So, I've long been skeptical about the use to which statistics/probabilities are put in contexts ranging from policy decisions to personal choices.
I think the underlying worry can be boiled down to something like the following:
Every event to which you might want to assign a probability either will or will not occur. There is an important sense in which, then, any probability we assign does not and cannot properly represent a fact. It does not represent any set of real circumstances. The only real set of circumstances related in an important way to the event we are interested in is that it has not, now, come to pass and that at any particular point in the future it either will or will not have come to pass.
This kind of approach to probabilities, I think, tracks the philosophical position broadly referred to as determinism. Now, I think we have to try to put a finer point on this classification--just what kind of problem can we have with probabilities (on this basis), and what sort of philosophical commitments does it require of us?
I need to think about this for a minute.
Going to re-read an old article by Michael Dummett and see what kind of light it might shed on the issue.
Librarian, You're a grand old
12 years ago

No comments:
Post a Comment