The point was that we have reason why we don't want contradictions: we want our inferences to be correct. So logic is instrumental there. Morality on the other hand seems to be something we wish to be a certain way, for its own sake.
Modus ponens works better than appeal to inappropriate authority wrt. making correct inferences. In the case of morality, what works better than what wrt. doing what?
You could ask why we would want correct inferences. But I don't see a point in reducing further.
The analogy to morality would make sense if say we already made up our minds to 'maximize happiness for the greatest number', for then we can check that utilitarianism would do this, and hence is correct. But morality seems to be more complex than that.
I'm agreeing with Marius in thinking that ethics is about achieving some sort of reflective equilibrium. I'm just rejecting the analogy with logic there.
To summarize, with logic, we already know what we want. With morality, we don't.
lukeprog gave a list of metaethics questions here:
Most of these questions make no sense to me. I imagine that the moral intuitions in my brain come from a special black box within it, a "morality core" whose outputs I cannot easily change. (Explaining how my "morality core" ended up a certain way is a task for evo psych, not philosophy.) Or I can be more enlightened and adopt Nesov's idea that the "morality core" doesn't exist as a unified device, only as an umbrella name for all the diverse "reasons for action" that my brain can fire. Either perspective can be implemented as a computer program pretty easily, so I don't feel there's any philosophical mystery left over. All we have is factual questions about how people's "morality cores" vary in time and from person to person, how compelling their voices are, finding patterns in their outputs, etc. Can someone explain what problem metaethics is supposed to solve?