Thursday, November 17, 2011

Is That Scientific Heretic a Genius—or a Loon?


Is That Scientific Heretic a Genius—or a Loon?

'For a profession whose product is new knowledge, science seems strangely resistant to novelty.' -John S. Dykes

The list of scientific heretics who were persecuted for their radical ideas but eventually proved right keeps getting longer. Last month, Daniel Shechtman won the Nobel Prize for the discovery of quasicrystals, having spent much of his career being told he was wrong.

"I was thrown out of my research group. They said I brought shame on them with what I was saying," he recalled, adding that the doyen of chemistry, the late Linus Pauling, had denounced the theory with the words: "There is no such thing as quasicrystals, only quasi-scientists."

The Australian medical scientist Barry Marshall, who hypothesized that a bacterial infection causes stomach ulcers, received similar treatment and was taken seriously only when he deliberately infected himself, then cured himself with antibiotics in 1984. Eventually, he too won the Nobel Prize.

Drs. Shechtman and Marshall are on a distinguished list. Galileo, Charles Darwin and Albert Einstein all had to run the gauntlet of conventional wisdom in the scientific establishment. For a profession whose very product is new knowledge, science seems strangely resistant to novelty.

In the 1840s, Ignaz Semmelweiss's lonely battle to get the medical establishment to accept that doctors were spreading childbed fever from mother to mother cost him his job and his sanity (though his prickly personality didn't help). Alec Gordon, a doctor in Aberdeen, Scotland, had failed in the same quest five decades before.

Next year will be the centenary of Alfred Wegener's theory of continental drift. By the time he died in 1930, few scientists had accepted the bizarre idea that continents could move like rafts. An especially vehement attack by the eminent evolutionary biologist George Gaylord Simpson in 1943 seemed to consign continental drift to history's waste heap. Only in the 1960s, with the discovery of plate tectonics, was Wegener rehabilitated.

I would hazard a guess that 90% of great scientists start out as heretics. The problem is that 90% of scientific heretics are talking nonsense.

For an instructive analogy, consider Meadow's Law, named after the pediatrician Roy Meadow's theory that one sudden infant death in a family is tragic, but two are suspicious and three means murder. The logical flaw here is that though it's true that the probability of more than one such death in a family is low, so is the probability of multiple murders. Likewise, it's irrational to argue from the high probability that a scientific genius was once a heretic to the conclusion that a heretic is probably a scientific genius.

After giving a lecture on scientific heresy last week, I was asked how you can tell when a scientific heretic is right rather than mad. I confessed that, as I've grown older, I've becoming more confused on this point. The problem is not just that vindicated heretics are rare, but also that the heretic who's right will be just as partisan—avidly collecting evidence to confirm his idea—as the heretic who's wrong.

Perhaps it's at least worth guessing which of today's heretics will eventually win a Nobel Prize. How about the Dane Henrik Svensmark? In 1997, he suggested that the sun's magnetic field affects the earth's climate—by shielding the atmosphere against cosmic rays, which would otherwise create or thicken clouds and thereby cool the surface. So, he reasoned, a large part of the natural fluctuations in the climate over recent millennia might reflect variation in solar activity.

Dr. Svensmark is treated as a heretic mainly because his theory is thought to hinder the effort to convince people that recent climatic variation is largely manmade, not natural, so there is a bias toward resisting his idea. That does not make it right, but some promising recent experiments at CERN (the European Organization for Nuclear Research) raise the probability that Dr. Svensmark might yet prove to be a Shechtman.

1 comment: