’s a long beach). You have no idea what process, if any, generated this string. Suppose that you assign equal probability to every possible binary string.
    Now, you observe. You see 010101010101010101. What prediction should you make about the next digit? Surely that it’s a 0, because surely it is right to conclude that the digits were probably generated by a process that alternates digits. But if your priors assign equal probability to every string of binary digits, then given what you’ve observed, the probability that the next digit is a zero is 1/2 and the probability that the next one is a one is also 1/2.
    In other words, equal probability assignments in a case like this make induction impossible.
    Instead, to make induction work in a setting with prior probabilities (to be honest, I am somewhat sceptical of the whole Bayesian setting, for reasons contained in my posts from Sept. 25-28 of this year on my own blog), we must assign higher probabilities to “simpler” hypotheses. It’s hard to do this in a rigorous way. One could fix a canonical language and use that to define the probabilities (basically, things that can be described more briefly get higher probability). But the notion of a canonical language is iffy.
    I think that in the bayesian setting one should work out all the priors without assuming the existence of conscious beings, and then conditionalize on this existence.

    December 15, 2009 — 21:19
  • Dustin Crummett

    We would not be these beings.
    If these beings wouldn’t be *us*, then why are we complaining? Perhaps God created these other beings as well. Surely he can create us, too, if he wants? Or have I misunderstood you?
    Also–I haven’t thought much about this and am only throwing it out there–with regard to the question, “Who would know?” well, God would. Is it possible that some action required by God in John’s scenario, or some piece of knowledge God has about our true histories, might, even in some slight way, disrupt our relationship with him?

    December 15, 2009 — 23:11
  • John A.

    Alexander
    “Assuming your beings are smart enough, they might think of your argument and start worrying about the veridicality of your argument. :-)”
    I wonder if ‘saved’ people could seriously entertain the possibility that they have been deceived. If they did question the veridicalness (word?) of their memories would this not be evidence that they were not saved? 🙂 Can a person who is saved become unsaved?

    December 16, 2009 — 9:50
  • M.

    In my example, I didn’t specify that the selection of bits was random or even independent. We only know that there are two very different computer programs being used. The algorithm each uses to generate bits is unknown (though they’re known to be quite different from each other). The first one might be programmed to influence the second somehow, but if so we wouldn’t know in what capacity. It still seems to be the case that simplicity considerations don’t lend weight to the hypothesis that my first number is 0. And this seems to be because we have reason to expect my first number and all subsequent numbers not to have been generated by the same process (though we’re totally in the dark as to what the two processes were): note that this doesn’t hold for your beach example. In other words, it’s the availability of a simple explanation (like being generated by a single process) that allows us to favor simple hypotheses (like all my numbers being 0). But if that’s the real reason, it seems we can say something important about perfections. I can’t currently conceive of a simple, viable, non-circular explanation E of a being B’s necessary omnipotence and omniscience such that E entails that B is morally perfect, as well. (I don’t think that “maximal greatness” would do the trick, since that’s just a conjunction of properties to be explained.) Therefore, it seems mistaken to imagine we should invest more prior credence in God than in God*, or that we could perform induction on different sorts of perfection.
    Is this coherent at all?

