Allow me to indulge in a probably futile attempt to clarify things.

(1) The original Katz post (according to memory) suggested that Noda's behavior seems irrational according to rationality defined by rational choice theory. My going-off-on-a-tangent addendum was that rational choice theory was not a theory of whether an actor's behavior was rational or not:
it is a theory of what happens if people interacted rationally; rationality defined in terms of the utility the actor derives from obtaining a certain goal. The value of the goal is assigned by the theorist as a postulate and has nothing to do with the actor. So, rational choice is not a theory that explains individual motives or preferences.

(2) If rational choice is not a theory of motives or preferences, it cannot be criticized or dismissed for being unable to explain Noda's behavior.
Might I add, ascertaining Noda's mixed motives is of little theoretical use unless it is applicable to other prime ministers in a certain situation, the payoffs and the situation (rules of the game) being specified by the theorist. If one assigns the payoff structure and clarifies the rules of the game and then tries to figure out what happens as a result of actor interaction, welcome to rational choice theorizing (in which, believe it or not,I don't take part).

(3) Finally, I couldn't understand what Aurelia was referring to by "personal utility (self-interest) defined in terms of a single, uniform variable." All I can say is, surely self-interest is not a synonym of selfishness and is not the opposite of societal well being or common good.

On a more positive note, recent studies by a new generation of spatial theorists (and thus in the Downsian tradition) such as Lawrence Ezrow (2010) and Bonnie Maguid (2007) have argued that it is rational for niche parties to stick to their principles, if they are willing to pay the cost of limiting their size, in the same way that it is rational for major parties pursuing vote/seat maximization (and hence power) at the cost of shifting their policy positions. Their theories have been tested mostly with West European data returning robust results. Being closer to my "thing" the impressive point about this revision of Downs' median voter theorem is the confirmation of the payoff between policy principles and vote (or seat) maximization in democratic party competition: note, however, that both are rational decisions and constitute an equilibrium of sorts when the different types of parties compete with each other. I think these studies question the claim that rational choice theory is incapable of analyzing principled actors.

Again for what it's worth and apologies if i bored your pants off.

Nobuhiro Hiwatari