I found this comment by Robert Vienneau interesting: "Williamson writes a lot about how economics is a science and how they follow a scientific methodology. A few months back, he told us he had to look up the meaning of the word "incommensurable". So we know Williamson is basically unlettered when it comes to the literatures on the philosophy, sociology, and history of science."
Richard Feynman once said "philosophy of science is as useful to scientists as ornithology is to birds". While he probably intended this to be more dismissive than I would agree with, I do agree strongly with the plain English of the statement. Williamson may not be well versed in the philsophy of science (or the economics of science, for that matter). So? Who cares? He doesn't need to know science as a subject of study to be an excellent practitioner or example of it. We study economic agents all the time who have very little knowledge of or use for the science of economics. It doesn't mean they're failing to behave effectively in the economy.
My interest in the philosophy/history/sociology of science is fairly circumscribed, personally - I'm mostly interested in it to the extent that it informs the economics of science and particularly the labor market for scientists. I've been trying to get a handle of how science is produced, which in turn of course helps understand the derived demand for scientists, which is my personal interest*. The philosophy of science that veers into methodology holds little interest for me, for that reason. I've found Kuhn particularly enlightening, and Vienneau's reference to incommensurability raises Kuhn too. But here again, what is the use of knowing about incommensurability to the scientist himself? I don't see any real use. Certainly most paradigm shifters probably had the sense they were shifting a paradigm. But do they really need to understand Kuhn (much less have a working knowledge of the term "incommensurability") to prosecute their paradigm shift? Of course not!
So, Vienneau may very wel be right. Williamson may be largely ignorant of the philosophy, history, and sociology of science. Certainly that of all things isn't what's going to call his own credibility into question.
UPDATE: Another way of putting this is that a scientist can do his work with a ridiculously naive view of "the scientific method" and even convince himself that that's what he's doing and that's how science progresses - and still be an excellent historical example of Kuhnian science.
* - I'm personally not very entranced by the bibliometrics craze that's sweeped the economics of science, but I'm not entirely sure what an alternative would be (or what could remedy the way people currently think about bibliometrics as a way of getting at scientific production.