When Atheists wish to engage in the moral language game they often use the utilitarian standard of pain and pleasure (understood as broadly as possible) as the ontological foundation for moral values. On this approach, a clever enough secular humanist can argue her way from the ‘is’ of our moral experience understood as reduced to the elements and relation of pleasure and pain, to the ought of moral values which are universally shared by human persons precisely because of the human situation. There may be some outlying exceptions, such as the psychopath, who doesn’t share those same values, but by and large this approach to the moral language game approximates to our moral intuitions (if they aren’t sharpened at least).
However, I just realized recently why Craig has always been so careful in his moral argument to talk about objective moral values and duties. The reason is because he believes God issues commandments to us which act as moral imperatives. I used to think this language was strange, as God’s nature is the ground of goodness, and that which we experience as pleasurable or good is such only insofar as it reflects God, the end (telos) of the human soul. I used to think that having an objective grounding of moral facts (in God) was sufficient for speaking about moral duties, but of course that would be to commit the fallacy of going from an ‘is’ to an ‘ought’ of which the secular humanist is so regularly accused. Therefore, in any ‘divine command’ theory, even if the commands themselves flow directly out of God’s Nature, which is ‘the good’ itself (it is the ‘form’ of the good), there is a requirement for God to issue moral injunctions as commandments to us.
Perhaps the act of commanding us, even without taking away our free will, is itself the inevitable act of Love, and also is the only way that we can even logically ‘disobey’. Issuing commandments, then, becomes an analogy to God’s giving us a tree in the Garden of Eden – in fact, now that I think on it, the tree came with the commandment. The arena of moral decision making plausibly requires commandments against which we can make morally significant distinctions – we can infer oughts from imperatives, not indicatives.