    December 16, 2009 — 12:38
  • M.:
    I guess the information that the process is different decreases the weight of the simplicity consideration. But I think you still have a tiny bit of support for the hypothesis that the string of bits you don’t know about is the same as the other string of bits.
    Here’s a fun experiment you can actually do (I haven’t done it, but I am relying on some fairly well known stuff). Put two people in separate rooms, with no communication. Ask each to write down a number between 1 and 100 and put it in an envelope. Then you open one envelope, and see a number. The following procedure works better than random at guessing what the other number is: Simply suppose that the other number is the same as the one you know. (I.e., if you follow this procedure, you have a better than 1/100 chance of getting it.)
    Why does this work? Because there are some numbers which people are more likely to pick when they’re asked to pick a number, because they pick a “random-looking” number. For instance, people will rarely pick 1, 10 or 50. They are likelier than random to pick a number that ends in a seven. Etc. But you do not need to know these facts to know that there is likely to be a correlation.
    Likewise, if you have two computer programs, they are likely to have been produced by two computer programmers. Suppose program 1 implements algorithm A. Then if B is a randomly chosen algorithm, there is a greater chance that program 2 implements A than that program 2 implements B.
    Now, you can control for all that and say that there is no similarity between the programmers, etc. Then your reason for prefering the simpler hypothesis to any given more complex one disappears. In fact, then you should prefer the more complex one. But that’s because you have additional information.
    As for the God case, one move is to identify a very small number of properties which entail all the others, and that makes for the God hypothesis being simpler. Swinburne thinks two will do: perfect freedom and perfect knowledge. Aquinas thinks one will do: absolute explanatory firstness. Anselm, Descartes and Leibniz also think one will do: maximal greatness.
    Maximal greatness is not just a conjunction of the great-making properties. It is the property of having all the great-making properties. It is thus defined in terms of one second-order property, the property of greatmakingness (or maybe: completepositivity–then we can use a Goedelian style axiomatization). x has maximal greatness iff (F)(greatmaking(F) → F(x)). Universal quantification over second order properties is not conjunction, just as universal quantification over individuals is not conjunction.

    December 18, 2009 — 9:04
  • M.

    Likewise, if you have two computer programs, they are likely to have been produced by two computer programmers. Suppose program 1 implements algorithm A. Then if B is a randomly chosen algorithm, there is a greater chance that program 2 implements A than that program 2 implements B.
    Remember, we’re conditionalizing on the fact that the two programs follow very different algorithms. If as far as we knew the two algorithms (1. and 2.) were randomly chosen by the programmer(s), then I’d agree with you that at the end of the experiment we’d have some slight evidence for the identity of my first number. But we’re distributing our credences for 2. only among algorithms which are highly dissimilar from 1., given some appropriate measure of dissimilarity; and presumably we have no reason to expect the programmer’s choice to be even slightly biased here towards one kind of output.
    It’s as if we’re asking the person in the one room of your thought experiment to pick a random two-digit number, and the person in the second room to pick the last two digits of his/her social security number. Now, it may be the case that there’s a reason out there to expect a positive statistical correlation between such numbers, but we have no information in that regard. All we know is that SSN’s aren’t assigned by anything like humans mentally picking numbers at random.
    However, I’m not sure how important this point is in the context of our earlier discussion. Assuming 1. we can agree that the evidence simplicity considerations lend us in situations of type T is at best negligible, and 2. I’m correct about simplicity of perfections being of type T, then 3. it follows that P(God) is only negligibly greater than P(God*). This, I think, would be enough to carry the claims I’ve made which ride on these matters, at least assuming they didn’t/don’t have separate problems.
    As for the God case, one move is to identify a very small number of properties which entail all the others, and that makes for the God hypothesis being simpler. Swinburne thinks two will do: perfect freedom and perfect knowledge.
    Yes, I’ve had to tacitly assume that arguments for moral perfection like Swinburne’s don’t ultimately work. My reason is that I’m not a moral cognitivist, so I don’t expect moral “truths” to really fall within the purview of omniscience. But that’s me.
    Maximal greatness is not just a conjunction of the great-making properties. It is the property of having all the great-making properties. It is thus defined in terms of one second-order property, the property of greatmakingness (or maybe: completepositivity–then we can use a Goedelian style axiomatization). x has maximal greatness iff (F)(greatmaking(F) → F(x)). Universal quantification over second order properties is not conjunction, just as universal quantification over individuals is not conjunction.
    O.K., that’s fair. But if maximal greatness can nevertheless be logically deduced from necessary omniscience, omnipotence and moral perfection (and vice-versa), I’m not sure it really counts as an explanation of any of them. As you can probably tell, though, I’m not very familiar with the literature on explanation, so I can’t press this point too forcefully.

