Keynes Presented No Concept of 'Radical Uncertainty' in the General Theory or in Any Other Published Work in His Lifetime : '…Even Though It Be on Precarious Evidence…' Does not Translate as No Evidence (Radical Uncertainty)
31 Pages Posted: 18 Jun 2018
Date Written: June 3, 2018
Keynes’s definition and concept of uncertainty in the General Theory has no connection with the terms fundamental uncertainty, radical uncertainty or irreducible uncertainty, which all mean complete and total uncertainty. Keynes’s concept of uncertainty is an inverse function of Keynes’s concept and definition of the weight of the evidence from the A Treatise on Probability. That definition is that the Evidential weight of the argument, V(a/h), is equal to degree w, where w has been normalized and is defined on the unit interval so that 0≤w≤ 1.A w=1 means that the relevant evidence is complete. Exact, precise, numerical probabilities can be specified. A w between 0 and 1 can only be represented by indeterminate (chapters 15-17,20,22) or imprecise (chapters 29,30) interval valued probabilities or decision weights, like Keynes’s original c coefficient of chapter 26 of the A Treatise on Probability. A w=0 means that there is no relevant evidence currently available. No probability, either ordinal, cardinal, or interval, exists. Keynes calls such a situation ignorance. It represented his fundamental disagreement with Laplace’s Principle of Non sufficient Reason, which was based on an equally balanced ignorance. Keynes replaced Laplace’s Principle of Non sufficient Reason with his Principle of Indifference, which requires equally balanced, symmetric Knowledge. Keynes’s Principle of Indifference has no connection with Laplace’s Principle of Non sufficient Reason. In the General Theory, ignorance is an issue only as regard the case of lonr run mec calculations.
Keynes made it very clear in his 1937 Quarterly Journal of Economics article that the case of w=0 applies only to the very special case of long run mec calculations about either the distant future (10 years) or very distant future (the rate of interest or price of copper 20 years from now). Keynes emphasizes that degrees of uncertainty, degrees of confidence or degrees of liquidity preference exist throughout the article. It is impossible to have such different degrees under ignorance.
In the Keynes-Townshend exchanges of 1937-38, Keynes agreed with Townshend’s summary of the General Theory as being based on the A Treatise on Probability concepts of the weight of the evidence and non numerical probabilities. It is not based on the belief of the existence of complete ignorance of the future.
It is the fallacious and deliberately misleading arguments of Joan Robinson, GLS Shackle, and Paul Davidson, who attempted to reinterpret Keynes’s different degrees of knowledge, based on his logical relation, V(a/h)=w, in terms of a completely different binary approach to uncertainty, where there is only knowledge and un knowledge. “un” is translated as no, so that h=0 in Keynes’s V relation, so that w=0 is the Robinson-Shackle concept of uncertainty.
Recent publications by Skidelsky, Kay, King, Fels and Krugman reveal the extent to which these authors are analyzing the Keynes-Knight concept of uncertainty through the mistaken Robinson-Shackle binary view of knowledge - radical uncertainty.
The modern restatement and updating of Keynes was done by Ellsberg in 2001 (1962).
Keywords: radical uncertainty, w= weight of the evidence, V(a/h) =w, estimates, Knight, Keynes, King, uncertainty
JEL Classification: B10, B12, B14, B16, B20, B22
Suggested Citation: Suggested Citation