Quote:
Originally Posted by ASEI
Hmm - correlation replacing theoretical constructs. It's interesting, but isn't equating correlation with cause, or ignoring cause altogether what got us magical thinking in the first place?
|
I think Hume made a pretty good philosophical (yet logical) argument that there can be no fundamental difference between correlation and cause. The effort to attribute cause that is more than strict correlation is simply its own form of magical thinking, albeit a more successful one.
Quote:
|
Maybe the new magic will work a bit better than the old, being based on mountains of data, rather than a few scattered points mangled by confirmation bias - but at the end of the day, doesn't it leave you just as ignorant?
|
Yes. Unfortunately, we have no other option. Attempts to understand "why" things are as they are does nothing but generate useful pictures that help us isolate unifying elements of the correlations we observe. We are seeking two elements then: unification and reliability. What you are calling magical thinking is strong on the former, correlation analysis is strong on the latter, and good science tries to find the best of both worlds. Who was it that said magic is indistinguishable from advanced technology? (They said it the other way around, but I like it better this way.)