A logical relation?
A "more fundamental criticism," to which we shall return later, is that "there really do not seem to be any such things as the probability relations he describes" (Ramsey, 1931, p. 161). In support of this Ramsey cites the alleged fact that others are "able to come to so very little agreement as to which of [these probabilities] relates any two given propositions" (Ramsey, 1931, p. 161). Of course, if Keynes is right about the structure of probabilities there is another and more natural obstacle to agreement, namely the lack of a vocabulary for naming these relations.
It must be admitted, as Ramsey (1931, p. 163) observes, that Keynes did waffle on the objectivity of the probability relation. Ramsey cites a passage in which Keynes make probability relative to the principles of human reason. This is a problem for any normative approach to human activities: our reach must exceed our grasp, or there is no function for the normative theory to perform; but it must not exceed our grasp by too much, or we will be unable to approach the ideal.
At this point Ramsey abandons his criticism of Keynes, and begins to construct his own theory. This theory has had profound and far-reaching effects not only in discussions of probability but in statistics and in the philosophy of science, as well as in other areas of philosophy and in economics.