I have a piece over at the ABC religion and ethics portal, in which I discuss the process of evaluating things - hammers, cars, novels, tennis players, friends, fellow citizens, or people to have around - as a context for talking about moral evaluations.
As I've argued on this blog in the past, there's nothing spooky about evaluating things as "good" and "bad", although the concept is quite tricky once examined, and it may be that even ordinary judgments of good and bad have a (largely innocuous) fictional element: we work on the assumption that the people we're talking to have shared requirements of things, when even with the simplest things our requirements are not totally shared.
From a very strict point of view, there is no "we". From a practical point of view, however, our requirements as individuals are often similar enough that there's nothing much wrong with saying, "My Honda is a good car" or even "Abigail is a good person." For all intents and purposes, such statements, spoken in a context, can be simply true.
In very many actual contexts "we" can say without being too misleading that something is a good example of its kind - i.e. it has the characteristics that are needed for it to meet "our" requirements for things of that kind.
Furthermore, our requirements won't be arbitrary - there's nothing arbitrary about what we usually want from hammers or cars or friends or even fellow citizens. We can have rich and quite rational conversations about this - and such conversations produce convergence and agreement in many, many cases.
With just that bit of slippage, creatures like us have a wealth of reasons to want hammers to be sturdy, cars to be reliable and fuel-efficient, friends to be loyal, and fellow citizens to be industrious and law-abiding (at least with respect to most laws, and we have a wealth of reasons for identifying some laws as more important for this purpose as others). There isn't even anything especially mysterious about our judgments of people's characters or of their choices in respect of actions that could affect others. Non-arbitrary, rationally defensible evaluations can be made of the kinds of things that we think of as falling in the "moral" domain.
As long as morality doesn't purport to be more than this, we can demystify it without debunking it.
What this won't get us, however, is a strictly objective morality. Whatever judgments we make do not compel all comers, regardless of their desire-sets, to act one way or another on pain of making a mistake about the world or something of the sort. Thus, when we think that moral judgments are special in this way - different from all other evaluations in being strictly objective - we make an error. To whatever extent that error is built into some of our moral language, that part of our moral language is systematically false (I've come to think that this won't be all or nothing, and I don't believe that John Mackie thought it was, either).
I don't go into the latter ideas in the ABC article. I.e., it doesn't explicitly address moral error theory. It does, however, show how evaluations in what we think of as the moral domain (the realm of character and choices) can operate much like our other evaluations - neither strictly objective nor merely arbitrary. The problem is that, very often, we really don't think that's good enough. We (or many of us, much of the time) want moral judgments to have more metaphysical grunt, and we don't want to be stuck with the kind of ambiguity and indeterminacy that pervades our evaluations of other things. Or so it seems to me whenever I discuss this with people out there in the forreal world.