Do these claims look defamatory to you?

Interesting to consider a different way of saying this that keeps you out of trouble. "The data cannot have been produced as described by Chandok"?

I mean, sure, "falsified" is a conclusion. It implies an intention to deceive and nobody can ever know with absolute assurance what the intent was. so depending on what the standard is....

within science, I think the decision as to whether one should claim falsification over error depends on whether it could have been a mistake. in some cases (detailed on these pages if I am not mistaken) there is a pretty clear case to be made that a result is very unlikely to be a mistake. in some cases it is not so clear cut, even if the data are clearly bad.

I think if there are reasonable data supporting the conclusion, regardless of whether they represent a circumstantial case, the statements cannot be defamatory. If the conclusions were based on faulty reasoning, I think they could be considered defamatory.

I donât think that any of them rise to the level of defamation but my reasoning derives from how a scientist should behave and not on the legal definition of defamation.

If you work with another scientist as a collaborator, you have an obligation to ensure that the data you generate together and publish together is correct and reproducible. If the data is not reproducible, then you have an obligation to either work until you do reproduce it, uncover the errors that made it non-reproducible, or admit that it is not reproducible and withdraw the data and conclusions from the literature as unreliable. To me, if you donât do those things, then you have committed some kind of scientific misconduct.

I think that if one decides to not cooperate with attempts to reproduce the data and the data is not reproduced, then you have forfeited the right to author the explanation that the scientific community is owed as to why the data was not reproduced. If those who did attempt to reproduce the data without input from you decide that you faked it that is a legitimate scientific conclusion. Your only recourse then is to reproduce the data and show that it wasnât faked, but no one has the obligation to work with you to do so, you forfeited that by not cooperating in the first place. You could also come up with a lame excuse that you remembered that you made a mistake, but that doesn't erase the scientific misconduct you did by not cooperating with the attempt to replicate the data.

Just so CPP doesnât flip out; this is the thought process I was using on the other thread which I left out because it was so obvious.

I think this is pretty clear, much more so than I expected from the previous post: if false, the statements would in my opinion clearly be sufficient to constitute defamation. So the only question is what level of evidence would allow for a 'truth' defense.

[I don't think that (as daedalus suggests) there's any *obligation* to cooperate with attempts to reproduce your work, although I agree that failing to do so is incompatible with scientific inquiry; so I don't see that as providing a basis for defense. I agree that it looks fishy, that it's not what one would want from a colleague or fellow scientist - but still not enough to be evidence of fakery.]

Level of evidence? Enough to go to whatever process the journal or institution has in place. And yeah, I'd be careful to say 'the conclusion drawn by panel [X] was that the data appear to be falsified' instead of making statements that claim such falsification to be a fact. Failure to reproduce is **not** enough to justify any allegation of malpractice; discovery that the claimed results from 108 rats came during a period when only 6 rats were ordered and used would likely be such: that is, clear evidence that something was impossible to be true, rather than a difference in experimental result that could have come from chance, however unlikely.

I don't think that just saying that you think that data is "shaky" and "unreliable" - which seems to be the terms used to people outside the institute - could reasonably be called defamation. It's not going to help your reputation much at all, but that's a judgment call about work quality.

"Misconduct" and "fabrication" are a bit more serious, especially when they're used in the scientific community. Those are accusations that can, if made publicly, wreck careers. I could see circumstances where using those terms carelessly could be considered defamation - if and only if the person making the claim knows that there is in fact no evidence of misconduct/fabrication.

IANAL

"The conclusion we draw from these results is that most or all of the data that Meena presented to us and in the Cell paper concerning the recombinant varP has been falsified."

but this seems pretty defamatory to me, no ifs, buts or maybes, he's saying pretty clearly that she falsified data and for an academic that's pretty much the 'kiss of death' in your career.

Assuming of course that she didn't falsify the data.

he's saying pretty clearly that she falsified data and for an academic that's pretty much the 'kiss of death' in your career

I want to live in your world.

