#949 M. Huemer’s Criticism of Divine Command Theory

Dr. craig’s response

I have already done so, Graham! In my debate with Erik Wielenberg on “God and Morality” I defended a Divine Command Theory (DCT) of ethics. In the published version of our debate Wielenberg tapped Wesley Morriston and Michael Huemer as commentators on the debate, to whom I in turn responded.[1]Although much more is said in the book than I can summarize here, I shall extract my responses to Huemer’s objections to my DCT.

In the debate I argued that DCT provides a good explanation of the objectivity of moral values and duties. I defended two subpoints: 

Let me say a word about Huemer’s responses to each of these points. First, concerning moral values, Huemer misrepresents the theory, for he speaks of grounding moral values, not in the maximally great being, but merely in a benevolent creator of the universe, which no DCTheorist would think adequate. Although Huemer scorns the ontological argument predicated on Anselm’s famous definition,[2] it is important to see that conceiving of God as a maximally great being in no way commits one to the cogency of the ontological argument. Even if one does not regard the ontological argument as a successful piece of natural theology, still Anselm’s argument can be of tremendous value as a guide to a robust concept of God as a being which is metaphysically necessary, omnipotent, omniscient, etc. Such a being serves as an explanatory stopping point that avoids arbitrariness concerns. 

Huemer complains that DCT never really explains why anything is good/bad or right/wrong. But this complaint is merely a repudiation of a paradigmatic approach to determining moral values.  For the reason something is good/bad according to the theory is its degree of resemblance to God, and the reason something is right/wrong is its accord with God’s commands. Since Huemer likes scientific analogies, consider the debate among palaeoanthropologists over what it is to be human. There is no consensus concerning necessary and sufficient conditions for classifying a hominin as human.[3] But if we take ourselves as paradigms of what it is to be human, then we can use the degree of resemblance to that paradigm as a way of speaking objectively of prehistoric human beings. A paradigmatic approach to moral values does not preclude describing the paradigm as loving, generous, faithful, kind, and so forth, without thinking that those properties make the paradigm good. As least Morriston sees this point, which Huemer does not. 

Finally, Huemer’s objection that DCT is not an objectivist theory of moral values is maladroit, since on DCT moral values are, in line with Huemer’s own definition of “objective morality,” observer-independent, being rooted in the divine nature, not the divine will, and so independent of divine attitudes.

So what about DCT’s account of moral duties? Here we recur to Huemer’s allegation that DCT is a subjectivist theory. We saw that that allegation is flatly false with respect to moral values, since these are independent of the divine will. But what about moral duties, which are constituted by God’s commands and so are dependent on his will? Truths about right and wrong are not independent of God’s attitudes. Now when one reflects that much of what God morally wills and commands is willed and commanded by him necessarily, being entailed by the divine nature, and therefore neither contingent nor arbitrary, then the dependence of right and wrong on God’s attitudes serves mainly to call into question Huemer’s characterization of “objective morality.” Attitude-dependence is supposed to express the contingency and arbitrariness of the feature so dependent, and that is what makes subjectivism in morality objectionable. But that fails significantly to apply to God. A better definition of objective morality would appeal to its independence of the attitudes of human observers or finite observers. 

Huemer does not like this move because “Physical facts – the paradigm of objectivity – are not constitutively dependent on any observers whatsoever.” Now even if we reject observer-dependent interpretations of quantum mechanics, still Huemer’s affirmation fails of Special Relativity theory, which holds that simultaneity and relations of earlier and later are, indeed, observer-dependent.  This most emphatically does not mean that such relations are subjective, but that they are relative to inertial frames. Moreover, on theism physical facts are no more independent of God’s attitudes than are moral facts—indeed, being contingent, they are less so—for they depend upon God’s will to create the physical objects and preserve them in being. Observer-dependence, then, ought not to have reference to God, lest the distinction between “objective” and “subjective” collapse.  Subjectivism in Huemer’s Pickwickian sense is no longer objectionable.

Huemer also complains that DCT “does not actually explain obligation, because no explanation has been given for why anyone should obey God’s commands.” But a theory which ultimately stops explanation somewhere may still enjoy the advantage of greater explanatory depth than its rivals. Explaining our moral obligations as arising from divine commands exhibits such explanatory depth, even if the theorist, in response to the question, “Why are we obligated to obey God’s commands?”, finally says, “Every theory must start somewhere, and my starting point is that obligations are constituted by imperatives issued by a qualified authority. As the paradigmatic Good itself, God is qualified to issue moral imperatives.”

