(wandering on a tangent)
Before I went full neuroscientist, I spent a fair bit of time as a philosophy of science guy. Inevitably, the demarcation problem (“how do you define the difference between science and not-science?”) got a lot of attention.
There have been many failed attempts at solving the demarcation problem. Popper was first, coming up with a solution typical of a philosophising physicist: simple, elegant and wrong. Kuhn introduced history into the mix, but paid too much attention to the revolutions and too little to the day-to-day. Feyerabend was a counter-enlightenment trolley; Lakatos was a confused attempt to weld Popper’s elegance to Kuhn’s complexity.
Any simplistic fortune-cookie definition of the scientific method or “what is science?” is easily and instantly cracked by counterexamples. Science is not just one single, simple thing.
The thing that comes out of it all though, once you combine the century of theorising with the reality of history, is that Science is the Stuff That Works.
Historically, if something proves to be a sufficiently reliable method of discovering successively closer approximations of an understanding of reality, it becomes adopted into the canon of “science”. The bits that don’t manage that cumulative-accuracy trick to a sufficient degree stay in the humanities; some of those do still build on prior knowledge in a significant way (history, literature, etc.), some not so much.
The usefulness of the various humanities disciplines does seem to correlate fairly well with how much of the cumulative-knowledge trick they can pull off. I suspect at least some empirically-based input is required for it to happen to any significant degree; theology is not notably advantaged in this respect.
