Rowe’s evidential argument from evil
October 7, 2010 — 13:44

Author: Alexander Pruss  Category: Existence of God Problem of Evil  Tags: , , , ,   Comments: 27

Rowe’s argument is basically:

  1. E is an evil for which despite serious study we can’t see a justifier.
  2. Therefore, probably, E has no justifier.
  3. Therefore, theism is false.

This argument is a bad piece of non-deductive reasoning. I will explain why.

Whenever someone gives us an inductive argument involving a particular case, it is appropriate to worry about selection effects in the choice of the case, and incorporate evidence arising from the nature of the selection procedure into the argument. In the case of (1), Rowe did not start with a list of the trillions of evils that happen in the world, and then pick out one, E, at random, and observe that we can’t see a justifier for it. There are many evils where people to whom they happened claim to know justifiers, and Rowe picked none of those. Rather, Rowe picked (or imagined) a particular case, E, which was such that we couldn’t see a justifier for it despite serious effort.

Thus, once we fill in the story about the selection method, Rowe’s argument really is:

  1. There exists an evil for which despite serious study we can’t see a justifier.
  2. Therefore, probably, there exists an evil that has no justifier.
  3. Therefore, theism is false.

Now it seems that I’ve made no progress in rephrasing the argument in this way. You existentially instantiate (4), thereby you get (1), thence you derive (2), and hence you conclude with (5). But this line of reasoning is fallacious in an inductive argument. Here is one way to see this. Let T be the theory that every large galaxy has a central black hole. Suppose we take T to be true. Then, I submit, we have very good reason to think that:

  1. There is a large galaxy that after serious examination does not look like it has a central black hole.

Why? Well, we know that large galaxies vary greatly in appearance. Even though we think every large galaxy has a central black hole, we have very good reason to think that large galaxies will differ in how evident that central black hole is. And we have very good reason to think that random variation will produce some large galaxies that don’t at all look like they have central black holes, despite having them. And so T gives us good reason to think (7) is true.

Consequently, we should be able to existentially instantiate (7) and get:

  1. G is a large galaxy that after serious examination does not look like it has a central black hole.

And then we conclude:

  1. Probably, G does not have a central black hole.
  2. Therefore, T is false.

But 7-10 is a bad argument against T, as can be seen from the fact that T is a serious astronomical theory while 7-10 is entirely a matter of armchair theorizing. But, I submit, neither would it be a good argument against T if after poring over DSS2 images we managed to find a particular large galaxy that after serious examination did not look like it had a central black hole. For armchair astronomy already told us that there would be such a galaxy (I suppose one can pore over DSS2 images in an armchair, too, but I guess that doesn’t make it be armchair astronomy).

Likewise, because we know that there are lots and lots of evils (over a hundred trillion if we do this quick calculation: 10 billion people, each living about half a century, each experiencing about one evil a day), we have good reason to expect that even if every evil has a justifier, some of the justifiers are going to be more obvious and some are going to be less obvious. And among those that are less obvious, there will be ones that are (roughly) the least obvious–and those are likely to be quite unobvious indeed. Thus, we have very good reason to expect that (4) is true even if every evil has a justifier. And hence (4) is not significant evidence against the claim that every evil has a justifier.

To put it briefly: Theories that apply to many cases are likely to have both weak and strong “anomalies”. C is a strong anomaly for a theory T if independently of the evidence for T, C looks like a counterinstance to T. C is a weak anomaly for T if independently of the evidence for T, it’s not the case that C looks like an instance of T. The mere fact that a general theory has an anomaly is not in general significant evidence against the theory. Scientific practice reflects this.

It is pseudoscience that latches onto unrepresentative anomalies. Think of pseudoscientific criticisms of evolutionary theory (I am not claiming that all criticisms of evolutionary theory are pseudoscientific). “Nobody has come up with an evolutionary explanation of biological feature F despite serious work. Therefore, F has no evolutionary explanation. Therefore, evolutionary theory is false.” This is a bad argument, because armchair biology would lead us to think that evolutionary explanations will sometimes be hard to find, even if all biological features have them.

The fallacy in question is an interesting one. There are two ways of trying to analyze it. One is that existential instantiation is problematic in inductive arguments. That’s an intriguing suggestion. The second is simply that the argument confuses what is we see independently of the evidence for T and what we see given the evidence for T. There are going to be cases where our only reason for thinking that they fit with T is the general evidence for T.

Anomalies are to be expected and the mere fact that a theory has an anomaly does not make us question it. Indeed, the lack of an anomaly can make one suspicious (if all of an experimenter’s data looks too close to some curve, one is worried about fraud). Of course, anomalies may pile up. Or they may turn out to be representative. If computer simulation shows that the black-hole-center theory predicts that only 5% of galaxies would look like they don’t have a central black hole, but 10% look like that, then that can be strong evidence against the theory. Interestingly, though, if the theory predicts that 5% would look like they don’t have a central black hole, but only 2% look like that, that’s also a problem for the theory!

Likewise, Rowe’s argument from evil can be replaced by new arguments, such as (a) that the percentage of apparently unjustified evils is greater than that which would be predicted by the theory that all evils have justifiers, or (b) that the number of evils is greater than theism predicts, or (c) that God couldn’t permit there to be a single evil for which a justifier is not evident. But each of these is a significantly different argument from Rowe’s original argument: (a) requires serious statistical investigation; (b) is messy; and (c) is deductive (I am grateful to Adam Pelser for this point), and hence needs only a defense to be refuted.

Observe that this response is not a sceptical theist response. It is not sceptical: on the contrary, it is thoughts like these that keep us from been sceptics about scientific theories. And the response is not based on any controversial theses about the realm of value. Rowe simply commits a fallacy of non-deductive reasoning, the fallacy of taking anomalies too much to heart. (Perhaps this is a fallacy that occurs elsewhere in philosophy.)

Notice also that while one might, and probably should, worry that (1) is apt to confuse absence of seeing with seeing of absence, even if one grants (1) with the stronger “seeing of absence” reading (which corresponds to “strong anomaly”), the point remains.

I do not dispute that the existence of a single anomaly is incremental disconfirmation of the theory. But the existence of a single anomaly–or even of a moderately large number of them (if 0.1% of the hundred trillion evils are likely to be anomalous–surely a not unreasonable estimate–the numbers of anomalous evils will be large)–does not significantly lower the probability of a theory. Such incremental disconfirmation should not worry theists. Rowe’s argument is, I think, over.

Comments:
  • Minor correction: We actually have a touch more information in 4: Among the evils we know about, there is one for which we don’t know a justifier despite serious examination. (I am grateful to Steve Kuhn for remarks that made me realize this.)
    This slightly affects the galaxy analogy. But given that we know of so many evils, the basic point remains.

    October 7, 2010 — 22:32
  • Gene Witmer

    Here’s one worry I have about your criticism of Rowe. Consider again the first premise of Rowe’s argument as you put it:

    E is an evil for which despite serious study we can’t see a justifier.

