The last post dealt with the relationship between identification and the real self. Let's here assume that identification can indeed specify the "real self", for the purposes of agency. Now I want to ask: what is the relationship between the real self (so understood) and acting freely (and thus moral responsibility)?
The obvious problem seems to be weakness of the will (sorry Neil and also Socrates). It seems that one can act from weakness of the will and be morally responsible for the weak-willed behavior; but weak-willed behavior presumably does not express the real self (the motivational states with which one identifies).
What has gone wrong? Frankfurt and Watson sought to give a more nuanced (and plausible) view of acting freely (and thus moral responsibility) than the simple view that one acts freely to the extent that one gets what one wants. On this view, in Frankfurt's terms, one acts freely insofar as one does what one *really* wants to do. Frankfurt goes on to specify what one really wants to do in terms of identification, and he gives successive elaborations of identification (in terms of elements including, but not limited to, higher-order volitions.) This suggests that one acts freely insofar as one acts in accordance with one's real or true self. But this implies that genuine weakness of the will--in which one freely acts against what one identifies with--is impossible, in my view an implausible result.
It seems to me that it is a mistake to tie moral responsibility so closely with action from identification and the real self. It seems to me that we need to distinguish between what might be called "autonomous" behavior (the term is not as important as the concept) and behavior for which one is morally responsible. Perhaps autonomous behavior is an expression of the true or real self, as specified by what we identify with; but one can morally responsible for behavior that is not (in this sense) autonomous.
(Note: I have developed these ideas, especially as regards the work of Frankfurt, Watson, Taylor, and Velleman, in some recent papers. Also, I have benefitted from reading work on these topics by Benjamin Mitchell-Yellin.)