Practically Rational

Jumping over point (1.) of Joyce's argument, let's take a look at point (3):

(3.) Practical rationality is the only source of statements that cannot be legitimately questioned.

This can be broken into two pieces:

   3a. Practical rationality cannot be legitimately questioned.
   3b. There is nothing else that cannot be legitimately questioned.

Joyce spends some time arguing for (3a), but the basic point is quite simple. The question, "Why should I care about practical rationality?" simply makes no sense. It amounts to asking for a reason I should care about reasons. This is obviously incoherent.

Oddly - given how central it is to Joyce's argument - he says very little about (3b). He merely points out that the argument in the previous paragraph doesn't work when "practical rationality" is replaced by any other normative system. Maybe this is enough, but it seems to me that such an important point needs more than a one-sentence support. (Of course, it may be that I am misrepresenting his argument in making (3b) so central.)

Anyway, it seems we could avoid moral error theory  if there was something other than practical rationality that could not be legitimately escaped. I don't see much hope for this escape route, though.

However, I wonder if we were to take the view of morality that I've been promoting - a social system that imposes constraints on individual behavior - could we argue that, while it is possible to logically step outside the moral system, there is no way to do so practically? That is to say, we are necessarily part of a society, and so are subject to the moral system of those around us, whether we like it or not. (Unless I am alone on a deserted island for the rest of my life - in which case there is, arguably, no need for morality.)

Joyce goes on to analyze practical rationality.

An agent S is practically rational to the extent that she is guided by her subjective reasons.

And,

S has a subjective reason to X if and only if she is justified in believing that S+ (S granted full information and idealized powers of reflection) would advise S to X.

The main take-away from this definition, for my purposes, is that practical rationality is agent-relative.