Los Angeles Times
September 26 2002
Prominent Physicist Fired for Faking Data Research: Bell Labs says scientist 'recklessly' misrepresented work on microprocessors.
By CHARLES PILLER
TIMES STAFF WRITER
An influential physicist whose work in superconductivity and molecular-scale electronics seemed poised to revolutionize his field has been fired by Bell Labs for falsifying experiments over a four-year period.
A panel of scientists appointed by Bell Labs found that Jan Hendrik Schon misrepresented data 16 times, publishing identical charts to support his thesis in several scientific papers even though the experiments were different. Schon was fired Tuesday, according to a spokesperson from Bell Labs' parent company, Lucent Technologies Inc.
"It is not possible that this set of curves represent real data, and therefore, this is a clear, unambiguous case of scientific misconduct," the panel concluded. It added that Schon "did this intentionally or recklessly and without the knowledge of any of his coauthors."
In a statement included with the panel's report, Schon stopped short of acknowledging fraud.
"I made various mistakes in my scientific work, which I deeply regret," he said. "I truly believe that the reported scientific effects are real, exciting, and worth working for."
Experts say the episode probably spells the end of his once-soaring scientific career. Schon could not be reached for comment.
The case represents what some experts say is one of the most egregious examples of scientific misconduct in recent memory, and it is the first case of fabrication at Bell Labs, whose researchers have won six Nobel prizes in physics since the facility's founding 77 years ago.
The affair follows a similar case of fakery at Lawrence Berkeley National Laboratory. Last month, an internal investigation at the lab criticized a team of its own scientists for failing to verify claims by a former colleague who three years earlier said he discovered element 118--an addition to the periodic table of elements and a scientific milestone.
Those claims later proved to be fabrications, according to the lab, which issued a retraction in July. The scientist implicated, Victor Ninov, was fired earlier this year.
Both incidents have stunned the usually staid world of physics and sparked a round of soul-searching over whether science has become too competitive, too ambitious or just too careless in its rush to discover the secrets of nature.
Until recently, Schon seemed to stand as a shining example of all things prized by science--curious, rigorous and relentless.
Schon, who did his graduate studies at the University of Konstanz in Germany, looked ready to continue the brilliant legacy of discovery at Bell Labs--remarkably producing dozens of scientific papers in two years.
His work focused on the daunting problem of how to wire together minute transistors so that they could function as miniature electronic devices--a major step toward the eventual goals of creating microprocessors many times smaller and more powerful than those based on conventional silicon chip technology.
"The claims were so extraordinary that if they have proved true, [a Nobel] would not have been out of the question," said Thomas N. Theis, director of physical sciences at the IBM Watson Research Center in Yorktown Heights, N.Y. "We were all over [Schon's studies] from the very hour" they appeared, he said, because the results could not be explained by known models in physics.
"If true, they would have been revolutionary," Theis said. "We spent many months trying to replicate the results, but were unable to."
Nor could any other researchers--some of whom noticed the identical-looking data curves and contacted Murray Hills, N.J.-based Lucent with their concerns.
Many of Schon's results between 1998 and 2002 were published in Nature and Science, two of the most important scientific journals. The panel indicated that Schon's actions made it impossible for them to determine the validity of his results.
Ironically, the process of trying to validate Schon's mysterious data on molecular "monolayers"--a single layer of molecules used to create an electronic switch, or transistor--may have indirectly advanced the field.
"By checking the claimed results and finding different results, [the Schon studies] pointed the way toward further possibilities," said Theis. "It's probably a wash--some people have wasted their efforts, yet the interest in the field continues to grow rapidly."
The Lucent panel exonerated Schon's 20 collaborators on the articles in question. But the panel's chairman, Malcolm Beasley, a professor of applied physics at Stanford University, said the extent of the misconduct raises troubling questions.
"Did [Schon's collaborators] meet their scientific responsibilities?" he asked. "We struggled with that."
Collaborators in complex research rely on the personal integrity of their colleagues, but this case suggests other checks and balances might be warranted.
"We have to soul-search about this," Beasley said. He noted that rigorous participation in data analysis by all co-authors in a given study--which did not occur at Bell Labs or the Lawrence Berkeley Lab--might prevent a recurrence of such episodes. Lucent has already moved to tighten its internal processes for reviewing scientific manuscripts.
Beasley suggests that journals such as Nature and Science could use this episode as a stimulus to examine their peer-review process by which outside experts comment on submitted papers before they are accepted for publication. He urged them to rethink the balance between reasonable precautions to prevent misconduct and the free flow of scientific information.
"I don't think the peer-review process was ever designed to be a guarantor against clever fraud, and I don't think we can create one that is," said Donald Kennedy, editor of Science. But he agreed that the Schon papers are "a signal that you need to go in and examine your procedures and make something work better."
The discovery of the problems in the course of further examination reflects the self-correcting nature of the scientific process, experts agree.
"Scientific misconduct occurred. That's not a good thing," Beasley said. "But the normal processes of science worked--they ferreted it out. There's some comfort in that."
"Carl Sagan said that extraordinary claims require extraordinary evidence," Theis said. "What I see here is that extraordinary results cause extraordinary scrutiny."