    I take it that the heart of your criticism is just that, given what else we know about how many evils there are and their likely characteristics, we should expect there to be some that meet the description in the first premise — those for which, despite serious study, we can’t see a justifier.
    But I don’t see why you think we should expect that. Here’s what look like the crucial move in your post, which you make right after giving the galaxy example:

    Likewise, because we know that there are lots and lots of evils (over a hundred trillion if we do this quick calculation: 10 billion people, each living about half a century, each experiencing about one evil a day), we have good reason to expect that even if every evil has a justifier, some of the justifiers are going to be more obvious and some are going to be less obvious. And among those that are less obvious, there will be ones that are (roughly) the least obvious–and those are likely to be quite unobvious indeed.

    But wait. Sure, we know that they are going to vary in that, among those that have justifiers, some will have more obvious and some less obvious justifiers. And we can suppose that some will have the least obvious justifiers. But this in no way gets us to the conclusion that the least obvious “are likely to be quite unobvious, indeed.” In particular, it doesn’t get us to the conclusion that they will unobvious in the way required to defuse Rowe’s argument — to be such that, despite serious study, we can’t see a justifier. Predicting that there will be cases that take the most work to uncover a plausible justifier is entirely different from predicting that there will be cases for which no one can propose a plausible justifier even after prolonged study.
    Of course, the theist can argue that given what we already know of the evils that exist, we ought to expect there to be evils that meet the conditions exemplified in Rowe’s cases — that is, are such that we find it hard even to conceive of some justifier for them consistent with other things we think we know. But my point is that you can’t get this result from the rather quick and breezy line of reasoning above. Rowe’s cases are not ones that are just the ‘least obviously justified’; they’re ones where centuries of reflection seem to offer no plausible route to justification.
    This is, I daresay, more like an anomalous galaxy that has been investigated with the best available equipment for decades and we still can’t come up with any hypothesis, consistent with much of what we think we know, as to how its central black hole might be hidden from us. Anomalies do become problems for theories, after all — when the investigation of them has gone on for a very long time, no resolution is in sight, and we have no reason to think ahead of time that we would be unable to find a resolution.

    October 8, 2010 — 17:18
  • “Rowe’s cases are not ones that are just the ‘least obviously
    justified’; they’re ones where centuries of reflection seem to offer no plausible route to justification.”
    Rowe’s argument does indeed become stronger the more one packs into the serious examination condition. Mere serious examination is not enough for the argument to be plausible–we are not at all surprised when a scientific theory has anomalies that evade us for a couple of decades. It needs something more than that. Let’s try “serious examination for hundreds of years”. Note that this isn’t any longer Rowe’s argument–it’s a new argument, with a stronger and hence more controversial first premise.
    Take Rowe’s two famous cases: Bambi and Sue. The case of Bambi has not been searched for justifiers for centuries. Deep concern about the suffering of animals as a species of the problem of evil is a fairly new development. The case of Sue, on the other hand, happened in 1986. There has, thus, been no more than 24 years of reflection on that case. Moreover, I do not even know if there has been serious investigation of the justifiers in that case. Have philosophers actually interviewed the individuals concerned? Have they done follow-ups to see what aretaic changes in their lives resulted from the events? Etc. My suspicion is that what has gone on is entirely armchair speculation.
    Now, maybe, other cases rather like that of Sue have been reflected on for centuries. But, again, it is not clear that any particular one of them was really made the subject of serious, focused investigation vis-a-vis the finding of justifiers, and almost certainly no one such case was made the subject of serious, focused investigation for hundreds of years.
    I suppose one might claim that hundreds of years of investigation of cases of this sort has not uncovered a plausible justifier for any imaginable case of this sort. But that’s false. It’s not at all hard to imagine cases of this sort (i.e., satisfying a description like the one of the Sue story) where there is a justifier. Or maybe we could say: hundreds of years of investigation of cases of this sort has not uncovered a plausible justifier for any actual case of this sort. But this is a pretty strong premise about lots of cases. Much stronger than Rowe’s premise. Moreover, even this doesn’t seem so problematic for the person who thinks that there is a justifier in each case. For it could be that for a case of this sort, one needs hundreds of years of investigation of a particular case to find a justifier.

    October 8, 2010 — 23:45
  • Here’s another way to think about this. Let T be the hypothesis that every evil has a justifier. Let E be a randomly chosen evil. What is P(a justifier for E would not be found after 500 years of study | T)? The sceptical theist thinks it’s not small at all–maybe it’s even large. But let’s grant it’s small. But just how small? It seems to me that it would take unreasonably great confidence in our intellectual faculties to suppose it’s less than one a million.
    But given that there are way more evils than a million, we shouldn’t be that surprised if some of them are such that a justifier can’t be found after 500 years of study.

    October 8, 2010 — 23:50
  • Gregory Lewis

    Is the reading of Rowe right, here? My reading of Rowe (1979) is that his strategy is not to show from a couple of instances of really-hard-to-justify evil the evidential premise of his argument, but rather that these give the starkest examples of evils being hard to justify in general (see esp the end of P337 here – his argument may well have changed since then). In other words, that there aren’t just anomalies to the theory, but that the theory is, across vast swathes of observation, to a greater or lesser degree defective.
    I doubt an Atheologian will accept that most evils have good justifiers barring the ‘theodicial nasties’ that get cooked up. For this would seem to imply that Theism is true! On Atheism, one would surely expect that almost all evils are not justified (given a morally indifferent universe, it’d be improbable to have goods and evils linked up in that sort of way). Thus the appearance that most evils (bar a few) are justified seems to argue against Atheism, and rather suggestive of the answer Alexander gives – that all evils are justified, and we are just seeing the ‘tail’ of the distribution of justification-seeability.
    Rather, atheologians should use ‘theodicial nasties’ as exemplars of the defects they see in theodicy – of deep faults running through the standard defenses which ramify to ‘easier’ cases. Focusing on the hardest cases like this doesn’t seem unreasonable – if you find evolutionary biology a bit dodgy, you’d want to focus on what you view the hardest problems for it to solve rather than stuff which you find a bit unsatisfactory. Given what Alexander has said, though, perhaps we should be careful not to pay too much attention to these cases on their own, but only insofar as they raise concerns for a much wider set of observations.

    October 9, 2010 — 9:01
  • Gregory:
    You may be right about Rowe 1979. But what would one mean by justifying a type of evil “in general”?
    Option 1: The task is to come up with a plausible theodicy for a type of evil, such that the theodicy would plausibly apply to all or almost all instances of that type? Call that a “uniform justifier” for the type. But while theism entails the existence of justifiers for every token evil, why think it implies that every type of evil has a uniform justifier? Compare our own justifications for actions. Sometimes one is justified in not fulfilling a promise. Is there, though, a uniform justifier for the class of all justifiably non-fulfilled promises? Surely not. Sometimes, a promise is justifiably not fulfilled because one was released from it; sometimes, it would be unhealthy for the promisee; sometimes, it would cause pain for the promisee; sometimes, it would be unhealthy for third-parties; etc. There are as many justifiers for the non-fulfillment of a promise as there are kinds of goods and evils, and there is no uniformity. (Of course, one can come up with a general description, like: “on account of evils”; but that can also be done in the Rowe case.)
    Option 2: The task is to come up with a plausible theodicy for some instances of a type of evil. But that’s not that hard. It just takes a bit of imagination.

    October 9, 2010 — 10:48
  • A point worth remembering is that the evidence for the theory that every large galaxy has a black hole at its center does not need that we have actually found the black holes in very many cases. The evidence could be that the theory is entailed by other theories that are independently confirmed, say.

