5
$\begingroup$

A visiting speaker in Economics recently happened to mention that John Maynard Keynes' A Treatise on Probability revolutionized probability theory. I have not heard any such claim before and it struck me as strange. The Wikipedia page contains some effusive praise from Russell but nothing specific. This leads me to ask:

1) Is this claim approximately true?

2) In what specific ways did it impact probability theory?

3) What are some specific citations which demonstrate this?

$\endgroup$
5
  • 2
    $\begingroup$ This seems like a stretch. See Aldrich (2008) for a nice review. There's also Robert (2011). $\endgroup$
    – dimitriy
    Apr 18, 2014 at 4:43
  • 1
    $\begingroup$ You can wait a while until the History of Science and Mathematics StackExchange site comes into public beta, then ask this question there. $\endgroup$ Apr 18, 2014 at 7:24
  • 2
    $\begingroup$ Was the visitor British? ;-) $\endgroup$ Apr 21, 2014 at 18:30
  • 1
    $\begingroup$ Yes, strange. If anyone can make a claim about revolutionizing probability theory, it is probably Kolmogorov. $\endgroup$
    – Michael
    Apr 23, 2014 at 2:24
  • 1
    $\begingroup$ You can find some comments in : Maria Carla Galavotti (editor), Cambridge and Vienna Frank P. Ramsey and the Vienna Circle (2006), with an essay by P.Suppes on Ramsey’s Psychological Theory of Belief, and also in Jan von Plato, Creating Modern Probability. Its Mathematics Physics and Philosophy in Historical Perspective (1994). I think that Keynes contribution was obscured by the "neopositivist" trend in pre 2WW philosophy and by von Mises and Kolmogorov contributions. Also, in England the figure Harold Jeffreys was of greater historical importance than Keynes in prob theory. $\endgroup$ Apr 23, 2014 at 15:12

3 Answers 3

2
$\begingroup$

If you have a copy of Kai Lai Chung's Elementary Probability Theory, you will find Keynes' picture next to that of Kolmogorov, Polya, Feller and other probability heavy weights. (I do not know if you would cosider this evidence.) According to my modest understanding, Kolmogorov formalized probability theory and practically re-wrote the field. As far as I know he provided the framework to study probability in a measure theoretic sense. Keynes was influential not in a mathematical sense, but in a philosophical sense. A situation that exemplifies his struggle between quantifiable probability and reality is that if you have a fair coin and toss it 20 times, say you get 15 heads in a row, it would be natural to argue the coin is not fair.

$\endgroup$
0
$\begingroup$

J M Keynes did not revolutionize probability theory.J M Keynes revolutionized decision theory by showing that the mathematical laws of the calculus of probabilities were a special case requiring that the weight of the evidence,w,where w is defined on the unit interval[ 0,1],must equal ,approximate ,tend,or approach 1 in order for the probability calculus to be operationalized where 0 is less than or equal to w ,which is less than or equal to 1.The weight of the evidence ,w,measures the completeness of the relevant evidence upon which the probability estimates are based.w=1 requires that the sample space of all possible outcomes be specified before any decision is made or,what amounts to the same thing,a specific ,unique probability distribution is known before any decision is made. A w <1 requires the use of interval estimates ,which contain an upper and lower bound. The first economist to specify an interval estimate of probability was Adam Smith in the Wealth of Nations on pp.106-109 of the Modern Library(Cannan) edition See p.714 for Smith's clearcut statement that probabilities can't be estimated precisely .

$\endgroup$
0
$\begingroup$

Thank you for your comment.Unfortunately, Kolmogorov's axiom system assumes linearity and additivity. His precise or point estimate approach to probability gives correct answers only in those fields where the weight of the evidence, w, equals, approximates, or approaches 1 in the limit. It is in many of the areas of physics, chemistry, biology ,and engineering where the weight of the evidence is close to 1,where w, the weight of the evidence ,is defined on the unit interval between 0 and 1.w=1 means you are able to define a sample space of all possible outcomes before you begin your experiment or analysis.

Keynes's approach is based on Boole's indeterminate approach to probability and uses intervals, not point estimates ,because in many instances, especially in social science, liberal arts , economics, business, finance , and public policy, there is substantial missing evidence that will never be available to the decision maker when he must make a choice. Boole's system, Keynes's version, and Hailperin's 1986 linear programming approach ,would treat Kolmogorov's approach as a special case where w=1.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.