Ethical naturalism is the doctrine that moral properties, such as moral goodness, justice, rightness, wrongness, and the like, are among the “natural” properties that things can have. It is the doctrine that moral properties are “natural” and that morality is in this sense an aspect of “nature.” Accordingly, it is a view about the semantics and metaphysics of moral discourse. For example, a utilitarian naturalist might propose that wrongness is the property an action could have of being such as to undermine overall happiness, where happiness is taken to be a psychological property. Unfortunately, it is unclear what the naturalist means by a “natural” property. For my purposes in this paper, I shall assume that natural properties are such that our knowledge of them is fundamentally empirical, grounded in observation. More precisely, a property is “natural” just in case any synthetic proposition about its instantiation can be known only a posteriori, or with the aid of experience.

The provenance of “naturalized epistemology,” so called, is too recent for the hand of Quine not to be still heavily upon it. But like its older relative, “naturalism,” it is an idea rich enough to be coveted, and protean enough to be claimed, by diverse comers with different things in mind. While Quine's version of naturalized epistemology of science inevitably furnishes the backdrop for current discussion of naturalizing moral epistemology, it is important to pause over what “naturalized epistemology” can and should mean in ethics. To what extent is Quine's example of an epistemology of science that helps itself to science the model for understanding knowledge of and in morality? Does it require a view of moral knowledge as reducible to, or in a fundamental way furnished by, science? Or a view of moral theory as sciencelike in some way? I argue that the appropriate analogy is instead a holistic and reflexive epistemology of morality that helps itself to moral judgments and standards seen as answerable to the experience of the kinds of shared lives they make possible and necessary. This approach neither privileges nor rejects wholesale what scientific inquiries might have to say.

Naturalized epistemology, as I understand it, is the practice of treating knowledge – human or otherwise – as a natural phenomenon, susceptible of investigation by the methods of empirical science. A naturalized approach to the study of knowledge differs saliently from more traditional forms of epistemology in taking the existence of knowledge for granted. Naturalized epistemologists do not concern themselves with skeptical challenges. Nor are naturalized epistemologists much concerned with questions about what counts as “knowledge,” properly speaking. They do not worry if a bird’s natively specified program for star-based navigation is “justified” for the bird, nor if the sub-personal data structures and algorithms posited by cognitive psychologists can be properly counted as “beliefs.” The naturalized epistemologist is interested in the explanation of anything that even appears to be a cognitive achievement, whether or not it passes muster as “knowledge” in some preferred sense.

Moral naturalism, as I use the term here, is the view that there are moral facts in the natural world – facts that are both natural and normative – and that moral claims are true or false in virtue of their corresponding or not to these natural facts. Moral naturalists argue that, since moral claims are about natural facts, we can establish the truth about moral claims through empirical investigation. Moral knowledge, on this view, is a form of empirical knowledge.

One objection to this metaethical view is that even if moral naturalists are correct in their claims about truth, they cannot answer the question of normativity. Jean Hampton, for instance, argues that it is not enough to explain the conduct's wrongness by showing it to be a property that necessarily supervenes on natural properties. For nothing in this analysis explains the relationship between these properties and us. The question is why should people care about these properties. Christine Korsgaard claims that moral realists take the normative question to be one about truth and knowledge.

The phrase “statements of fact” has a clear, unequivocal ring. It speaks of a stable place untouchable by contests in epistemology and in more secular places, around questions of constructivism, subjectivism, and the politics of knowledge. It offers fixity, a locus of constancy in a shifting landscape where traditional certainties have ceased to hold, maintains a vantage point outside the fray, where knowledge-seekers can continue to believe in some degree of “correspondence” between items of knowledge and events in the world. Within the social institutions and practices where knowledge is an issue, it designates a secure starting place for deliberation, a way of ensuring that processes of decision-making remain cognizant of the “realities” they have to address.

In the institutions of knowledge production and knowledgeable practice that generate the examples I appeal to in this essay – medicine and law – “statements of fact” appear to comprise an incontestable core around which interpretive strategies may indeed have to be enlisted, but which itself functions as a kind of interpretation-exempt zone. There are good reasons for its retaining that status, for resisting the instabilities and sheer whimsicalities that would follow from dislodging it.

If moral epistemology can be naturalized, there must be genuine moral knowledge, knowledge of what it is morally right for someone or even everyone to do in a particular situation. The naturalist hopes to explain how such knowledge can be acquired by ordinary empirical means, without appealing to a special realm of moral facts separate from the rest of nature, and a special faculty equipped to detect them. Various learning mechanisms for acquiring moral knowledge have been proposed. Most, however, have the following deficiency: What they actually explain is moral acculturation with respect to accepted or author-preferred moral norms, not the acquisition of moral knowledge. Of course, an additional premise to the effect that accepted moral norms or author-preferred norms embody moral truths would deal nicely with this problem, but at the expense of the distinction between opinion and knowledge, or true belief, in which epistemologists are necessarily interested.

The error theory of moral judgment says that moral judgments, though often believed to be objectively true, never are. The tendency to believe in the objectivity of our moral beliefs, like the beliefs themselves, is rooted in objective features of human psychology, and not in objective features of the natural world that might exist apart from human psychology. In naturalized epistemology, it is tempting to take this view as the default hypothesis. It appears to make the fewest assumptions in accounting for the fact that humans not only make moral judgments, but believe them to be, at least some of the time, objectively true. In this paper I argue that from an evolutionary perspective, the error theory is not the most parsimonious alternative. It is simpler to suppose that mental representations with moral content arose as direct cognitive and motivational responses to independent moral facts.

Recent work in cognitive science highlights the importance of exemplar- based know-how in supporting human expertise. Influenced by this model, certain accounts of moral knowledge now stress exemplarbased, non-sentential know-how at the expense of rule-and-principle based accounts. I shall argue, however, that moral thought and reason cannot be understood by reference to either of these roles alone. Moral cognition — like other forms of ‘advanced’ cognition — depends crucially on the subtle interplay and interaction of multiple factors and forces and especially (or so I argue) between the use of linguistic tools and formulations and more biologically basic forms of thought and reason.

Professor Clark's splendid essay represents a step forward from which there should be no retreat. Our de facto moral cognition involves a complex and evolving interplay between, on the one hand, the non discursive cognitive mechanisms of the biological brain, and, on the other, the often highly discursive extra-personal “scaffolding” that structures the social world in which our brains are normally situated, a world that has been, to a large extent, created by our own moral and political activity. That interplay extends the reach and elevates the quality of the original nondiscursive cognition, and thus any adequate account of moral cognition must address both of these contributing dimensions. An account that focuses only on brain mechanisms will be missing something vital.