    October 9, 2010 — 10:57
  • Gene Witmer

    Alex,
    As I see it, the key issue here concerns the following two propositions:
    (1) Such and such specific evil E is such that: despite serious study, no justifier has been found for E.
    (2) Given what we antecedently know about the number and variety of evils there are, even if all them have justifiers, we should expect there to be some such that, despite serious study of them, no justifier has been found.
    If (2) is true, that defuses the significance of (1). Granted. What I was worried about was that the strength of the “such that, despite serious study, no justifier has been found” will need to be interpreted in a rather weak way to make (2) true, but that it can be understood in a much more demanding way and (1) can still be true. Your response makes it plain that you think that a stronger reading will not render (1) true.
    But there are different kinds of “serious study” that may be at issue. Your own discussion makes it plain that you have something that requires careful empirical investigation. For instance, in talking about the case of Sue, you write:

    The case of Sue, on the other hand, happened in 1986. There has, thus, been no more than 24 years of reflection on that case. Moreover, I do not even know if there has been serious investigation of the justifiers in that case. Have philosophers actually interviewed the individuals concerned? Have they done follow-ups to see what aretaic changes in their lives resulted from the events? Etc. My suspicion is that what has gone on is entirely armchair speculation.

    That’s right. But this, I take it, is because we normally presume that the empirically ascertainable facts are a certain way, and no real investigation of those is needed. If theists who addressed these issues disagreed really thought that such an investigation was needed, I wonder why they didn’t react to the Bambi case by insisting that we have no evidence of it ever happening. At least as the example is presented, it seems that the fawn’s suffering and death occurs without leaving any traces that might come to human attention. So why believe such worrisome cases even occur?
    Well, the answer is obvious: given what else we — both theist and atheist — already believe about the world, it seems incredibly unlikely that there are no such cases. Insisting that we need to investigate to find out whether there are Bambi cases would be like insisting that we need to take seriously the suggestion that perhaps the person who appears to be suffering horribly from cancer is only exhibiting the behavioral signs and is really quite tranquil. It’s possible, but not something I can take seriously, anyway. And I expect most theists wouldn’t take it seriously either — I hope.
    This point is connected to your next comment:

    I suppose one might claim that hundreds of years of investigation of cases of this sort has not uncovered a plausible justifier for any imaginable case of this sort. But that’s false. It’s not at all hard to imagine cases of this sort (i.e., satisfying a description like the one of the Sue story) where there is a justifier.

    One way to put the point I was making above is to say that Rowe’s argument could be understood in the following way (whether this was his actual intent, I don’t speculate):
    We can describe a kind of case such that (1) nobody seriously doubts that cases of that kind occur and (2) in thinking about merely possible justifiers for cases that fit that description, we can think of no justifier J such that (a) J is both adequate as a justifier and (b) given our general background beliefs about the world, is one that has any significant chance of being in play in the actual cases of this kind.
    Perhaps we can imagine cases that fit the relevant description and where there is a justifier; but if the imagined justifier itself conflicts radically with other things we generally believe about the world, we may then be reasonable in rejecting it as too improbable to take seriously. For example: perhaps in the case of Sue, her attacker had to be free to inflict such suffering and loss in order to allow him to be free to do some other act that would later bring about enormous good. Perhaps during the rape he thought of a cure for cancer, and wrote it down after for others to use later, and for some reason he couldn’t think of it under any other circumstances, and for some reason God couldn’t just inspire this solution in others under much happier circumstances, and so on. Even here I have to invoke “for some reason” because I can’t think of any way there could be such constraints on God, but it doesn’t matter. The point is that you don’t want the imagined justifier to require us to take seriously hypotheses that are so blatantly unlikely given the background beliefs we have that are relevant — e.g. that it is extremely unlikely that medical discoveries could only occur to such people in such circumstances, or that it is extremely unlikely that forest fires never catch random creatures in their path and lead to their suffering and death.
    I expect that when you say that it’s “not at all hard to imagine cases of this sort (i.e., satisfying a description like the one of the Sue story) where there is a justifier” you don’t mean to invoke scenarios of the sort that are as outlandish as the above. But if you mean that the sorts of suggestions one finds in the usual attempts at theodicy might actually apply to these cases, it is my turn to say just “But that’s false.” The serious examination of cases like this is really the serious examination of proposed general strategies for finding justifiers for cases like this, and one may, with Rowe, come to think that none of them work. You might think they work, after all, but then the disagreement here is about whether any of those familiar proposals have any merit.
    Someone who has considered those familiar proposals and found them plainly inadequate will not, of course, have examined them for hundreds of years, since none of us live that long. But they will have the benefits of the thoughts of previous thinkers that have accumulated for centuries. And if, like me, you agree that philosophy makes progress, at least in the abilities of its practicitioners to evaluate arguments and be clear about the options, this would give us just what we need to say that centuries of reflection on possible justifiers, combined with what is hopefully noncontroversial about the actual cases of evil, make plausible a version of (1) —

    (1) Such and such specific evil E is such that: despite serious study, no justifier has been found for E.

    where the “despite serious study” clause is as hefty as I’ve described.
    But does (2) seem plausible when the “despite serious study” clause is this hefty? Consider it again:

    (2) Given what we antecedently know about the number and variety of evils there are, even if all them have justifiers, we should expect there to be some such that, despite serious study of them, no justifier has been found.

    Surely not. What we antecedently know just about the number and variety of evils hardly gives us reason to think there will be evils that resist explanation in the way described above. I agree it gives us reason to think that a fair amount of imagination, creative thinking, and careful philosophical thought are needed to be confident we’ve not overlooked any serious possibilities — but that, I take it, is exactly what has been done in the long history of thinking about the problem of evil.
    -Gene