    December 18, 2009 — 14:21
  • Yeah, so the big question is the unity of the concept of God.
    I don’t actually know whether maximal greatness can be deduced from omniscience, omnipotence and moral perfection (ooamp). For instance, here are some other greatmaking properties: (a) being perfectly beautiful; (b) being creator of everything contingent; (c) being necessarily existent; etc. It is far from clear that (a), (b) and (c) follow from ooamp. For no list of greatmaking properties would we have much reason to think the list complete.
    In any case, it seems quite possible that p should explain q even though q entails p. Here are some cases:
    – That snow is white explains why snow is white or 2+2=5.
    – That x caused y explains why y exists even if essentiality of origins makes the existence of y entail x’s causing y.
    – One can have mathematical explanations. In these, p explains q, even though both p and q are both necessary truths, and hence in particular p is entailed by q.
    The other move I can make is that even if the concept of God is defined by the conjunction of three properties, that’s still a pretty simple concept. Competing deity concepts will likely have to have free parameters. For instance:
    – Omniscient, omnipotent, but of moral goodness level L.
    Here, L is a free parameter, and free parameters tell significantly against the simplicity of a theory. Maybe this would be better:
    – Omniscient, omnipotent, but perfectly evil.
    However, I am not sure this makes sense or is coherent. It seems to me more plausible that there is no perfection in evil than that there be no perfection in goodness.
    Moreover, to be omnipotent, one needs to be an agent. Agency requires acting on what one takes to be reasons. To take something to be a reason is to take it to be a consideration that in fact favors doing the action, and that entails believing it to be a consideration that in fact favors doing the action. But if one is omniscient, then one only believe p if p. So, an omniscient agent always acts on the basis of considerations that in fact do favor the action. But that means that if an omniscient agent does A, there always is a consideration in favor of doing A. But then there is some value in doing the action, and we do not have perfect evil.
    You might, of course, dispute that there are facts about what is and is not a consideration in favor of an action. That will be a deep disagreement here.
    Alternately, you might say that the considerations the agent acts on might be prudential rather than moral. Here, it is I who hold the controversial view. All prudential reasons are moral reasons. (It’s kind of odd that this so extremely controversial. After all, it is a consequence of consequentialism, virtue ethics and natural law. Moreover, while Kant will deny that prudential reasons are moral reasons, he will say that reasons of benefiting self are on par morally with reasons of benefiting others, and that just as one has duties to benefit others, one has duties to benefit self.) There is no difference in kind between the reasons I have to benefit myself and the reasons I have to benefit another.
    But suppose we make the distinction. Now we have the puzzle: What sort of prudential considerations could move an omnipotent and omniscient being to do actions that are perfectly evil? A prudential consideration is based on an apparent good to oneself. (As Heda Segvic pointed out to us in class once, even Hume thinks that we always act for the sake of an apparent good.) If one is omniscient, one can delete the word “apparent”. So, the prudential consideration is based on an actual good to oneself. But what sorts of goods to self would get an omnipotent being always to behave in a perfectly evil way? If it’s for the sake of pleasure, the omnipotent being could surely get pleasure in another way.
    Maybe, if one has a distinction between prudential and moral reasons, we could try to suppose a being that is omniscient, omnipotent and amoral–a being that never acts on moral reasons, but only on prudential ones. But here is a puzzle: Why does this being never act on moral ones? (It’s not as if the being is ignorant of the moral reasons.) Is it just chance, or a matter of character? If it’s just chance, then the hypothesis is improbable, because it is unlikely that every time the agent would choose a prudential rather than moral reasons, when aware of both.
    So suppose it’s a matter of character. The agent is so constructed as to know both the prudential reasons and the moral ones, but to be entirely unmoved by the latter. But now our hypothesis is more complex than the following theistic hypothesis: there is an omniscient and omnipotent being that is moved by all the reasons that there are, and that always acts on decisive reasons when there are such.
    For there are infinitely many ways of specifying the subset of reasons that the being might be moved by. Maybe the being is moved by prudential reasons only; maybe the being is moved by non-prudential reasons only; maybe the being is moved by prudential reasons on Tuesdays, and on other days the being is moved by the subset of non-prudential reasons that are not grounded in the dignity of Alexander Pruss; maybe the being is moved by aesthetic reasons only; etc. All these hypotheses are more complex than the hypothesis of being moved by all the reasons there really are. And there is a pleasing simplicity about the hypothesis that whenever there is a decisive reason, the being acts on it.
    I am guessing that you’ll reject the cognitivism about reasons that makes them into considerations. Maybe you’ll say reasons are desires or something like that. That, of course, will be another debate. (My throwaway line is: Why should the fact that I want something be any reason to go for it?)

