Wednesday, February 3, 2010
What I mean by provisionality is the idea that nothing in science is final. Scientific explanations depend on evidence and must be open to their own falsification as new evidence is collected. As I before said, if the new contradicts the old, then the old must be revised to reflect the new. As a result, even the most solid scientific conclusions are never more than "the evidence shows X". In the end, every conclusion is open to revision should the evidence demand it.
I think this little account, however commonplace, of provisionality is an important part of what makes science distinctive and made its advent revolutionary . And I think so too does the account, however commonplace, of science as a whole that I briefly sketched, though it's not original with me.
I don’t mind characterizing the measuring of scientific truth by the phrase “what works”. What I am having trouble with is understanding why that characterization is profoundly different from the way one might typically think about successful scientific findings.
When we say a scientific finding or conclusion is true it’s because that truth corresponds to the world. We can measure any specific correspondence or causal explanation between 2 phenomena by testing for them, testing for any possibility of their lack of explanatory power. Saying scientific proposition x is true because it corresponds to out there is not the same thing—does not land us in the same boat—as saying x is true because it corresponds to God’s will. These two statements seem incomparable.
We know nothing, can know nothing, about God’s will, but we know about out there; and what science exactly does is attempt to verify according to its protocols what out there is. I don’t understand how it’s new or radically consequential to call that verifying “what works.” How is “what works”, in other words, different from what science traditionally says it needs to do to, to test hypotheses?