A difficulty with non-reductive naturalism

Discuss:

The difficulty of classifying any view as non-naturalistic is, at bottom, the cost of allowing non-reductive versions of naturalism.

It seems to me that if you are going to call yourself a non-reductive naturalist, then you leave yourself open to the same challenge. Can you give us an example of a non-naturalistic ethics?

Well, the short answer is that a non-naturalistic ethics is one that is not grounded in moral observations. In section 4.2 of my thesis, I explore the possibility of moral observations. I note that if we think of an observation as a bare percept, then it seems clear that moral observations are not possible, because we only percieve facts as such, not values. However, an observation involves more than a bare percept; it also involves an interpretation of that perception that infuses it it with a point of view. This point of view can consist in an intuition or value that couples with a perception, but it could also be something as complex as a theory. For example, when we see one billiard ball striking another, we perceive the two separate events of one ball ceasing motion, and the other ball commencing motion. From this, we seem automatically to infer the notion of a cause that links the two events, but we cannot perceive the cause itself, as Hume observed. Nevertheless, as an intuition coupled with a perception, we seem legitimately to claim that we have observed the cause. In fact, a physicist who looks at a streak of light fall across the window of a cloud chamber seems also legitimately to be able to claim that he has observed a pi meson. There is an enormous body of theory that renders this observation possible, but that it is a legitimate observation seems hard to deny. Therefore, surely a legitimate moral observation could theoretically consist of a perception coupled with some kind of value judgement, or even a moral theory.

But see, there's the difficulty: You already allow that ethics cannot be reduced to anything other than ethics, and you already allow that a perception coupled with a moral theory can legitimately be called an observation. But if that is the case, then it seems to me that any moral theory at all could form the basis for a moral observation, which means that any theory at all is surely 'naturalistic'. For example, what if Kant claimed to observe an act of virtue occurring in the world? (However, he would have to qualify this observation by reiterating that one can never be sure when any act has moral worth, hence was performed out of virtue.)

The difference with a philosopher like Kant is that he is very obviously a rationalist, claiming to justify his ethics a priori. Therefore, any moral observation that he makes presupposes a theory that does not itself derive from observation. Whereas, it is possible theoretically to start off with one's moral observations, and from there derive a theory of ethics. This way, one is forcing one's theory to fit one's judgements, rather than one's judgements to fit one's theory, which seems inherently a more sound approach. In other words, a naturalistic theory of ethics starts by cataloguing our moral observations as they already exist, and inferring general principles from them, if such are to be found.