How Coherent is Hume's Attempt to Disprove the God of Theism?
The first thing that must immediately be stated about this question is that it perhaps isn't completely fair to the task that Hume is trying to perform. In Dialogues Concerning Natural Religion Hume is not attempting to disprove the existence of God - far from it, he is trying to formulate a logical, reasonable, and above all justified system for believing in God. Rather than having a particular bias or axe to grind, Hume is attempting to present a balanced account of the issues involved (and I would say he goes a long way towards doing this), so we should not consider any developments that conclude that there is no evidence proving the existence of God to be a success of what Hume is trying to achieve. I would suggest that even before entering negative ideas (and the one that I will focus on with be 'The Problem of Evil') he has failed to produce a satisfactory justification for belief in the existence of God in the more positive first part of the Dialogues, and thus has already failed to prove the existence of the God of Theism.
One reason why this is particularly important is that the problem of evil is only really a problem once we have concluded that the weight of evidence points at classical ideas of religion being correct. It seems slightly strange, rather than using the problem of evil as a way of showing that the already established ideas about God might need more work, to attempt to disprove God to start with. The reason that I say this is that 'evil' is usually a very religious word, and one that cannot easily be used in a secular context. An eruption of a volcano or a murder only cease to be 'an unpleasant thing' and become a 'manifestation of evil' in a religious context. It does not seem logical for Atheists to use what would be classed to them as mere misfortune to claim that there can be no God, any more than it would make sense to say looking in the bible that Genesis 6 vii (plunked entirely at random) proves that God doesn't exist. It should be fairly obvious that the bible itself isn't going to be the primary tool to disprove God.
The problem of evil is that it doesn't seem possible that if there were an all powerful, all knowing and totally benevolent god, that there would exist evil in the world - surely such a god would not allow it. Therefore it does call into question whether the traditional view of God could hold, or whether at least one of these three qualities must be abandoned in order to preserve any deity with evil present in the world. Usually it is thought that we understand quite clearly what to be omnipotent or omniscient might mean (though perhaps we cannot imagine being so ourselves), and thus it is usually the third constituent parts that is debated when arguing against the traditional form of deity. 'Perfectly good' is a quality that is very difficult to pin down, especially in terms of behaviour, and the burden is clearly on someone attempting to claim that God is such to explain what they mean by it, rather than someone trying to point out that such a deity is somewhat incompatible with ideas of evil. Many people, however, question whether it is fair to take the anthropomorphic step of claiming that for God to be ultimately benevolent he must act in a way that humans would label good of other human actions. The argument continues that omniscience will by definition mean an understanding of 'the bigger picture' and thus such a being would not simply be thinking of short term implications, but of goodness in eternity as a whole.
Even Le Poidevin finds himself rationalising the theistic position stating 'Job dares to ask why this apparently unwarranted suffering has been visited upon a virtuous man', but it isn't clear in what sense he is using the word 'apparently'. There is no need for the suffering to be warranted - apparently or not - it may be that God understands that the only way to true goodness is through totally random punishment. Whilst it is a pretty weak argument, in that we can never prove it either way, it is entirely reasonable for a theist to claim that God understands things that humans can never comprehend, and thus truly understands how to use punishment.
Therefore such a question as 'Why are we not perfectly happy now?' falls down on three issues - why should we be the ideal good beings for whom god has created a world, and towards whom his benevolence is directed; why should happiness, in terms of our satisfaction, or simply pleasures, be the measure of goodness; and why must now be the time which God is aiming at, rather than his intentions being that we should eventually understand true goodness and one day achieve God's intended utopia. Gaskin too quickly dismisses the argument that a perfect world is not necessarily conducive to good or happy behaviour - and he seems almost naïve to believe that short term perfection must ultimately be the best solution.
