Noticed: Sleuths are starting to get credit for retractions

Nosyrevy/iStock

Pseudonymous sleuth Claire Francis has flagged thousands of papers over the years, so they rarely see something new. But an email from Frontiers about an upcoming retraction on a paper Francis originally flagged offered just that: The option to be acknowledged in the retraction notice.

After years of publishers not routinely – or even often – naming sleuths despite many asking for their often unpaid and risky work to be acknowledged, the trend of acknowledging who identified issues in papers may be gaining momentum. Frontiers is one of several publishers developing such policies, and the Committee on Publication Ethics (COPE) plans to release new guidelines in August that would recommend this practice. 

Frontiers began offering acknowledgements last year, a spokesperson for the company told Retraction Watch. “Once investigations are complete, the third party is informed of the outcome and, if a retraction is to be published, offered the option to be recognized in the notice with a standardized statement,” they said.

In the email Francis received, they were offered the option to be thanked by name, or as “the concerned reader.” (Francis chose the latter.)

A retraction published in March gave a shout-out to another sleuth familiar to our readers: “Frontiers would like to thank Alexander Magazinov for contacting the journal regarding the published article.”

While some retractions now contain this standardized statement, the publisher uses it only when the investigation originated from a third party raising concerns, the spokesperson told us.

Other publishers plan to follow suit. Both Springer Nature and Wiley representatives told us they are working on policies to acknowledge third parties across their journals. 

“There are a number of important areas that we believe need careful consideration,” Tim Kersjes, head of research integrity, resolutions, for Springer Nature told us, “including potential legal risk to those who are named in notices, permanence of records of investigations, thresholds for inclusion, an effective process for opting in or out of recognition, and a process for launching it across over three thousand journals.”

For Wiley journals, “if the concern was shared with us in confidence, we will only include the source’s name if they have provided permission, and if the journal’s editor also agrees,” a spokesperson told us. As this process varies by journal, “we are working toward a more standard approach across our portfolio.”

We asked Francis, as well as sleuths Alexander Magazinov, David Bimler, and Dorothy Bishop, for their thoughts on new policies to name third parties in retraction notices. All four favored this type of policy because it increases transparency into the process. 

“Ideally publishers should link to a PubPeer thread and contributors’ pseudonyms, if they were notified through public channels,” said Bimler, a retired perceptual psychologist from Massey University in New Zealand.

He and the others noted that giving credit should not come at the expense of offering anonymity. “It is extremely important not to disclose the third party’s name if they don’t explicitly consent to that,” said Magazinov, a software engineer in Kazakhstan. 

The sleuths emphasized that getting credit was secondary to the main goal, “decontamination of the scientific literature,” as Magazinov put it.

“It might also help forge a more positive relationship between sleuths and publishers,” said Bishop, an emeritus professor of developmental neuropsychology at the University of Oxford. These relationships have been strained in the past, partially due to policies explicitly denying sleuths credit. Thomas Tischler, a senior editor at Springer Nature, told sleuths last year that “by policy, we do not mention how the case came to us and do not acknowledge any external parties.”

The Frontiers spokesperson said the initiative “has been positively received by third parties and the broader publishing community.” 

COPE Chair Nancy Chescheir told a committee of the National Academy of Sciences in June her group would be revising its guidelines for retractions later this summer. “If concerns about the article are raised by a third party, their name could be included in the retraction notice, if relevant, with permission,” she said at the time. 

Holden Thorp, Editor-in-Chief of the Science family of journals, told the same committee that Science would prefer to acknowledge sleuths, but institutions and authors “are often very hostile to the idea of including the sleuths.” He said he was happy to hear about the new policy because “if we can fall back on a COPE guideline that allows us to put that in there, we’ll definitely start doing it.”

Sleuths aren’t the only ones receiving credit under the new guidelines. “If concerns were raised by an institutional investigation, this information should be included,” Chescheir told the National Academies committee. 

