Thin Skins and Fraud at M.I.T.

By Philip Elmer-DeWitt

The case should have been settled nearly five years ago. That is when an obscure postdoctoral fellow at M.I.T. first charged that a celebrated scientific article signed by some of the university's leading biologists -- including Nobel laureate David Baltimore -- was based on data that had been fudged. But rather than reopen the experiment (which involved introducing foreign genes into a mouse and observing the effect on the animal's own genes), the scientists, led by Baltimore, closed ranks. The junior researcher, Irish-born Margot O'Toole, was asked to give up her place in the lab. The senior scientist accused of misconduct, a gifted Brazilian immunologist named Thereza Imanishi-Kari, went on to win a prestigious appointment at nearby Tufts University.

But the story did not end there. Seized on by some tenacious watchdogs at the National Institutes of Health, the case became a symbol of the fallibility and arrogance of modern science -- and of government attempts to police science. The affair reached a critical point last week when a preliminary NIH report of the latest investigation was leaked to the press. That draft asserts % that Imanishi-Kari faked her results and that Baltimore failed to take the allegation seriously enough.

Those conclusions come only after probes by two different NIH committees and three separate congressional hearings over the past three years. The highlight was an icy confrontation in May 1989 between Baltimore and John Dingell, the powerful chairman of the House Subcommittee on Oversight and Investigations. At the time, the scientific community rallied behind Baltimore, one of its brightest stars, calling the hearings a "witch hunt" and Dingell a "new McCarthy." Dingell called in the Secret Service, which began going over lab notebooks with the forensic equivalent of an electron microscope.

What the Secret Service found, according to the NIH draft report, was a pattern of data falsification that began before the 1986 paper was published and continued, in a clumsy effort to cover up earlier misdeeds, into the late 1980s. The report raised questions about whether some crucial experiments were ever performed at all. Faced with the evidence, Baltimore has finally moved to distance himself from the work done by Imanishi-Kari. In a statement issued from Rockefeller University, where he is now president, he acknowledged that "very serious questions" had been raised, and for the first time asked that the original paper be retracted. He left it to Imanishi-Kari -- who faces a possible cutoff of federal research funding -- to explain what went wrong.

Baltimore and his former colleagues at M.I.T. owe O'Toole an apology, if not a job. And like other scientists currently facing critical scrutiny -- including AIDS researcher Robert Gallo and cold-fusion gurus Martin Fleischmann and B. Stanley Pons -- they owe it to themselves to take a close look at their thin-skinned response. Making mistakes is part of science. But blindly denying the possibility of error goes against the heart of the scientific method. Baltimore seems to have worried more about a colleague's reputation than about the truth of a junior researcher's complaint. In the end, he damaged not just his own reputation but science's as well.

With reporting by Sam Allis/Boston and Dick Thompson/Washington