It is generally accepted that to some extent Keynes changed his ideas over time, and he did say in his essay 'My Early Beliefs' that rational behaviour was far less prevalent than he had originally thought. However I am concerned here with the argument that Keynes implicitly went much further than he actually admitted, and that before he wrote the 'General Theory' he had recanted the very possibility of rational behaviour, at least in the sense in which this is commonly understood. This paper is concerned only with the historical question of what Keynes said, but the matter has contemporary significance because of what is called 'the crisis in macroeconomics', or the inability of economists to explain Keynesian phenomena as the outcome of rational behaviour. If Keynes recanted the possibility of rational behaviour, then the 'General Theory' must be the economics of irrationality, in which case we would not expect to explain macroeconomic phenomena as the outcome of rational agents.