With PubPeer and other online resources (such as PubMed Commons and our comment section), it’s never been easier for readers to raise public suspicions of published papers. But what should we do when we learn about potential problems in papers that are decades old, and all of the authors are deceased?
That’s the question we’re asking ourselves, after reading “Due process in the Twitter age,” this week’s editorial from Science editor Marcia McNutt.
She presents this compelling scenario:
Should there be a statute of limitation on retractions? Two panelists had experiences as editors with requests to retract papers that were published more than 50 years ago. This clearly raises the question of due process. None of the authors were alive to respond to the charges of misconduct. Only incomplete records survived regarding how decisions about those papers were made. The requests were declined. Although panelists did not come up with a fixed amount of time beyond which a paper would be too old to retract, consideration of a paper’s current influence and whether evidence exists to provide due process should weigh into the course of action in such cases.
Some rules exist — the U.S. Office of Research Integrity, for instance, will only investigate allegations of misconduct that occurred within the previous six years, with some exceptions (such as if the alleged misdeeds had a potentially major impact on public health).
Tell us what you think in the poll below. If you answer “yes,” you can post a comment explaining how long you believe the period of limitation should be.
Like Retraction Watch? Consider making a tax-deductible contribution to support our growth. You can also follow us on Twitter, like us on Facebook, add us to your RSS reader, sign up on our homepage for an email every time there’s a new post, or subscribe to our new daily digest. Click here to review our Comments Policy. For a sneak peek at what we’re working on, click here.
I vote no.
” U.S. Office of Research Integrity, for instance, will only investigate allegations of misconduct that occurred within the previous six years”.
6 years does seem a bit short as it may take a few years for problematic data to come to light. People would react negatively if degrees, professional qualifications were only valid for 6 years. There should be some symmetry between how long professional qualifications last and how far back problematic data can be investigated.
Even if somebody is dead problematic data is still problematic data. A statute of limitations would mean that we would still be stuck with what the Ancients published.
Here, here!
I refused to respond with Yes or No because I believe any systemic limitation on retractions should be conditioned on several factors other than a fixed amount of time.
I’m with Juan Reza on this. When I first saw the question, I thought there shouldn’t be any statute of limitations. However, as he notes, it really needs to be case-by-case.
Consider, if we say it is 50 years, would we expect a university to launch an investigation and potentially waste significant resources on an “old” paper for which the error/fraud hat may never be determined?
On the other had, if we find information that clearly shows an “old” publication was fraudulent, would we say it need not be retracted because it’s beyond the statue of limitations? And if a journal retracted an article “beyond its statue of limitations”, could the author sue them?
Placing a statute on limitation is purely for administrative expediency. It would only limit the access to the truth. Fraud in science should be handled like murder. There’s no statute of limitation on murder.
Here, here!
The Science article misses the point about retractions — they are NOT about punishing authors or determining if misconduct has been committed. They are about making sure readers aren’t misled. So, until there is a statute of limitations on reading or citing an article, there shouldn’t be a statute on retractions (or expressions of concern if the case isn’t conclusive). Don’t confuse journal responsibility (to clear up the literature) with institutional responsibility to investigate (and, if necessary discipline) researchers.
Here, here!
No, but with the caveat that the passage of time should not lessen the standards by which the retraction is judged. If misconduct or negligence can be *proved*, why on earth would you not retract something? There’s no value to something being a matter of scientific record if it can be proved to be suspect, just because it timed out for retraction. If it can’t be, then it can only be challenged experimentally and with fresh evidence, to accept any less would be to open the door to requests motivated by reasons “other than the pursuit of scientific rigour”.
But of more practical consequence is whether institutions should be obliged to retain lab notes, emails or similar for a period of time? Say 20 years? Without the audit trail it’s difficult, if not impossible, to show impropriety, and I’m not aware of this being something that most institutions have a good grasp on.
According to recent retractions I think that a period of 15 years is suitable. However, if all the authors are deceased, before 15 years after publication of their work, retraction should be done only for severe misconduct that could have serious consequences on future research.
A couple of examples might inform the debate.
1999, last century. Mol Cell Biol.
http://retractionwatch.com/2013/05/30/five-kato-papers-subject-to-an-expression-of-concern-plus-a-statute-of-limitations-on-correcting-the-literature/
In 2012, ASM was notified that the University of Tokyo Institute of Molecular and Cellular Biosciences is conducting an investigation of possible scientific misconduct by Dr. Kato. Among the publications under investigation for possible data manipulation are these five MCB articles. The Expression of Concern is only for papers coauthored by Dr. Kato that have been published since 2007, in accordance with the DHHS/ORI six-year limitation on research misconduct (http://ori.dhhs.gov/sites/default/files/42_cfr_parts_50_and_93_2005.pdf).
Two 1999 Mol Cell Biol papers:-
1. Mol Cell Biol. 1999 Sep;19(9):6427-40.
Functional analysis of H-Ryk, an atypical member of the receptor tyrosine kinase family.
