Modern divine command theorists say that God’s commands flow from his essential nature. Since God is loving, we can know that he would never command horrible things, like the torture of a small child.
On the other hand, nearly all theists believe that God allows evil things to happen while simultaneously accepting the claim that God would prevent any gratuitous instance of evil. In other words, theists are committed to the following principle:
(E) Every instance of evil that occurs is such that either (a) its occurrence is necessary to prevent the occurrence of something equally bad or worse, or else (b) its occurrence is necessary to bring about some greater good.
If there is any instance of evil that does not satisfy either (a) or (b) (or both), then it is gratuitous. A perfectly loving God would eliminate any unnecessary evil.
Thus, what a theist believes is that if a person suffers, it is better that they suffered than had they not suffered because either (a) their suffering was logically required to prevent something worse from happening, or else (b) their suffering was logically required to bring about some greater good.
Suppose I have a close friend or family member who is addicted to heroin. He comes to me broke and suffering from severe withdrawals because he has been unable to purchase the drug or acquire it through other means. He asks to borrow some money and it is obvious to me that he intends to use the money to buy more heroin. Now, I submit that the right thing to do is to refuse to give him the money. I should do this even though I know that if my friend does not get his fix, he will continue to suffer greatly as his withdrawal symptoms get worse. This is because it would be better for my fried to go through the withdrawal on the path to losing his addiction than to relieve his suffering temporarily by feeding the addiction. That is, I should allow my friend to suffer because it is in his best interests to suffer (even if he doesn’t agree that it is). For the purposes of this discussion, the important uphsot of this example is that it is not the case that if I love someone, then I will prevent any instance of their suffering that I can prevent. However, we can even go further. Indeed, it is plausible that the best thing for me to do in this instance is to take my friend to a detox clinic where he can endure his withdrawal symptoms in a controlled environment. That is, the best thing to do is something that will cause my friend more suffering.
So, in general, it is not the case that if I love someone, I want to prevent every instance of their suffering. Rather, what I want is what is best for them. And, in at least some cases, what is best for them is that they suffer. Now, typically what is best for a person is that their suffering is minimized (at least that is what I think we have most reason to believe; I am not sure that a theist can accept this claim), but at least sometimes, a loving person will allow those they love to suffer.
Here is what is important for the present argument: That God is loving does not imply that he will want to prevent all suffering; it implies that he wants what is best for us. This is what underlies principle (E).
Now, if we are not in a position to know whether, for any instance of apparently gratuitous suffering, the suffer is better off than she would have been had she not suffered, then we are not in a position to know whether a loving God would command torture.
Atheists typically believe that cases of apparently gratuitous suffering really are cases of gratuitous suffering. The suffering endured by a dying cancer patient appears gratuitous. There does not appear to be any greater good such the the patient’s suffering is necessary to bring about that good; nor does there appear to be any evil equal to or greater than the patient’s suffering that the suffering is necessary to prevent. The atheist says that things are exactly as they appear; such an instance of suffering is gratuitous. The theist, however, has to believe that appearances are deceiving. The theist believes that the patient’s suffering is not gratuitous because he believes that God exists and that God is loving.
Given this, what reason can a theist give for believing that God would not command the torture of an infant? That God loves the infant? Well, we just saw that, in general, being loving does not entail wanting to prevent every instance of suffering. Rather, it entails wanting what is best for the beloved. So that God loves the child is not evidence that God does not want the child to be tortured. If the torture of the child will bring about what is best, then God, being loving, will command it.
So, what the theist needs is a reason for thinking that it is never best that children be tortured. Without that, we cannot know that God would not command the torture of an infant. But can the theist provide such a reason? I don’t see how. If theists are willing to believe that the cancer patient’s suffering is not gratuitous, that somehow the world is better off for that instance of severe pain, how can theists consistently maintain that they know that an instance of infant torture is gratuitous?
It is common for theists, during discussions of the problem of evil, to point out that, given our epistemic limitations, we are not in a position to know that God does not have reasons for failing to prevent the many horrendous and apparently gratuitous evils in our world (the name for this position is Skeptical Theism). Well, the same would seem to apply to any instance of child torture. If our epistemic situation is so limited that we cannot know that God does not have reasons for permitting the slaughter of children, then it must be so limited that we cannot know whether child torture is sometimes necessary to bring about some greater good or prevent something equally bad or worse. Thus, the theist cannot know that a loving God will not command the torture of infants.