[NOTE: The foregoing is excerpted from an online class discussion for Patricia Greenspan's metaethics course at the University of Maryland. This is out of context, but I thought some of you might find it interesting. The topic is whether we should consider the task of moral philosophy to be the systematization of our internal/first-person intuitions about normativity, or an external/third person investigation into the natural bases of our moral discource and activity. In a previous message I was, not unreasonably, interpreted as defending the external task, and a deflationary metaethics, and sent this clarification to the list.]
I certainly don't wish to give the impression that I'm just ruling out the internal perspective from the beginning, or even from the middle, or at all. I think it takes a lot to get to the point where an attitude of skepticism toward the first-person perspective makes sense. At the point that it does make sense (if it ever does for a given individual), it's not really a last resort, but it's certainly a later resort, because we are right to try to preserve as many as the appearances as is possible within a good general picture of the world.
An epistemological excursion may help me convey some of what I have in mind.
Roderick Chisholm presents a nice distinction between methodism and particularism in epistemology. A methodist begins with a method or criterion for determining rational or justified belief, and then tests our commonsense beliefs against the criterion to see what lasts. The arch-methodist, Descartes, finds that nothing survives the method of doubt, save god and the self. Hume, trying to reconstruct everything according to the theory of ideas and impressions, finds that we can't really know some things that we think we know (that causal relations are necessary, for example). Reid, my particularist hero, in a proto-Moore's-hands move argues that our reasons for accepting Descartes and Hume's theoretical methods for determining justified belief are much weaker than our reasons for believing that the world exists independent of consciousness, that I really see a dog out there, and not just an internal representation of a dog, that laws of nature really are laws, and so on and so on.
So the particularist just catalogs all the things we are sure that we know, including moral propositions. Beginning from our list of certainties, we try to construct a theory that unifies and systematizes them. In the process of theorizing, we find that items on the list come in conflict. How do we adjudicate the conflicts? Well, first, we note that every item on the list doesn't get equal weight, and that we'd be perfectly willing to give up one item to save another. But our weightings (the “prior probabilities” we assign to items) can be idiosyncratic. So we try to devise a method for determining rational beliefs that will deliver most of the list, but will be useful for helping us assign weights and adjudicating conflicts in a non-idiosyncratic way. Our initial standards of reason and evidence when applied over time helps us to erect general scientific methods, which we discover to have the capability of adding new items to the list that are weighted with very very high probabilities.
So, although we didn't start out with the propositions of the bacterial theory of infection, say, we find that they come to have a fairly priveleged place on the list (to be near the center of the web of belief, in Quine's terms.) Now, if one keeps going in this way, one finds that the methods one has devised to unify and systematize the list end up pushing us to radically revise the list, perhaps assigning low probablities to certain theological and ethical propositions that had begun with very high probabilities. As a matter of fact, there is no way to get everyone to converge on a list (well, Bayesians tell us we should, but we don't–we don't update probabilities in the “right” way), because of differences in priors [and differences in techniques of updating]. And some people may choose to reject the methods rather than some items on the list (one man's modus ponens…), which may or may not be rational (I'm not sure).
The charge of “scientism” I think applies to a priori scientific methodists who just begin with the standards of science and see what survives scientific scrutiny, much as Descartes begins with the method of doubt and sees what's left. This is I think quite different from a naturalism wherein scientific standards emerge through a reflective process of attempting to unify and systematize the materials of common sense over time.
The reason I have come to be a bit skeptical of the internal perspective is that I've come to put higher weights on the work in psychology that tells us that we are very often victims of self-deception, systemic bias, and unreliable introspection than on my own (and others') first-person judgments about, say, the nature of agency and moral obligation. As a consequence, I assign a higher weight to the existence of curved space-time, something that I don't directly experience, than to the existence of supremely authoritative, rationally binding moral imperitives, something I do find some basis for in my first-person experience.
Now, I agree that it would be dogmatic and premature to simply rule out the existence of moral reasons that are normative in just the way it seems to us that they are. I think it is rational and extremely worthwhile to attempt to preserve these appearances within a theory that relates well with our best overall picture of the world. But it happens that I find the external explanatory task, that of making sense of the empirical grounds of our moral experience, to also be extremely useful and interesting. And it's possible to do both at once, I think.
One of Rawls's points in “The Independence of Moral Theory” is that metaethical questions are hard, and that throughout a very long history of inquiry, we haven't nailed down the answers, but we still need to figure out how to live with each other, so we mustn't wait on the metaethics to get to the practical ethics. Rawls seems to think we can build something useful largely out of the matter of our standing first-person moral conceptions. He does say that science comes in during the process of wide reflective equilibrium, and that the “theory of human nature” places constraints on our ideal of the moral person/citizen of the well-ordered society. However, I don't think he takes these constraints seriously enough. This is what really motivates me to follow up on the “external” project. But I really don't mean to disparage the aims of traditional metaethics. I do find that I sometimes have trouble remembering what set of questions I'm trying to answer, and so I'll slip into external/descriptive mode when the question at hand is internal/normative, and so I'll seem to be debunking when I don't really mean to be.