Katso RM1, Russell RB, Ganesan TS.
Author information
1Molecular Oncology Laboratories, Imperial Cancer Research Fund, Institute of Molecular Medicine, John Radcliffe Hospital, Headington, Oxford OX3 9DS, United Kingdom.
https://pubpeer.com/publications/10454588
Figure 2B.
http://i.imgur.com/eyk9mzm.jpg
Figure 4C.
http://i.imgur.com/7FTYvCr.jpg
Figure 6A.
http://i.imgur.com/F7yORHg.jpg
Figure 6C.
http://i.imgur.com/61OIjKB.jpg
Figure 7B.
http://i.imgur.com/FasYnuC.jpg
Figure 8A.
http://i.imgur.com/vIyDKGs.jpg
2. Mol Cell Biol. 1999 Jun;19(6):4289-301.
A network of mitogen-activated protein kinases links G protein-coupled receptors to the c-jun promoter: a role for c-Jun NH2-terminal kinase, p38s, and extracellular signal-regulated kinase 5.
Marinissen MJ1, Chiariello M, Pallante M, Gutkind JS.
Author information
1Oral and Pharyngeal Cancer Branch, National Institute of Dental and Craniofacial Research, National Institutes of Health, Bethesda, Maryland 20892-4330, USA.
https://pubpeer.com/publications/10330170
Figure 2A.
http://i.imgur.com/nOb54t1.jpg
Figure 2B.
http://i.imgur.com/gEMdlbL.jpg
Figure 2C.
http://i.imgur.com/HSbehKu.jpg
Figure 2D.
http://i.imgur.com/BLgWAzM.jpg
Figure 3D.
http://http://i.imgur.com/45kIvmj.jpg/BLgWAzM.jpg
Figure 7.
http://i.imgur.com/rwkRawF.jpg
Science stands for truth, and truth knows no time. So if an author decides that – with hindsight – his work of 20, 30 or 40 years ago was flawed, he should “do the right thing” and retract it. So the retraction or EOC should stand independent of time. Whether the ‘punishment’ in cases of scientific fraud should have a statute of limitation is – as Liz Wager pointed out – a separate issue.
The world is not perfect and trying to make it better by applying principles no matter what usually does more harm than help.
Trying to get the more recent scientific world better is difficult enough. Trying to make the 50 or 100 year old scitific record better is ridiculous and simply bound to fail.
A 50 year old paper has either been proved correct by generations of new scientists by replication and building new knowledge on top of it or it has been proved wrong and it is long forgotten by now. Or it is completely unimportant and noone cares.
While most of us would agree that a 2 year old paper in nature that presents a completely wrong conclusion due to an honest scientific error should be retracted once the authors realize their error, the same paper should IMHO not be retracted if it is 100 years old. Otherwise, especially considering todays knowledge you would have to retract a lot of the old papers.
With misconduct, it’s not that different. Whether the 50 year old paper was generated by misconduct or not, you know by now what the data is worth. So what purpose does the retraction fulfill?
I very much disagree with MannyHM who claims that “Placing a statute on (sic) limitation is purely for administrative expediency”. The legal philosophy behind a statute of limitations both in the common law and the civil law systems is mainly that proving what has happend becomes increasingly difficult with time.
Other point: What do you do when the journal does not exist anymore? Not every old perished journal still has a legal successor. Who will ultimately decide on the retraction?
No, let the old stuff rest in peace. You can always discuss the time line, but a statute of limitation on retractions of max. 50 years, probably more like 20-30 seems reasonable to me.
The clock and the calendar should not be a barrier for science to pursue the truth. Let’s leave the statute of limitation defense to the cleverness of the lawyers. A scientific lie doesn’t become less of a lie with time. Like a tumor, it should be removed.
Unfortunately, MannyHM, several misconduct allegations are now heading to court and this has become a popular hunting ground for lawyers. Whether we agree or not, given the current atmosphere surrounding retractions, we should start thinking about statutes of limitation.
Thank you, my thoughts exactly!
Like so many things, “it depends.” At some point the scientific record is part of the human historical record, and the foibles of scientists should not be scrubbed even if clearly wrong or abhorrent. Think of eugenics, phrenology, and craniometry. Yet fraud is fraud. Think of Piltdown Man.
Note that the six-year time limit begins when the misconduct is first alleged, not when the misconduct occurred: “Six-year limitation. This part applies only to research misconduct occurring within six years of the date HHS or an institution receives an allegation of research misconduct. “
This is a misreading of the ORI regulation, which requires that the allegation be made (as seen by HHS/ORI) no longer than 6 years after the misconduct occurred (with some exceptions to prolong this time period, when the research is reused to benefit the perpetrator — see my comment below).
I can see no reason why there should be such a statue, so I voted no.
In practice there will of course very rarely see retractions of 40+years-old papers, since it will be almost impossible to ascertain whether there is anything was done wrong, so the realities of the business will put in a soft cutoff whether we want to or not. Let’s just keep it at that.
