The third way of establishing a dependence relation of morality on God is by means of the argument from justification. What Korsgaard calls the “normative question” is the issue here: Why should I be moral? Or, why should I accept morality as a proper demand on me? By the second Critique, Kant didn’t think he could give a justification of the moral demand. Earlier, in Groundwork of the Metaphysics of Morals, he attempted a proof of freedom. But he seemed to come to think such efforts suffered from circularity. Hare thinks the argument from justification is weaker than the first two arguments because it would take a lot of work to show the DCT answer to the normative question is the only or best answer. (Hare does some of that work in chapters 6-9 of Why Bother Being Good? as does C. Stephen Evans in ch. 4 of God and Moral Obligations.)

A DCT answer to the normative question is that one should accept morality as a proper demand because it’s God who places this demand. A justification of a normative claim can’t be derived from a factual claim alone. To say that I ought to live a certain way because God tells me to do so requires, for completeness, the claim that I ought to do what God tells me to do. This has led some to accuse DCT of begging the question. Why should we do what God tells us to do?

If obedience to God is itself a moral obligation, then to justify moral obligation by appealing to it is viciously circular, but if it’s not a moral obligation, then it seems no justification is available by this route. If we consider Socrates’ attempt to show Crito there’s an obligation to obey the laws of the city, all of his various arguments have the form of deriving a justification of a claim that we have an obligation of a certain kind by deriving it from a higher obligation. But then the project of justifying our moral obligations as a whole seems hopeless.

Hare thinks there’s a reply to this difficulty, though. He uses the distinction Scotus draws between natural law strictly speaking and natural law in an extended sense. Scotus thinks the command to love God given in the first table of the Ten Commandments is natural law strictly speaking, known to be true just by knowing its terms. But he thinks the second table, which concerns our various duties to the neighbor, is natural law only in an extended sense. It’s contingently true, whereas it’s necessarily true that God is to be loved, because we know that, if God exists, God is supremely good, and we know that what is supremely good is to be loved. It’s also true we know that to love God is at least to obey God. There’s scriptural warrant for this, but in addition, it’s plausible to say that to love God is at least to will what God wills for us to will. God is not “another self,” as there’s an obvious disproportion between us and God, and there are some things God wills that God does not will for us to will. So what we repeat in our wills is God’s will for our willing. But willing what God wills for our willing is obedience. So it is necessarily true not just that God is to be loved, but that God is to be obeyed. The appeal to God’s command as authoritative thus doesn’t require justification (except in so far as we have to justify the claim that God exists), so this means that divine commands do not generate all our obligations, because there is one important exception, namely, the very obligation to obey divine commands. But this is not a troubling exception once one accepts the necessary truths (if God exists) that God is to be loved and that God is to be obeyed.

Now, what is the relation between the divine command and the obligation? Quinn offers a causal account, whereas Adams offers a constitution account. Hare thinks a causal account is wrong because it suggests two different events, in logic separable. And constitution seems wrong because obligation is not a natural kind like water.

Hare instead notes the way a king makes a law by declaring: “The king wills it.” This is an “explicit performative,” in Austin’s terminology, like “I promise.” Searle claimed that the case of promising shows that it’s possible to derive an ‘ought’ from an ‘is’. But we don’t have a case in promising of a prescription following from a description without any intervening prescriptive premises. There’s an endorsement of the institution of promising, which is itself a kind of prescription.

Now consider the case ‘God commanded Jones to pay Smith five dollars’. If DCT is true, it follows that Jones ought to pay Smith five dollars. Is this deriving an ‘ought’ from an ‘is’? No, because the truth that God is to be obeyed plays the same role in the argument as the institution of promising did in the earlier one. If I affirm that God is to be obeyed, I am prescribing obedience, and so prescribing that, if God commands Jones, Jones ought to do what God commands. Different from causation or constitution, it’s better to see the relation as resembling that of performative acts like a priest’s baptizing, or a president’s signing a bill. In all these cases, there is an internal conceptual connection between the ‘producing’ and what is ‘produced’, but making the inference requires the endorsement of an institution or a necessary truth.

There’s a reply here to an objection to DCT by Wolterstorff, who denies that all moral obligations are generated by God’s commands, since some are generated by human commands, such a parent’s command to a child to clean up his room. Hare thinks Adams’ answer that these are merely pre-moral obligations is wrongheaded. Hare instead agrees that sometimes human commands generate obligations, but he says we should see DCT as operating in answer to the normative question why we should hold ourselves under those obligations. For why should we keep our promises? To draw the implication from my having said ‘I promise’ to my obligation, I need to endorse the institution of promising, and the fact that God requires this faithfulness of me gives me a reason for this endorsement. [My question: who’s drawing the implication here? Me? If I don’t happen to endorse the institution of promising, that doesn’t vitiate the obligation for me, just because I don’t see myself as having an obligation.]

