Does a retraction constitute defamation of your coauthor?

I’m used to reading about cases of alleged scientific misconduct in science-focused publications and in major media outlets like the New York Times and the Boston Globe. I’ve had less occasion to read about them in law journals. But today, on the front page of the New York Law Journal, there’s an article titled “Scientists Defamation Claims Over Colleagues Efforts to Discredit Her Research Are Dismissed”. (The article is available to paid subscribers. This may be a good time to make a friend with access to a law library.)
The legal action the article describes was brought by a scientist who argued she was being defamed by a collaborator who no longer stands behind work they jointly published. The defendant says the published results are not reproducible; the plaintiff says, stop defaming me!
The judge says, your case doesn’t meet the burden to prove defamation.
From the article:

By “interjecting herself” into a public dispute, Meena Chandok is considered a “limited public figure” who bore the burden of demonstrating that her one-time colleague, Daniel F. Klessig, defamed her by acting with falsity, actual malice and “some degree of fault” when calling into question her research, which purported to demonstrate a breakthrough in plant biology with implications for human disease resistance, said District Judge Joseph M. Hood.
Ms. Chandok failed to carry that burden, Judge Hood held in granting Mr. Klessig’s motion for summary judgment of the defamation claims in Chandok v. Klessig, 5:05-cv-01076.
Ms. Chandok “willfully interjected herself in a public controversy by way of creating the very subject of the controversy, and the controversy and community are sufficiently public to invoke the constitutional protection of free speech,” wrote Judge Hood, a senior judge in the Eastern District of Kentucky sitting by designation in the Northern District of New York. “In few other spheres is the need for a free marketplace of ideas as indispensable to the very operation of the endeavor as it is to scientific research. The public good would be ill-served by the interjection of such a murky field of law.”

What went down here?
According to the article, Chandok was a scientist under Klessig’s supervision (one assumes as a postdoc or a technician or something along those lines) at the Boyce Thompson Institute for plant research (BTI) at Cornell University. In the course of work on isolating plant proteins, she reportedly discovered nitric oxide synthase activity in Varient P protein. Apparently, this discovery was a big deal to the plant biologists, and Klessig and Chandok (plus some other coauthors) reported it in a paper published in Cell.
Now, before the work was submitted to Cell, Klessig had one of his doctoral students try to verify it, and this attempt was at least good enough not to put the brakes on the manuscript submission. Later (although it’s not clear from the article whether it was after the submission of the manuscript or after its publication), Klessig directed more researchers on his team to work on duplicating Chandok’s study.
As these attempts at replication were underway, Chandok moved on from BTI to a position at University of Maryland – Baltimore. Klessig got in touch with her to let her know that the attempts at replication were not successful, and to invite her to BTI to help them figure it out. Chandok declined.
Then, according to the article:

a committee was convened at the institute to investigate the veracity of the study and whether the institute should withdraw its support of the findings. Officials from the National Science Foundation and the National Institutes of Health also were notified by Mr. Klessig of his suspicions that Ms. Chandok had committed scientific misconduct.

