Skip to main content
v2026.11,986 entries · CC-BY 4.0

Editorial · CASRAI · Research integrity and misconduct

Nature: Retractions of Decades-Old Papers Are on the Rise

Nature reports 358 retractions and 120 expressions of concern for papers over 20 years old, and what it means for researchers and integrity offices.

Published 14 Nov 2026· 5 minute read

On 11 November 2026, Nature published a news feature by freelance journalist Dyani Lewis (additional reporting by Miryam Naddaf) reporting that decades-old papers are increasingly being pulled from the literature. According to the feature, an analysis of Retraction Watch data found 358 retraction notices and 120 expressions of concern issued for works more than 20 years old at the time of action, and the oldest retractions now make up a larger share of the total than they did before 2010.

What the Nature feature reports

The feature, titled “Retractions of old papers are on the rise: when should bad science be culled?” (Nature 659, 286-289), says standardized retraction processes emerged in the 2000s, with the Committee on Publication Ethics (COPE) issuing its first retraction guidelines in 2004. It reports that around 30 retraction notices were published each year in the early 2000s, while more than 10,000 were recorded in 2023, and that the Retraction Watch database now holds more than 70,000 retractions and expressions of concern. It also says digitization has let journals reach further back into their archives.

Nature’s analysis of Retraction Watch data

The feature states that Retraction Watch data capture only a subset of all retractions. Within that data, it reports 358 retraction notices and 120 expressions of concern for papers or book chapters that were more than 20 years old when the notice was issued. Since 2015, the oldest retractions account for about 0.7% of all retractions, up from 0.2% before 2010. About 3% of expressions of concern relate to works older than 20 years, which the feature suggests may reflect publishers preferring that notice type when historical records are hard to investigate. The longest gap it records is 80 years: a 1923 hoax case report in the Nederlands Tijdschrift voor Geneeskunde, retracted by an editor in 2003.

The cases the feature examines

Vishwa Jit Gupta and the Himalayan geology record

The feature recounts that in August 1987 palaeontologist John Talent reported anomalies in the work of Vishwa Jit Gupta, and that investigations over the following seven years found fabrication in more than 450 publications. It reports that none of Gupta’s papers have been retracted; a 1991 evaluation of 19 of them in the Journal of the Geological Society of India called the data “incomplete bordering on disinformation” without using the word retraction. Much of his work sits in defunct periodicals or volumes not indexed by Scopus, Web of Science or Dimensions, and the feature says AI search summaries still surface his papers, giving Google’s AI overview for the Tandi Limestone as an example. The journal’s editor-in-chief, Ghulam Bhat, plans to correct the record. Springer Nature’s research integrity director, Chris Graf, did not confirm a re-examination of Gupta’s 1960s paper in Nature but said an editorial expression of concern is likely the preferred mechanism when information is limited.

Other cases

  • Barry Milborrow (2005): he sought to retract five papers from 1980-1982 whose findings later work had disproved. The Biochemical Journal retracted its two, another journal issued a correction, and the remaining two received no notice.
  • Svante Pääbo (1985): a report of DNA cloned from an Egyptian mummy was later attributed to modern contamination. Pääbo argues the paper should stand, because correcting old papers “would create more confusion than clarity.”
  • Hans Eysenck: a 2019 King’s College London investigation deemed 24 articles co-authored with Ronald Grossarth-Maticek “unsafe.” The feature reports 16 retracted, 4 with expressions of concern and 4 unflagged, and says 94 Eysenck papers overall have been retracted or flagged.
  • Caesium therapy papers: two 1984 papers in Pharmacology Biochemistry and Behavior were removed by Elsevier in April 2026 after Marcel van der Heyden raised concerns. The feature notes COPE guidance allows removal when content poses a serious health risk.

The feature also notes inconsistent journal responses to fabricated trials by Yoshihiro Sato, reported by Andrew Grey, and that the John Darsee scandal of the 1980s produced 17 retractions. Gráinne McNamara, a COPE council member, told Nature that journals increasingly ask COPE how to handle concerns about very old papers.

What it means for researchers, librarians and research-integrity offices

The points below are CASRAI’s interpretation of the reporting, not claims made in the feature.

  • Researchers: a paper’s age is not evidence that it is sound or that it has been checked. The Gupta case, as reported, shows flawed work can sit unretracted and outside major indexes while still being surfaced by automated summaries, so reading the original source and checking for notices matters more than relying on a summary.
  • Librarians and information specialists: notices may be spread across several types. The feature’s finding that expressions of concern are relatively more common for old works means checks for retraction status should also look for expressions of concern and not only formal retraction statements.
  • Research-integrity offices and editors: the Eysenck and Gupta cases show that investigation outcomes and journal action can diverge for years. Requests to correct or retract historical work are, according to COPE’s McNamara, becoming more frequent, so a documented approach for old material is worth having. Related CASRAI coverage includes COPE’s retraction guidelines, the scale of the Retraction Watch database and research on PubPeer comments that precede retractions.

Open questions

The feature frames the central tension as when historical errors should be removed and when leaving a paper in place, as Pääbo argues, is less confusing than correcting it. It does not present a single rule, and the cases it describes were handled differently by different journals.

Sources

Ask CASRAI · full access

Ask about this story

Full access: large monthly usage, up to 10 source passages behind each answer, and MCP and API access for Claude Code, Codex and other CLIs. Cancel any time.

An AI assistant for research: funder rules, compliance, methods and publishing. It links the sources behind its answers and labels answers that come from a web search.

Answers draw on CASRAI's guides and dictionary plus the federal and funder documents we index: Federal Register, Grants.gov, Regulations.gov and UKRI.

Works on this site and inside Claude, Cursor and the AI tools you already use.

Everything CASRAI publishes — this page, the dictionary, the guides and the news — stays open to read, with no account and no card.

Referenced across the research world

University of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logoUniversity of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logo
  • University of Cambridge logo
  • Columbia University logo
  • Crossref logo
  • University of Edinburgh logo
  • Harvard University logo
  • University of Oxford logo
  • Princeton University logo
  • Stanford School of Medicine logo
  • University College London logo
  • ORCID logo

View CASRAI adoption →