In an ongoing dialogue in this journal (Sophia), Robert Larmer and I have been discussing whether the undisputed occurrence of certain conceivable events — for instance, astonishing healings — could require all honest, thoughtful individuals to acknowledge that God has supernaturally intervened in earthly affairs. I have not denied that a theist (or nontheist) could justifiably consider the occurrence of certain possible (or even actual) events to be strong evidence for theism — for the existence of a God who benevolently intervenes in earthly affairs. But nontheists, I have argued, can justifiably maintain that evil — that the amount and nature of human pain and suffering — stands as strong evidence against God’s existence. Furthermore, I have argued, nontheists can justifiably maintain that the evidence against God’s existence generated by evil would outweigh any amount of evidence for theism that might be produced by any conceivable set of events. And for this reason I have continued to deny that there exists any conceivable context in which a person who did not acknowledge that God has intervened in earthly affairs could justifiably be accused of having conducted herself in a nonrational manner.

In his latest response to this line of reasoning, Larmer no longer denies that a nontheist can, at present, justifiably maintain that evil is strong evidence against God’s existence. In fact, he no longer challenges my claim that a nontheist can, at present, justifiably deny that the evidence for theism furnished by any set of conceivable events would outweigh what this nontheist can justifiably consider to be the evidence against God’s existence generated by evil.

But there are, he still wishes to argue, conceivable contexts in which it would no longer be rational for a nontheist to claim that the evidence generated by evil outweighs the evidence for divine intervention. Let us assume, he asks, that “Basinger’s sophisticated naturalist approaches my miracle-worker and informs him that, although [a set of] ‘miracles’, considered in isolation, provide strong evidence for theism, she feels compelled to interpret them naturalistically, since she feels that the disconfirming evidence from the existence of evil is so strong.” And let us further assume that the miracle-worker “asks her to draw up a list of the evils she has in mind and proceeds to work down the list, remedying each evil as he goes.” Surely in this case, he argues, we would “arrive at a point, even before the bottom of the list is reached, that the naturalist would have to revise her views in favour of theisms.” Different naturalists, he acknowledges, would “revise their views at different points, but this in no way demonstrates that there will not reach a point at which every rational person would have been convinced.”

This example, he readily admits, “seems farfetched.” But it does, as he sees it, exemplify a valid point: that “unless the naturalist is prepared to argue that it is logically impossible that God would allow any evil whatsoever . . . she cannot argue that disbelief in God is rational now matter how small in principle the class of evils becomes.” And this is enough, he maintains, to demonstrate that I am wrong to conclude that “no matter the ‘miracles’ in question, and no matter the circumstances in which they occur” the body of evidence generated by these miracles “could never, even in principle, be such that it would persuade all rational persons.”

But I remain unconvinced. First, I am afraid that Larmer has (once again) misinterpreted my basic position. I have no idea how all or even most people would actually respond to a given set of very unusual occurrences that seemed to point, in isolation, to divine intervention. Thus it has never been my contention that some conceivable set of occurrences would never, as a matter of fact, convince all rational people that God exists and has intervened directly in earthly affairs. What I have continued to argue, rather, is that, regardless of what any number of people do in fact decide to believe, a nontheist could, in the face of any hypothetical ‘miracle’, justifiably continue to deny that this event is the result of God’s direct supernatural intervention. That is, the contention I have been defending is that

(1) No event — no alleged miracle — could ever, even in principle, be such that it would require all rational persons to acknowledge that God has directly intervened in earthly affairs. However, this mistake is not fatal to Larmer’s case. It is Larmer’s contention that

(2) There is, in principle, some conceivable point at which every nontheist would have to acknowledge that the evidence against God’s existence generated by evil is so weak that it can no longer justifiably be considered to outweigh the evidence for theism.

And if (2) is true, then he has in fact offered a valid counter to (1) — to my claim that there exists no conceivable context in which no rational individual could justifiably deny that God exists and has directly circumvented or modified the natural order. However, I continue to deny that (2) is true — that the evidence against God’s existence could, even conceivably, become so weak that no rational individual could at that point justifiably continue to consider it to outweigh any conceivable evidence for God’s existence. Larmer bases his affirmation of (2) on this belief that…

(3) No nontheist can justifiably maintain that an all- powerful, perfectly good God could allow absolutely no evil whatsoever.

But (3) is in a very important sense a dubious. The evil with which the vast majority of nontheists are concerned is the pain and suffering experienced by innocent individuals. And it is, as a matter of fact, held by many nontheists that the innocent suffering of even one person in a world alleged controlled by a being who has the power to prevent it is enough to demonstrate that no perfectly good being with such power exists. Many resonant, for instance, with the type of rhetorical question posed by Dostoyevsky’s Ivan:

Imagine that it is you yourself who are erecting the edifice of human destiny with the aim of making men happy in the end, giving them peace and contentment at last, but that to do that it is absolutely necessary, and indeed quite inevitable, to torture to death only one tiny creature…, would you consent to be the architect on those conditions?

And Larmer, as far as I can tell, has given us no reason to conclude that this nontheistic position — which stands as a counterinstance to (3) — is “indefencible.” But there is, I believe, an even more serious, systemic problem with (2) — with Larmer’s contention that the evidence against God’s existence could, even conceivably, become so weak that no rational individual could at that point justifiably continue to consider it to outweigh any conceivable evidence for God’s existence. (2) can be affirmed only if

(4) It is conceivable that we will at some point in the future come to possess a set of neutral, nonquestion-begging criteria for belief assessment in relation to which we can objectively determine the extent to which any evil (or set of evils) can justifiably continue to be considered strong evidence against God’s existence.

However, (4) is exactly what I continue to deny can be affirmed. I continue to believe, in agreement with such diverse philosophical company as Hume, Kierkegaard and Alvin Plantinga, that the question of whether evil can justifiably count as strong evidence against theism will, by its very nature, always remain primarily a subjective, relative matter. And Larmer has still given us no reason to believe otherwise. That is, Larmer has still given us no reason to believe that the question of whether evil can count as strong evidence against God’s existence is the type of issue that could in the future (anymore than now) be subject to objective adjudication.

Thus, while I admire Larmer’s perseverance and ingenuity, I continue to deny that the undisputed occurrence of certain conceivable events could require all honest, thoughtful individuals to acknowledge that God has supernaturally intervened in earthly affairs.