One of the most powerful arguments against orthodoxy theism is William Rowe's evidential argument from evil. The argument relies on what has come to be known as the controversial “noseeum” inference which goes as follows;
P: We are, upon careful reflection, unable to find any God-justifying goods for E
Q: (therefore) There are no God-justifying goods for E
P is said to provide strong evidential support for Q, which in turn entails the first premise of Rowe's syllogism. The most popular strategies of attack against Rowe's argument undermine the support his noseeum inference provides to the first premise. The theodicist denies P by building a theodicy; a plausible explanation of the sorts of goods that could justify God's permission of E. The skeptical theist, on the other hand, may very well accept P but wish to dispute the strength of support it lends Q. It has been argued elsewhere that this inference can't fail altogether and thus the skeptical theist needs to say that P does not offer Q any significant support. In fact, due to the vast quantity of evils that we can plug into this inference (assuming the theodicist's project isn't terrible persuasive), the sort of support P would in general have to provide Q would be miniscule as to ultimately be negligible.
A common objection to skeptical theism is that the skepticism invoked cannot be contained. In this the skeptical theist can defeat Rowe's inductive step only at the cost of undermining the justification of other beliefs he would naturally hold dear. Of all the things skeptical theism could undermine, our judgements of moral value seem the most obvious—after all that's what it's proposed for. But does it go too far and lead, as some say, to an overly extreme moral skepticism? The claim is that because the skeptical theist has given up his knowledge of the realm of all things considered value, he is in no position to make judgements about what sorts of things they ought or ought not obtain, permit, or prevent. If this is the case then skeptical theism would lead to moral paralysis; ones actions could never be motivated by moral considerations.
The skeptical theist might respond in saying that, while it's true he has no confidence in the all things considered value of his actions, he nevertheless knows what things are good and what are evil and can base his moral judgements on that. After all only the strictest of consequentialists would say that the consequences of ones actions are all that is relevant to moral decision making. The skeptical theist could still find reason to help a stranger in need because of how it makes the world a better place here and now, and within the foreseeable future. But what if skeptical theism threatened not only our ability to make judgements about all things considered value from how they appear, but also prima facie value? As Piper suggests, “it well might be the case that some of the things we strongly consider goods are, in a similar way, overturned by unknown higher goods and their entailment relations with putative goods (and evils) within our ken. If this is the case, our knowledge of the nature of goodness might be in large part erroneous.”2 How is it that our grasp of putative goods and evils could be threatened by beyond our ken greater goods or evils? Consider a case of child abuse—beating a child is uncontroversially regarded as an evil thing. Why is it an evil thing? Because of how it stands in relation to other evil things; the suffering the child experiences, the potential for long term physical or psychological damage, and so on. There are other properties such an event could hold, such as occurring in North America or there being over one hundred thousand hairs on the child’s head—as the the latter suggests even properties which would typically be beyond our ken. There is of course always a great deal more going on in any daily event than we're aware of, just consider how vastly complex human affairs would be on the cellular level. It wasn't long ago when these sorts of features of the world were well beyond our ken, much like the supposed greater goods that justify God's permission of all evils. But if we're going to admit that any event contains a great many inscrutable features, and that for all we know there is a vast realm of beyond our ken goods and evils, it follows that for all we know any of these inscrutable features themselves hold moral value—maybe even great enough to render our perception of the moral value of an event false. On skeptical theism there very well could be some vastly good thing that is among the many evil things related to child abuse such that it's value overwhelms the dis-value of suffering and physical or psychological damage. Notice that what's being said here is not merely that on skeptical theism the all things considered value of an event could be drastically different from how it appears, but that even that the moral value of the event itself could differ drastically from how it appears to us.
For a more concrete example suppose there was a slave owner whose knowledge of morality was restricted in such a way that he had no understanding of the human right to freedom. Nevertheless he knew that humans need sustenance, shelter, and understood that providing for another human's needs is a good thing. The slave owner might look at his situation and think to himself that he ought to be praised for feeding and clothing these people, and in truth he might actually be taking very good care of them. Regardless the state of affairs is, as we know, not a good thing. It's not just that forcing other persons to work for him has terrible (unbeknownst to him) consequences, but that the evil of denying another persons right to freedom overwhelms whatever goods might be involved in his particular case of slavery. The skeptical theist should admit then that for all he knows he's like the ignorant slave owner, thinking his actions are benefiting rather than harming others when really that is not the case.
How could one determine the right choice when they're completely out of their depth in judging the moral value of either option? Helping a little old lady across the street because it appears as if doing so would be good for her, for example, is irrational. For all the skeptical theist knows it would not benefit her at all or worse would harm her. The skeptical theist might say that regardless of whether his actions will be good for others, he can still know they would be good for himself as doing what appears to him to be evil has a degrading effect on his nature. It may be, for all the skeptical theist knows, that his murdering someone would be a very good thing, but as long as he believes it to be evil then his actually going through with murder would wreak havoc with his own conscience. In this the skeptical theist has prudential reasons to refrain from the sorts of behaviour that appear to him to be evil, and to indulge in the sorts of behaviour that appear to him to be good.
