While this was just a petty aside in a blog post, it could be safely ignored. But now the same points have been made in the august JEL, a house journal of the American Economic Association, they acquire a cachet which means they can't be so easily dismissed as Aunt Sallies.
I'm pleased to see this problem get a bit more attention because I've always been frustrated at how economists and rational choice theorists in general play bait-and-switch with the rationality assumption. Obviously if anything can be rational, then rational choice cannot make substantive predictions. But at the same time, the claims that are usually made in the name of rationality are far tougher - ie that people are self-interested, rational and capable of complex calculations about their material interests. The tautological rationality assumption is only wheeled out when the latter claims run into trouble (which they generally do when we try to explain non-trivial problems).
So, as Quiggin says, whenever we want to make this kind of point, we can just cite Williamson 2011 in the JEL.
As a footnote, I love Williamson's introductory paragraph, which states that 'in economics, the 2008-9 financial crisis has resulted in renewed appreciation of advances made in information economics, the theory of financial intermediation and monetary economics'. Which is like saying that collapse of the Berlin Wall brought renewed appreciation of advances in our understanding of Soviet communism through the 1980s, or that 9/11 alerted us to our most cutting-edge findings in the study of international terrorism.