Think about statements (1)-(3)....
(1) Bob seems tall to me.
(2) Bob is a tall guy.
(3) Bob is 6'5.
....and compare them to statement (4):
(4) Murder is wrong.
Of (1)-(3), which one is most like (4)?
If your answer is (1), then you're endorsing a sort of extreme moral relativism whereby the only properties tracked by moral statements are individual preferences. If I think abortion is OK, and you think that it's wrong, we aren't really disagreeing. It's just wrong for you, but not for me. (As Kang says in the classic Simpsons Halloween segment, "abortions for some and miniature American flags for others!")
If your answer is (2), then we've moved to a more moderate and interesting form of relativism. It is, after all, possible to be mistaken about tallness and to be shown your mistake with evidence and arguments. (E.g. "Wow, that guy's tall!" "No, he isn't, man. He just looks tall because of the angle that photo was taken at. Plus, look at his shoes...") Still, exactly what (or, more to the point, "vaguely what") "tall" means surely varies depending on the time, place and context of utterance. Tall-in-12th-century-China and tall-in-21st-century-America are not the same.
The view that moral statements like (4) should be thought of on the model of (3)--combined with the claim that at least some such statements are true is, of course, objectivist moral realism. This is an attractive view for a variety of reasons, but, at least from a naturalistic point of view, it's often hard to make sense of. The easiest, most natural way of cashing out objectivist moral realism--that moral facts exist eternally outside of space and time--is immediately confronted with obvious epistemic difficulties, more or less equivalent to Benacerraf-style worries about numbers-as-abtract-objects. If there were such non-naturalistic moral facts, how would we ever know about them? What reason would we ever have to think that our moral intuitions, refined by some sort of process of reflective equilibrium, even roughly, non-coincidentally tracked them?
One sort of story one could tell to try to reconcile objectivist moral realism with some kind of respectably naturalistic metaphysical and epistemic framework would be to identify moral properties with whatever cluster of properties "out in the world" happen to provoke the right sort of intuitive responses in people under the right circumstances. When cashing out this sort of meta-ethical stories, it's fashionable to talk a lot about evolution, and to invoke analogies to Chomskyan "language engines." This provides at least a rough model, and lets the whole enterprise bask in the reflected glory of more scientifically rigorous disciplines.
Now, at this point, the obvious line of skeptical reply involves questioning the evidence, unfavorably comparing the rich body of rigorous empirical research backing up Chomskyan ideas about language with the paperclips and chewing gum that tend to be used to hold together any remotely plausible-sounding meta-ethical story about a "moral engine." The next interesting question is whether the end result of that line of attack is that the naturalistic objectivist moral realist needs to work harder, and that for now we should all be cautious agnostics about the whole thing, or that the "moral engine" research program is dead in the water....but I don't want to get into that now.
Instead, let's just assume for the sake of argument that the "moral engine" concept is right on the money. Evolution programs us to have certain instinctive reactions to certain things in reliable ways (perhaps excepting sociopaths and other genetic abnormalities, people whose cultural brainwashing has overwhelmed and overridden their instinctual moral reactions, etc.), and that's what our moral language tracks. Just to finish off the picture, we can throw in some Kripkean stuff about "rigidification" here to explain just how concepts like "wrongness" relates to various items of external reality. (Now, our view is basking in the combined reflective glow of Chomsky, Darwin and Kripke. It's unstoppable!) So far, so good.
Now, this means we can be mistaken about moral claims not just because we make logical mistakes in moral arguments, but because the core intuitions that lie at the justificatory base of the whole structure are flawed in some way (genetic abnormality, cultural brainewashing, etc.) One can argue flawlessly, be aware of every relevant counter-example, etc., have thought of everything, but can have intuitions that differ from the ones that fix our common moral language and thus be morally mistaken.
This is an important point, because without this, the whole project collapses into a kind of relativism.
Now, surely moral language existed two thousand years ago. (I.e. post-Plato, post-Aristotle, just before the rise of Christianity, etc.) That said, a lot of judgments of right and wrong that seem clear-cut now were nowhere to be seen then.
Slavery existed, but hardly anyone (if anyone) thought that slavery was universally morally wrong. Different people had different views about men, women and gender, but if anyone thought that men and women should have completely equal rights, then not a lot of them were writing this down. (The notion of "rights" itself doesn't seem to track much of anything in e.g. ancient Greek moral philosophy.)
Etc., etc., etc.
Now, all of this is plausibly explained by empirical progress in terms of common knowledge of relevant non-moral facts, by cultural blinders and prejudices that have been overcome, by bad moral reasoning that's been corrected, etc., and all of that's fair enough. Imagine, however, the following (admittedly silly and fantastical) discovery:
Some substance that was, up until a couple of thousand years ago, trapped deep beneath the surface of the earth, slowly started to leak upward at that time until it eventually made it into the world's supply of drinking water, and we've all been drinking it ever since. This substance causes a specific sort of brain disease that doesn't in any way impair or impede people's power of *reasoning* but does have one major consequence, but does slightly alter people's moral intuitions in a specific way: it makes them identify more with the plight of those who are different from them in various respects, in a way that makes a process of reflective equilibrium more likely to end up with them being sympathetic to the plight of slaves, women, religious minorities, etc.
Thus, given the sort of naturalistic objectivist view sketched out above, we should all (upon the discovery of this substance and its effect on our moral intuitions) realize that we were mistaken in coming to the conclusion that, for example, "slavery is morally wrong." Not because of any logical flaw in our reasoning, not because we weren't clever or imaginative enough in formulating examples and counter-examples and considerations, but because our deepest moral intuitions were defective, because they differed from the ones that the genetics would have predisposed us to if all else had been equal, the ones involved in the rigidification of moral language. Remember, earlier, our recognition of the possibility of these sorts of mistakes was essential. Without it, the view collapses into relativism.