    October 10, 2010 — 13:26
  • Gene:
    These are very helpful, and challenging, comments.
    “If theists who addressed these issues disagreed really thought that such an investigation was needed, I wonder why they didn’t react to the Bambi case by insisting that we have no evidence of it ever happening.”
    The Bambi case could be different from the Sue case. In the Bambi case, we may well have a right to think we have most of the relevant information. There are no deep complexities to the moral lives of deer, most of us think. (Myself, I think a more work about the connection between pain and cognition is needed. It could be that cognitive features present in humans but not present in deer are needed for pain to be very bad. But that is a long-shot.) However, the Bambi case is also one we haven’t collectively thought about for hundreds of years, because there hasn’t been serious concern about animal pain in the context of the problem of evil until pretty recently.
    In the Sue case, however, one really does want to know more. Without bringing in remote possibilities like cures for cancer, one still would like to know more about how things morally went for Sue shortly before death, for Sue’s relatives over the years succeeding the death, and for the assailant.
    “you don’t want the imagined justifier to require us to take seriously hypotheses that are so blatantly unlikely given the background beliefs we have that are relevant — e.g. that it is extremely unlikely that medical discoveries could only occur to such people in such circumstances, or that it is extremely unlikely that forest fires never catch random creatures in their path and lead to their suffering and death.”
    Here I do want to raise one minor concern. What does the atheist think God should do in such cases? One option is: produce laws of nature such that no animal can be painfully burnt. But we don’t know that such laws would satisfy all the other desiderata God might reasonably have, like simplicity. Another option is: miraculously prevent such cases. If the atheist thinks that God, if he existed, should have opted for this second option, then by the atheist’s own lights the “blatantly unlikely” becomes relevant, and evidence needs to be provided that God did not, in fact, do this.
    “I expect that when you say that it’s ‘not at all hard to imagine cases of this sort (i.e., satisfying a description like the one of the Sue story) where there is a justifier’ you don’t mean to invoke scenarios of the sort that are as outlandish as the above.”
    Actually, I was thinking here of outlandish scenarios, because the atheistic proposal I was imagining (and I think Rowe’s wording could suggest this) was that it is logically impossible to have a justifier for a case that falls under the description of the Sue case, and to a logical impossibility claim, outlandish scenarios are quite relevant. (I think Plantinga’s transworld depravity is likewise an outlandish scenario. In fact, I think a variant of transworld depravity would do the job here. If the transworld depravity story is possible, it is likewise possible that the conditionals of free are such that in every feasible world with significant freedom other than those in which the Sue case occurs, all the significantly free choices are wrong, and in some feasible world at which the Sue case occurs, only the choices involved in the Sue case are wrong, and they are followed by an infinite future of wonderful significantly free choices.)
    In invoking hundreds of years of study, I think you may be thinking that if there are justifiers in cases like those of Sue, they are likely not to depend on fine details of the case. For if they might very well depend on fine details of the case, then we are at most entitled to invoke 24 years of study.
    But now take this analogy. There are many cases of justified promise non-fulfillment. However, observe that the justifiers for the non-fulfillment of very similar promises can be very different. That’s part of the complexity of life. For instance, x promises to pay back a loan to y in n years, and he fails to do so. There are many possible justifiers, and I bet that in the history of the human race, each of them has been used:
    – y releases x from the promise
    – while x is carrying the money to y, x is robbed
    – y becomes a drug addict, and giving y the money would result in y sinking deeper into addiction
    – y goes off on a long journey and despite reasonable effort can’t be located
    – x’s mother needed emergency treatment in year n-1, and the costs of the treatment made it impossible for x to gather the money by year n
    – in the meanwhile, y defrauded x of a greater sum of money
    etc.
    If all we were told was that x failed to pay back the loan to y in n years, we wouldn’t have nearly enough information to tell if there is a justifier or not, and if there were, what it could be. The vicissitudes of human life are too varied for that. Similar actions are often done for very different reasons.
    Likewise, it seems quite plausible that different cases sharing the same general pattern as that of Sue could have very different justifiers, depending on fine detail of the situation. For instance, the following story is compatible with the scenario as given by Rowe, and not particularly remote:
    (*) Just prior to losing consciousness, due to divine grace, Sue looks with horror-stricken but forgiving eyes at her assailant. Her assailant is haunted by this look for years, until finally it brings moral transformation, while this act of graced virtue gives texture to Sue’s afterlife over an infinite future.
    Do we have any evidence independent of theism that (*) was the case? No. But neither is (*) a particularly remote possibility. Such things, surely, do happen. And nothing in the story as given by Rowe rules out (*).
    And, probably, with a bit more thought we can come up with other such non-remote scenarios compatible with the Sue story.
    Now, one might claim that there surely were cases like the Sue case where nothing like (*) happened. But now the argument becomes weaker in two ways. First, the alleged occurrence of the evil is more controversial–it would be crazy to deny that no case like the Sue case occurred, but to claim that no case like the Sue together with the stipulations ruling out scenarios like (*) would be more controversial. Second, the more complex the case is–the more stipulations have to be included in it–the less confidently can we say that we’ve thought about cases like that for hundreds of years.
    Going back to the science cases, however, I also think it’s worth saying that nothing analogous to the sort of richness of empirical analysis that is brought to bear on scientific anomalies gets brought in these problem of evil cases. In the case of scientific anomalies for major theories, all sorts of little details of the situation get carefully examined in case they might carry a clue to what is going on. This is done even if, ahead of such minute empirical investigation, we cannot think of any way in which the detail might bear on the case. Thus, even if I grant that some case was substantially thought about for hundreds of years, the amount of empirical research brought to bear on the case falls very significantly short of the amount of empirical research brought to bear on a “hot” scientific anomaly over, say, 30 years. But the persistence of an anomaly over 30 years should not be taken to be significant evidence against a theory. (I think the issue here may in part be this. Most philosophers aren’t very good empirical researchers. But the argument from evil has a very significant empirical component.)
    Another complicating feature of the problem of evil case is that I do not know that the problem of evil was taken all that seriously by a sufficient number of theistic thinkers over several centuries. First of all, historically a lot of theologians and theistic philosophers thought that they had perfectly good answers. If you asked them, they would say something about how the clay has no claims against the potter, or how evil is a privation, or how evil flows from free will, or how there has to be a solution because the existence of God has been proved, or even how the existence of evil proves the existence of God. Whether these points work or not is irrelevant to my argument. What is relevant is that a lot of the smart theists thought that these points settled the issue, and hence did not think any more about the finding of justifiers. Given these points, they no longer thought of evil as an anomaly.
    But the kind of concerted effort that an anomaly receives in science may well be dependent on a near-consensus that it is an anomaly. And even with that, it is not particularly strong evidence for a theory if the anomaly remains anomalous for a couple of decades.

    October 10, 2010 — 16:31
  • Of course, in the last sentence I meant: “evidence AGAINST a theory”.

    October 10, 2010 — 16:32
  • It’s also worth noting that not at all rarely, a person to whom something terrible has happened will, years later, have a theodical story to tell about the event, talking of how the event made her who she is, taught her such and such, etc. Serious investigation of a particular needs to check if the people involved have such stories, and whether the stories make for plausible theodicy. (An obvious question will be whether God couldn’t bring about the relevant goods equally well without such terrible evils. But to answer that question, we would need to look very carefully at the particular goods claimed, and whether they are the sort of goods to which the causal history is really important.)

    October 10, 2010 — 19:30
  • Mike Almeida

    (2) Given what we antecedently know about the number and variety of evils there are, even if all them have justifiers, we should expect there to be some such that, despite serious study of them, no justifier has been found.
    For what it’s worth, I’ve argued along Alex’s lines that the chances are extremely low that we’d find a God-justifying good for all evils, ‘On Stone’s Evidential Atheism’ Theoria Vol. 76, No. 1 (2006b) 5-22

    October 11, 2010 — 9:00
  • Gene Witmer

    Mike –
    Thanks for the reference to your article; I’ve not yet read it, but certainly plan to.
    Alex –
    This reply is rather late in coming, but here it is nonetheless…
    I am a bit confused by some of your comments, especially these:

    Actually, I was thinking here of outlandish scenarios, because the atheistic proposal I was imagining (and I think Rowe’s wording could suggest this) was that it is logically impossible to have a justifier for a case that falls under the description of the Sue case, and to a logical impossibility claim, outlandish scenarios are quite relevant.

