Deirdre McCloskey's eureka moment, on realizing the explanatory power of rhetoric, in Times Higher Education:
I realised with a jolt that economists are fibbing when they say that they follow a positivist, hypothetico-deductive method-of-science-I-learnt-in- secondary-school. The new awareness came partly from noting that when I put forward any proposition in Chicago School economic-historical science (wholly correct, I assure you), I got from other economic scientists not counter-evidence but anti-Chicago School ideology. I had already perceived that my Keynesian teachers at Harvard in earlier years had exhibited in their ways of arguing such a disgraceful lack of devotion to the plain truth. Shame on them. But around 1977, hiding in Hyde Park, I realised that the Chicago School folk had it, too – in spades, redoubled and vulnerable.
A professor of English at Chicago, Wayne Booth, asked me – on the strength, I think, of a reputation I had for being marginally more open than other economists – to give a lecture to some undergraduates on the “rhetoric of economics”. Sure, I said. But what's that? Wayne suggested that I read books such as The Uses of Argument by Stephen Toulmin and The New Rhetoric by Chaim Perelman and Lucie Olbrechts-Tyteca, and I suddenly got it. Oh, my God! Even a science such as economics has a rhetoric – that is, a means of unforced persuasion! And the claimed “scientific method” ain't it!
This blindingly obvious point led me to start doubting that economics was the queen of the social sciences, or at any rate that the queen had any clothes on. First I realised (I recall the day, by then based at the University of Iowa, talking to my colleague Richard Zecher) that a crucially important technique used by economists, “statistical significance”, was rhetorical rubbish. Close fit is not the same thing as scientific or political importance. It just isn't.
[H/t: Mark Blyth]