In his commentary on John Rigden's new book Hydrogen: The Essential Element (New York Times Book Review, 14 July 2002), Lawrence Krauss makes a wonderful distinction between the "... speculative musings dominating the physics sections in bookstores ..." and the way science actually advances --- from complexity to simplicity and back again, via deep analyses of extraordinarily accurate observations.
Take the Lamb Shift (please!), or the hyperfine structure of spectral lines, or the Mössbauer Effect, or any of a myriad other parts-per-million (or better) tests of the laws of Nature. That's where the real action is ... not in handwaving about fuzzy notions that aren't ever subject to refutation.
The simplest of atoms offers a lovely archetypal example. As Krauss writes:
[Rigden] demonstrates elegantly midway through Hydrogen --- when one senses his own interest in the subject really begins to peak --- how minute disagreements between theory and experiment, which otherwise would have been completely ignored, had to be taken seriously precisely because of the underlying simplicity of the hydrogen atom itself. Indeed, at a time when many books and news reports describe speculative theories that hope to probe deep cosmic mysteries but so far have failed to touch base with a single observation or experiment, it is a pleasant change to find a book on a humble topic that demonstrates the remarkable beauty and subtlety of nature, and of the experiments scientists have developed to explore it.
Precisely! And that's why the old canard "All science is either physics or stamp collecting" is so unfair. Good science begins with stamp collecting --- the meticulous accumulation of detailed knowledge. (And philately is technically just a branch of numismatics, as a numis-bigot friend (SK) once noted at a coin club meeting. (^_^))
(see also VulnerableTheories (17 May 1999), OnSilence (30 Dec 1999), ExposureAndEncapsulation (7 Jan 2000), WebsOfEvidence (15 Feb 2000), ScienceVersusStampCollecting (20 Jun2000), ScienceAndPseudoscience (6 Oct 2001), UniversalKnowns (13 June 2002), ...)
TopicScience - Datetag20020716