He first runs into the problem that one may answer the question with just anything (by the principle of explosion) if one assumes that the question contains an impossible antecedent. He anticipates this response and says that on the contrary such is not “inconceivable.” The problem is this, however: if one considers God to be morally perfect and the antecedent is not morally perfect, then the question is in fact inconceivable. To say it isn’t is just to admit, in the case of the omnibenevolent God (which the argument assumes in order to show its untenability), that such a God’s commanding the actions is possible, and hence the conclusion is false. My point here is that the worst that follows is that Abraham was mistaken and the Bible is not inerrant (a hefty price, to be sure, but one that is somewhat less than an omnibenevolent God’s existence!).
Of course, he may rightly question whether I would take the above route. The answer is that I would not. So, what are the three options? The first (1) is to answer, “No, because this would be murder.” If we do that, then we accept the premise that a supposedly good God commanded murder (and murder is always wrong). (1) is obviously not a viable option.
(2) is to answer “yes,” we would be morally justified in obeying. However, the reasoning is because what “God commands is by definition good.” He offers a two-pronged critique. I will quote his own words here: “If what is good is that which God commands, then, presumably, He may command and perform any act which, ex hypothesi, must be good, in which case morality may be said to be arbitrary and capricious; entirely contingent upon what God may at any time decree.” The other prong of the critique is to say that God’s commands and the good are simply identical, and thus the question of what “good” is really becomes meaningless for the theist here.
The third option (3) is to say that God is a necessarily good being, and thus it is always good to follow commands given by a necessarily good being. His criticism seems to mirror the second prong against (2) above, followed by supposed counterexamples (i.e., examples of the Abrahamic God behaving in ways different than we would expect an omnibenevolent being to behave).
It occurred to me he did not quite tease out the third option, for it will be this teased-out third option that I will claim. If one will recall, the second option is that of a causal chain: God commands X, X becomes good. This is indeed arbitrary, and it is for this reason I think it should be rejected. However, (3), properly understood, should reverse the causal direction: X is good, God commands X.
The author claims, “To say God is by necessity ‘good,’ and for such a pronouncement to be meaningful, the theist must be able to delimit actions that God cannot perform because they are bad.” The only way I know to take this indicates he thinks we ought to be able to list all the actions God cannot perform before we are able to say God is a necessarily good being, and that I don’t think has been shown. In fact, he doesn’t really argue for it at all. All we mean by “necessarily good” is that he is the ground of objective moral values; it is a part of his nature. That does, by definition, mean there are certain actions he cannot perform, but it does not at all follow that we cannot say God is a necessarily good being without knowing all the actions he cannot do. I certainly do not need to know all of the false answers to 2+2 in order to claim 4 is correct. While it is true that the value of 4 ontologically delimits the number of values that can be sufficient for combining a pair of 2s, it does not follow that we must delimit a list of things it cannot be to know what it is!
However, I think the author’s point is that if we say “yes, we are justified in killing our children if God so commands because God is necessarily good,” that this necessary goodness also necessitates God cannot perform evil. Killing children is evil. Therefore, (3) is not an acceptable answer. However, this just assumes what it seeks to prove. Using words like “genocide,” “murder,” “torture,” and “slavery” is just question-begging. This is because the way we use such words entails morally evil content. Now, if the author merely means to be descriptive of the events, let him use synonyms or sentences to describe them. In many cases, they lose some (or even most) of the force.
Next, he attempts to demonstrate the biblical record contains these atrocities. Aside from ignoring Paul Copan’s critique (which defends the biblical accounts quite well), those who take (3) as an option have plenty of avenues to explore. The argument runs like this:
A. No act commanded by a necessarily good being is evil
B. A perceived command is evil.
C. Therefore, it is either not the case that the command was given or the command was not given by a necessarily good being.
In neither option is the theist committed to saying that holding to an omnibenevolent God is untenable. In the first scenario, possible (and even plausible) solutions include: commands being misunderstood, Biblical inerrancy’s falsehood, the command is not to do evil (in the case of
Copan’s defense), God’s not being obligated to extend life, etc. In the second, solutions include: command misunderstood to be from God but from other source, made up command, infused genuine command with meaning to kill all, etc. I don’t even have to pick any one of these (especially since some are decidedly less plausible or palatable than others), but it remains that holding to an omnibenevolent God is not “untenable.”