In responding to these criticisms, many proponents of divine command theory "bite the bullet", agreeing with the point the critic is making but arguingthat it is not a problem with the theory. For example, writers like William of Ockham argue that if God had commanded murder, then murder would indeed have been morally obligatory. Indeed, Ockham goes so far as to say that God could change the moral order at any time. Thus Ockham embraces divine command theory wholeheartedly; his view has been characterized as being that "God's command is good" is analytically true. He can be thought of as saying: "God could have commanded us to commit murder, and then it would have been obligatory - but he didn't, so it isn't ... so what's the problem?" It is also possible to bite the bullet regarding the naturalistic fallacy by arguing that defining morality in non-moral terms is not a fallacy at all. Other writers disagree more directly with these criticisms. Duns Scotus is responsible for one approach that has been influential in modern times. He argues that, for one set of moral values at least, God could not have commanded otherwise because they are necessary (omnipotence, of course, means being able to do anything, but the logically impossible is essentially nonsensical, and not part of anything). However, this would mean that necessity, not God, is the source of objective morality. God is reduced to a passer-on of moral laws.
Some moral values, on the other hand, depend on particular decisions of God, and so he could have commanded otherwise. Thus, for example, that murder is wrong is a truth, and though God commanded us not to murder He couldn't have done otherwise, nor can he revoke his command; keeping the Sabbath day holy, on the other hand, is only contingently wrong, and god could have commanded otherwise, and could revoke his command. This is similar to a more recent approach developed by Richard Swinburne. In developing what he calls a Modified Divine Command Theory, R.M. Adams distinguishes between two meanings of ethical terms like "right" and "wrong": the meaning that atheists can grasp (which in fact Adams explains in roughly emotivist terms), and the meaning that has its place in religious discourse (that is, commanded or forbidden by god). Because god is benevolent, the two meanings coincide; God is, however, free to command other than he has done, and if he had chosen to command, for example, that murder was morally right, then the two meanings would break apart. In that case, even the religious believer would be forced to accept that it was correct to say both that murder was wrong and that God commanded us to commit murder.
Thomas Aquinas claimed that God creates moral norms that reflect his own essence, meaning that his demands are not arbitrary. In this case, it would become necessary to examine the essence of God. There have been two prominent responses to the problem of knowing God's commands. Some writers have argued that the metaethical divine-command theory leads to a normative theory which gives the required moral guidance; that is, God's command gives us the definition of "good" and "bad", but does so by providing practical criteria for making moral decisions. For example, John Gay argued that God had commanded us to promote human happiness, thus marrying divine command theory with a version of utilitarianism. Another response to the epistemological problem was made by the utilitarian Jeremy Bentham. He wrote, "We may be perfectly sure, indeed, that whatever is right is conformable to the will of God: but so far is that from answering the purpose of showing us what is right, that it is necessary to know first whether a thing is right, in order to know from thence whether it be conformable to the will of God." In other words, because God's will and what is right are identical, if we find out what is moral we necessarily discover what God's will is. However, once a human has the practical criteria for determining what is moral, this removes the need for a God to dictate morality through divine command.