Science Is Not a “Brooding Omnipresence in the Sky”

An article in today’s Chronicle of Higher Education is entitled “Despite Occasional Scandals, Science Can Police Itself.” The author is Alan Kraut, executive director of the Association for Psychological Science, and he describes the fallout from the revelation that Dutch social psychologist Diederik Stapel had simply fabricated research papers and the social science data upon which the papers had claimed to be based. The extent and duration of his deception is a major embarrassment to the social science world and to “science” generally.
Yet Kraut assures us that “science” can police itself. In a world where faith is often contrasted unfavorably with the reliable objectivity of “science,” Kraut’s faith is touching, but misplaced. “Science” is not a “brooding omnipresence in the sky” — to use Oliver Wendell Holmes’s contemptuous reference to mistaken views of the common law. “Science” can do nothing. Only scientists can correct mistakes. And it is true that scientists often have strong motives to identify and correct the mistakes of others — from raw jealousy to an altruistic desire to better the lot of mankind.
But it must also be recognized that along with the strong motives to correct mistakes there is also the human tendency — characteristic of scientists as it is of every human being — to resist the assertion of claims that conflict with the self-interest of the individual scientist. The history of the practice of science is replete with instances of bad ideas persisting until the scientists who benefitted from the bad ideas were replaced by other scientists with different interests.
This is not to adopt a Marxist interpretation by which scientific propositions are simply a reflection of the material interests of the scientists. But let’s not be na�ve — as I’m afraid Mr. Kraut has been — in believing that some mystical force spares scientists the unpleasant task of admitting the same kinds of mistakes that grocers, governors and athletic coaches have been forced to acknowledge.