Keynes on probability and decision: evidence from the correspondence with Hugh Townshend

In the literature on his philosophical ideas the correspondence Keynes had with Hugh Townshend over the just-published General Theory has attracted significant attention. Excerpts from the exchange have been used as a relevant piece of evidence by scholars who claim that Keynes came to reject rational decision criteria, thus focusing on the necessity for economic agents to form expectations on market sentiment, rather than fundamentals. This note concentrates instead on the whole correspondence and tries to show that a comprehensive reading of the exchange between Keynes and Townshend, unfolding through the years 1936-1938, suggests that its discussion thread was more technical than usually understood. It is argued that the correspondence provides evidence for the fact that Keynes still had a keen interest in a problem left unsolved in the Treatise on Probability, namely, the definition of an alternative to what he termed «normal ethical theory» in the Treatise and identified with «strict mathematical calculation» in the General Theory. The correspondence reveals that the issue of whether a useful decision rule can be devised under uncertainty still appears central in Keynes’s thought in 1938.