The article doesn’t lay out precisely what the committee concluded. Minimally, it seems to have concluded that the Varient P study couldn’t be replicated, as the Cell paper reporting this research was retracted (as was another one in the Proceedings of the National Academy of Sciences. Klessig expressed his view that the research was unreliable both in an address at a scientific conference and in emails to a number of his scientific colleagues.
On the basis of these statements about her work, Chandok sued Klessig for defamation.
Here, I imagine a lot could turn on the actual content of Klessig’s communications about this work.
Was he noting that serious efforts to duplicate the work had failed, and thus the claims in the two retracted papers must be recognized to be on shaky ground, scientifically speaking? If so, that seems like exactly what the scientific community should want to see happening in these circumstances. It’s even possible that Klessig was going beyond the normal levels of diligence in emailing scientists he had reason to believe might be attempting related work, helping them not to waste time and effort on what Klessig had become convinced was a dead end.
If Klessig were not permitted to communicate with other scientists about his concerns with these published findings, how much worse might that be for the scientific community?
Was he also “expressing suspicions that Ms. Chandok had committed scientific misconduct”? This might not be an unreasonable thing to do, especially if Klessig was identifying them as suspicions, being clear about the facts on hand to support these suspicions, and being clear about the facts on hand that might argue against these suspicions. Trouble reproducing a result does not in itself prove that the result was fraudulent. Sparse lab notebooks do not in themselves constitute scientific misconduct. A schedule (or budget) too tight to permit travel back to the BTI lab to help troubleshoot replication attempts does not mark one as a cheater. But the three together, perhaps in combination with other facts not mentioned in the New York Law Journal article, might be enough of a pattern to make a scientist nervous.
Remember here that scientists are not just working with isolated reports of scientific findings. They are also working with other scientists. To the extent that the scientific division of labor requires that they trust their fellow scientists, scientists might have an interest in sharing factual information about other scientists who have not been as accountable to their results and each other as might be hoped.
Now, if Klessig was communicating that he knew Chandok was a cheater, that would be crossing a line (at least in the absence of definitive evidence that she had fabricated, falsified, or plagiarized the work; if such definitive evidence was in the offing, I’m betting the committee investigating the matter at Cornell would have done more than simply retracting a couple of papers). But from what I can glean from this article, he didn’t do that.
So, is it defamation if someone raises questions about the goodness of a particular piece of research you conducted? In the event that other researchers who have tried to replicate that research cannot replicate it, the scientific verdict on that piece of research is that it is not good. In fact, the scientific judgment needn’t even include a judgment on the scientific knowledge and skills of the researcher whose results are in question. Weird stuff happens in the lab all the time. Artifacts can be mistaken for robust effects. Mistakes are made in setting up equipment and reagents. This is why replication is so important — other scientists help you sort out the really robust results from the glimmer of hope that turned out to be a fluke.
I reckon it’s a good job skill for a scientist to be able to distinguish her scientific integrity and personal sense of worth from whether every single result she reports ends up standing for all time.
But, to the extent that your results reflect on your scientific integrity and skills, I’d think you’d want to take an active role in assisting other scientists trying to replicate those results, whether to convince themselves that the research you reported really works or to use it as a staring point for new research. If you are firmly convinced of your results — to the point of being ready to argue that it is the scientists who cannot reproduce the results who are making a mistake — you should be ready, when asked, to answer questions, and maybe even to revisit the experimental work yourself. Otherwise, you run the risk of holding on to a possibly false conviction (that the results of the experiment are just what you reported) while ignoring mounting evidence against that conviction.
A serious scientist is supposed to be on guard against self-deception and open to new evidence, after all.
The judge in this case seems to have been working with about this picture of scientific activity, judging by his ruling:

Judge Hood rejected the contention that Mr. Klessig’s efforts to discredit her research showed actual malice sufficient to carry her defamation claims.
“Defendant reported the discrepancies and the failure to reproduce the results, as he was required to do,” the judge wrote. “He invited Plaintiff back to BTI to attempt to replicate the results herself or to more thoroughly explain how she arrived at her results, something her sparse lab notebooks failed to do. Defendant gave Plaintiff every opportunity to help explain the inability of other scientists to duplicate her work, efforts that are far from a clearly convincing showing of actual malice.”

I don’t think we can generalize from this ruling to say that erstwhile scientific collaborators never try, with actual malice, to discredit each other’s research (or to discredit each other’s scientific integrity). However, in the judge’s opinion, the facts in this case don’t support such a conclusion.
Indeed, I’m left wondering what the most effective recourse would be in a case where a former supervisor or graduate advisor or colleague with a bigger scientific reputation than one’s own was mounting a concerted effort to damage your scientific reputation. Maybe, in the event that research you did together turns out not to be so reproducible, it seems like he’s trying to pin all the blame on you. Maybe he has other nefarious reasons for wanting to undercut your scientific credibility. Maybe he’s just a jerk. If you can line up evidence that this other scientist is acting with malice and spreading untruths about you, is a defamation suit the best way to address the problem? What other options would be worth pursuing — and likely to work?
Thanks to the reader who scanned the article and emailed me a PDF!

facebooktwittergoogle_pluslinkedinmail
Posted in Current events, Ethical research, Institutional ethics, Tribe of Science.

12 Comments

  1. Sounds like the judge was schooled very clearly and convincingly by the defendant’s attorneys about how these kinds of things routinely occur in the scientific community, how they are absolutely necessary to ensure the integrity of the scientific enterprise, and how allowing a defamation suit to proceed under these facts would have a serious chilling effect on the free expression of critical opinion required for scientific progress.