    December 18, 2009 — 16:08
  • DL

    Some belated comments:

    Alex: God could produce the better character directly

    I’m not so sure about that. I think the free-will defence applies here: for what is “character” but a collective term for all our volitions? (All right, it probably also includes some understanding and memory of our past acts of will, but surely you can’t have a character independent of your free choices.) So God presumably could override our character, but only by violating our free will.

    Alex: (What if one is in heaven and the perpetrator is in hell? I don’t know what to say then. Maybe one runs a Tertullian line about joy at divine justice, but that line makes me uncomfortable.)

    Don’t worry; after a few hundred millennia in heaven, you’ll get over it. =)
    But it’s also worth noting that the benefit doesn’t have to be directly connected to the pain, as far as I can see. A horrible crime may have all sorts of effects down the road that we can’t possibly see from our mortal perspective (though there are also cases of good coming out of evil that we can see). Perhaps in that case you could say something like the pleasant memories are gained posthumously, but I think your basic idea will work either way.

    John A: How would we be able to tell the difference? If God did a good job they would seem veridical to us, would they not.

    Indeed. But come to that, how do you know God didn’t do a good job, so good that you only think you actually experienced anything painful, when really your whole life is an implanted memory?!

    M: But why should God have allowed the experience to have been painful at all? Why not arrange it so that I get to have pleasant, veridical memories of at-the-time pleasant event, instead?

    Because evil creatures should get to feel evil? Not only do trials build character, but they may be necessary to build character; that is, it may be necessary for us as free beings to be able to sense evil in some tangible way, to experience it in order to understand the evil of our own bad acts. On top of that it is a suitable way for us to learn

    What about to Christian theists? They’ll have to wrestle with the reverse phenomenon in hell, where plausibly memories of one’s Earthly existence make things horribly, horribly worse.

    Actually, doesn’t that explain, or help explain, why Hell is our own fault? God wants to heal everyone, but some people reject that offer, and are thus stuck with their pain and suffering for eternity. It’s not a punishment in a retaliatory sense, but something the damned do to themselves by not letting God remove their pain.

    A.P. Taylor: Suppose that in order to be sure that I would believe the memories given to me were authentic, God crafted them to be consistent with not only my own non-Everest memories, but also the memories of everyone else. How then should I ever discover that they were not veridical? And if I cannot discover this, how could the memory be any less valuable than a genuine veridical memory (to me)?

    Sure, to you; but God would still know. How would He be able to sleep at night knowing that everyone’s life was a lie? Or, to turn it the other way around, even supposing that God could do that, I can’t see that He’d be obligated too. As long as Alex’s idea of healed pain works out in the end, then surely God could work the universe that way if He wanted to.

    John A:Does God violate our free-will by making us experience things we would not knowingly and freely choose to experience had we been given the chance?

    No. God is not obligated to provide us whatever world we think we might like.

    Gordon Knight: This does not remove the problem of evil, but it may help us think of it in a new light. Our perspective now is as of one in a dream.