IANAL, and have never played one on TV. But I have seen some guidelines for English (& Welsh) defamation. Going on that, it is not clear that the public statements (15 and 19) even start to be defamatory: nobody is mentioned. Things might be a bit different if some of the audience would have known who had carried out the experiment.

Assuming that Chandok could be identified from the public statements, they could be interpreted as "Chandok carried out some experiments, and the results were unreliable". Assuming the methods are robust, this implies that Chandok is at fault, and hence is defamatory (according to the guidelines I'm looking at).

I assume Klessig's defence would be the justification defence - i.e. that was was said was true. The first part of statement 15 is easy to verify, the second would need the additional data to be presented in court: but here Klessig would just have to demonstrate unreliability (not fraud). This would then cover statement 19. I guess then that the public pronouncements would be fine.

I was under the impression that defamation covered public statements, and none of the other statements are. OTOH, does FoI make them de facto public? Dunno.

Anyway, as with the public statements, the other statements would make one think worse of Chandok, and hence would be defamatory. The defence would be the same (i.e. truth), but I'm not sure if he would have to demonstrate that the evidence indicated that some of the data were falsified (that would cover statements 1-8), or that the data were falsified. I think statements 12-14 might have to be justified by showing that the data were fabricated, it may not be enough to show that there is good evidence (but IANAL, and I don't have any experience).

Statements 21 to 23 are clear: Klessig (and Coughlin in 22) would have to provide the evidence of fabrication.

Overall, it's clear to me that the statements are defamatory, according to the guidelines I've seen. But he can only be successfully sued if he can't show evidence of fabrication he's fine.

To demonstrate his claims, I'd be expecting Klessig to show his evidence, firstly that the results were extremely unlikely, but then either direct evidence of fraud (e.g. notes in Excel sheets say "hahaha this will show them"), or strong evidence that they could not have been produced by accident, e.g. through mis-reading a machine's output.

In addition to being factually false, don't you have to prove they were stated when known to be false (that is with maliciousness)?

Some of the statements are definitely defamatory if untrue, given the legal definitions in use.You can't just accuse people of faking data without evidence.

I do not know what evidence he had for his statements. I would allow a margin of appreciation for him however. As long as he had some evidence, even circumstancial I would also have dismissed the charges. It seems to me the judge got it right, good for him.

By the way, can you post the ruling? There is no copyright in federal court opinions.

Most of them you can't really tell, as they're outside their context. For example, #1 "Unfortunately, the evidence still argues that she falsified at least some of the data on the recombinant varP."

That's not an entire email. The evidence referenced isn't presented as part of the quote, but I would be surprised if there wasn't some in that email. If he emailed Brian Crane and laid out a bunch of evidence... and then ended with that sentence, that's not defamation even if Brian (or anyone else reading that email) would disagree with the conclusion based upon the evidence: Klessig's conclusion can be wrong and not defamatory if he's including that evidence in his correspondence (Klessig's summarization is an his opinion), as it gives Crane the opportunity to evaluate the evidence himself.

It's perfectly okay (at least under the legal terms of defamation) to accuse someone of something with pretty flimsy evidence.

Usual disclaimer: I'm not a lawyer.

It seems to me that the case will hinge on this one:

Statement 21: To Dr. Blissard and Dr. Wang, via e-mail
"Therefore, her results with recombinant AtvarP protein made in baculovirus expression system had to be falsified because she could not have made the protein." Ex. 109.

That's a pretty simple summary of the reasoning behind his belief that the results were falsified. In order to show defamation, the plaintiff would have to show that Klessig is *wrong* (that the protein can be made) and that Klessig knew that he was wrong.

I don't think there's much of a leg here, personally.

One of the elements of a successful defamation claim is the following:

If the defamatory matter is of public concern, fault amounting at least to negligence on the part of the publisher

Since the reliability of scientific data is absolutely a matter of public concern, Klessig could have been absolutely wrong in his falsification assertions, yet so long as he was not negligent in making those assertions, there is no defamation. Since the plaintiff presented no evidence that Klessig was negligent as to the truth of his falsification assertions, the judge correctly entered summary judgment for the defendant.

This was an open-and-shut case.

In order to show defamation, the plaintiff would have to show that Klessig is *wrong* (that the protein can be made) and that Klessig knew that he was wrong.

This is not quite right. Because the plaintiff would only have to show that Klessig was negligent as to the truth of his false assertion, it would not be necessary to show that Klessig knew he was wrong; a showing that Klessig reasonably should have known would be sufficient.

The legal issue at play here is not whether these statements are defamatory (i.e., harm her reputation); at least enough of them are that it's not a legal question that determines the outcome of the case. The issue is whether her supervisor (and/or other co-defendants) can be forced to pay her damages. Truth is a defense: it doesn't change a defamatory statement into a non-defamatory statement; it just lets you avoid having to pay the plaintiff. It may seem like a nit-picky point, but it's hard to follow a legal opinion in a defamation case unless you understand it.

IIRC, there are 3 types of plaintiff in a defamation case: a private person, a public figure, and a "limited public figure", and your status as a plaintiff determines what level of mal-conduct you have to prove the defendant engaged in to win damages.

For a private person, the threshold is pretty low: even a good-faith belief the statement was true doesn't save the defendant; the plaintiff just has to prove you knew they are "defamatory", and that you "published" them. And it doesn't matter what the topic was.

For a public person, the plaintiff generally has to prove the defendant knew the statement was false, or was so reckless about his/her knowledge that it was grossly reckless for them not to have known they were false (that is essentially what is meant by the legal term "actual malice"; it's not a question of whether the defendant intended to harm the plaintiff). And it doesn't matter what the topic is about. Bill Clinton is a public figure (as are most all politicians, celebrities, etc), and for the most part you can probably get away with just about any claim about him. (And again, it's not that you can't make defamatory statements about him, he just has an enormous burden of proof to get you to have to pay him).

Finally, there is the "limited public figure", which roughly means that they are a private figure with respect to most issues/matters, but because of their involvement in a "matter of public concern" becomes a public figure for matters relating to this issue. In litigating these types of cases, the key plaintiff goal is to be declared a private person, and the key defense goal is to have the plaintiff declared a limited public figure. (because whether you win or not usually hinges on their status). Usually in matters where you are not the topic, you can't (in theory!) become a limited public figure involuntarily; a will court look for evidence that you "interjected" yourself into the controversy. For example, if you are pro-choice, you would ordinarily be a private person with respect to defamation involving that issue, but if you attended a pro choice rally and had a big sign and yelled loudly to the cameras, then you probably became a limited public figure for that topic (but not, for example, on the issue of whether you are burglar).

I'm not an expert on this area of law, and didn't read the opinion. But looks to me like the court declared the plaintiff to be a limited public figure merely because she "interjected" herself into the controversy by doing the experiments and co-authoring the paper. I not sure I'm comfortable about that, because it seems to me the issue at play is whether she falsified the data, not whether the data and the findings are true or not. But these types of cases are always very tricky, and people have been found to have "interjected" themselves into a controversy by doing much less.

@ Comrade PhysioProf

> It would not be necessary to show that Klessig
> knew he was wrong; a showing that Klessig
> reasonably should have known would be sufficient.

Yes, that's a valid point. My misstatement.

I'm not an expert on this area of law, and didn't read the opinion.

Don't let that stop you from blithering idiotically!

One must assume that Klessig had the benefit of examining the raw data and protocols of Chandok's experiments. His assertion of possible falcification is probably based, at least in part, on examining the protocols and the raw data. Ms. Chandok's refusal to engage in attempts to reproduce the data indicates, at least in my mind, that she knew what would be the outcome of such attempts. Her lawsuit claiming defamation is a typical reaction by cheaters who caught cheating. Her best chance to rehabilitate her name was to cooperate with Klessig on reproducing the data. Her refusal is an admission of guilt.