  2. It’s $295/year to subscribe to the NYLJ where this report was published.
    Yet the judge is a public servant, the case law set applies to everyone, the lessons valuable for the scientific community.
    When is the law going to enter the public domain?

  3. Sounds like the right decision, though I can’t but wonder what actions, if any, Ms. (Dr?) Chandok took before suing?
    Vince: perhaps just a reference, lest we find out how he defines “fair use”.

  4. Mr. Beezer:
    The New York Law Journal is a private, commercial newspaper for lawyers and other legal professionals. The results of court proceedings themselves, including orders, opinions and decisions, are public and free. They are available at the courts they are pending in. If court determinations have broader legal import, they will often be reported in a variety of other publications, public and private, including books and electronic data bases, many of which are accessible without charge.

  5. Unfortunately UK defamation laws are notoriously plaintiff friendly. They do not provide the protection of truth. That is, in the US, if I say something that is true it is not defamation. In the UK, that doesn’t matter. That’s why there was the kerfuffle from Wakefield suing Ben Goldacre. :( The scary thing is, I think I’ve heard a few cases in the US that are ignoring the fact that statements made were true, even if they were damaging. However, those rulings should be overturned on appeal.

  6. A major difference between plants and animals is that plants have fluids with mM/L levels of nitrate (they are only microM/L in animals). The levels of NO that are important are pM/L, about 9 orders of magnitude less. It can be very difficult to remove nitrate, nitrite and other NO/NOx species from plant materials during purification sufficiently to not get artifacts. The tests that were done are not as specific and free from potential artifacts as I would like.
    When this was first reported, there was considerable controversy. I never thought it was all that credible because all plants have considerable nitrate reductase. The major source of nitrogen to plants is nitrate which they reduce to ammonia. There are lots of enzymes that have some degree of nitrate reductase activity and produce NO.
    It is good that the court threw this out.

  7. I was working at the Institute right before this all happened (this is all a matter of public record, so I don’t think I’m speaking out of turn here). About two years before that, there was a case of some fairly serious scientific misconduct by another postdoc at BTI – involving actually going into Photoshop and changing microarray data. I know both of these people and I did not see a situation between the two of them ever getting this serious. But Dan Klessig was probably erring far to the side of being open about problems with Meena’s work given that BTI had just gone through an ORI investigation recently. Having had more experience with irreproducible data/scientific misconduct than I really want to, the PI may have suspicions of scientific misconduct, but those are not actually verbalized to the wider community (the way Dan did) until an investigation is complete.

  8. It is good that the court threw this out.

    Dude, I know you are a nut, but that may be the most outlandish non sequitur I have every encountered in my entire life. It *is* good that the court threw this out, but the appropriateness of that outcome has absofuckinglutely *nothing* to do with whether the results of the original paper were actually right or wrong. This outcome is good because scientists who *reasonably believe* a result to be erroneous and/or fradulent should be completely free to publicly express that belief.

  9. CPP, I didn’t base my agreement with the court for throwing it out on whether I thought the result was correct or not. The throwing it out was the correct process. Disagreements between collaborators about scientific matters (which the validity of data is) should never end up in court.
    I just threw the stuff about NO in there to give people some background on why it would be easy to make a mistake and find something you thought was a NOS in plant stuff.

  10. Sad, but I’m not surprised to see it in court: if there really was a bad job of falsifying data, it suggests a personality that has an inadequate grasp of the likely impact and consequences of actions and may genuinely believe that they are being treated unfairly.
    In the one case of misconduct I’ve ever known to be close to home, a colleague who was expelled from grad school because of blatant invention of data and faking of images (massive PhotoShop again) later turned around and sued the school, quite possibly on similar grounds (although I think it was ‘denial of livelihood’ or some such?)

  11. Could the plantiff have won the case by proving that:
    A: Her irreproducable data was a result of sloppiness, not fraud;
    B: Her supervisor had proof (or at least strong evidence) to show that the sloppy or inappropriate techniques were the cause of the problem;
    and
    C: The supervisor described the work to colleagues as fraudulent, despite knowing that the irreproducibility was a result of insufficient skill, as opposed to malice.
    Of course, the plaintiff may not be willing to testify under oath that they are a terrible scientist, just to win this case.

Leave a Reply

Your email address will not be published. Required fields are marked *