We have seen that Keynes developed an ethical philosophy which would begin from Hume's identity of probabilistic and ethical reasoning, but would blur Hume's division between ethics and fact. A famous response to Hume's ethics had already been made by Kant; but, whereas Keynes believed that Kant was right to say that a rational theory of ethics was possible, nevertheless, Kant had failed to understand Hume's parallel between probability and ethics. Decades later, Keynes came to believe that there was a particular reason for Kant's failure.
In the course of 1937 Keynes and his former protege, Pietro Sraffa, ascertained that an Abstract of Hume's Treatise on Human Understanding, which it had been thought was written by Adam Smith, was in fact written by Hume himself, and that Hume had disguised his authorship. The discovery was not revolutionary, but it clarified some aspects of the Treatise, and it was decided to reprint the Abstract with a joint introduction. Soon however Keynes meant to write an introduction that would go well beyond the Abstract, using it as a springboard to criticize Hume for being a congenital liar and for misleading Kant with his lies. This intention was thwarted by Sraffa, who insisted that an introduction to a newly discovered work was not the proper place to accuse the author of being a liar, unless the facts were undoubted, and especially not unless the matter was of great significance, which Sraffa thought was not the case.
There ensued an exchange of correspondence between Keynes and Sraffa and between Keynes and various authorities on Kant, the drift of which is illustrated by the following excerpts.
Keynes to Sraffa, 1 January 1937
Tell me if you think that my footnotes are getting too irrelevant. But I feel this is a production where irrelevance is not altogether out of place.
I now return to my notes on pages 2 and 2A for which I plead. The story of the disappearance of the Treatise during the eighteenth century is of first class interest and importance, scarcely paralleled in the history of philosophy. It comes in I thought, very apropos with the disappearance of the Abstract. I do not believe there is any doubt about the facts.
Keynes to the Kantian Scholar, Dr E. Rosenbaum, 17 November 1937
I am already having difficulty with Sraffa who is trying to delete everything I have written on Kant on the ground that it is not relevant to put footnotes about Kant in the reprint of a Hume. I am battling with him to preserve at least a number of really interesting details.
Hume had not only disguised his authorship of the Abstract, but had tried for most of his life to suppress the Treatise. The Treatise failed to attract interest on publication; it was extremely rare even in Hume's lifetime, and no second edition was published until long after. Hume disowned the Treatise, which in his eyes was superseded by the publication of his Inquiry Concerning Human Understanding. However, the Inquiry is now generally regarded as an inferior book, being more readable but less than adequate concerning the logical foundations of Hume's system. On 27 November 1937 Keynes wrote to Sraffa under the heading Hume's Veracity:
The letter continues with Keynes detailing various of Hume's lies concerning the publication of the Treatise. Here were two philosophers—Hume, who said that there is no rational basis for action, and Keynes, who said that truth is the rational basis for action—and Keynes was accusing Hume of being a liar. What Keynes was driving towards was that not only did Hume lie, but that the spread of his philosophy was advanced by his lies. Keynes's enquiries concerning Kant led him to believe that Kant did not know of the Treatise at the time that Kant wrote his (p.201) famous replies to Hume. Hume suppressed the Treatise and favoured the Inquiry, and he went so far as to say in the preface to the Inquiry that it was the first book that he had ever written.
I have always regarded his [Hume's] statements in regard to the Treatise as unreliable. The effect of the violent disappointment he had about that was to ensure a complex in him which prevented him from being sensible on the matter for the rest of his life; his chief object being to cover up and to excuse. On looking into Birbeck Hill I find I am not alone in attributing this side to his character. Hill writes on p. viii of his preface ‘Hume, with a levity which is only found in a man who is indifferent to strict truthfulness.’
Keynes to Sraffa, 27 November 1937
As regards Kant . . . my point was that he was not acquainted with the Treatise a good many years earlier when he was writing on Hume in the Kritick and the Prolegomena. The point was that Kant was unacquainted with it when he wrote his famous passages on Hume's philosophy.
The point was that Kant had not been able to respond to Hume on Hume's own terms, because Kant had not known properly what Hume's terms were. Kant was familiar with Hume's way of thinking from the Inquiry, but he was unaware of Hume's logical structure. Two decades earlier, in the Treatise on Probability, Keynes had accused Kant of error, and now it transpired that Hume's deceptions were the reason for Kant's error. (p.202)