Author: Calamia Joseph
Institution: Physics & Science Writing
Date: July 2005
When Bell Laboratories' star physicist Hendrik Schön started work in 1998, he quickly showed his talent at stirring up excitement. He revolutionized computer parts and confirmed hotly contested theoretical physics. Colleagues described him as "a hard working, productive scientist" and soon whispered of a Nobel Prize. But as Schön's flawless data continued to appear (and reappear) in leading science journals, and frustrated scientists failed to recreate Schön's stunning results, the physics community began to hear other, less-savory, whispers.
Hendrik Schon and coauthor Zhenan Bao researching organic transistors in 2001. Copyright © 2001 Lucent Technologies Inc.
Spurred by scientists across the nation, Bell investigated. In 2002, the lab reported its verdict: "At a minimum, Hendrik Schön showed reckless disregard for the sanctity of data in the value system of science." He lied - his Nobel dreams in quantum field theory and organic superconductors proved to be nothing more than the perfect graphs of a computer simulation.
But the 127-page report released by Bell and the flurry of articles published during the investigation hint that Schön did much more. He wasted money: funding from Lucent Technologies (Bell's parent company) and outside investors like the United States Department of Energy. He squandered time: the two years when physicists attempted to recreate his impossible results. But above all else, he hurt reputations: of the laboratories where he manufactured the props for his play, of the coauthors listed above his fabricated graphs, and of the oldest science, once known as the benchmark of objectivity.
Somehow, Schön slipped through the cracks of the "scientific method," the journal editing process, and the verification of peer-reviewed research - his trail of fake data and fried computer parts leading to serious questions about scientific standards of review. As the smoke cleared from his transformation from idol to embarrassment, he left colleagues to point fingers to find who failed at separating science from sideshow.
The Laboratory
Some wondered whether the prestigious laboratory might be to blame. In its public statements, Bell attempted to hide neither Schön's transgression nor the lab's longstanding reputation. "We are deeply disappointed that a case of scientific misconduct has occurred at Bell Labs -- the first in our 77-year history," stated Bill O'Shea, Bell's president, in a 2002 press release. "That's an enviable track record, but we take this one exception very seriously."
But despite its "enviable" record, according to some, Bell was far from stable even before the curtain rose on Schön's magic show.
"Psychologically, it was a blow to morale at the Labs because Lucent was already falling on hard times," says Judith Reppy, professor of Science and Technology Studies at Cornell University. "One might speculate that some of the pressure to have exciting, breakthrough results came from a desire to reaffirm the Labs' importance to the company."
The Coauthors
This drive for results might explain Schön's sloppiness and coauthors' acceptance of his phenomenal conclusions. When Bell's investigation team, led by Stanford's renowned physicist M.R. Beasley, reviewed lab techniques, questioned coauthors, and collected evidence, they found only a mystery of missing data books, deleted files, and destroyed lab equipment. As the circumstantial evidence piled up, the committee concluded that Schön had few hard facts, despite many exact figures. Bell blamed Schön for his "mistakes," firing him in 2002, but cleared all his collaborators. Coauthors who, despite the appearance of their names alongside Schön's in leading journals such as Science and Nature, apparently never witnessed Schön's controversial lab "techniques."
In the "Responsibilities of Coauthor's" section of their report, Bell's committee establishes a distinction between, "scientific misconduct" and "professional responsibilities" - implying that though coauthors may have acted excitedly, they did not tamper with science. Perhaps this separation of research roles and responsibility was the committee's attempt to avoid what chairperson Beasley describes in an interview as the most challenging part of their investigation, "the issue of relative responsibilities in collaborations." Or maybe, as Reppy explains, this conclusion just hints at a more minimal role of peer-review in scientific work.
"The main check doesn't come so much from peer review," she states, "but rather that an exciting result causes other scientists to try to reproduce it as a step to doing further work."
The Publication
Yet if not the lab or coauthors, what must stop distribution of erred data to the scientific community at large?
Literally, journals provide the last check before data distribution. Yet in an interview, R. Brooks Hanson, Physical Sciences Editor at Science, explains that this final check is necessarily tied to those before it. A journal's reviewers, swamped by 800 to 11,000 papers a year, must rely on the peer review that produced the original paper. Also, the editing process itself centers on the same competition as the initial scientific discovery. Journals select articles based on superlatives such as "potentially the most interesting," and it is these "interesting" articles that editors then send for review by researchers in the field.
Hanson stresses that this is only one part of a multi-leveled review system that starts with the authors ("all of them"), their department and granting agency, and finally the journal. "None of these can catch clever fabrication always, but I have seen all of them work well in some cases." Hanson also agrees that competition is the final check when it comes to new research: "What works, if these don't catch all cases, is exposure (publication) and attempts at replication by other scientists."
Schön's case uncovers a hidden process of scientific review - a process that seems driven as much by a competition to out-discover as by some abstract search for nature's mysteries. Some, such as Reppy, worry about the implications of this science contest. "If scientists are viewed as no better than used car salesmen, ready to say anything to get a sale, they have lost a lot."
References
easley, MR. Interview. 1 Mar. 2005.
Hanson, RB. Interview. 26 Feb. 2005.
Lewenstein, B. Interview. 23 Feb. 2005
Lucent Technologies. 2002. Bell Labs announces results of inquiry into research misconduct.
Retrieved 20 February 2005.
Lucent Technologies. 2001. Bell Labs scientists create record-breaking, high-temperature organic superconductor out of carbon bucky balls. Retrieved 22 February 2005.
Lucent Technologies. 2002. Executive Summary: Report of the Investigation Committee on the Possibility of Scientific Misconduct in the Work of Hendrik Schön and Coauthors. Retrieved 20 February 2005.
Lucent Technologies. 2002 Report of the Investigation Committee on the Possibility of Scientific Misconduct in the Work of Hendrik Schön and Coauthors. Retrieved 20 February 2005.
Nature. "Publication Policies." Retrieved 3 March 2005.
Reppy, J. Interview. 1 Mar. 2005.
Service, RF. 2002. Bell Labs Fires Star Physicist Found Guilty of Forging Data. Science. 30-31.
---, 2002. Pioneering Physics Papers Under Suspicion for Data Manipulation. Science. 1376-1377.
---, 2002. Winning Streak Brought Awe, And Then Doubt. Science. 34-37.
- Joseph Calamia