A Recursive Trap For Evaluation

Jerry: I’m gradually wrapping my head around your Book. Are you sure that there will be much left? I am beginning to get subversive thoughts about statistical inference, thanks to you. I wonder whether there is a recursive trap here, which Science escapes. The trap is in the evaluation; what are the criteria for evaluation, and what are the criteria for their evaluation, etc. It all gets more artefactual as you slide down the recursive helix into mathematical idiocy. Silvio and I saw it in the calculi for managing uncertain information, and so we decided to go informal early on, when it made sense, rather than N loops down the recursive helix. Now, Science escapes it in that blatant form, because Science seems to Work, at least in the short run up to now. Of course, philosophers tie themselves in knots about the validation/justification of science, but ‘normal scientists’ can do their puzzle-solving, hoping that in their lifetimes the anomalies won’t escape ‘evasion or suppression’ and The Revolution won’t affect Me.

Jason: Ah, BUT. I agree with everything you say, including your point that science escapes the problem, at least to some extent. But how do we know that our current scientific methodology works better than an alternative? To answer that question I think we have to get back into the recursive trap. And I don’t know how to get back out of it. I do know (I think) how to show that the Frequentist way of evaluating a scientific methodology is particularly unacceptable, intuitively, but I don’t know how to find out which method is best (assuming there is such a thing).

Don’t hesitate to phone me when you have great ideas like this. It’s no longer expensive to phone Australia (well, depending on your phone company … anyway, it’s cheap for me to phone you back). +61 425 263 912.