Mainly for the sake of being provocative and antagonizing a good philosophical friend, I recently posited what I'm now calling the strong thesis of normativity's empire: all concepts, in themselves, are essentially normative. So normativity is an empire covering everything, or, at least, everything which might belong under a concept.
Stated baldy, it can be a pretty implausible-sounding hypothesis, I'll admit. There are clearly classes of concepts--moral concepts, most obviously--which seem intrinsically normative; that is, as carrying with themselves an idea of reason, where the idea of reason is itself ineluctably connected with the idea of what might or must be a proper way for the world or its parts to be. Then there are concepts in a gray area--concepts like etiquette, and law, and (even) grammar--which, although bound up with the idea of rule-governed activity, perhaps seem to be dignified too much if regarded as, in themselves, really normative. These concepts, we might think, are uncertain cases worth arguing about. (The argument on the pro- side would go like this: to say that a thing is lawful, or grammatical, or polite--and just that alone--is to purport to offer some weak reason to prefer it. But that just means that the ideas in question are essentially normative.)
But then what of concepts like 'red'? Surely, you say, a concept like that is purely descriptive, and it would have to be the height of extravagance to suppose that it was intrinsically normative in any full-blooded sense of 'normative.' This is the burdensome hard case for the strong thesis. Nevertheless, maybe the burden can be met. Restricting our conversation just to 'red,' what can be said in favor of the strong thesis of normativity's empire? This, I think:
To say that something is red is to say that it ought to produce a certain subjective reaction in a subject. So, in fact, the idea of 'red' is an idea connected essentially with a certain idea of how stuff should go, and how the world should be. It is, in other words, an essentially normative concept.
The question is whether this account actually makes any kind of sense of 'red.' It seems to me that it might. So: we can make sense of the idea of red-blindness because we have the idea of the way in which, in certain (red) circumstances,* a subject ought to be phenomenologically affected. We can identify a certain frequency-range of lightas red only because we have a prior idea of how subjects should, in certain (red) circumstances, be affected: the frequency-range in question, then, is just the range which "matches" the appropriate response in the appropriate circumstance. So, in short, the normative account of the concept 'red' generally makes sense of the ways we might use the concept (in medicine! in physics!). So, that's saying that it's a Pretty Good Account.
So, maybe the strong thesis of normativity's empire is true after all. At the least, it appears at present to be an undefeated conjecture. Anyone out there want to offer the knock-down reason that buries this conjecture?
* - I know what you're thinking, and you're probably right, insofar as I'm thinking it too: there is no way of individuating the circumstances in question without having an idea of red already. So, it turns out that we only get the idea of red with the idea of red. This, I aver, seems intuitively both empty and right. It is a problem which occurs for other accounts of red. Suppose we say that 'red' is that (pointing now at something red). How do we individuate 'that'? Not just any arbitrary way-- it must be individuated in the red way. But then we have the same problem. I do not know the way out here; and yet! 'red' means something!