    Sure. But the atheistic argument need not take that form at all, and as I was imagining it, it didn’t. Without taking that form — of claiming that it’s logically impossible to have a justifier for a case of a certain sort — the evidential argument can still have considerable force.
    Maybe it will be helpful to return to your tidy representation of the argument
    1. E is an evil for which despite serious study we can’t see a justifier.
    2. Therefore, probably, E has no justifier.
    Therefore, theism is false.
    The entire dispute here focuses on whether 1 can justify 2, and whether it can will depend on how much is packed into “despite serious study.”
    Now, the specific proposal that you were commenting on is an attempt to move away from the fine details of a particular case. So, let’s consider a somewhat different way to set up the evidential argument that incorporates that move away from particular details.
    1. There is a type of evil (type A) such that: despite serious consideration, we cannot imagine any justifiers for type A evils except those such that, given our background knowledge, we should consider them very unlikely to exist.
    2. We know that there are many instances of type A evils.
    3. Hence, probably, there are instances of type A evils for which there are no justifiers.
    Hence, theism is false.
    Supporting premise 1 does not, happily, require us to investigate the fine details of particular cases. Of course, it does require that we consider (i) what we take to be the background knowledge that limits the likely varations in fine detail and (ii) the variations consistent with that constraint insofar as they might make a difference as to whether there’s a justifier to be imagined.
    I think at this point there are two questions to consider: whether your initial criticism of the evidential argument has force against this version, and whether premise 1 of the above is plausible.
    First: Does the initial criticism of the evidential argument have force against this version? The initial criticism gets its appeal largely, I take it, because it exploits the fact that, of course, philosophers have not extensively studied the fine empirical details of cases of actual evil. That such extensive empirical study has not been done is undeniable. The point of the above is to bypass the need for extensive investigation of particular cases.
    Whereas the initial criticism relied on this claim:
    (a) Given what we antecedently know about the number and variety of evils there are, even if all them have justifiers, we should expect there to be some such that, despite serious study of them, no justifier has been found.
    I take it for the above version of the evidential argument the critique would need to rely on this somewhat different claim:
    (b) Given what we antecedently know about the number and variety of evils there are, even if all them have justifiers, we should expect there to be several that belong to a type T such that, despite serious consideration, we cannot imagine any justifiers for evils of that type except those such that, given our background knowledge, we should consider them very unlikely to exist.
    I can imagine at least two things you might say in trying to make (b) plausible.
    You might say (1) that we should expect our background knowledge to be limited or misleading in important ways, so that it’s not surprising that it would lead us to think things very unlikely even when they aren’t all that unlikely. You might say (2) that we shouldn’t expect our imaginative abilities, even after extensive exercise, to be up to the task of surveying the relevant possibilities.
    These are worth thinking about, but I guess I find it odd to try to justify either of these by reference to what we antecedently know about the number and variety of evils. If they’re justified, it’s going to be by reference to something else: something about our cognitive abilities generally, I would think
    In any case, as you concede, the more that is packed into the ‘despite serious study’ condition, the more potent the argument becomes; the suggestion above is designed to make it easier to see how we could in fact have given very extensive study to the relevant questions without having done extensive empirical research on particular cases. So setting things up this way makes it easier to defend the claim with a hefty ‘despite serious study’ condition.
    The second question to consider is just this: Is the first premise plausible? That is, *are* there types of evil that meet the description given — that are such that, despite serious study, we cannot imagine justifiers for tokens of that type that don’t involve outlandish scenarios?
    Well, let us consider the Sue case, and suppose the type at issue is relatively coarse-grained. You suggested the following about this type of case:

    Likewise, it seems quite plausible that different cases sharing the same general pattern as that of Sue could have very different justifiers, depending on fine detail of the situation. For instance, the following story is compatible with the scenario as given by Rowe, and not particularly remote:
    (*) Just prior to losing consciousness, due to divine grace, Sue looks with horror-stricken but forgiving eyes at her assailant. Her assailant is haunted by this look for years, until finally it brings moral transformation, while this act of graced virtue gives texture to Sue’s afterlife over an infinite future.
    Do we have any evidence independent of theism that (*) was the case? No. But neither is (*) a particularly remote possibility. Such things, surely, do happen.

    But — no; this won’t do. Even if it is not a terribly remote possibility that the sequence of events in (*) happen, that simply isn’t adequate for showing that there’s a non-outlandish justifier. Merely showing how an evil might lead to a good is not enough, of course; the evil needs to have been necessary for the good result, and the good result needs to be good enough to warrant the evil. So, if we consider your (*) to even be relevant, it has to be part of a story that makes this evildoing unavoidable. And what story could that be? Perhaps an appeal to Plantingan transworld depravity?
    If so, however, that doesn’t help. After all, you wrote — and I was glad to see you concede this point! — that you think Plantinga’s transworld depravity “is likewise an outlandish scenario.” So, if, for (*) to count as a justifier, we need include in it the supposition of transworld depravity, then (*) does not meet the conditions. It is rather a justifier that our background knowledge should lead us to count as very, very unlikely.
    There are other stories one can appeal to, of course — other ways in particular to make the appeal to free will. But we can’t do the ‘serious examination’ in this one post. I do hope, though, that the problem with the quick appeal to (*) makes plain why someone could think that a strong version of the first premise — that there are types of evil for which we cannot, even after very serious consideration, imagine a non-outlandish justifier — is in fact well supported.
    One other comment before I shut up. You commented:

    It’s also worth noting that not at all rarely, a person to whom something terrible has happened will, years later, have a theodical story to tell about the event, talking of how the event made her who she is, taught her such and such, etc. Serious investigation of a particular needs to check if the people involved have such stories, and whether the stories make for plausible theodicy. (An obvious question will be whether God couldn’t bring about the relevant goods equally well without such terrible evils. But to answer that question, we would need to look very carefully at the particular goods claimed, and whether they are the sort of goods to which the causal history is really important.)

    I take it the intended significance of this observation is supposed to be this: Given that many people have such stories to tell, there is some evidence that those evils have justifiers, and hence we’re obligated to examine those stories and their plausibility as justifiers before concluding that we’ve done the ‘serious study’ that has been discussed above.
    I don’t think this point has any such force, however. And this will likely sound terribly harsh of me, but, I doubt that any more than a vanishingly small proportion of the people who give these folk theodicies ever even consider the point you mention in the parenthesis — whether the evil was needed for the good outcome. To put it bluntly, these sorts of personal folk theodicies have just about zero credibility given the circumstances of their production. Indeed, you yourself make the nice point that throughout history a large number of theists have not taken the problem of evil as sufficiently serious to give it a thorough examination; how much more true is this of believing non-philosophers? A random sampling of such folk theodicies shows, I think, that there is almost no hope of finding plausible justifiers among their ranks.

    October 14, 2010 — 9:45
  • Thanks, Gene, for these insightful comments. Let me break up my response into two parts. First, some comments on your proposed type-based argument.
    There is some ambiguity in that argument as it stands, particularly in the premise: “Despite serious consideration, we cannot imagine any justifiers for type A evils except those such that, given our background knowledge, we should consider them very unlikely to exist.”
    One way to get at the ambiguity is this. Suppose that I came up with a justifier for a type A evil that was very unlikely to exist in the case of any particular token of a type A evil, but given how many type A evils there were, it was not unlikely that it existed in the case of some type A evil. Would that refute the premise?
    If yes, then I think this is not so difficult a task, as per my example (*) (which I’ll discuss in my next comment), unless the description of A is gerrymandered to rule such a thing out. And if the description is thus gerrymandered, then it is no longer true that we’ve been seriously examining them for a long time.
    If not, then I think we once again may have to go back to examining particular evils of type A. For once we’ve shown that one can come up with a non-outlandish justifier for a type A evil, the question is whether every type A evil has a justifier. And it is prima facie fairly likely, on the hypothesis that every evil has a justifier, that among the many different type A evils there are many different justifiers.
    I should note that the type approach also connects with the science cases. For it is quite possible for a well-confirmed scientific theory to have not only individual anomalous cases (those often get put down to experimental error anyway and not even considered as anomalous) but types of anomaly that contain many token anomalous cases.

