In The General Theory of Employment, Interest, and Money (1936), John Maynard Keynes quietly drew attention to the link connecting the volume to his earlier Treatise on Probability (1921): both books are centrally concerned, albeit from different perspectives, with the rationality of behavior under conditions of uncertainty. The relationship between the two books did not come under sustained scrutiny until after the publication of his collected writings in the 1970s, but by now there is a considerable literature on the topic.
In a new book, Keynes on Uncertainty and Tragic Happiness: Complexity and Expectations (2021), Anna M. Carabelli offers a clear and bold interpretation of the relationship between the philosophy of the Probability and the economics of The General Theory, taking the position that Keynes introduced his own original method of investigation in the Probability that he later applied in The General Theory: “In this book I will argue that the true novelty of Keynes's theory is his method, which represents a new way of reasoning in economics. . . . Keynes's A Treatise on Probability is his ‘essay on method’” (3). Although Carabelli claims that her interpretation in Keynes on Uncertainty and Tragic Happiness is “new and unconventional” (9), the argument resembles that presented in her first book, On Keynes's Method (1988), itself among the first monographs to focus attention on Keynes's Probability. The author does not locate the new book relative to the earlier one, so it is not clear to what extent her thinking may have developed or evolved.
Keynes on Uncertainty and Tragic Happiness makes its core theoretical argument in three chapters. In each case it marshals an impressive number of interesting passages in which Keynes comments on methodological issues, especially uncertainty, rationality, and incommensurability. The book distills general principles from the passages and takes the set of those principles as the substance of Keynes's method. It identifies evidence of the method so constructed in the Probability and in The General Theory. This systematic approach brings together much that Keynes wrote about method, and this is a definite strength of the book.
A drawback of this approach is that it provides no support for Carabelli's claim that Keynes's method was original with him. In order to determine the originality of his method, it would have been necessary to compare Keynes's ideas with those of his predecessors and contemporaries. Keynes on Uncertainty and Tragic Happiness does not do this.
Keynes did not present his Probability as his own methodological innovation, and that is not how it was perceived by his peers. The preface to the Probability makes clear his own dissatisfaction with the book. His contribution was to have been a new understanding of probability as concerned with logical relations and therefore objective and reliable rather than purely subjective and doubtful. He did not feel that he had satisfactorily established that conclusion, but he abandoned hope of doing so: “I propound my systematic conception of this subject for criticism and enlargement at the hand of others, doubtful whether I myself am likely to get much further, by waiting longer” (Keynes  1973: xxiv).
In his editorial introduction to the Cambridge Economic Handbooks series, originally published in 1922, soon after the Probability appeared, Keynes wrote, “The theory of economics does not furnish a body of settled conclusions immediately applicable to policy. It is a method rather than a doctrine, an apparatus of the mind, a technique of thinking, which helps its possessor to draw correct conclusions” (Keynes 1983: 856; cited on p. 151 of Carabelli's new book). Carabelli cites this twice as evidence that Keynes introduced an original method. The context of the introduction, however, indicates that Keynes was championing the method of Alfred Marshall.
Marshall had already described economic theory as an abstract method of investigation to be applied in the discovery of sound conclusions: a “machinery of universal application in the discovery of a certain class of truths” (Marshall 1885: 19) or an “engine for the discovery of concrete truth” (24). For Keynes as for Marshall, economic theory is an objective but abstract aid to reasoning or way of reasoning that exists independently of the various materials to which it might be employed. Keynes restated this position in 1922 for a new generation, but it was not his idea originally.
The authors of the books in the Handbooks series were all under Marshall's influence and shared his method. Keynes described them as “orthodox members of the Cambridge School of Economics,” including Hubert Henderson on supply and demand (1922), M. E. Robinson on public finance (1922), and Denis Robertson on money (1922) and the “control of industry” (1923). Keynes (1983: 856) explained that it had been the goal of the authors “to expound . . . the most significant elements of economic method.”
Keynes did not view this apparatus as Marshall's innovation either. On the contrary, he continued the introduction with a description of its historical development in terms of Marshall's view of the history of economic thought as following a path of continuous improvement: “Before Adam Smith this apparatus of thought scarcely existed. Between his time and this it has been steadily enlarged and improved. . . . It is not complete yet, but important improvements in its elements are becoming rare” (Keynes 1983: 856). Keynes could not have seen himself as the author of a method predating his birth.
It is possible that Keynes might have introduced a new and original method without being aware of it, but Carabelli's description of Keynes's method refers to ideas that were widely held. For example, Carabelli writes, “Keynes maintains that what matters is the reasonableness, which is neither the absolute rationality nor the truth, of judgement and action” (14). Marshall and the authors in Keynes's book series all agreed that action and policy should be reasonable. Individuals might differ over what is in fact reasonable, but they agreed that judgment, action, and policy should be reasonable and that unreasonable policy should be rejected. This could not have been Keynes's original contribution. Or again, Carabelli emphasizes that Keynes held that the world is complex and it should be treated as if it were complex: “The material of economics . . . is . . . complex and requires complex theory” (153). It is hard to imagine that Marshall or any of Keynes's authors might have argued for a simple approach to a complex problem or that interdependent things should be treated as if they are independent. This was not original to Keynes. According to Carabelli, Keynes's method called for avoiding tacit hypotheses and logical fallacies: “A general theory is a theory that does not introduce tacit assumptions. . . . It should take account of logical fallacies” (53). But, like scholars everywhere, the members of the Cambridge school tried to avoid tacit assumptions and logical fallacies. This is not and could not be Keynes's contribution.
In the early 1920s, then, Keynes had not invented a new methodological approach—he was working to make an incremental change to an approach that he thought had already existed for a long time. It was only later, not before the late 1920s, that Keynes began to be impressed with the novelty and gravity of his own contributions, as he moved toward The General Theory.
The final third of Carabelli's new book contains chapters on a series of ambitious and interesting topics not addressed in the first book, including the moral complexity of tragedy, Aristotle and the meaning of happiness, and international economic and political relations. It is true that tragedy involves making choices with consequences over time and therefore has a limited formal similarity with decision-making under uncertainty, but the claim that Greek tragedy played any significant role in Keynes's thought seems to have very little support in Keynes's writings beyond occasional references in a few of Keynes's Apostles' papers. I think that it is too strong to say that “Keynes accepts the Aristotelian notion of the good and happy life” (111), but Keynes's reading of Aristotle does seem to have influenced his thinking. The chapter on international relations provides a very interesting account of Keynes's alternative to globalization, in which strict controls would be used to encourage autonomy and “policy space” (128) giving nations the opportunity to conduct social experiments in social organization as they wished without risking harm to anyone but themselves.