These methodological points bear on Wilhelm’s first challenge to my metaethical conclusions. In the target paper I argue for a form of internalism (roughly, the view that moral judgments are essentially motivating). Wilhelm points out that my evidence for this claim will not satisfy many externalists. I primarily rely on evidence that moral judgments always co-occur with emotional states, but few externalists will be impressed; they will say that such findings cannot address questions about whether it is necessarily the case that moral judgments are motivating, even if they always happen to be motivating.
Wilhelm helpfully replies to this objection on my behalf, using Jon Tresan’s distinction between de dicto and de re internalism. The former is a thesis about the concept of moral judgment (viz., it is a conceptual truth that when that concept applies, motivation applies as well). The latter is a claim about moral judgments themselves (viz., moral judgments do in fact carry motivational force). Wilhelm concurs that my evidence can contribute to a defense of the de re claim. He suggests that I abandon the case for de dicto internalism, since naturalists should not concern themselves with conceptual claims.
I welcome Wilhelm’s suggestion, and I am inclined to endorse it. Let me mention, however, a strategy available to the naturalist whose heart is set on defending the de dicto claim. Returning to Wilhelm’s discussion of methodology, let’s imagine that naturalists wage a successful campaign against the a priori. Properly pursued, such a campaign might also undermine metaphysical necessity. Metaphysical necessities, unlike nomological necessities, are alleged to be true in virtue of conceptual entailments rather than laws of nature or natural facts. The critique of a prioricity threatens metaphysical necessity because it advances the view that truths about concepts are open to empirical revision. Let’s suppose that concepts are mental representations garnered through experience with the function of classifying things in the world. So construed, concepts are susceptible to improvement through empirical inquiry. Initial concepts are rough and ready pointers that we use to carve up the observational world, and revised concepts are carvings that remain after observation. Now let us define a “robust conceptual truth” as the conceptual entailments that survive after a concept has been subjected to empirical fine-tuning. Such truths would more or less coincide with how the world is, together with certain pragmatic assumptions that go into theory construction. Thus, they would coincide with truths that emerge from our study of the things themselves (which are also constrained by pragmatic assumptions). On this picture, de dicto collapses into de re. Rather than bypassing de re internalism, we can try to defend it by naturalizing conceptual truth.
Wilhelm might reply that this defense of de dicto internalism would not persuade non-naturalists. The defense is based on the assumption that the naturalist critique of a prioricity goes through, but that is just what non-naturalists are inclined to deny. Thus, it might appear that the debate over the de dicto position is hostage to unresolvable disputes about the nature of philosophy.
Here I’d balk at the claim that such disputes are unresolvable. Those who believe in a prioricity may dislike naturalism, but they certainly believe that their views require evidential support. Naturalists offer an account of what concepts are (mental representations) and an explanation of conceptual intuitions (introspection of mental representations). Non-naturalists are obliged to provide an alternative account of both, and the two accounts can then be compared by agreed upon standards. I venture that the naturalist account will find a resounding victory in such a head-to-head match. It is a more parsimonious view, since both sides must grant the existence of mental representations, and I suspect it can fully account for our conceptual intuitions.
These are, of course, big debates, which I cannot settle here. My point is simply that we can imagine a two-stage process that begins with broad issues about naturalism, and then moves on to first-order views. On my prognosis, we won’t end up abandoning the notion of conceptual truth, but rather revising it. If so, de dicto naturalism might turn out true. Wilhelm may be right, however, that until we come to greater consensus on the nature of philosophy, naturalists might be on firmer ground if they try to bypass conceptual questions. He is also right that, from a naturalist perspective, de re internalism may be the more interesting thesis. Conceptual claims lose their distinctive interest if concepts are revisable and, ultimately, coincident with empirical theories.