    October 14, 2010 — 10:13
  • Gene:
    Now on to (*). This was my proposal “Just prior to losing consciousness, due to divine grace, Sue looks with horror-stricken but forgiving eyes at her assailant. Her assailant is haunted by this look for years, until finally it brings moral transformation, while this act of graced virtue gives texture to Sue’s afterlife over an infinite future.” I said that we did not have evidence that (*) happened and we did not have evidence that (*) did not happen.
    Your objection is that I have not shown the evil to be necessary for the good.
    But the good here is something like:
    (G) Sue’s assailant is morally transformed, from being morally horrific to being good, by Sue’s fast forgiveness of his unspeakably sadistic assault on her, and Sue’s forgiveness gives texture to her afterlife over an infinite future.
    The existence of G entails that Sue was unspeakably sadistically assaulted, and hence the assault is necessary for G.
    Now, you can say that there are other similarly valuable goods for which the assault is not necessary:
    (G1) Fred is morally transformed, from being morally horrific to being good, by Sue’s forgiveness of what she vividly believed to be a very recent unspeakably sadistic assault on her, and Sue’s forgiveness gives texture to her afterlife over an infinite future.
    G1 is a good that does not entail that Sue was assaulted. However, G1 does not contain all that is of value in G. While there is a value in forgiving what one believes to be a grave harm, if the belief is false, the forgiveness lacks the value of “veridicality”.
    Moreover, if a part of Fred’s moral horrificness is embodied in the assault on Sue, a statement of the good that leaves out the assault also leaves out an aspect of the good of moral transformation.
    Now, we can argue whether G was worth it. Maybe it would have been better to have G1 and no assault: we lose the veridicality of the forgiveness and one aspect of Fred’s moral horrificness, but Sue isn’t assaulted. But it is at least not obvious that G1 + no assault, call it G1*, is at least as good as G. Consider, for instance, that the occurrence of G1* entails that Sue vividly believed herself to have just been unspeakably sadistically assaulted. That is a horrific memory to have. Granted, it is probably preferable to have that horrific vivid memory falsely implanted than to have to suffer the assault, but it is still pretty bad. Moreover, in G1*, the “added texture” in her infinite life does not contain a veridical memory of veridical forgiveness. It is far from clear that G1* is at least as good as G.
    I think the important point here is that plausible justifying goods are goods that are essentially thick and storied, with the stories including the causal history, and hence such as to entail the evils. The ordinary person says: “Suffering this has made me humbler.” Of course, God could make one humbler in all sorts of ways, including by directly and instantaneously making it be that one have the dispositions constitutive of the habit of humility. But to be made humbler by God’s direct willing and to be made humbler by patient suffering are different goods.

    October 14, 2010 — 10:42
  • Gene Witmer

    You’re right that there’s an ambiguity in the premise:
    “Despite serious consideration, we cannot imagine any justifiers for type A evils except those such that, given our background knowledge, we should consider them very unlikely to exist.”
    So let’s distinguish two distinct premises:
    (1a) Despite serious consideration, we cannot imagine any justifiers for type A evils except those such that, given our background knowledge, we should consider it very unlikely, for any particular evil of type A, that a justifier of that kind exists for that evil.
    (1b) Despite serious consideration, we cannot imagine any justifiers for type A evils except those such that, given our background knowledge, we should consider it very unlikely that a justifier of that kind exists for any of the many instances of type A evils.
    Suppose only (1a) is granted. Then we are not in a position to conclude that for any particular evil E of that type, it probably doesn’t have a justifier. That’s right. Does that mean that we’d have to go back to examining particular evils of type A? You write:

    …I think we once again may have to go back to examining particular evils of type A. For once we’ve shown that one can come up with a non-outlandish justifier for a type A evil, the question is whether every type A evil has a justifier. And it is prima facie fairly likely, on the hypothesis that every evil has a justifier, that among the many different type A evils there are many different justifiers.

    Well, it’s at least not unlikely, anyway, that among the many different type A evils, there are many different justifiers. I don’t see why that should compel the move to looking at particular cases empirically, however, since we might think we have good reason to buy the stronger (1b). My guess is that you think (1b) is just too strong a claim to make plausible. I don’t. But then that’s likely because I think the only imaginable justifiers require conditions that are not just unlikely in this or that particular circumstance, but unlikely in any circumstance. The example of transworld depravity has this character. It’s not just that we find it outlandish to suppose that God couldn’t have picked a possible person to exist in the circumstances leading up to Sue’s death who would have refrained from such evils; we find it outlandish to suppose God couldn’t have picked a possible person to exist in just about any of the circumstances that people actually confront who would have refrained from serious evildoing. Or, at least, that seems outlandish to me.
    In general I suspect we’d just disagree about the remoteness of various proposed justifiers. This comes out in your remarks on your imagined justifier for the Sue case, hypothesis (*). In your remarks, you propose that the great good that is at issue in (*) is, as you put it, “essentially thick and storied, with the stories including the causal history, and hence such as to entail the evils.”
    Here’s what it seems to me is going on. The part of the justifier that is extremely remote is just being shifted around, if I may put it that way. On this proposal, I can agree that the evil is necessary for that particular good, and I can agree that the story of Sue’s look leading to the attacker’s eventual transformation is not especially unlikely; but now we need to consider the plausibility of this moral claim that allowing such evil for the sake of that particular moral good is something a morally perfect being would do. And that moral claim strikes me, as I’m sure it does many other atheists, as outlandish. We are to think that the value of such repentance and forgiveness is so high as to permit such vicious evil and suffering that is otherwise unnecessary for any good thing? That’s as crazy sounding as universal transworld depravity! There’s the empirically outlandish (say, the hypothesis that Sue’s attack will bring about a cure for cancer), the metaphysically outlandish (transworld depravity) and the morally outlandish (such forgiveness is so good it’s worth the evil), but is there any imaginable justifier that avoids any of these pitfalls?
    Of course, I expect you will not find your own proposal morally outlandish, so there is disagreement about that sort of point. But those of us who have reflected on these cases and proposed justifiers and remain convinced that they all require something massively improbable are still in good shape for mounting an evidential argument for atheism where it’s not, as you first put it, just “a bad piece of non-deductive reasoning.”