    That is a very good point, and surely heaven is “more real” compared to this world than this world is compared to a dream. Yet I’ve never heard of a bad-dream argument against the existence of God!

    December 21, 2009 — 20:17
  • dguller

    Just wondering how any of this applies to natural disasters?
    I mean, the Free Will Defense does not apply, because it is only God who allegedly freely chooses to kill hundreds of thousands of human beings in a tsunami, for example, including thousands of children who horrifically drowned.
    How does one reconcile a God that is capable of causing such suffering with a God who genuinely loves human beings, is ever looking out for their best interests, and honestly wants to minimize their suffering?
    The fact is that God has the potential of creating an existence of neverending bliss and happiness. It’s called Heaven or Paradise. So, He COULD HAVE created it for us already, but chose to create an existence of impersonal and interpersonal suffering and pain.
    If my father claimed to love me, but punched me in the face on a daily basis in order to teach me to appreciate the moments of peace in life and to have empathy and solidarity with others who suffer, then I would be hard pressed to agree that he loves me at all, except in a perverse and pathological way. And it doesn’t help my father’s case if he gives me $10,000 after every punch. He still comes off as a real jerk.
    As has been known since antiquity, there are really only three possible solutions to the Problem of Evil:
    (1) There is no God.
    (2) There is an all-powerful God, but He is not benevolent.
    (3) There is a benevolent God, but He is not all-powerful.
    I don’t know. (1) seems the most plausible to me, because the world just appears to be one that does not occur under the benevolent guiding hand of an all-powerful deity. There is just too much needless suffering for that to be plausible for me.

    January 1, 2010 — 14:24
  • Actually, the standard Christian view is that our exercise of free will has significantly distorted the arrangement of the physical world or at least our place in it. So on that view, our free choices are behind either the tsunami or our being in its path. Of course, that raises the question whether it would be right for God to permit us this much freedom to distort the physical world.
    The memory story I give suggests that heavenly life may be more valuable for memory of sufferings on earth.

    January 1, 2010 — 17:28
  • dguller

    Or maybe the “standard Christian view” is just plain ridiculous. I mean, seriously? Our choices resulted in the shifting of the tectonic plates in the Indian Ocean that caused the 2004 tsunami? How exactly does THAT work?
    And sure, happiness today is often enhanced by comparison with past sufferings, but that is only because that is nature of our psychology, which God created. He could have easily created our psychology to not make such comparisons, and just automatically placed us in a heavenly paradise where there is only bliss and no suffering at all.
    The bottom line is that God freely chose to make humans suffer, both through natural causes and interpersonal relationships, when He did not have to, which appears to negate His qualities of benevolence and mercy.

    January 1, 2010 — 22:34
  • It’s not so much “by comparison” with past sufferings that our happiness is enhanced, but by thicker and more significant connections between the past sufferings and present happiness that our happiness is enhanced.
    When one has achieved a goal through adversity, this is objectively valuable independently of whether we experience it as such. That we do experience it as valuable is a contingent fact about us, just as it is a contingent fact about us that we can see green light as green. But if we did not experience it as valuable, we would be insensitive, just as if we did not see green light as green, we would be visually insensitive.

    January 2, 2010 — 11:58
  • And as for the Fall, there are many views we could have of it. Version 1: We were protected by God from various physical dangers, but by our free choice we put ourselves outside of that protection. Version 2: We had knowledge that protected us from various physical dangers (e.g., God would have told us when a tsunami was about to strike), but our disobedience to God removed that knowledge. Version 3: God had engineered a delicately balanced system such that tsunamis, earthquakes, etc. would never happen in inhabited areas. However, a delicately balanced system can be disturbed by an event that is not within the design specifications of the system. Sin was not within the design specifications of the system, but was a matter of our free choice. There is some reason to think that various climactic and geological phenomena are chaotic: that large changes in outputs can result from small changes in inputs.