An alternate answer can be developed by re-assessing the implications of skeptical theism. There is a case to be made that the skeptical theist need not give up his grasp of the all things considered value of to deflect the noseeum inference. To see this we must distinguish between an evil's being gratuitous and God's permission of the evil being gratuitous. The skeptical theist wants to say that for all we know God has justifying reasons for permitting the evil he does. As Anderson points out, unless we assume that God is justified in permitting an evil only if the evil is itself not gratuitous, the skeptical theist need not doubt the all things considered value of the event itself. He continues to argue that, “[w]e might all agree that the occurrence of horrendous suffering is at least sometimes bad ATC without committing one way or the other on whether God’s permission of it always is. Without the very tight connection between gratuitous suffering and gratuitous permission that the assumption assumes, I see no reason for a general commitment to skepticism about the ATC badness of all cases of suffering.”3
The question, then, is whether there is any reason to think that God's permission of a gratuitous evil can be justified. Freewill of the libertarian sort allows for such an explanation. Suppose that God wanted to give John the freedom to choose between right and wrong and John chose wrong. Johns making the wrong choice was not necessary for his being free as he could have possibly chosen otherwise. In such a case the world might have been better off had John's choice never been freely made. If God was going to give John freedom then he had to run the risk of there being unnecessary evil, and yet God's taking this risk doesn't appear obviously unjustified. Yet if God can be justified in permitting gratuitous evil then it is of no consequence to orthodoxy theism that gratuitous evils do in fact occur. This defeats Pipers objection as the skeptical theist can then say that he has a very good grasp of the all things considered value of most events, while still thinking himself out of his depth in assessing the all things considered value of God's permission of those events.
I think a further objection can be made following the same spirit of Piper's, but being more threatening to Anderson's argument. Suppose now that this wasn't the first time John has chosen wrongly in this way, but that he in fact has been living a very sinful life for quite some time. Instead of just eliminating John, God might instead want to offer him a chance at redemption. Because of this God causes John to suffer terribly as both a punishment and to help John find motivation to turn away from his pattern of sinful behaviour. Now John's suffering is not necessary for the greater good God is trying to achieve as John, being perfectly free, could choose to reform his behaviour regardless of the state he finds himself in. Likewise it might be that despite God's stern parenting John never turns from his sinful ways and his life comes to a very sad end. In such a case it would have been better, all things considered, if he had not been made to suffer terribly. Despite this God would still be justified in permitting Johns suffering, assuming this was the best method available to him to motivate John to change his ways.
It seems that if God has this sort of reason for permitting or causing someone to suffer gratuitously, then it follows that we also have a reason to permit the gratuitous suffering. The notion that we ought not impeded the will of God is at home with many religious traditions, but even if it wasn't this claim would still have a great deal of plausibility. It's often considered a parents right to justly punish their child. Likewise it is simply not anyone else's place to lessen or end the suffering a good parent justly inflicts on their child for punishment's sake. If we can extend this analogy to our heavenly father then it would follow that for cases such as these we have an all things considered reason to permit gratuitous evil.
Now let's suppose a very different situation. God permits John to suffer, but not because it has the potential to benefit John in any way. Suppose there is another individual Stacy who is terribly immature and self centred. Gods purpose in permitting John to suffer could very well be so that Stacy will get involved and lend help where she can. In doing so Stacy might gain some maturity and a valuable understanding of how insignificant her own problems are compared to others. Now again John's suffering isn't necessary for any greater good as it's wholly possible that Stacy will not choose as God intends her to. Once again despite the gratuitousness of the suffering, God seems perfectly justified in permitting it.
In the previous case Stacy undoubtedly had an all things considered reason to prevent John's suffering. In the one before that Stacy (if she were in a position to) would have had an all things considered reason to permit John's suffering. If Stacy were a skeptical theist, how could she distinguish between the two? What we have is one possible God-justifying good that entails that we ought to permit a gratuitous evil, and another which entails we ought to prevent a gratuitous evil. These examples need not be taken as anything more than a demonstration that either possibility is live for the skeptical theist. As long as the skeptical theist is completely out of his depth in judging what sort of greater good might justify God's permission of an evil, he will be completely unable to discern between a gratuitous evil that ought to be prevented and one that ought to be permitted. In this we fall back to the original conclusion; that skeptical theism reduces all moral reasoning to considerations of what is good for ourselves. I suspect that for many, if not most skeptical theists will not be happy with this result—I certainly would not be. If all we have are prudential reasons to favour one choice over another, we can hardly say we're choosing right from wrong. Our choices, in this case, shouldn't have any more significance than a choice to brush ones teeth or to study for an exam. Like studying, it's in our best interests that we don't go about murdering others but there's nothing more to it than that.
In the face of such corrosion of moral reasoning the obvious thing to do is to G.E. Moore shift away from skeptical theism and in favour of a non-egoistic meta-ethic. With that Rowe's noseeum inference appears to be a good one, and the burden falls upon the theodicist to find objections to the evidential argument from evil. One might wonder if it matters whether God's actions are ultimately justified or not, maybe God is above this sort of requirement and can be good regardless. Bergmann speaks strongly about this supposition; “When we see or learn of utterly horrific suffering, the sensible and appropriate response is to be extremely upset that it has occurred and, with deep feeling, to think 'There had better be a good reason for God, if he exists, to permit that suffering; if there isn't, then there is no perfectly good God.'”4 Indeed, I can't imagine any other appropriate response. While I recognize that there are some interesting theodicies (such as Hicks soul making theodicy) I think there is significant tension between the more plausible ones and the traditional tenets of at least two of the Abrahamic religions. As these theodicies are motivated a great deal by traditional Christianity and Islam I believe a great many of these sorts of theists will have hard time finding successful answers to the evidential argument from evil. That being said the more bare theistic or open theistic individuals may very well form a plausible cumulative case against the argument with number of thoroughly developed theodicies.