Another step
We propose a modification of the Scientific Method.
Our tutor’s students in the science subjects sometimes start out their syllabus with a lesson or two on the Scientific Method, before getting down to the material they’ll be responsible for on tests. (Not all science classes; for some reason, it appears mostly in the lower-level Chemistry courses.) The Method goes something like this: the scientist gathers observations; from them, forms an hypothesis; constructs an experiment to test the hypothesis; then as a result of the experiment, modifies or rejects the hypothesis, or submits it to further testing. In Statistics class one learns that an hypothesis is never actually confirmed, but at best avoids rejection.
Of course the real practice of science is never quite so clear-cut. Some time ago our astronomer looked into the history of stellar structure, that is, how our present understanding of stars came about. He proposes another step after forming the hypothesis: looking for agreement.
To oversimplify the matter, Sir James Jeans postulated that stars were liquids (or at least made of incompressible material) powered by nuclear fission, and formed by the fragmentation of gas clouds under gravity. Sir Arthur Eddington held that stars were ideal gases or something close to that, powered by some not yet identified nuclear fusion reaction. Starting with his idea, Jeans identified objects in the sky that looked like gas clouds collapsing into stars; Eddington derived a relationship between mass and luminosity that seemed to fit the few stars for which that information had been determined. Thus encouraged, each worked out more details and sought more data.
Of course a handful of stars do not make a universe. And the objects that Jeans identified as infant solar systems are nothing of the sort, instead being vastly larger spiral galaxies. But if neither had found anything to match his ideas, they would have been discarded straight away. And discovering agreement can say much about how one’s hypothesis actually works.
The next step might be called looking for disagreement. Here one looks for places where the hypothesis doesn’t work. You can normally count on other people to do it for you. It takes a lot to work up a good theory nowadays, and you might not have time (what with filling out requests for grant support and telescope access) to do it yourself. But again the step is not as clear-cut as the Chemistry textbook depicts. Jeans calculated that ideal-gas stars powered by temperature-sensitive reactions would blow themselves up almost immediately, a result that stood for something like twenty years before it was redone correctly. For a similar time, the simplest steady-state universe was calculated to be younger than the stars in it, until a systematic error in the data was discovered. What should have been conclusive evidence wasn’t enough to kill the theories; which is good, because they turned out to be correct.
Perhaps we shouldn’t place much emphasis on the looking-for-agreement step, nor on the inconclusiveness of apparent disagreement. Humans are very inclined to look for evidence that they’re right and ignore any that they’re wrong. If anything we probably should discourage them doing it. But maybe there are scientists out there who need encouragement.