    January 2, 2010 — 12:03
  • dguller

    Yes, many natural systems are governed by non-linear dynamics and, as such, are exquisitely sensitive to minor variations in initial conditions. However, the onus is still upon you to demonstrate how a free choice that happens to be sinful is capable of initiating a chain of events that ultimately shifts tectonic plates. Also, wouldn’t virtuous choices also affect initial conditions? Wouldn’t choices that God approves of also cause tsunamis?
    And something else that I do not understand: how can sin not have been anticipated by an all-knowing being? I thought He knew everything?
    I think that despite the multiple epicycles that are appended to theism to make it consistent, it all becomes quite silly beyond a certain point. I think that we have reached that point a long time ago. I also think that the clear conclusion is that the traditional conception of God is completely inconsistent, both within itself in many respects, and especially with the empirical world in which we live, and therefore must be amended or utterly rejected.

    January 2, 2010 — 12:57
  • The suggestion in question was that God designed a system such that things would be guaranteed to go well if the system were interacted with virtuously. Such a system, if implemented using simple laws of the sort we think our world is governed by, would be unlikely to be robust in its response to inputs–it would be likely that non-virtuous interaction would lead to bad stuff. Consider a different system that precariously stands between human beings and entropy–an airplane. Especially if we’re talking of a simpler airplane, one without much computer guidance, most possible interactions between the pilot and the plane would result in disaster. Only a very small minority of possible interactions preserve the safety of the human beings.
    In any case, onus of proof questions are tough–they require a very careful description of a particular dialectical situation. If you’re arguing that the Fall did not happen as described, then the onus seems to be on you.
    As for the anticipation of sin, that would require knowing lots of propositions of the form “If x were created and put in circumstances C, then x would freely do A.” But it is a controverted question whether such propositions are true when their antecedent is false. And if they are not true, then even an omniscient being doesn’t know them, just an omniscient being doesn’t know that 2+2=5.

    January 2, 2010 — 20:03
  • dguller

    First, it looks like you’re trying to make an analogy between the human input in running an airplane properly and the human input that goes into causing tectonic plates to shift. I think that this is a false analogy, especially since the former condition is clearly understood with respect to HOW a human being can behave in a way that will cause a plane to fly or crash. I still want to know how a human action, whether virtuous or sinful, can eventually cause tectonic plates to shift.
    Second, how can the onus possibly be on me to disprove the idea that the Fallen nature of humanity can have dramatic physical effects upon natural phenomena, such as earthquakes? I mean, doesn’t the fact that it violates the laws of nature suffice to put the burden upon you?
    Third, I still do not understand why God would not know what our free choice would be before we made it, especially since His knowledge should encompass the totality of our neurobiological inputs, processes and outputs. Therefore, He should know our choices before we make them, no? Unless there is something over and above our neurobiology that is both necessary and sufficient for our decisions? I hope you’re not going to bring the soul into this! 😛

    January 2, 2010 — 22:55
  • My intuition is that given laws like ours, there will be a very narrow range of inputs such that nothing bad will happen to humans. Any departure from this narrow range of inputs will result in bad stuff. How? Presumably by some sort of non-linear amplification effect of the sort we get in chaotic systems. I see no reason, even prima facie, to suppose this violates the laws of nature.
    As for knowledge of our actions, you are assuming determinism, which we do not have good reason, empirical or philosophical, to assume.

    January 3, 2010 — 19:12
  • dguller

    First, no more abstract philosophical handwaiving. I’d like to know how a sinful or virtuous choice can cause tectonic plates to shift. And how could you identify which specific choices were responsible for the tectonic shift?
    Second, the “laws like ours” were given by God, and thus he is responsible for their outcome.
    Third, regarding determinism, I wonder what “empirical” evidence you have against our neurobiology being responsible for our mental states.

