The Non-Contingency Condition (NC): No moral obligation, R, is objectionably contingent.
The Accessibility Condition (AC): Necessarily, for any moral obligation, R, R is binding only if R is cognitively accessible in principle.
However, it appears that divine command theories tend to violate one or the other: that is, accounts that preserve NC violate AC, whereas accounts that preserve AC violate NC. I think this is a promising line of argument, and probably some version of it can be made to work.
I don't think it's quite in shape yet, though, in part because I think the argument against command accounts, which forms the backbone of the argument that divine command theories preserving AC violate NC, is not quite sophisticated enough to handle all divine command accounts. A key element of the argument is the following scenario:
For consider a possible world, W, in which there exists a community of five people: a man (call him ‘Zed’), and four children, none of which is older than ten years old. Suppose further that this community has no practice of commanding. Finally, suppose that Zed regularly sodomizes these children. Since, by hypothesis, this community has no practice of commanding, command formulations of DCT entail that none of Zed’s acts are morally wrong. But, surely, moral wrongness supervenes on at least some acts in W. But then moral wrongness doesn’t require the existence of a community practice of commanding as a part of its supervenience base. And if not, then command formulations of DCT are false.
This is actually fairly similar to an argument found in Catherine Trotter Cockburn in her attack on Warburton's divine command theory -- well, similar in function, since Cockburn's example didn't involved pedophile sodomy, but just a man trapped in a pit in a society of atheists. (I quote part of that argument here.) But there are some similarities in how the example is used.
The problem there, as here, is that it's actually not wholly clear whether this is a problem for a divine command theorist like Warburton, because part of Warburton's argument is that what we call 'morality' is actually a complex of things -- a threefold cord -- and that when we talk about something being 'morally wrong' an adequate account of what we mean has to take into account all three parts of this complex. Warburton could perfectly well agree that there's moral wrong in Zed's case; what he will deny is that Zed has violated any obligations. Zed's action can be wrong in the sense that it's the wrong sort of action to commit if you are considering all the consequences of your actions and want to avoid detrimental ones as much as possible. The damage he does to the children is imprudent, and could return on him someday. Further, Zed's action can be wrong in the sense that it's disgusting and distasteful, and doesn't fit with our notions of what a civilized person does. So it's wrong in the sense that it's bad sense and bad taste. What Warburton will deny is that there can be anything more to it in the absence of the expressed will of a superior; Zed's not violating any obligations because no one has obligated him. Thus, Warburton will say, you can have moral wrongness, i.e., wrongness relevant to morals, in the absence of such a command; but if you condemn someone on the basis of it, your condemnation of it only differs in degree from a condemnation based purely on prudential calculation and/or aesthetic taste, and even the union of the two can't yield the normative force we attribute to obligations. (It certainly is utterly stupid and utterly distasteful to murder someone, but that, Warburton would say, does not constitute an obligation not to murder. It just means that murder is senseless and icky.)
So there are divine command theorists who can escape the dilemma, at least as far as the post lays it out. Since the dilemma laid out there is only a 'quick and dirty' one, there's the possibility that a more developed version can close the loopholes. Cockburn argues that Warburton is wrong about his claim that rational appropriateness can't constitute an obligation, and while it's an uneven argument, it's promising as well. Perhaps a combination and development of the two would do the trick.
