FOM: RE: Chaitin

[Raatikainen Panu A K]

On 30 Mar 01, at 10:49, Harvey Friedman wrote:

> One point of clarification. You do need that if T proves K(a) > b then K(a)
> > b is true, and that appears to be a form of soundness. However K(a) > b
> is always a Pi-0-1 sentence, and so it is just consistency, and no more.

RE: Yes, this is the crucial point - I even knew this but I had 
already forgotten it (it was 3-4 years ago when I really worked with 
these issues) - now I was just thinking of the form of the definition 
of K(x) which is apparently Sigma-0-2, and that is why I thought 
that one needs more substantial soundness assumptions ... 
Good that this issue became clear ! 

Panu Raatikainen