Joan Robinson, first with the aid of GLS Shackle, and then with the help of R. Skidelsky, attempted to convince academics, after Keynes had died in 1946, that Keynes had made major changes in his ideas on uncertainty and expectations in his 1937 QJE article that directly conflicted with what he had stated in the General Theory. The major change that was claimed to have occurred was that Keynes had supposedly changed his definition of uncertainty from being a range of possible different degrees of uncertainty, between the extremes of Complete Knowledge and Complete Uncertainty, to becoming a situation of “fundamental uncertainty” (fundamental uncertainty is also called radical uncertainty, irreducible uncertainty, deep uncertainty, genuine uncertainty, and real uncertainty by heterodox economists), which Skidelsky defines as “When a person cannot …compare the probabilities of two arguments, he may, using a terminology drawn from recent discussion, be said to be in a state of radical, or complete, uncertainty.” (Skidelsky, 1992, p.81). Skidelsky’s conclusion makes no sense because comparable probabilities must be exact, precise numerical probabilities a la Frank Ramsey. Of course, what Keynes had actually done, in Part II of the A Treatise on Probability, was to show how interval valued probability, using upper and lower bounds, could be used to deal with incomparable, non measurable, and incommensurable situations, where exact, precise, numerical probabilities could not be specified, by means of what Keynes called approximation and inexact measurement. Keynes NEVER, EVER discussed any such things as fundamental uncertainty anywhere in any publication in his lifetime. What he did discuss on pp.309-315 of chapter 26 of the A Treatise on Probability, was the case of IGNORANCE, which can only occur in the distant or far future (long run), and NOT in the near or immediate future (short run). His discussion in the A Treatise on Probability of ignorance is by far the most detailed and specific discussion of the impact of ignorance on decision making ever made by Keynes. Keynes’s discussions in chapter 12 of the General Theory or in the 1937 QJE article provide nothing new or novel and are, at best, small footnotes made to the discussions in the A Treatise on Probability. This, of course, is what Keynes told Townshend in late 1938 in response to Townshend’s five page letter that is contained on pp.289-293 of the CWJMK in Vol. 29 (1979). The goal of Robinson, as well as of Shackle and Skidelsky, was to eliminate any and all connections between the A Treatise on Probability and the General Theory, due to her severe mathematical ineptitude and innumeracy. This is why the exchange between Keynes and Townshend on pp.289-293 in Volume 29 of the CWJMK have been covered up by Robinson, Shackle, and Skidelsky, who cites only the last page of Keynes’s reply on page 294, but not page 293, where Keynes accepts Townshend’s assessment that there was a major and direct connection between Keynes’s liquidity preference theory of the rate of interest in the General Theory and Keynes’s nonnumerical probabilities and weight of the evidence from the General Theory, a connection that Skidelsky had desperately tried to undermine and destroy in his Shacklesque attack on the A Treatise on Probability on pp.60-100 of his 1992 second volume of his biography on Keynes.
Read full abstract
Apr 3, 2021
Deborah M Figart 
Open Access