One answer to this might be to claim that it should be clear that happiness through all time must be better than happiness in part of it, and thus clearly to be aiming at it in the future overall will not reflect ultimate benevolence since, however good / happy the world is in the end, if it were that happy all through time it would surely be better, and reflect a greater benevolence. It is at this point of the argument about evil that the theists invoke the most important defence of the claim that evil is compatible with their beliefs. I find this response entirely satisfactory, and I think that it is on this basis that attempts to use the problem of evil to argue against theism fall down. The defence is simply that in order for God's divine plan to be enacted, humans must choose to be good, for to be good without any choice would make them automata, and not in any way moral beings. I find this explanation entirely satisfactory - and whilst fundamentalists might attempt to claim that it is in fact God's intention that people should be good all the time and that they will burn in hell if they are not - there is considerable evidence for this view in the bible. There seems, for example, little explanation for the story of the fall from Eden unless it is to show that God wants humans to be good through choice. Whilst Adam and Eve disobey God's direct command, at least they have the choice to do so, and it is only this form of choice that will allow humans to be good, and to distinguish themselves from other amoral beings. Once they have fallen from Eden they have the opportunity (as well as the motivation) to choose to be good, rather than simply being good with no choice. There seems little point in the story of Abraham and Isaac if he had felt forced to sacrifice his son - the story that is actually presented is that he realises that worshipping God and doing as he commands is good by himself, using his moral faculty and free will. One of the major (and frequently missed) points in the old testament is that the Red Sea does not part until the Israelites actually step into the water - it is not simply God blindly helping out the right people, so that good automatically occurs, instead he allows them to take a test of faith, and by choosing to put their trust in God they do indeed survive.
I find, therefore, the argument that free will is absolutely essential for humans to be moral beings to be perfectly persuasive, and thus that God could in fact not be entirely benevolent without allowing humans to do evil - though presumably with some understanding that they will hit upon the realisation that they're going down the wrong track. I do not intend to go into the argument about whether we have free will or not, but I would like to quickly note the difficulty for a modern proposer of determinism, with the entirely random possibilities of quantum theory surely making it pretty impossible to suggest that the entire future is mapped out by the current state of the universe. If a single electron can go one way rather than another surely it could ultimately upset the whole balance, and produce an entirely different time-line. Le Poidevin attempts to debate whether determinism causes problems for those attempting to use free will to evade the human side of evil, but I am uncertain how such a debate would help - surely the theist could protect his ideas about God by putting human evil down to free will, and it would be up to the determinist to state why this would absolutely not be possible (rather than simply claiming this might not be the case, if determinism were true).
In the context of the Dialogues, however, we should not be as forgiving of the theistic attempt to justify the strange behaviour of God in such a way. The Dialogues is a positive, not a negative, work, and we are trying to establish for certain particular faculties of God. The reason that the problem of evil is especially pertinent in the case of the Dialogues is that it isn't clear at all that we should be able to conclude from the evidence of the deities supposed result (the universe) anything like as much as the theists would like to say about the cause. As Gaskin points out:
'It is for Hume just a special case of the general illegitimacy of inferring more power, more goodness - more anything - in the cause than is actually manifest in the effect, the Universe itself'.
Whilst we can't say for certain that God doesn't have a special insight that as humans we are not aware of, and is not exercising ultimate benevolence, even though we can't tell, what we certainly cannot say is that all the evidence says he does do this. Hume is attempting to prove the things that we can be certain about with regards to God, and he points out quite clearly that using the Argument to Design it is not possible to infer anything more than that what is directly suggested by the evidence of the universe. The problem of evil is indeed very problematic in these circumstances, since, rather than suggesting that there is a benevolent god, as Philo quite correctly points out the evidence would suggest anything but such a being. Few philosophers (though some such do exist) would attempt to argue that in fact we are mistaken, and that we do not in fact experience pain in our daily life - and at a basic sensational level I cannot see how such a claim could be justified. So, as Gaskin points out, we are left to believe that either the world is all bad (which is going too far, in my opinion, and is just as against all the evidence as saying the world is all good would be), or to claim that it is mixed - good and bad. But from the evidence the final option does actually seem by far the easiest - simply to believe that the world is neither. The belief that the world was not created with concepts of 'good' or 'evil' in mind is by far the strongest from the point of view of evidence, and to claim that in fact a certain order was imposed to do with levels of good and evil is simply complicating matters in a way for which there is no justification.
Overall, therefore, Hume is quite right to spot that the problem of evil is a major obstacle for the proponent of the Argument to Design, at least in terms of believing that God is as traditional Christians would hold him to be. There is little or no evidence in the universe that God is exercising an attempt to regulate the good and the bad, and if anything evidence would suggest that he is happy with the existence of a good measure of each.