    October 14, 2010 — 17:15
  • Gene:
    I agree that your modified versions of the argument don’t seem to have the problem of the original formulation, once enough is packed into the serious search.
    I think the modified versions do have one problem that I alluded to earlier. That many theists find morally plausible certain kinds of justifiers that you don’t find morally plausible reduces the evidential force of the centuries of serious investigation. For once one finds an explanation of an anomaly to one’s theory that is satisfactory to oneself and to many people in one’s community, the seriousness of one’s search for an explanation is apt to decrease significantly, or one might stop searching altogether. (One might, of course, keep on searching for apologetic reasons. But if one is not existentially invested in the search, the seriousness of the search is apt to go down.)
    As for the disagreement on the moral plausibility of something like (*) being a justifier:
    1. I wonder sometimes if one of the differences between the intuitions of Christians and atheists isn’t a difference in how one compares moral goods to suffering. Christians tend to find plausible the Socratic idea that moral goods greatly outweigh the bads of suffering. Now, if we take the Socratic intuition on board, that does not settle the Sue question by itself. For in addition to the badness of Sue’s suffering, there is also the great moral bad to her assailant of his assault. But there are still some reasonable ways of weighing things so that G comes out worthwhile. For instance, one can weigh the value of Sue’s forgiveness and her memories of the forgiveness over and against the badness of Sue’s suffering, and there the value of the forgiveness will far outweigh the badness of the suffering; and one can then weigh the value of the assailant’s later moral transformation over and against the disvalue of his earlier moral decay, and the two will either cancel out or, if one has the kinds of intuitions I do, the moral transformation will win out in value. Or, one might simply have the reasonable intuition that forgiveness in some important sense defeats moral evil: that value of forgiving an evil is significantly greater than the disvalue of the evil forgiven. And the excess (the value of forgiving the evil minus the disvalue of doing the evil), because it’s a matter of moral value, may well far outweigh the badness of the suffering.
    2. I think the power of infinity is not to be underestimated. Suppose that the experience of forgiving a horrendous evil adds only a little bit of value to Sue’s memories. Nonetheless, if she lives for an infinite amount of time, as (*) claims she does, that little bit of memory is apt to get multiplied by infinity, both because she might very well recall the forgiveness infinitely often and because it enriches her character over an infinite amount of time. Now, one might worry that the memory of the suffering will also be infinitely multiplied. But I say that it will be remembered as overcome by the forgiveness.
    Of course, you might think that an afterlife is outlandishly unlikely. But then we don’t need to go into the evidential argument from evil, but you can just argue directly for atheism from your reasons to disbelieve in an afterlife, together with the claim:
    A. If God exists, there is eternal life available to persons
    which is very plausible given the kind of dignity that persons have.

    October 14, 2010 — 19:12
  • David

    I am pondering Rowe’s argument, but have a different issue than the one raised in the original post. I was hoping some of you might help me think through this.
    Rowe says (in Peterson’s “Contemporary Debates in the Philosophy of Religion”), “…if there were only a few such examples as these, perhaps it would not be unreasonable to believe that…[there are justified reasons]…It is the enormous amount of apparently pointless, horrendous suffering…that grounds the claim in the first premise that there are pointless evils in the world [that a triple-O God could have prevented without forfeiting some greater god, etc.].”
    Rowe says he could live with a few instances of (apparently) gratuitous evil….but not the “enormous amount” that we actually see in this world. Let’s say we could turn a knob and decrease gratuitous evil incrementally. Is there any objective basis for stopping the knob at a particular point and saying, “ok in this possible world we can reasonably believe God might have justifying reasons for the few (apparently) gratuitous evils we find.”
    Am I missing something here? Isn’t Rowe really appealing to human intuition here? I’m not talking about the noseeum assumption, I’m talking about what apparently amounts to too-many-of-um assumption. Apparently the noseeum assumption only gets off the ground once we accept the too-many-of-um assumption (for Rowe at least).
    Thoughts?

    October 20, 2010 — 7:00
  • EcomGeek

    I very much enjoy the discussion here.
    As I understand it, your objection hinges on the premise that we have good reason to believe that there are goods that would be unknown and would justify God in permitting E1 and E2, should God exist. I don’t think you’ve argued for this premise so much as you’ve argued that it’s the case that we have good reason to believe that there are some evils for which we wouldn’t know the justifiers, should God exist. That is, I don’t see an argument for the E1 and E2 justifiers also being unknown justifiers.
    However, even if your claim is that we have very good, very very good, reason to believe that there are unknown justifiers for E1 and E2, the objection still fails. That’s because so long as the Probability (no known good does justify God in permitting E1 and E2 given that God exists and all we know)

    October 25, 2010 — 8:13
  • That’s an interesting question. I take it your worry is this. I have given reason to think that there exist many evils with unknown justifiers. But this does not give us reason to think E1 and E2 would have unknown justifiers. That’s true. But here is where sampling error comes in. E1 and E2 are chosen as paradigm cases of evils with unknown justifiers. They aren’t simply randomly chosen evils. So there is no need to give a specific reason to think E1 and E2 would have unknown justifiers.
    Compare this. Evolutionary theory (let’s suppose–I’m oversimplifying) says that all major features of organisms have evolutionary explanations (not all of them based on natural selection). But a bit of armchair biology also shows that it is likely that in some cases serious study will be unable to identify the evolutionary explanations. The opponent of evolutionary theory then searches for two cases, C1 and C2, where it’s particularly hard to identify the evolutionary explanations. Now, the theory predicts that there would be cases like that. But it doesn’t predict that C1 and C2 would be cases like that. Nonetheless, C1 and C2 cut no ice, because of sampling bias.
    Here’s a different way of putting the point. If naturalism is true, some evils would look like they might well be justified and some wouldn’t. Naturalism doesn’t predict that specifically E1 and E2 would be among those that wouldn’t look justified. It only predicts that there would be a lot of evils among those that wouldn’t look justified. So neither theism nor naturalism predicts that E1 and E2 would be among those that wouldn’t look justified, and so we should not take the specific facts about E1 and E2 as significant evidence for naturalism over theism. (And what I said about naturalism applies to other alternatives to theism.)

    October 25, 2010 — 8:38
  • EcomGeek

    The last part of my comment was cut off because it contained an angle bracket. So, let me sum up what the last bit said before I take up your response.
    However, even if your claim is that we have very good, very very good, reason to believe that there are unknown justifiers for E1 and E2, the objection still fails. That’s because so long as the Probability(no known good does justify God in permitting E1 and E2 given that God exists and all we know) is less than 1, the Probability (God does exist given that no known goods justify God in permitting E1 and E2 and our all we know) will always be less than the Probability(God does exist given all we know). That is to say, the addition of the fact that no known goods justify God in permitting E1 and E2 to the total background knowledge of a perfectly agnostic deliberator will always make it more probable that God doesn’t exist. In order for your objection to succeed, I believe that you have to show that the Probability(no known good does justify God in permitting E1 and E2 given that God exists and all we know) = 1.

    October 25, 2010 — 10:16
  • David

    I don’t think my previous comment was very clear.
    Going along with your analogy, exactly how many missing black holes does it take to dis-confirm theory T? It only takes one empty pantry to discredit the theory that “all pantries contain crackers.” Obviously our certainty about the facts (such as “G is a galaxy” and “G is a galaxy without a central black hole”) is less than perfect, such that we need to stack up more than one counter-example to dis-confirm theory T.
    Arguing about whether E1 or E2 are actually evils would be a fruitless endeavor; and you’ve already raised concerns about whether we have indeed studied gratuitous evils seriously enough. However, my point lies in a different direction.
    My question is simply how do we known when a theory has been dis-confirmed by evidence to the contrary? I am not a professional philosopher, so it wouldn’t surprise me if the answer is, “uhh that’s just how induction works.”
    I am just confused how to evaluate this argument properly without knowing what possible world would be acceptable. Would half the amount of gratuitous evil still provide enough disconfirmation of theism that Rowe could say it is more rational to accept atheism? I just don’t see how to quantify arguments of this type at all.

