SCIENCE & TRUTH

Edward Edelson

"Finagle" is not a word that most people associate with science. One reason why science is so respected these days is that the image of the scientist is of one who collects data in an impartial search for truth. In any debate—over intelligence, schooling, bias, energy—the phrase "science says" usually crushes the opposition.

But scientists have long acknowledged the existence of a "finagle factor"—a tendency by many scientists to give a helpful touch to the data to produce desired results. The latest example of the finagle factor comes from Stephen Jay Gould, a Harvard biologist, who has examined the important 19th century work of Dr. Samuel George Morton.

Morton was famous in his time not only for gathering a huge collection of skulls but also for analyzing the brain size as a measure of intelligence. He concluded that whites had the largest brains, that the brains of Indians and blacks were smaller, and therefore, that whites make up a superior race.

Gould went back to Morton's original data and concluded that the results were an example of the finagle at work. "I have reanalyzed Morton's data," Gould wrote last week in the journal, Science, "and I find that they are a mixture of assumption and finagling, controlled, probably unconsciously, by his prejudiced way of ranking—his folks on top, slaves on the bottom."

Morton reached his conclusions, Gould found, by leaving out embarrassing data, using incorrect procedures, making simple arithmetical mistakes (always in his favor) and changing his criteria—again, always in favor of his argument.

Left alone, that finding would not be particularly disturbing. Morton has been thoroughly discredited by now. Scientists do not believe that brain size reflects intelligence, and Morton's brand of raw racism is out of style.

But Gould goes on to say that Morton's story is only "a noticeable example of a common problem in scientific work." Some of the leading figures in science are believed to have used the finagle factor.

One of them is Gregor Mendel, the Bohemian monk whose work is the foundation of modern genetics. The success of Mendel's work was based on finding a three-to-one ratio in the dominant and recessive characteristics of hybrid plants he was breeding. He found that ratio. But scientists recently have gone back to his data and have found that the results are literally too good to be true. Like Morton, Mendel gave himself the benefit of the doubt.

So, apparently, did Isaac Newton. Gould says that "Newton put the data together falsely to support at least three central statements that he could not prove.

And so, apparently, did Claudius Ptolemy, the Greek astronomer whose master work, the Almagest, summed up the case for a solar system that had the earth at its center. Recent studies indicate that Ptolemy either faked some key data or resorted heavily to the finagle factor.

All this is important because the finagle factor is still at work. In the saccharin controversy, for example, it was remarked that all the studies sponsored by the sugar industry found that the artificial sweetener was unsafe, while all the studies sponsored by the diet food industry found nothing wrong with saccharin.

No one suggested that the scientists were dishonest; it was just that they quite naturally had a strong tendency to find data that would support their beliefs. The same tendency is observable in almost every controversial area of science today—the fight over race and intelligence, the argument about nuclear energy, and so on.

It is only occasionally that the finagle factor turns into pure dishonesty. One example seems to be the research of Cyril Burt, the British scientist whose studies were used to support the belief that intelligence is mostly inherited. It now appears that Burt invented not only a good part of his results but also made up two co-workers whose names appear on his scientific papers.

The moral that Gould draws from his study of Morton is not that scientists are wicked but that they are just human beings, like the rest of us, and so should be subject to doubt like the rest of us. "The culprit in this tale is a naive belief that pure objectivity can be attained by human beings rooted in cultural traditions of shared belief—and a consequent failure of self-examination," Gould said.

In other words, listen to what science has to say, but never get far away from a grain of salt.