F. P. Ramsey: Critical Reassessments [Hardcover]

This set of essays on Ramsey's contributions to philosophy,economics,probability,etc.,could have earned a five star rating.The reason it has not is that none of the authors is willing to deal with the fact that Ramsey's ideas on mathematical probability are a very special case of Keynes's general theory of logical probability which applies not only to point estimates(Ramsey's theory)but to interval estimates and ordinal rankings as well.Ramsey is primarily concerned with degrees of belief(not degrees of rational belief which is the major concern of Keynes's ) where he believed that "...there was an authentically logical interpretation of the (PURELY MATHEMATICAL)laws of probability...[as]consistency constraints on the distribution of partial belief".(Howson,p.145).A major confusion of Ramsey's concerned Keynes's conditional probabilities.Only in the special case of the weight of the evidence,w,equaled one(w=1,where w is an index that measures the completeness of the relevant information,data,knowledge,etc., upon which the probability estimates are being based) would Ramsey's conditional probabilities(say P(A/B)=C) be identical to Keynes's conditional probabilities(say p(a/b)=c).Otherwise,different conditional probability estimates will result.Ramsey's subjectivist approach is thus a special case of Keynes's more general logical approach.The reader should note that Tversky and Kahneman made the exact same mistake as Ramsey did for about 20 years in their continuing series of exchanges with L Jonathan Cohen in the journal Brain and Behavioral Sciences over the rationality of decision makers in experimental setups between 1975 and 1994.Tversky's support theory was a belated attempt by Tversky to fix up the giant hole in his theory in an ad hoc manner since support theory has not been integrated into cumulative prospect theory.Keynes would ,of course, agree that ,in those special cases where the purely mathematical laws of probability were applicable(w=1), agreement with these laws is a coherence and consistency requirement that the rational and reasonable decision maker would want to accept ,provided that his probability preferences were linear.The problem ,then, is that Ramsey's theory is so narrow(only linear preferences and probabilities with weights of 1 can be used) that it is practically inapplicable to problems in everyday decision making,social science,liberal arts,business,economics,etc.Its applicability is resticted to areas in the life and physical sciences where the weight of evidence is equal to one or approaches one in the limit.
The second major flaw in Ramsey's approach is the requirement that ALL probabilities MUST be precise single number estimates.This directly contradicts Keynes's approach where the probabilities are primarily interval estimates based on Keynes's path breaking work in chapters 15 and 17 of the TP.All of Part III is based on Keynes's analysis IN Part II of the TP.Ramsey never read Part II or Part III of the TP.Ramsey reviewed chapters 2 and 3 of Keynes's book.Ramsey had no idea at all about the fundamental interval estimate foundations of Keynes's approach.The only conclusion that is possible is that Ramsey's two book reviews of Keynes's TP,in 1922 and 1926,are completely worthless in so far as they purport to deal with Keynes's logical theory of probability.These two reviews have continually misled literally thousands of economists,philosophers,statisticians,psychologists,and logicians over the last 80-85 years.

› Go to Amazon.com to see the review  3.0 out of 5 stars