My earlier post outlined the varieties of semi-subjective reasons that can be derived by presupposing objective reasons. But I must now address the prior question: what are objective reasons? What sorts of facts "count in favour" of performing one action rather than another?
Now, I don't think it makes any sense to claim that "oughts" or "reasons" are real free-floating entities that exist apart from the natural world. So I want to instead understand them as some sort of formalist construction in relation to a framework of natural components. For example, we might stipulate a formal system called "morality" which is concerned with human well-being, or another called "prudence" which is concerned with self-interest. But there's nothing metaphysically special about these particular frameworks - we might just as well construct an arbitrary system "bluedence" that aimed to maximize the amount of the colour blue in the world.
Now, it's easy to get instrumental normativity out of these frameworks. We can say you "prudentially ought to visit the dentist", or you "bluedentially ought to paint your house blue." But as the latter example demonstrates, these framework-relative considerations don't always provide us with reasons to act. The mere fact that I "bluedentially ought" to paint my house blue in no way counts in favour of my doing so. After all, I have no reason to care about the bluedence framework, and hence can reasonably ignore the "demands" of bluedentiality.
We want morality to be different. We want to say that morality is genuinely reason-giving -- moral rightness counts in favour of an action, whether you care about morality or not. But what's the difference? What determines whether a framework is genuinely reason-giving or not? I see four options: non-cognitivism, pragmatism, realism about 'instrinsic values', and rationalism.
Let's start with 'rationalism', by which I mean a substantive, non-instrumental account of rationality. This would allow us to 'transcend' frameworks, and make (objective) judgments between them. I have no idea how this could work, and would really appreciate it if someone could explain this for me. My understanding of rationality is thoroughly instrumental. That is, I see rationality as being inherently relative to some presupposed ends; the job of rationality being to identify what means would best achieve those ends. So it's prudentially rational to visit the dentist, etc., but I can't make any sense of the notion of 'rationality, simpliciter'. I can't see how rationality alone could specify ultimate ends. In fact, that's more or less what this entire problem is about, since rationality might be defined as "acting on reasons", or something along those lines. So this option doesn't seem promising.
'Non-cognitivism' is at the other extreme. On this view, the reason-giving frameworks are those that we personally endorse. So I am bound by prudential and moral reasons, because I happen to care about myself and others, whereas "bluedence" does not provide reasons, as I don't care about it. This is a very neat and easy option, but also unsatisfying in some respects. For example, it denies that the amoralist has any reason to act morally. But perhaps that is just the sad truth. I will explore this option in greater detail in a future post.
Next, consider 'instrinsic value' realism. This presupposes that there are some things in the world that have instrinsic value, e.g. human well-being, or perhaps more objective things such as beauty, knowledge, etc. The claim is then that human well-being (or whatever) has intrinsic value, and thus we have objective reason to promote it. This would validate prudential and moral reasons, since both are concerned with well-being. Though it's not clear that epistemic reasons would survive, since it isn't obvious that truth and knowledge have instrinsic value (quite apart from any subjective desires we may have for them). But this option might do the trick, if we can defend a realist account of instrinsic value. So this answer would push the problem back a step.
Lastly: pragmatism. We might forget about the metaphysical problem and simply recognize that certain frameworks are more conducive to the needs of society. So we might say the amoralist has reason to act morally, because morality is a framework that's useful to society. Or something like that. I'm not sure if this is an answer or an evasion, but never mind.
Are there any other options? Which do you think is best? I'm afraid I might end up having to settle for non-cognitivism...