We abandoned the idea that (physical) science needs apodictic certitude centuries ago, but we transferred our desire for it to the scientific method, which is seen as giving absolute certitude so far as the only remedy for bad science is more science, and never anything extra-scientific.

All this assumes that we can discern the structure and exercise of the scientific method well enough to recognize it in controversial cases, but it’s pretty clear that we can’t. Verification and falsification don’t work well enough to, say, keep whole departments of scientists from working on string theory or even to keep the occasional scientist from dabbling in NDE’s, telepathy, or Intelligent Design. Neither criteria seems to be able to solve our on-again-off-again beliefs about whether to repressed memories or racial science are scientific. This is as it should be, since one can give verificationist and falsificationalist criteria for Acupuncture, packets of sexual virility pills sold in gas stations, or even Jack’s magic beans without suggesting that any of them are scientific. If I could run a double-blind placebo test of claims made for Jack’s magic beans, I have no doubt I would show the claims false, but verificationism and falsificationism both demand that this would simultaneously show that magic bean claims were scientific.

We can have interesting and robust conversations about whether a claim is true or false, but arguing over whether it is scientific or not is all leaves and no fruit. This is probably some corollary to Gödel’s proof, perhaps a variant of our inability to give formal criteria for what will count as a formal system.

%d bloggers like this: