Today, Paul Krugman argued that freshwater economists (you know, the folks who don't see any evidence of neo-Keynesian confirmation in the last few years of reality) are playing a Chicago version of Calvinball. Not only do I think Bill Watterson is among the best sociologists of childhood around (I know, I know: you think of him as a cartoonist), but Calvinball was a particularly beautiful representation of Calvin's essential anarchism.
Ah, but in the hands of researchers who don't want to admit that the evidence doesn't fall their way, Calvinball offers a way out: change the rules! This isn't anarchism but a way of cheating death oneself and one's audience. A few years ago I devoted a week of this blog to noting my own errors, big and small, and I should do that again in the near future (but with Brand! New! Errata!). The truly sad thing about academics who play Disciplinary Calvinball is that you don't get to learn anything new in your career if you not only are 100% correct but you always were and always will be.