Consider this metaethical naturalist view:
- Every moral property is identical with a natural property.
- (Premise) For any natural property N, it is possible for there to be a being that lacks N but has M.
- Therefore, M is not identical with any natural property.
- (Premise) M is a moral property.
- Therefore, there is a moral property that is not identical with a natural property.
The controversial premise is (2). One way to (2) is this. We can imagine a non-natural agent that lacks all natural properties but nonetheless has M. We can imagine non-material beings that have no energy, no charge, that do not occupy space and time, but that, nonetheless, have moral properties like M.
But there is, I think, an interesting response to the argument. Maybe a property P can be natural in a being x but not natural in a different being x. For instance, the property of being a cause might be a natural property of a physical event, but a non-natural property of a supernatural being. Thus, one might argue that our non-natural being x that has M but "lacks all natural properties" can still have P, because P might be natural in us, but non-natural in x.
If so, then (1) should be modified to:
- Every moral property is identical with a property that is natural as found in beings like us.
- Every moral property of a being like us is identical with a property that is natural as found in beings like us.
But (7) is subject to a worry. If the argument is right, then moral properties like being obligated to do something or not being guilty of anything can only be identical with those properties natural in us that a radically supernatural moral agent could also have. Those would be complexes of very general properties like being a cause, but not of properties like having a brain. This constrains the ingredients for a naturalist metaethics to a subclass of the natural properties—namely, those that radically supernatural moral agents could have—and thereby makes naturalist metaethics harder. And the idea that naturalness is relative to the being in which the property is found is going to be at least somewhat controversial.
The same argument applies in the case of mental properties.