    October 25, 2010 — 13:17
  • Ah, I see. The Rowe point. So my previous reply won’t apply.
    Now, first of all, it’s really easy to get tiny bits incremental confirmation in either direction. I just looked under my chair. I didn’t see a sound argument for atheism written there. P(there is no sound argument for atheism under my chair | theism) = 1. P(there is no sound argument for atheism under my chair | atheism) < 1. So, my observation provided me with some incremental evidence of theism. But it’s such a tiny increment that it doesn’t really matter. And maybe the same is true here.
    But nevermind that. What we really want to look at is not P(no known good does justify God in permitting E1 and E2 | theism) but P(no known good is such that it would justify God in permitting E1 and E2 | theism). Here’s one way to see this.
    Let K1 = Some known good does justify God in permitting E1 and E2
    K2 = Some known good is such that it would justify God in permitting E1 and E2
    T = God exists.
    You (with Rowe in another piece where I think he got things badly wrong) think we should compare P(T | ~K1) with P(T), and say that P(T | ~K1) < P(T). Since P(~K1|T) < 1 and P(~K1|~T) = 1, you’re right, assuming P(T) < 1. But observe that K1 entails K2. Therefore, ~K2 entails ~K1. Now, our evidence is not just ~K1, but also ~K2. Hence by using ~K1 as the evidence, you are using less than the total available evidence. To use the total available evidence, you need to use the stronger claim ~K2. But it is not true that P(~K2|~T) = 1, so the argument no longer works with ~K2 in place of ~K1.

    October 25, 2010 — 13:24
  • David:
    It would take one unjustified evil to disprove the thesis that all evils are justified (and hence to disprove theism if theism entails the thesis). But given the quantity of evils that we observe, we have good reason to think, whether theism is or is not true, that some won’t look justified.
    Likewise, consider the theory every pantry in the world has a cracker in it. That theory is contradicted by a single pantry that doesn’t have a cracker. But the theory is not contradicted by a pantry that doesn’t look like it has a cracker.
    Given that there are hundreds of millions of pantries, even if every pantry had a cracker, it would be unsurprising if there were some pantries which looked like they didn’t have any crackers in them. Surely it would be likely that in some pantries the cracker has been put by a kid in a cereal box, or the cracker has slipped under the pantry lining, or it is a very small cracker and it is wedged in between the edge of the shelf and the wall of the pantry, and so on. It is not unlikely that among the hundreds of millions of cracker-containing pantries, at least a few pantries would be such that a moderately serious search would fail to find a cracker.
    The more serious the search, or the greater the number of pantries apparently empty of crackers, the more plausible the argument. In the case of crackers, we are also aided by the fact that there is a minimum cracker size (anything smaller than that is a crumb, not a cracker) and that crackers are by their nature visible.

    October 25, 2010 — 13:38
  • David

    I re-read Rowe’s argument as well as Howard-Snyder and Bergmann’s response to it. It seems that in this case, they were indeed only arguing from specific instances of evil. I was quite confused since they hadn’t been too explicit about it. Rowe only made a brief mention of how the “enormous amount” of evil helped ground his claim. His interlocutors didn’t respond to it at all.
    For instance, Snynder and Bergmann say, “Other noseeum arguments from evil are just like this except that they focus on the amount of suffering rather than on particular instances of intense suffering or horrific evil.” (Ch 1 response) Presumably the skeptical response applies equally to arguments using specific instances and ones leveraging the amount of suffering.
    Also it was interesting to note that they had considered something along the lines that you mentioned in your considerations: the amount of time that humans have applied serious study to some evil.

    “Knowledge has progressed in a variety of fields of inquiry, especially the physical sciences. The periodic discovery of previously unknown aspects of reality strongly suggests that there will be further progress…it wouldn’t be surprising if there is much we are currently ignorant of. Now, what we have to go on in charting the progress of the discovery of fundamental goods (like freedom, love, and justice) by our ancestors is meager, t say the least. Indeed, given the scant archeological evidence we have, and given paleontolical evidence regarding the evolutionary development of the human brain, it would not be surprising at all that humans discovered various fundamental goods over tens of thousands of years separated by several millenia-long gaps in which nothing was discovered. Hence…it would not be surprising if there were goods of which we are ignorant, goods of which God-in his omniscience-would not be ignorant.”
    (from same reply in Ch 1)
    Cheers,
    David

    October 25, 2010 — 20:24
  • I don’t find that plausible the idea that the discovery of fundamental goods is so slow. At least when we’re talking of goods-for-humans, the fundamental kinds of goods are ones that we have a natural tendency to pursue. We seek to be free, to love and be loved, to be treated justly. We do not seek these things because we have made the great intellectual discovery that these things are good. We made the intellectual discovery by reflecting on what we pursue. Now, it may be that there are goods-for-humans that we have a natural tendency to pursue that we have not yet discovered due to insufficient self-reflection. But if they are important goods, then I doubt that they are going to be goods that we are entirely ignorant of. The great novelists and spiritual writers surely would have caught glimpses of them in their penetrating analysis of our behavior.
    Of course, there can be goods-for-other-beings (angels?) that are beyond our ken. But it is not clear whether God would permit horrendous evils to humans simply for the sake of benefits to other beings. Such goods beyond our ken could play a role in justifying evils, but I do not think we should have them play a primary role in this.
    The response to Rowe that I was defending does not require any ignorance of kinds of goods. All it requires is that it be unsurprising, even given theism, that among the vast multitude of evils there should be some where we do not see any of the familiar kinds of goods as justifying them. For the connections of the familiar kinds of goods to the evils can be subtle and hard to find. It can be hard to see that the evil that x suffered was a punishment, because x’s sin might have been secret; it can be hard to see that the evil that y suffered led to deep moral transformation, because we didn’t know y very well before that transformation; etc. None of this requires any kind of scepticism about our knowledge of the range of goods. All it requires is uncontroversial cognitive limitations that we all agree on.
    Compare this: The police after searching for many years for a murderer give up on the case. They can do this while still having a justified belief that a murderer exists (or at least existed). To rationally hold on to the belief that there is a murderer even though they haven’t found him despite serious investigation they do not need to posit hypotheses of types of murderers beyond our ken (supernatural murderers; murderers acting on really weird motives; etc.) All they need is a sober reflection on ordinary limits on our cognitive abilities. Quite possibly, the murderer is the ordinary-looking guy or gal next door.

    October 26, 2010 — 11:36
  • David

    I’m not sure if you read Dr. Vallicella’s blog, but there is an interesting discussion going on (far above my head!) regarding the the use of the evidential argument against the Anselmian concept of God.
    http://maverickphilosopher.typepad.com/maverick_philosopher/2010/10/god-possibility-and-evidential-support-for-non-contingent-propositions.html

    November 2, 2010 — 20:28