Probability

A Treatise on Probability is a difficult book to place in the intellectual development of Keynes. Practically ignored till the 1980s, some economists now see it as the epistemological foundation of his later work. Its date of publication, 1921, is misleading since the book grew out of a fellowship dissertation begun in 1905 and first presented (unsuccessfully) in 1907. Reworked and resubmitted in 1909, it earned its author a fellowship at King’s. Despite the fact that, as Harrod notes, Keynes soon saw that ‘salaried drudgery’1 (i.e. government work and academic teaching) was not for him, he yet worked on the dissertation until 1914 to turn it into a book, his labours being interrupted by his work at the Treasury and at Versailles. In 1920 he put the finishing touches on the work, the preface being signed 1 May 1920. In the Royal Economic Society edition we have a book of 500 pages dealing with one of the most fundamental topics in the theory of knowledge to which, however, the author seldom made a reference in the spirit of a scholar who sees a continuity between it and his later works. On the contrary, a well-known reference to it in The General Theory is designed to warn the reader away from Probability, to distinguish the concept ‘uncertain’ of that book from the term ‘improbable’ in the Treatise? In 1938, his controversy with Tinbergen brought about an oblique reference to the Treatise when Keynes noted that since his youth little progress had been made to strengthen the foundations of the theory of inference. There are also brief references to Probability in correspondence with Hugh Townshend (1938) and in a note on Frank Ramsey (1931) that many take as a repudiation of his earlier views.