A COPE spokesperson said the updated retraction guideline is being finalized and requires trustee approval, but they expect it to launch in August. 


Like Retraction Watch? You can make a tax-deductible contribution to support our work, follow us on X or Bluesky, like us on Facebook, follow us on LinkedIn, add us to your RSS reader, or subscribe to our daily digest. If you find a retraction that’s not in our database, you can let us know here. For comments or feedback, email us at [email protected].


Processing…
Success! You're on the list.

16 thoughts on “Noticed: Sleuths are starting to get credit for retractions”

  1. Also, sleuths continue to be blamed for :
    https://www.theanalyticalscientist.com/issues/2025/articles/july/who-watches-the-watchers/
    The authors of that piece haven’t yet responded to my inquiry about their role in this article.
    Perhaps they are comfortable that errors and fraud remain in the scientific literature, in order to protect some hypothetical innocents. They don’t seem to bother to suggest a better way to clean up the crap.
    ¯\_(ツ)_/¯

    1. That article is one of the most insane ones I have read in a long time. That article seems nothing bit a tool to spread the science guardians message. Furthermore, how can Damia Barcelo not know about COIs? He was an EIC for many years, if anyone would know the policies, it would be him. It’s a shameful article and it grotesquely shows what that platform, the analytical scientist stands for.

    2. Can you share what specifically you take issue with in that piece?

      Also, they do indeed suggest means of improvement, if you had bothered to read the full article:

      > More broadly, we propose:
      > Verified accountability for all post-publication commentary, including platforms like PubPeer.
      > Strict transparency requirements for publishers – fair retraction protocols and independent appeals.
      > Institutional oversight of online integrity platforms – to distinguish critique from coordinated abuse.
      > A decisive end to anonymous mass accusations, particularly by individuals with no credentials, oversight, or transparency.
      > And most critically, a new generation of independent integrity platforms, governed by scientists, technologists, and legal experts – not influencers or mobs.

      It would be faulty to assume that any system is perfect, and I welcome critique and improvement of the post-publication peer review system.

      1. Not a single legit scientist agrees with the science guardians. It’s a platform to defend fraudsters. Damia Barcelo was the EIC for many years for many journals and he doesn’t know what a COI is? Common. The rest of the article is just pure dribble to look like a legit piece but it’s not. The science guardians are a front for conspiracy people and frauds.

        1. Thanks for not actually answering my question or providing any concrete justification for your non-specific, circular and baseless attacks.

          I genuinely want to know what you think makes science guardians bad. The fact that you can’t come up with a real reason makes it hard to take you seriously at all.

          1. Demanding transparency and mentioning legal experts in the same listing, doh? It is because of people like you “why we can’t have nice things”.

            But your post also reiterates the message from the Springer’s representative. On that note, the SLAPP law in Europe should be presumably extended and strenghtened.

          2. For some reason the RW commenting platform won’t let me respond to your comment so I’m responding to my own instead.

            You still didn’t really provide any legitimate clarification on why you think science guardians is bad. You’re handwaving about a dissonance between legal experts involvement in establishing a communications platform and the need for transparency without providing any basis for that association. And I’m not really familiar with SLAPP, but from a cursory wikipedia read it seems like this is precisely why they would need legal advice — to provide guidance on these matters.

            What Springer message are you referring to? If you’re accusing me of being a shill for big publishing, at least cite your sources!!

            If your comment is indicative of the quality of unverified post-publication peer review, then I deeply question the value it brings.

          3. Did you actually read the article? Springer is there, and you should learn, you know, how to click hyperlinks.
            Regarding SLAPPs, I am glad you disclosed your motives openly. Good luck! Many have tried and all failed, not least because scientific debates, including about research integrity, are in the very deepest core of the freedom of expression.

      2. “> Verified accountability for all post-publication commentary, including platforms like PubPeer.
        > Institutional oversight of online integrity platforms – to distinguish critique from coordinated abuse.
        > A decisive end to anonymous mass accusations, particularly by individuals with no credentials, oversight, or transparency.”

