The encyclopedia · R&D & Science · Technical decision · 2002
A Bell Labs physicist faked breakthroughs across 28 papers — physics' biggest fraud
Jan Hendrik Schön published organic-electronics 'breakthroughs' in Science and Nature at a rapid rate — until identical noise in two experiments gave him away.
Bell Labs · Lucent Technologies · 2002-09-25
What happened
Between 1998 and 2001, Jan Hendrik Schön, a young physicist at Bell Labs, published a stream of papers in Science and Nature claiming breakthroughs in organic semiconductors — materials that could, he said, be made into transistors, lasers and even superconductors. At his peak in 2001 he was adding a new paper roughly every eight days. The results promised to extend Moore's law beyond silicon and were celebrated worldwide; Schön collected prizes including the Otto-Klung-Weberbank Prize and was widely tipped as a future Nobel laureate.
No one could reproduce any of it. The first cracks appeared when researchers Julia Hsu and Lynn Loo, working on an unrelated patent, noticed that Schön had duplicated some numbers. Then physicist Lydia Sohn spotted something damning: two experiments run at very different temperatures had identical 'random' noise in their graphs — something two real measurements can never share. Paul McEuen found the same duplicated noise in a third experiment. Others dug in and found more copied data across multiple papers.
Bell Labs convened an independent committee, chaired by Stanford's Malcolm Beasley, which reported in September 2002. It found that Schön kept no laboratory notebooks, that his raw data files had been erased (he blamed limited hard-drive space), and that his samples had been discarded or destroyed; his coauthors had never even watched him take a measurement. Schön was fired. Science and Nature went on to retract more than two dozen of his papers, and the University of Konstanz revoked his PhD.
Why it happened
- Bell Labs and top journals rewarded spectacular results, so extraordinary claims were celebrated before anyone demanded the raw data or replication.
- Peer reviewers trusted Schön's productivity and the prestige of the lab instead of scrutinizing methods — and no group could reproduce the work, a red flag that went unheeded.
- Schön kept no notebooks and erased his data, so there was nothing to audit until outsiders compared the graphs and found them copied.
- The pressure to sustain a legendary institution's reputation for innovation made colleagues reluctant to doubt a rising star.
The lesson
Extraordinary claims need extraordinary evidence. Peer review and a famous lab's name are not proof — demand raw data and independent replication before treating a breakthrough as real.
Aftermath
Bell Labs overhauled its research-integrity practices, and the scandal — called the biggest fraud in physics in at least fifty years — sharpened demands across science for data retention, replication and statistical checks. Schön's prizes were rescinded and his doctorate revoked after years of court battles. The case is now the canonical warning that a famous name and a flood of papers are not evidence, and that the cheapest fraud detection is simply asking for the raw data and trying to reproduce it.
Sources
spotted an error? The club wants to know.
Comments · 0
Sign in to join the comments.