Me: "I'm not an expert on this area of law, and didn't read the opinion."

What I meant by saying I'm "not an expert in this area" is that it is not an area I practice in, so don't steep myself in the subtle nuances of these types of cases on a regular basis. I was basing my comments on my recollection of basic tort law and additional courses in con law and first amendment law. And I didn't read the opinion because I could not find it in a (admittedly non exhaustive) search, so was relying on the excerpts posted in the preceding thread.

PhysioProf: "Don't let that stop you from blithering idiotically!"

Well, since I gather that, by saying that, you are an expert in this area of law and did read the opinion, perhaps you can post a link to it for me. It might help me make sense of what you said, particularly that this is an open and shut case. I am curious about the source for your clear claim that the mere fact that it is a matter of public concern alone determines the criteria by which the defendants actions will be judged, regardless of any facts about the plaintiff. This is a new (and radical) development in the law in this area, overturning a reasonably well settled principle. Since (I gather) that is now the law, I can see where what I wrote would be irrelevant to the discussion of this issue.

Thanks in advance to the reference to the opinion in this case and to the SCOTUS opinion that changed the law in this area.

Divalent,

Eventually, you are new to this sphere of science blogs and aren't familiar with our resident expert in everything. ;)

BTW, do not hold your breath, waiting for an upcoming source on which CPP's claim is based.

Thanks for the heads up, S. Rivlin. I'll take that into account in the future.

But to the extent maybe others find my post a bit blathering, I'll be a bit more succinct here to summarize:
- with respect to the question posed in the title of this thread, the answer is (for most of the statements) a clear "yes": he did defame her. (truth or falsity only affects whether the plaintiff or defendant can prevail)
- the state of the evidence is probably that, on the one hand, the defendant can't prove absolutely that the allegations are true (so would prefer to avoid litigating the issue), but OTOH the plaintiff can't prove he knew, or should have known, that they were false.
- Thus, plaintiff has a chance of winning if she is a "private figure" but clearly loses if she is a limited public figure. Thus (further), although I don't have a copy of the opinion, I'll bet a lot that the key inquiry (and the legally interesting one) is: what was her status, and why?

BTW, the other question posed in above is "What can scientists say to each other about such matters?" is also interesting. It might not be wise to take the result of this case (if it survives appeal) as generally applicable to all situations. It may be a different outcome if the papers were not published and retracted: but rather was a series of incidents at a much earlier stage of the research. Like: suppose the PI suspected, on the basis of the same evidence, that falsification occurred, but the evidence surfaced well before a paper is even written (i.e., when it was just an "internal matter", and was being handled as such). Would that have been a "matter of public concern" and would a post doc have "interjected" themselves into the controversy in a way that made them a limited public figure? (That's why it would be interesting to read the court's analysis of her status.)

And what if the evidence was less conclusive on whether someone falsified data, but was strong enough to convince the PI that the person could not be trusted and so was no longer wanted as an employee? What could you say to others then? In other words, when is the person you are speaking about a private person and when do they become a limited public figure? Because that distinction determines what statements you can be held liable for.

I'm not sure that the distinction between private and limited public persons is that complicated here -- the person in question was a co-author on several published papers. Given that the validity of those papers is the matter in question, the fact that she put her name on it and published it should make it clear that she belongs in the second class.

By the way, Divalent, what country do you practice in? I ask because we've heard in several places on these blogs about the difference between English and American law on these sort of cases.

"the person in question was a co-author on several published papers. Given that the validity of those papers is the matter in question, the fact that she put her name on it and published it should make it clear that she belongs in the second class."

That is one way to isolate the issue. Another way (and IMO, the correct way) is that the matter in question is whether the plaintiff committed scientific misconduct. Remember, the comments are not defamatory because they merely question the validity of the results, they are defamatory because they clearly imply she fabricated the data.

