From The Edge:
History abounds with examples of how instinct, not data, led to discoveries. Even Einstein’s theory of relativity had to wait decades for verification, says Ian McEwan. Some science appears true because it is elegant – it is economically formulated, while seeming to explain a great deal. Despite fulmination against it from the pulpit, Darwin’s theory of natural selection gained rapid acceptance, at least by the standards of Victorian intellectual life. His proof was really an overwhelming set of examples, laid out with exacting care. A relatively simple idea made sense across a huge variety of cases and circumstances, a fact not lost on an army of Anglican vicars in country livings, who devoted their copious free time to natural history.
Steven Weinberg describes how, from 1919 onwards, various expeditions by astronomers set out to test the theory by measuring the deflection of starlight by the sun during an eclipse. Not until the availability of radio telescopy in the early Fifties were the measurements accurate enough to provide verification. For 40 years, despite a paucity of evidence, the theory was generally accepted because, in Weinberg’s phrase, it was “compellingly beautiful”.
In James Watson’s account, when Rosalind Franklin stood before the final model of the DNA molecule, she “accepted the fact that the structure was too pretty not to be true”. Nevertheless, the idea still holds firm among us laypeople that scientists do not believe what they cannot prove. At the very least, we demand of them higher standards of evidence than we expect from literary critics, journalists or priests.
More here.