^zhurnaly   -   Random   -   Recent   -   Running   -   Mantra   -   Tarot   -   Help


What makes a theory good? What differentiates the most powerful ideas from the also-rans? One key factor is self-exposure to refutation.

A great hypothesis does more than just tie together disparate phenomena in a startling, beautiful, simple way. A great hypothesis also is vulnerable to disproof from a thousand directions. It makes specific predictions, in detail, without any wriggle-room to permit redefining failure as success.

General relativity, Einstein's theory of gravitation, is a good example. As Misner, Thorne, and Wheeler describe in their textbook Gravitation, relativity stands out from competing theories in that it has no free parameters, no knobs to adjust. It has just one chance to succeed. If it doesn't predict the right orbit for Mercury, the right amount of light-bending near the Sun, the right amount of time-delay for signals going to and from distant spacecraft, then relativity fails. (So far, it hasn't ... though it still could, and some unconfirmed experiments are always around to raise doubts!)

Non-scientists sometimes wonder why scientists seem so conservative. A new theory comes along, gets some coverage in the popular press, and sounds great --- why isn't it taken more seriously in Academia? Or, even more frequently, an individual amateur proposes something radically different in high-energy physics, or medicine, or cosmology --- why do all the "experts" ignore his or her letters?

The answer is multifaceted. To start with, most new theories are simply wrong. They disagree with well-understood evidence from observation and experiment. Many other novel proposals are expressed in too fuzzy a language to make specific predictions; they lack sharpness or specificity, and shouldn't even properly be called "theories".

Those few radical ideas that aren't in disagreement with the body of scientific evidence, and do predict something, usually don't offer anything new and testable. They don't make any different predictions that would expose them to refutation --- or if they do, they have a host of fudge-factors available to explain away disagreements. So they're just not "interesting" to an expert working in the field.

There are exceptions, new theories that come from outside the mainstream ... scribblings by a patent clerk in Zurich, for instance. But it takes time for such ideas to be accepted. They have to land on fertile ground, in circumstances where existing theories aren't entirely satisfactory. Then, scientists do take new ideas seriously --- until disagreeing evidence piles up and still newer ideas come along to push the previous revolutionary theory aside!

Monday, May 17, 1999 at 21:16:23 (EDT) = Datetag19990517


(correlates: 1 Comment on AsIfSoManyMinutes, AppropriateUnits, 1 Comment on Standard Model, ...)