Huemer acknowledges that we must all have some unexplained facts, but he insists that “not just anyclaimed obligation is a suitable starting point. The brute, unexplained moral axioms should be things that seem obvious when one thinks about them.” Now I see no reason to think that on atheism, the moral axioms to which Huemer appeals would be obvious. But never mind. What about the facts to which DCT appeals? Here is where Huemer’s failure to grasp adequately the concept of God comes home to roost: “But it just isn’t obvious in that way that, if there were a creator of the universe, we’d be obligated to obey that creator. So what if He made us? What if you found out that you were actually created by Satan – would that show that you are morally obligated to obey Satan’s will in all things?” 

This pièce de résistance is predicated on Huemer’s inadequate conception of God as merely the creator of the universe rather than a maximally great being. 

Huemer rejoins, “You could say that we have to obey God because God, unlike Satan, is perfectly good. But this requires there to be independent truths about goodness. But if we have facts about good and bad, then we don’t need God to explain morality; we could just directly explain morality in terms of good and bad.” But as we have seen, on DCT there are, indeed, attitude-independent truths about goodness rooted in the nature of God himself.

Moreover, why think that the DCTheorist must stop explanation in brute facts? In answer to the question, “Why are we obligated to obey God’s commands?”, the most straightforward answer is, “Because God commands us to!” At first blush such an answer might seem to launch us on an infinite regress, for now the objector will demand, “Why are we obligated to obey God’s command to obey his commands?” But there is no need to embark on such a regress. Rather DCT’s definition of moral obligation should be understood as an impredicative definition, one that includes itself. Impredicative definitions are not inherently paradoxical. For example, the set of all things mentioned in this QoW is itself mentioned in this QoW. No paradox there! So suppose God has issued to us the following command: “Obey everything I command you!” If we ask, why obey that command, the answer lies not in a further command (leading to an infinite regress) but in the command itself. The command to obey everything God has commanded us is one of the things God has commanded us to do. So if God’s commands constitute our moral obligations, we are obligated to do everything God has commanded us, including obeying his command to do everything he has commanded us.

Huemer’s objection that this last account is “obviously circular” fails to reckon with the nature of impredicative definitions. There is nothing objectionably circular about saying that we are obligated to do everything that God commands us—including that command. Huemer gravely warns that such “self-referentiality is a well-known source of paradox, so one had better be careful in appealing to it.” But self-referentiality or impredicativity is not inherently paradoxical. One has to construct the situation in a particular way in order to generate a paradox. One thinks of Russell’s illustration of the village barber who shaves everyone who does not shave himself or the set of all sets which are not members of themselves, which brought down naive set theory. Huemer is certainly correct that one can stumble unawares into paradox, but he has not demonstrated that there is any paradox in DCT as I have articulated it, nor do I see any paradox looming. 

Huemer further says that if this answer is acceptable, “then the following would also be an acceptable explanation: you’re morally obligated to obey my commands, because I command you to do so.” That all depends on who is speaking in the first person. If it is God, then the theory is acceptable; but if it is Michael Huemer, the theory is utterly implausible, since it is not the paradigm of goodness who speaks but a finite and morally imperfect being. The problem with Huemer’s parody is not self-referentiality but the nature of the source of moral value and duty.

Huemer’s final objection to DCT’s account of moral duty is that it leads to moral absurdity. For in the absence of a divine command, refusing to help a burning child would not be wrong, which is absurd. This objection trades on the rhetorical ploy of appealing to present moral intuitions with which the theist agrees and then interpreting them counterfactually. This is plain in Huemer’s personal affirmation:

The relevant counterfactuals in each of Huemer’s three cases have impossible antecedents. For God exists necessarily and is essentially loving, fair, kind, and so forth.  There are therefore no possible worlds in which God does not exist, or is indifferent to child torture, or is hateful. Intuitions about what is in fact right and wrong, however deeply held, do not tell us what would be the case under any of these impossible conditions. It may be that none of these counterfactuals has a non-trivial truth value, or, if they do have non-trivial truth values, that the counterfactuals are true in their descriptions of states that cannot possibly obtain. In answer to Huemer’s question as to “why we should believe in morality in the first place,” the answer is, as he himself affirms, that we can, indeed, trust our ethical intuitions about what is right or wrong.

In sum, these objections fail to subvert DCT’s account of the objectivity of moral values and obligations. I conclude that theism grounds the objectivity of both moral values and moral duties in an explanatorily deep way.  Hence, it is evident that if God does exist, we have a sound foundation for the objectivity of moral values and duties. For a critique of Huemer’s own moral theory, see the book!