(BTW, all my comments refer to US law.) In any event, I don't think it really makes a difference. The standard for what makes a limited public person out of a private person *requires* that the private person take *voluntary* actions that change their status *after* the controversy arises. (Calling a press conference to dispute the claims will do it, but merely saying "I didn't do it" to a reporter who asks her won't do it.) That's why I am uncomfortable with the idea (which the court apparently applied here) that merely being a coauthor of a scientific paper is sufficient.

Even for the innocent, it's pretty hard to escape unscathed from a charge of fabricating data. I'm not so sure the mere fact that you publish a paper should be sufficient to allow anyone to, with impunity, accuse you of what is arguably the worst crime in science with as little as a few unreliable bits of evidence and a cockamamie hypothesis to tie them together.

Divalent, I gather you are not a scientist either. You are really distorting the facts of the case and extrapolating it into regions that make no sense.

It wasnât just âanyoneâ who raised these issues, it was a co-author. He didnât do it on the basis of a âfew unreliable bits of informationâ, he did it following a conserted effort to replicate the results, an effort to which the plaintiff did not cooperate or contribute. It was not a âcockamamieâ hypothesis that she fabricateed the data, it was a reasonable hypothesis which explained both the lack of replication of the data and why she was unable and unwilling to cooperate in its replication.

The âcontraversyâ arises when the paper is submitted to a journal. If you donât want to be considered a public figure with respect to data you publish in a scientific journal, donât publish it in a scientific journal. If you do, expect to be held to the same standards as every other scientist and if your data isnât replicable, and if you donât have records, and donât cooperate with trying to replicate or understand why it isnât replicable, donât go crying to the court that your co-authors are defaming you when they communicate with other scientists about what could have gone wrong and why your data is unreliable.

The incremental damage to her scientific reputation because her co-author raised the possiblity of data fabrication is tiny, de minimus. The big damage comes from her not cooperating in trying to replicate the results and then suing for defamation instead of behaving as a scientist. In contrast, if her co-author did not raise the possibility of data fabrication when he considered it a possibility would have damaged his scientific reputation. As co-author, he is in the best position to evaluate the reasons why the data was not replicable. If he has reasonable suspicions that the data was fabricated, then the scientific community wants and expects him to voice those suspicions, and will fault him if he doesn't. Non-experts in the field donât have the background to evaluate the data, they rely on experts in the field to do so. Inhibiting the free exchange of scientific ideas by threats of legal action would greatly damage the scientific enterprise.

Divalent, I agree that someone shouldn't be allowed to throw a charge of data falsification around without being able to back it up. This has happened too many times in the past, actually. On the other hand, as daedalus2u pointed out, the person who is making the claim was also a co-author. Thus, he had an ethical obligation to retract the paper if he felt that it wasn't reliable. In order to do that, he had to explain why.
I also find it hard to understand why an attorney thinks it's appropriate to conclude, without hearing any testimony, what level of evidence there was. Assuming that the statements in question accurately depict the situation regarding further testing, we have unspecified "discrepancies" to deal with -- the sort of thing that will surely come out in court. It is true that simple failure to replicate the data is not sufficient to indicate that she falsified the data -- but it seems that all of the early statements (assuming that they're in chronological order) state that the failure to replicate instigated a further investigation -- not a definite conclusion. That appears to have come later.

psweet and daedalus2u, I was doing what the court did regarding the evidence when ruling on a motion for summary judgment (that is, before there is even a trial). In that situation the court *has* to assume that the plaintiff will be able to support her claims at trial (since the motion he is ruling on would deny her the right to a trial). Essentially, the judge said it makes no difference: even if the facts are as the plaintiff claims, SHE STILL LOSES! That is, even if she didn't falsify the data, and can prove that at trial, she cannot win her suit against the defendant(s). Because she is a limited public figure.

My postings on this are not related to the question of whether or not she falsified the data in the papers. It is about the legally interesting issue of whether or not merely doing research and being an author on a paper (even a "minor" author, I wonder?) makes you a limited public figure.