Dear Anthony, I accept the difference between Intellectual History and the History of Economic Analysis. To avoid writing nonsense, the one requires serious competence in history and philosophy (including at least until Marshall, as you point out, theology!). To avoid writing nonsense, the other requires a serious competence in economic theory. Right. But I think you would agree that the two need to interact. For example, an internalist history of the sort that Stigler practiced, and Blaug and Samuelson practice, is often bad history, and precisely because of its timeless and ahistorical premises results in a particular theory of how to do economics. If you take the internalist view you think of Keynes and Menger as spoiled examination scripts. We have it right today. An example is Bob Lucas' macro course at Chicago, in which he announces on the first day that they will be reading no work older than five years. The kids at Chicago literally do not know how to pronounce "Keynes" (no exaggeration). Lucas' practice appears to come out of a theory that economics is "like physics" (people who say this commonly have no idea how physics actually works, by the way, and are not eager to find out). You cite Feyerabend. He would have supported my counterargument here: one needs to study the actual history of Galileo's findings and astrology and so forth, said Feyerabend, in order to establish that there is no such thing as a timeless Scientific Method. This is something that scientists need to know, somehow, so that they are not enchained by a limitation of arguments acquired from their high-school chemistry teacher. The best scientists (I think of Feynman and Samuelson) acquire it on their own. Dolts like me have to have it shown to them before they can get over the powerful and erroneous claim current in the culture that Science is something quite different from human argument. So a physicist or a chemist does need a true history of science. Another example of the same problem is the practice in departments of philosophy of treating Plato as a contemporary. I had a colleague once who taught Plato to graduate students from the old Victorian translations and didn't know a word of Greek. He argued that it he didn't need to know Greek because it was the philosophical arguments that mattered (and therefore he didn't realize for example that the differential uses in Attic Greek of verbs of seeing and verbs of hearing in indirect discourse affected---indeed, achieved---Plato's proofs). People like Rorty and Toulmin and MacIntyre have argued that philosophers should be treated historically, and that doing so changes how one does philosophy. In Rorty's case for example a historical study arrives at the conclusion that novelists are as important philosophically as official philosophers for philosophy. In our field the analogy would be the historically-acquired grasp that, say, "pre-analytic vision" a la Schumpter and latterly Bob Heilbroner of blessed memory is very important in economic analysis. Regards, Deirdre McCloskey