Another reason not to have a hard cutoff can be found if you go to the list of most highly retracted authors on this page. Several of those are cases that turned out to go back for decades once people started looking.
There’s no statute of limitation on citations. So why on earth would there be a statute of limitations on retractions?
A criminal statute of limitations for scientific fraud, sure. But definitely not a statute of limitations for retracting flawed or fraudulent literature.
Good editorial that does make you think and makes me support a statute of limitations (but maybe not for fraud), but it also posits a quesion: “Is the journal Science, and its editor, worried about the something?” Why would these people on the twitters and internets pay more attention to them? It’s not like over the last 50 years authors are rushing work and taking shortcuts to publish more in certain journals over others…
“. . . the U.S. Office of Research Integrity, for instance, will only investigate allegations of misconduct that occurred within the previous six years, with some exceptions (such as if the alleged misdeeds had a potentially major impact on public health).”
Actually, the most applied exception is (1) in ORI statute of limitations at 42 CFR 93.105:
“(a) Six-year limitation. This part applies only to research misconduct occurring within six years of the date HHS or an institution receives an allegation of research misconduct.
(b) Exceptions to the six-year limitation. Paragraph (a) of this section does not apply in the following instances: (1) Subsequent use exception. The respondent continues or renews any incident of alleged research misconduct that occurred before the six-year limitation through the citation, republication or other use for the potential benefit of the respondent of the research record that is alleged to have been fabricated, falsified, or plagiarized.”
I observed from misconduct cases I have reviewed that this exception has often been applied to the simple citation of a publication that contained some fabricated, falsified, or plagiarized figure, data, or words (citations in curriculum vitae, biosketches, methods section, or bibliography). In my view, this exception should only apply if the citation was to the actual “research record . . . fabricated, falsified, or plagiarized” — that is, referring to to actual figure, data, or words that constituted the research misconduct.
Statute of limitation on what? On, yet to be, criminalized research fraud? on misleading the public? on returning grant money? on depriving more qualified applicants of a chance or opportunity? on ill gotten awards………On precisely what? It seems that in certain cases, the statute of limitation is an assurance of impunity for some fraudsters, which is akin to an invitation to more dishonesty, plagiarism and fraud.
Yes or no? How to respond? Much common sense, much good judgment …
The paper is invented? It is a fraud? The result depends on the techniques used?
The author is dead? Date of publication?
Each of us can add even more questions.
A rough example:
Aristotle: the Sun rotates around the Earth (geocentric system). It’s true ? No.
We can retract the writings of Aristotle? We study this great philosopher at school …
“We can retract the writings of Aristotle? ”
It is generally known that Aristotle’s laws of motion are incorrect.
For a long time his ideas were accepted, but when people could eliminate the effects of friction it become obvious he was incorrect. These “laws” are in effect retracted.
Yes, in agreement with you. But here we talk about retractions. 1,500 years ago these laws were The Laws (physic phylosophy).
Science is constantly changing. What is true today, it will probably be wrong in 50 years .. In 2066 we will ask the retraction of today’s scientific work?
Science does not have time limit. What the science can do is update and correct. Mistakes must not be hidden. Fraud, but it’s always fraud (no expiry date).
Correction: Fraud, it’s always fraud (no expiry date)
Yes, and let´s retract the flogisto theory writings! Lavoisier was a softy by not demanding this in no uncertain terms!!
When I was 11 we did study the Phlogisten theory of combustion in physics at school. We learned that it could explain many observations, but that there were better theories, which could explain more. The take home message was that at first sight a theory might seem to explain observations, but that when it could not explain all observations, or if better theories came along, it need to be discarded. Lavoisier was a victim of the French Revolution so we do not know if he would have demanded the retraction of the Phlogisten theory.
” We study this great philosopher at school…”, but we do not study him as a physicist who can predict motion. Some may study his ideas of physics in comparison to the ideas of physics of others, but few take his physics in itself seriously. The fact that he is considered one of the more important philosophers inhibited the rejection of his ideas about physics. His general reputation was an obstacle to scientific progress because many thought he was correct about everything.
Knowledge evolves, and today’s firm scientific consensus may be overturned tomorrow by new findings. The papers that state the “incorrect” conclusions are part of the scientific record, and are themselves important sources of data about how knowledge has evolved in the field. For that reason, papers should not be retracted for being “wrong”, even when they have been proved so by later evidence. However, for the same reason, fraudulent papers can distort the scientific record and do harm even decades after publication – and they should be retracted. That means that simple metrics like “time passed” can’t be used to make decisions about whether papers should or shouldn’t be retracted. Instead we should use a more sophisticated metric: “It depends”
I think it’s important to note that the passage of time can make it more difficult for someone accused of misconduct to defend themselves. When people move to different institutions or careers they may no longer have access to their notebooks, records can get lost, and memories fade. It would be very easy for an angry coworker to discard your records and a year later bring you up on misconduct charges – and then what?