Such a view answers the normative question by referring back to something that is good in itself. More now needs to be said about goodness. Relying on Stump’s take on Aquinas, ‘being’ and ‘goodness’ are the same in reference and differ in sense, and the sense of the good is that what is good is desirable. Goodness supervenes on the natural property of the actualization of a specifying potentiality. This means that a human being is good to the extent that the human is rational (if we take rationality to be the specifying potentiality of human beings, that is, the potentiality that makes something belong to the species ‘human’). Hare makes a few qualifiers at this point: That’s probably not a good way to specify what makes something human, because it leaves too many people out, and also probably goodness is not to be confined in its reference to degrees of natural-kind actualization. Pleasure is good, for example, and so is aesthetic beauty, but neither of these forms of goodness is amenable to this kind of analysis.

Aquinas’s account of the sense of ‘good’ is separable from this account of its reference, though. ‘Desirable’ is a term of notorious difficulty: able to be desire, tends to be desired, or worthy to be desired? From the fact that all people desire something, it does not follow that it is desirable in that last sense. On the other hand, to say that something is good is not simply to say that it is worthy to be desired or loved, but also to express one’s desire or love for it. There is something odd about the combination of saying sincerely of a thing that it’s good and being indifferent to it.

At any rate, the good is the desirable in the following sense: If I say something is good, I express the fact that I desire or love it, and I claim that it merits such desire or love. To say sincerely that something is good is to express that one is drawn to it, and to endorse the claim that it deserves to draw one in that way. This offers an account of the supervenience of goodness on being: To say something, like a strawberry, is good is to say it’s good because of its natural properties, but this description is not entailed by the evaluation that the strawberry is good, without the prior endorsement of a set of descriptive criteria. Hare admits Stump’s is the more plausibly Thomist account, but her view seems to Hare too narrow an account of the criteria for goodness as a whole.

Now we can return to the divine command answer to the normative question, that the demand of morality is a proper demand on me because it is God who makes the demand. When I say that something is good, I am expressing my desire or love for it and I am claiming that it is worthy to be desired or loved. This account escapes Moore’s objection (in the ‘open question’ argument) to accounts of evaluative terms that attempt to reduce them to descriptions. The theist claims that what is worthy and deserves our love is, supremely, God, and in saying God is good she expresses that love. She may add, secondarily, what is worthy of love is her own love of God, and others’ love of God. Scotus suggests that loving God is itself not only good but our destination as final end. Thirdly, the theist may say that what promotes the love of God or draws us toward it is good. These references are not part of the meaning of ‘good’, but are claims about what satisfies the criteria for worthiness to be desired or loved.

On this account, there can be two seemingly opposite priority relations between what’s obligatory and what’s good. On the one hand, the good has priority over the obligatory, because the justification relation is as just explained: I should try to meet my moral obligations because God gives them to me, and I prescribe that God is to be obeyed because God is to be loved as the supreme good. (Also, the commands God gives are to do good things, and their goodness gives me a reason for doing them. [A thought: when God tells us to do the lesser of two evils, the goodness of the action doesn’t apply, but the first condition still would, so perhaps there’s an advantage to Hare giving pride of place to the first condition.]) On the other hand, the obligatory has priority over the good, because there is an enormous number of good things possible for me to do, and God, in prescribing some obligation, selects some of these goods and neglects others. Only the ones God selects for prescription are obligatory. The two priority relations are the opposite way round, but there is nothing contradictory in this, because we have two different kinds of priority. The first kind of priority is what Aristotle calls priority ‘in account’, and the second is something more like ‘veto’ or ‘overridingness’. The good has priority to the right because everything that is right is good, though not vice versa. [Note: not sure about that: again, what of the duty to do the lesser of two evils?] The right has priority to the good because the goods that God selects as mandatory for us are, so to speak, trumps.

This helps us answer the claim that DCT makes morality arbitrary (see note below). This is often traced to the Euthyphro, but Hare has argued elsewhere that the dialogue is weak support for the opponent of DCT, because the opponent needs an explicit argument that the gods love the holy because it is holy, and the dialogue does not in fact provide one. But Socrates speaks a truth here, Hare thinks, but one consistent with DCT. We know that God’s commands are not arbitrary, because we know that what God commands is good, and the goodness is not produced by the command. This does not, however, make God’s command redundant, because only those good things that are commanded are obligatory.

*Note: The term ‘arbitrary’ has different senses. It can have a negative evaluative force, meaning that a decision ignores some relevant consideration, and this is how the term is used in the objection from arbitrariness to DT. But it can also mean that a decision is within the agent’s arbitrium or discretion, and in that sense Hare wants to agree that God’s commands are arbitrary.

Image: By Phillip Medhurst - Own work, CC BY-SA 3.0, https://commons.wikimedia.org/w/index.php?curid=20116195