        The authors clearly do not like PubPeer. By what legal or other means could these three points possibly be accomplished and/or enforced, especially on the global, international scale required? I could point out other problems with their arguments (who gets to decide if someone has “credentials”?), but what’s the point? PubPeer and anonymous commentary (occurs on reddit and blogs, for example) are here to stay.

        1. The whole article is a front for pubpeer bashing (among other things). I was however surprised about this website, it seems that this website is less legit than I figured. Their article about how Barcelo was going to save the planet seemed odd and I figured it was just an honest mistake (naive editors of the website) but now this? I can’t phantom how this is possible unless the analytic chemist is also part of some rather obscure attempt to undermine sleuths and pubpeer etc. Barcelo saving the world: https://theanalyticalscientist.com/issues/2024/articles/dec/the-planet-protector

        2. Thanks for your well thought out reply! I definitely see your point that there are some systemic barriers regarding validating credentials! But have you heard about the Peer Community In… platform (https://peercommunityin.org/)? Although it’s not really post-publication it does a pretty decent job at enabling community members to select reviewers, and it’s worth looking into, if not as _the_ solution, then as something that can be iterated upon and improved to operate at scale.

          I also think that there’s a difference between professional and non-professional venues for review. For instance, PCI reviewers are strongly encouraged to include their names and affiliations, and those reviews have been the most helpful and well thought out comments I’ve ever received. It behooves a reviewer to be constructive and insightful in their contributions when their name is attached. As for blogs and reddit, I can’t imagine anything good comes from reddit (my personal opinion), and for blogs it really comes down to who the author is. I look at the about page to determine whether they have any legitimate expertise, and base my reception of their comments on that. So in effect, anonymous review does not exist in my experience with those platforms, and I’m ok with that!!

          And if we’re counting journal clubs and lab meetings as post-publication peer review (I certainly do!!!) then that is another great kind of non-anonymous reviewing that tends to be extremely constructive.

          Regarding anonymity being here to stay… not all comments are valuable, not all are good quality. With no much noise, it becomes very easy to just ignore the nonsense (as I’ve done with reddit). But if I’m sitting in a loud bar with my friend, I can focus on our own constructive conversation while ignoring the crowd around us as white noise. Which introduces its own problems or echo-chambers, which is why it’s imperfect. But this may be improved through systematic community-led review efforts like PCI, perhaps.

    1. I agree: it is a good move. Eventually, things like this initiative would also reduce a need for pseudonyms, and in the future people might be also rewarded from their volunteer research integrity work. New initiatives like CoARA have already taken small steps toward this direction. While I acknowledge a risk of petty feuds, there are already some “sleuths” who have done more for science than many full professors.

  2. the Frontiers publisher earns 1,471,050 CHF by accepting the papers first (without giving a chance to reject a paper) and then retraction the papers, putting bad names to reviewers and editors saying they manipulate the process. The truth is, they do not allow the reviewers and editors to reject the papers and do not allow them to say something in retract (hey, you play the whole game? Everything is under your control? What is Frontiers role in publishing? Can we retract the integrity team?). After that, just leave a bad mark to editors and reviewers that they manipulate the publishing process, who on earth is manipulating the publishing process? Should the reviewers and editors put the case to court and ask for compensation? All publishers like this should get punished. Not just say they have done good to the journal. It only exposes the journal is problematic: you should act as a good gate keeper before the paper published. There is no second chance.

  3. When Elena Vicario Orri, Head of Research Integrity at Frontiers, and her team keep retracting papers—over 500 in the last three years—it raises concerns for authors like us about the risks of submitting to journals like this. How many papers have they published in that time? The number seems alarmingly high. What does this imply? It suggests that the publisher may have serious issues.
    Honestly, journals that retract papers so frequently often lose their status in SSCI, SCIE, and Scopus. It’s a harsh reality for a journal that should be striving for integrity before accepting the papers. Just look at Hindawi—what’s going on there?

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.