There have been many unfortunate (and incorrect) statements made 
already  in this fledgling discussion about risk and uncertainty. I 
have often felt that the discussions on this list are of such a low 
(and often anti-scholarly) character that it is simply useless to get 
involved. There is already a large and complex literature on many of 
the questions being discussed here and I am uncertain how to begin 
addressing the many aspects of that literature that could bear on the 
discussion.

I admire Ross Emmet for joining in the discussion, however, and 
trying to bring a more serious and informed tone to it.

Thus, it is out of my respect for his work and his effort to improve 
this discussion that I write to say that he is simply wrong that 
Keynes's theory of probability in his Treatise on Probability was 
disconnected from action (unlike Knight's). Keynes began his study of 
probability because of a question in ethics raised by the moral 
philosopher G.E. Moore in his Principia Ethica. Keynes does develop a 
much more sophisticated (philosophically sophisticated) theory of 
probability than Knight (even if Keynes did abandon it later in the 
face of criticism from Frank Ramsey), but this is not because it is 
somehow about epistemology or driven by "philosophy" whereas Knight 
was just a "pragmatic economist" dealing with "actual human 
decisions". Keynes was keenly interested in human actions and their 
ethical ramifications and that's why he developed his theory of 
probability. To use Ross's vocabulary, it is just as important to 
Keynes that his theory address the needs of the  ACTOR as it is that 
it address the needs of the OBSERVER.

I have written about this in many places (as have others), but the 
easiest place to see the full ethical argument that Keynes was 
involved in trying to investigate and how he believed he had solved 
it would be in my book Keynes's Uncertain Revolution. While there is 
a healthy and robust  secondary literature on Keynes's theory  of 
probability , and my own interpretation is only one of many, I do not 
believe that there is any serious dispute about the ethical problem 
that Keynes was trying to solve when he first undertook to study 
probability or the fact that he felt that his theory had provided 
answers that would allow for more ethical ACTIONS.

Brad Bateman