On Being Mugged by Reality

by F.

We all pick goals to strive for. They occupy our time, distract us, and let us forget about ourselves—all good things. But sometimes we choose unreachable goals. Not a good idea, unless we can avoid being mugged by reality. But it’s surprisingly easy to avoid this mugging. This is what cranks do so expertly.

In 2002, Jennifer Kahn wrote a piece in Discover about cranks. She talked with a number of real scientists, some of whom

when they talk about cranks, evince something close to envy. “There’s curiosity, excitement, a kind of purity of purpose,” Geoff Marcy says. Unlike conspiracy theorists, science cranks inhabit a happy universe: one that’s accessible to those who plumb it (“Dear universal adventurer!” one postcard about quantum gravity begins). To read their ideas is a vicarious thrill, Arkani-Hamed admits, “but eventually you go back to what you were doing. In the end, the thing that makes science so amazing is that it works.”

It would be nice to be emancipated from the duty to be right—the duty to have your work actually work. The problem is, it’s often hard to refute cranks. Kahn gave one cranky paper on physics to a real physicist, who had typical difficulties actually refuting what was obvious nonsense:

“As I read this, I kept thinking: ‘How hard can it be to prove that this paper is incontrovertibly wrong?'” he said. “But it is hard. Not because his ideas are right. They’re not. But because he’s created a self-consistent system of arguments.”

As Kahn says,

Self-consistency is not in itself a valuable trait—the theory that aliens created Earth and continue to control its evolution is a self-consistent system—but it can make things hard to refute.

Kahn’s physicist friend continues:

“I’d love to find just one equation in here and say, ‘We have observations proving that’s not correct,'” the physicist said. “But there’s no mathematical progression. He starts with some very basic equations from classical mechanics. He mixes, stirs, spends some time hypothesizing in a very general way about physics, and out pops another familiar equation: E=mc2. But really, he’s just waved his hands. He could never have gotten to that next equation if he didn’t already know what it was—and he knew what it was only because other people had figured it out for him using the traditional framework of physics….”

“First, he’s talking about gravitational radiation, which is a real but minute effect; now he’s talking about the solar system being sensitive to small changes,” the physicist said. “It’s true that if you moved the Earth a little bit today, its position and velocity in a month would become quite different. But that doesn’t mean the shape of the current orbit is going to fall apart. We have simulations showing just the opposite, actually: that the solar system is stable over an incredibly long timescale. But that’s what I mean. Every error you find, he’s just going to change the subject. It’s never ending.”

Cranks use the same moves, over and over. After Kahn’s article was published in Discover, the magazine printed this letter—which was, I bet, representative of a larger sample:

In “Notes From a Parallel Universe,” Jennifer Kahn writes that “cranks are pretty much never right.” There are two notable exceptions. One was Robert Stirling, who invented the Stirling, or external combustion, engine. The other was Thomas Bayes, who developed the Bayes theorem, one of the most significant principles in probability theory. In particular, Bayes suffered the ridicule of mathematicians until his idea was eventually accepted.

Another favorite example of “putative crank-actual genius” is Ramanujan, whom Kahn mentions in her article:

In 1913 Ramanujan was a clerk at Madras Port Trust—”a short uncouth figure,” in the words of one contemporary, “stout, unshaven, not over clean, with one conspicuous feature: shining eyes.” Although largely self-taught in mathematics, Ramanujan had the audacity to mail 120 of his theorems to the British mathematician Godfrey Hardy at Cambridge University. Hardy dismissed the pages as gibberish at first, only to find, upon careful consideration, that some of the theorems were truly revelatory. Five years later Ramanujan was elected to the Royal Society of London.

Bayes, Stirling, Ramanujan—there are lots of folks who seemed to be cranks but turned out to be right. Usually, however, cranks are wrong, so to point to the one successful crank is like pointing to the one lottery winner and saying, “See, you said the odds of winning were one in a million, but look, he won!” The odds were still one in a million.

Note: Kahn’s article first appeared in DISCOVER Vol. 23 No. 4 (April 2002) and is cached here.