Do all theodicies fail?

I'll comment on part of a thread by atheist Jeff Lowder:

@SecularOutpost

Likewise, when atheists argue that facts about evil, pain, suffering, imperfection are evidence against God's existence, it's a complete nonstarter to talk about how God is logically compatible with those facts. 

i) That depends. Mere logical compatibility might be a makeshift explanation. That's not sufficient. If, however, logical compatibility means evil, pain, and suffering are not surprising given the overall tenets of Christian theism, then that's a legitimate explanation. If that fails to satisfy the evidential argument from evil, the failure is not in the explanation but in the way the evidential argument is formulated. 

ii) It's unclear what Jeff means by "imperfections". For instance, it's not a design flaw that I don't have fireproof fingers. If I accidentally burn my fingers, that's not an imperfection. Fingers need to be sensitive to perform many functions. Fireproof fingers would be numb. 

For parallel reasons, all known theodicies for the arguments from evil fail. They provide a possible explanation for which we have no independent reason to believe is true and/or the explanation is not probable on the assumption theism is true.

i) That's ambiguous. We often resort to explanations that are reasonable even though we lack independent evidence that they are true. Why is someone late for work? Maybe they had a flat tire, accident, or family emergency. We don't require corroboration for that conjecture to make it a legitimate conjecture. We know that those kinds of things happen. We know he's a responsible employee. 

ii) It's often rational to provide a possible explanation when we have no independent evidence that it's true, because it's not necessarily about having direct evidence for the explanation, but indirect evidence given the character of the agent, as a competent and benevolent agent who has good reasons for what he does.

iii) On the assumption that Christianity is true, it's not merely probable but inevitable that some of God's actions will be inscrutable given the complexities of historical causation. Normally it's wrong to inflict pain on a young child. The child doesn't understand why the doctor is performing a painful procedure. He doesn't understand why his dad is standing by, allowing that to happen. 

(Aside: I also forgot to mention another requirement: the theodicy or atheodicy has to make the fact to be explained probable. Many theodicies and atheodicies also fail this requirement.) 

Probable in relation to whom or what frame of reference? An atheist? 

Once again, an explanation needn't be probable to be legitimate. To recur to my previous example, an employee may be late for work because they had a flat tire, accident, or family emergency. Since I don't know for a fact why they are late for work, I can't say which explanation is probably the correct explanation. And it may be an explanation I didn't consider. 

But we could put this in reverse: it's improbable that he decided to play hooky, given his track-record as a responsible employee. 

For example, the pain a terminally ill patient feels in the hours or days before death does not aid in survival or reproduction. Now, if theism is true, then God must have a morally sufficient reason for allowing all pain, including pain which does not aid in survival or reproduction. 

i) Because the human nervous system is fairly coarse-grained. It wasn't designed to be that discriminating. But that cuts both ways. It wasn't designed for us to enjoy chocolate gelato. That doesn't aid in survival or reproduction. But why is that the only justification? 

ii) Because Jeff is locked into attacking generic theism, he overlooks distinctive assumptions, resources, and explanations provided by Christian theism. He acts like all pain must be a design flaw, as created. But in Christian theology, the natural world always contained dangers and potential sources of pain. The difference is that in an unfallen world, humans might expect special providential protection from certain kinds of harms. 

The basic idea of UPD is that God exists, and God may have a morally sufficient reason for allowing pain, suffering, imperfection, or evil, and that reason is unknown to (or unknowable by) humans. 

That reply is good as far it goes, but it doesn't go far enough to defeat the atheistic arguments from pain, suffering, imperfection, and evil. Yes, God may have unknown reasons for allowing such things, but he might also have unknown reasons for preventing such things. 

There is no antecedent reason why God-permitting reasons are more likely than God-preventing reasons, and so both of those reasons cancel out. What we're left with is what we do know. 

i) Those aren't mutually exclusive explanations. It's not a choice between God permitting every evil and God preventing every evil. Some evils are necessary sources of second-order goods, but too much evil swamps the good. There is no general principle that God-permitting reasons are more likely that God-preventing reasons, or vice versa. God-permitting reasons ought to be more likely that God-preventing reasons. That assessment involves striking a balance between competing goods that humans lack the information and intelligence to appreciate.  

ii) In addition, the preemption of evil is invisible. A nonevent leaves no trace evidence. So we can't do a comparative assessment of how often God permits evil in relation to how often God prevents evil. We only have one side of the comparison.

iii) However, both divine permission and divine prevention of evil have a disruptive impact on the future. So these represent alternate world histories. A world where God prevents more evil will have a different history than a world in which God permits more evil. 

iv) That, however, doesn't mean a world with less evil is a world with more good. Some evils are necessary evils insofar as some evils provide the necessary conditions for certain kinds of goods.