Note: I was working on an addition to my “About” page wherein I attempted to briefly state my meta-ethical views when I noticed that I had gotten carried away and written far more than was appropriate for the page. Thus I have decided to share it as a post.
Among metaethical options, I think that non-cognitivism has no merit due to the available evidence, and constructivism/subjectivism is impossible if realism isn’t true. Thus, as I see it, moral judgments refer to independently existing moral properties, but the question is whether those properties truly exist. If they do, then realism is true. If they do not, then the error theory is true.
But even if realism is true, the search for answers does not end there, as there are two very viable options: naturalism and non-naturalism. Much of my research has been on how best to draw this distinction, in order to adequately test the merits of both. Distinguishing the two views has been made difficult by Cornell realists arguing that moral properties are natural but irreducible properties; before the Cornell school only non-naturalists thought moral properties were irreducible, due to their being sui generis. A common distinction made is that naturalists think that synthetic property identities between natural and supervening moral properties will be discovered through empirical researches, while the non-naturalist thinks that these synthetic property identities are impossible, and we have knowledge of how moral properties obtain through moral intuitions, which are fleeting glimpses of a priori moral knowledge.
The way I like to think of this difference is that naturalists think that moral properties are not directly observable, but are confirmed by coherent theories, whereas non-naturalists think that moral properties are observable through a moral sense. Given this controversial reading, it might seem surprising that I lean towards non-naturalism.
My thinking is that naturalism seems to entail that we do not have a real grasp on how morality works because we do not have the correct synthetic property identity at present, and that it might turn out when we uncover that identity relation that we have been going about morality all wrong. For naturalism, we are waiting on empirical data to vindicate or disprove what we think we know about morality. That is a pretty wild claim considering that it seems to me that I have a pretty good grasp on what is good and wrong about actions even if I do not know the synthetic property identity between moral goodness and some natural descriptive property. This knowledge seems to speak in favor of the direct moral perception that non-naturalism allows and that naturalism denies. Thus, that I seem to know about good and bad because I seem to have an immediate sense that a priori goodness supervenes on this set of natural properties, has a lot more explanatory power about how I experience the moral richness of the world than naturalism.
This is not to say that the empirical methods proposed by naturalists are not without their explanatory benefits. That the correct synthetic identity relations between moral and non-moral properties have yet to be discovered could explain why there is such widespread moral disagreement on planet Earth. The argument from disagreement is one of the most commonly cited reasons to be skeptical of realism; I am not however impressed by it. The naturalist has an open response with their insistence on synthetic property identities that have yet to be discovered.
The non-naturalist can also contend that moral intuitions are not infallible, and that even claims that are a priori true can take a while to see as self-evidently true. Much work has been done exposing how our intellectual intuitions (of which moral intuitions are a sub-group) of how to interpret data are smeared and obscured by biases; the same failures of reasoning might hold true of moral intuitions. In fact, it seems that given moral claims are action-guiding, that we can expect more biases than in other areas of life, such an scientific research. If I have a desire to stay as the dominating figure in my family, then this patriarchal desire might bias how I understand how I am to treat women in the greater society.
This gets at another fair response to the argument from disagreement: that often times what we disagree on are the non-moral facts and not the moral facts, such that there is not an irresolvable moral conflict, but a dispute on the meaning of terms. Consider the debate on abortion. Both sides think life is morally valuable, but the sides disagree on when life begins. This is a disagreement about the meaning of a non-moral term, “life”, such that it is possible that if the meaning of the term was agreed upon one side would concede to the other. Unfortunately, our beliefs are normally supported by an interconnected web, even when those beliefs have no true logical connection but really are just connected due to committed associations, and it makes it difficult to give up one belief when we think it is supported by another. This is just another form of bias.
So the argument from disagreement is no harder for the non-naturalist to respond to than the naturalist.