It is also true, as the Austrian economist Menger has remarked, that sciences have been created and revolutionized by those who have not stopped to analyse their own method of enquiry.--J.N. Keynes (1891) The Scope and Method of Political Economy, p. 5.
I was reminded of Keynes's treatment of Menger while listening to a lovely paper by the young scholar, Jens van 't Klooster, at a recent HOPOS conference. J.N. Keynes -- an author of a logic text-book -- is the father of the more famous philosopher-economist, Keynes. Menger is often associated with the idea of a scientific revolution because together with Jevons and Walras, Menger is taken to have revolutionized formal economic theory by introducing marginal analysis. Keynes reports that Menger was familiar enough with the concept of a scientific revolution. In fact, in the remark attributed to Menger, Menger recognizes both the establishment of sciences and their (intermittent) revolutions.
As an aside, we should not assume ideas about revolutions are original with Menger. Menger was a close student of Adam Smith's "History of Astronomy," (here) which offers a historical narrative about successive revolutions among (psychologically) incommensurable systems of thought in the sciences, which have regular patterns of development between each revolution (recall).
Despite early mention of the idea of revolution in science, Keynes does not elaborate at first. But it's clear that Keynes thinks of economics as a less mature science than, say, "physics and astronomy," (or less "definitive"). For Keynes relies on the idea that an a science develops through different stages, including a "descriptive or classificatory" stage before reaching an ultimate (or "definitive"), "deductive" stage. (These stages, themselves, are not to be confused with the three step/stage deductive method that Keynes adopts from J.S. Mill.) In fact, within this progressive framework of stages, Keynes inscribes a now familiar ideal:
If political economy regarded from the theoretical standpoint is to make good progress, it is essential that all extrinsic or premature sources of controversy should be eliminated; and we may be sure that the more its principles are discussed independently of ethical and practical considerations, the sooner will the science emerge from its controversial stage. The intrusion of ethics into economics cannot but multiply and perpetuate sources of disagreement. (Ch.2)
Good scientific progress is itself characterized as moving from an immature "controversial stage," in which the wrong sort of disunity occurs to a more mature stage, which is characterized by the possibility of agreement over matters intrinsic to theoretical science (even if, of course, disagreements are possible then, too). Keynes relies on the idea that when it comes to "fundamental ethical questions that have long been the subject of controversy" no agreement is to be expected. Keynes proposes a robust separation between positive and normative economics (he also recognizes an 'art' of economics): by making "positive" science "independent" of ethical issues it can hope to reach the mature stage more readily. This positive science is primarily focused on economic facts and generalizations about these.
Now, this is not the place to enter into an analysis of Keynes's claims about the nature of positive economics and its independence from ethics. In many ways, Keynes is just echoing his more illustrious Cambridge contemporary, Sidgwick, who is explicit that his aim is to salvage "the really sound and valuable results of previous thought" from "the waves of disputation;" in other words to "eliminate unnecessary controversy." (recall) Sidgwick's project within economics is to create a technical apparatus that produces consensus among the experts. For, without this the political economist has no special standing as the expert worth listening to by policymakers. All existing further disagreement has either to be caused by conflicting facts (as Sidgwick acknowledges in context) or by fundamental differences in values (as Sidgwick describes in great detail in his discussion of the legacy of Adam Smith in the next chapter).
Keynes's key move is, thus, not the distinction between (i) a realm of facts and their generalizations that may lead to unanimity and (ii) a realm of values that are sources of disagreement; rather the key is that he deploys that distinction in the context of a not-entirely-explicit theory of scientific progress in which a field develops from immature, conflict-ridden stages to mature stages characterized by considerable agreement. As the nod to Menger suggests, this immature-mature picture of scientific progress is also able to accommodate scientific revolutions. Among Keynes's keenest readers were two Chicago economists, Milton Friedman and George Stigler (who calls Keynes's book "excellent"), who spent a career criticizing his son, while accepting these elements of his philosophy of science. Not surprisingly, Stigler also actively promoted Kuhn's version of these ideas (recall).
The advance, if it is an advance, of Keynes's position over Adam Smith's idea a century earlier is that one must make economics independent of fundamental ethical questions and judgments. In context, Keynes does not reflect on all the difficulties with the idea of making theoretical or positive economics independent of ethical judgments. That there is a fact-value distinction is one thing; that one can make the concepts that are meant to track the facts and their generalizations purely factive without either presupposing or tacitly deploying various ethical judgments is not an easy matter to establish. But we should not ignore the fact that such an idea is very attractive to those economists that wish to get on with their research, and sell its fruits to others, as well as to those philosophers, which are enamoured by the idea that they are the experts of ethical matters, and that have their agenda for us.