David Efird's post on whether Divine Command Theory (DCT) can be shown to be false or circular contained the following argument:
(1) Necessarily, if God commands that subject S commit action A, then it is ethically right for S to commit A.
(2) Possibly, God commands subject S to commit an act of wanton cruelty.
So, (3) Possibly, it is ethically right for S to commit an act of wanton cruelty.
David wanted to deny (1) and had to deal with a charge of circularity. My inclination was to deny (2) and say that God couldn't command something contrary to his nature, but David's worry about that was that you then have denied DCT. DCT is supposed to explain morality in terms of God's will, but then if you deny (2) it seems to be not God's will but God's nature that grounds morality. I think this is a mistake, and William Alston's "Suggestions for Divine Command Theorists" gives a way to do this without that consequences. In the comments, I'd mentioned Alston's solution, but now that I've gone back and read his paper I think I did his suggestion some serious injustice, so I'd like to explain how Alston's response would handle the issue at hand.
Alston's main strategy is to separate the notions of moral obligation and moral goodness. Moral obligation is the subject of DCT. Moral obligation is derived entirely from God's commands. Alston doesn't think God has any moral obligations for the same reasons he doesn't think God has beliefs. Obligations admit of the possibility of not meeting them, and God doesn't have such a possibility. Obligation derives from God's commands, which derive from God's nature, and God's nature includes moral goodness. Moral goodness is therefore a prior notion to moral obligations, which only we have and not God.
It would be circular to define moral obligation in terms of God's will and then say that God's will is the way it is because God's nature is morally good, if moral goodness itself derives from moral obligation. But the account Alston offers says moral goodness is prior to moral obligation. God has the former but not the latter, and we have the latter in virtue simply of God's commands.
This solution has the benefit of retaining DCT's central tenet, that obligation derives merely from God's commands. It avoids the problem of having morality independent of God's existence, thus subjecting God to external requirements or standards. It also avoids the problem of arbitrariness, since God's commands have limits due to God's nature being morally good. God couldn't command the torture of children for the fun of it, because it's against his nature. Finally, there's no circularity because the ground level of explanation is God's metaphysical goodness. He doesn't ground that in something further, particularly not grounding it in God's commands as the standard introductory philosophy class version of the Euthyphro has DCT doing.
This last point brings up the main objection that I had, which Alston does deal with. The objection says that God's goodness as the ground level of explanation here means we haven't explained anything, since we now haven't explained God's goodness. Alston replies that with any explanation we have to stop somewhere. I'm not convinced that this reply would convince the agnostic or atheist in the face of a moral argument for God's existence, since they could claim that we could just stop somewhere higher up, e.g. with either brute (i.e. contingent but without any explanation) or necessary moral truths.
That's a fair point if the argument is to convince the atheist or agnostic of the existence of God on the basis of the existence of morality, though I do think brute moral truths should sound really spooky to most metaphysicians, and necessary moral truths not grounded in God's nature don't strike me as any better explained simply by stipulating that they're necessary. Still, even if this is a good response to a moral argument for God's existence, all Alston is arguing here is an account of God's relation to moral obligation and moral goodness. He has done that, and it avoids the arbitrariness of God's commands and the circularity of interdefinition. If it turns out not to help in an argument against the agnostic or the atheist, that doesn't threaten what Alston was up to in this paper.
"all Alston is arguing here is an account of God's relation to moral obligation and moral goodness."
We are still left with this: God's command is binding on us only because God is good. Thus, the obligatory force of divine commandment doesn't "derive[] merely from God's command[]." It derives from the prior postulate that those commands are issued by an omnibenevolent god.
Thus, it seems to me, this modification of DCT renders DCT neither an explanation of obligation nor or of goodness. There is still a prior and independent notion of good that is inelimable from the theory.
Incidentally, Alston presupposes the necessity of God and of God's goodness. While that might be the prevailing view in modern Christian theology, one needn't go outside Christianity to find skepics on that point.
Yes, Alston realizes this. As I said in the post, this doesn't threaten DCT itself but rather (at best) the use of it as a moral argument for the existence of God (conceded by me only with the cautionary suggestion that brute and necessary moral truths without God may still be spooky to a naturalistic metaphysician). I don't see why it's a a problem for Alston's account. of the relation between God and morality, since he admits that explanation has to stop somewhere. The fact that it's merely an account and not an argument for that account also mutes the second point a bit. The existence of another account doesn't undermine this one very much.
If we are careful about how we speak when speaking of DCT, the dependence of God's will on his nature should not bother us. It is, according to DCT, God's will that explains obligatoriness, and what God wills is explained by his nature. You only get obligatoriness explained by God's nature if you take the explanatory relation to be transitive. My money is on the denial of that, as it is regarding causation...
I can see denying transitivity of explanation if you're not talking about sufficient explanation. Is that what you have in mind? Presumably it's what you have in mind for denying transitivity of causation, since not every cause is sufficient. Or is it more controversial than that?
Yes, I think even sufficient causes/explanations fail of transitivity. In a deterministic universe, there's lots of sufficient causes/explanations, but I don't think the original state of the universe should be thought to cause everything throughout history.
I think the temptation to affirm transitivity here is because of the word 'sufficient'. Logical sufficiency is transitive for first-order theory as well as for modal logics, but even in logic things are messy: think of counterfactual sufficiency, for example. Transitivity fails for counterfactuals. Take A implies B and B implies C. Assume there are closest worlds for simplicity. Then here's a simple model on which transitivity fails: the closest A world is a B world and a ~C world, while the closest B world is a C world and a ~A world. To the extent that you are tempted to see counterfactuals and causation as related, this is motivation to deny transitivity of causation, even sufficient causation. Even if you don't like the counterfactual theory of causation, the failure of transitivity for counterfactuals still gives reason to doubt transitivity for (sufficient) causation.
It does seem a little weird to say that I caused my children's death by leading to their existence, though by transitivity of causation I did. This was something that bothered Lewis in one of his later papers on causation. So you're just denying the primary/secondary cause distinction and saying that what people call secondary causes aren't causes at all? I'll have to think about whether the nearest possible world trick would work with all the primary/secondary cause distinctions it would need to work with to match up the failure of transitivity with all the cases needed to absolve God of moral responsibility or to prevent earlier but near conditions from causing me to do something if I'm determined.