The Open Question Argument

(For those regular readers, I apologize that this post is a day late. Unfortunately, a cold waylaid me this past couple of days.)

In this post, I want to discuss G. E. Moore's famous open question argument. As I understand it, the open question argument allegedly creates a problem for certain reductive theories of genuinely normative properties (such as the goodness or ought-to-be-doneness of an action) to certain descriptive or non-normative properties (such as caused pain-minimization). The idea is that we have principled reasons for ruling these theories out because it is always an "open question" whether, say, the goodness of an action can be identified with the pain-minimization that the action caused.

(I should say that I think that the open question argument can be formulated in such a way as to not only theories that identify genuinely normative properties with some descriptive or non-normative properties, but also theories that say that, on a particular possible occasion, the having of these genuinely normative properties will be nothing more than the having of some or other descriptive or non-normative properties, in the sense that the latter will explain, without remainder, why the former is instanced. To formulate the "open question" argument in this way, one suggests that it is always an an "open question" whether, say, the goodness of particular action on some occasion can be constitutively explained by the pain-minimization that the action caused even in light of what other descriptive or non-normative properties the action had on that occasion.)

The open question argument seems related to Hume's suggestion that one cannot "derive" an 'is' (that is a descriptive fact) from an 'ought' (that is a normative fact) or vice-versa. Critics of the open question argument might accuse proponents of begging the question on this front. Perhaps some of these questions are not open, but only seem so because we are not fully rational enough to grasp how the question is closed (because our grasp on genuinely normative properties is a bit tenuous). The idea is that there is no further information (gained, for instance, via ethical intuition) that we need to see that the question is closed; we simply need to do better at processing the information that we have. Call this "the first response." (A second response would be to claim that although the question is open, still some sort of reduction is possible.)

I think that the first response fails, and I want to give some reasons why before opening up the discussion to others. The reason that I think that the first response fails is that it simply ignores the motivational aspect of the genuinely normative concepts that pick out these genuinely normative properties. Note: The word 'genuinely' here is supposed to make it clear that we are talking about just those concepts (or just those properties that these concepts pick out) that have this motivational aspect, and not any others that don't.

The motivational aspect that I have in mind is just this: The use of a genuinely normative concept incurs certain rational commitments when it comes to action. For instance, if I use this genuinely normative concept to judge that I ought (all-things-considered) to be nice to my elders, then if there is something rationally defective about me if I intentionally go on to act in a not-nice-way to my elders. (Basically, I am not functioning properly in this kind of scenario.) To avoid this rational defectiveness I need to either relinquish my judgment or stop failing to be nice to my elders.

The motivational aspect of genuinely normative concepts seems to suggest that these concepts already incorporate one way of picking out genuinely normative properties. Genuinely normative properties just are those properties that make it irrational to judge one way or act contrarily (as described with me and my elders).

My reasoning, then, is that given that we already have one way of picking out genuinely normative properties (as the properties that account for why it is rational to act in certain ways), then it must be an open question whether we can also pick them out in some other way by certain descriptions that have nothing to do with motivation and proper functioning.

Questions: Is this reasoning right? If so, what is the upshot for naturalism about genuinely normative properties?