    January 3, 2010 — 21:57
  • Sorry, here handwaving is all I have. However, note that I am not claiming that the free choice is responsible for tectonic shift. Perhaps the free choice is responsible, instead, for humans being present in the area where the tectonic shift occurs.
    The laws were given by God, but it might well be that either (a) no other laws are compatible with the existence of human beings, or (b) the same holds for any other set of laws that are simple enough to make human-like science possible. I find at least (a) quite plausible: if the laws were different, I don’t think we would have the same particles, and likewise the kinds of beings inhabiting the world would be different. But then whom has God, allegedly, wronged by creating these laws? Not human beings–for if he created different laws, there would be no human beings, and that’s worse (at least assuming an appropriate afterlife for human beings). Nor the beings of the possible world with different laws, since one cannot wrong a non-existent being.
    As for determinism, I do not see why philosophical argumentation does not count. Moreover, I do not think determinism is a consequence of the claim that neurobiology is responsible for our mental states. One needs an auxiliary hypothesis that neurobiology is deterministic. And that, I think, it’s too early to claim.

    January 4, 2010 — 9:34
  • dguller

    First, you are now revising your claims. Earlier, you wrote about the Christian view that “our exercise of free will has significantly distorted the arrangement of the physical world or at least our place in it. So on that view, our free choices are behind either the tsunami or our being in its path.” Now, you are denying that “our free choices are behind … the tsunami”, likely because you have realized that this is absolutely ridiculous, non-linear dynamics or no non-linear dynamics.
    Second, you appear to be arguing that God is limited in the laws that he can create, and not just in the trivial sense that the laws must be logically and analytically consistent. You are now saying that the universe HAD to be created with the exact natural laws and material entities for human beings to exist. Ordinarily, this would also be trivial, but theists have the belief that God can INTERVENE and ALTER the laws of nature at will, which seems to imply that they are not as hard and fast as you are describing. If you are a deist, then this objection does not apply to you, but then again, you would be unable to believe in the majority of Christianity or any revealed religion, either.
    Third, I never said that “philosophical argumentation does not count”. You stated earlier that there were “empirical” reasons against determinism. I was wondering what those may be. Perhaps you will cite quantum mechanics, which is the paradigmatic indeterministic theory. However, we were previously discussing my contention that our “free will” can only occur within our neurobiology, and thus must have a deterministic component. Quantum mechanics has nothing to do with neurobiology, except possibly by having small effects on terms of the function is neurons, which could be magnified into larger effects since the brain appears to be a non-linear dynamic system. However, this does not help libertarians at all.

    January 4, 2010 — 14:23
  • I think this discussion, interesting as it is in various ways, has come close to its natural end. I may not respond beyond this…
    1. If I first affirmed “p or q” and later “p”, I have not revised any claims. On the contrary, by arguing for p, I am a fortiori arguing for p or q.
    2. God cannot do what is metaphysically impossible, and it is, perhaps, impossible that we exist within a system with different laws than the ones we have. Of course, there could be beings very similar to us in a system with different laws, but they would not be us, just as the pieces in a game with rules slightly different from those of international chess would not be international chess pieces. Whether God can intervene in the laws is a different issue. I happen to think the laws all hold ceteris paribus, and so intervention does not entail a change of law, any more than a president’s pardoning a criminal entails a change of criminal law.
    3. On fairly standard interpretations of quantum mechanics, everything ends up being indeterministic. And that includes neural phenomena. Granted, some phenomena will be “approximately deterministic”–i.e., the probability of a particular outcome will be overwhelmingly high. However, if something is merely approximately deterministic, then it is, literally, indeterministic. As for the question whether quantum indeterminism helps the libertarian, that’s beside the point. You claimed that God should know the neural outputs if he knows the neural inputs. I pointed out that you appeared to assume that the system was deterministic. This point is independent of the question whether libertarianism is true, a question on which I have commitments, but which commitments I was not relying on here.
    It is an interesting related question exactly which morally relevant neural phenomena are approximately deterministic. I think at this point we simply do not know the answer.

    January 4, 2010 — 16:02