One month ago, a paper was retracted from Frontiers in Psychology. It was called “Recursive fury: conspiracist ideation in the blogosphere in response to research on conspiracist ideation”, from Australian psychologists Stephan Lewandowsky and colleagues.

Most retractions are valuable corrections to the literature, taking flawed science or plagiarised work out of circulation. I have myself been responsible for getting two papers retracted and I think science is a better place because they’re gone.

But the Frontiers retraction was… not such a good thing. In fact I think it sets a very dangerous precedent.

The retracted paper is still available in numerous places, e.g. here. As a point of principle I’ll send a copy to anyone on request, because I don’t believe it should have been retracted.

A paper should be retracted if new information comes to light concerning either scientific or ethical flaws. The scientific qualities of the Lewandowsky et al paper are not even in question; ethically, it has been queried but, I believe, it passes every applicable test.

The article might well be considered distasteful, but this is a matter of opinion, not ethics, as I will explain.

*

The story of Recursive Fury (RF) began with a previous paper from the same authors, Lewandowsky et al (2012). This first paper reported an online survey study and the results showed that, amongst survey respondents, belief in the idea that climate change is a myth, was correlated with belief in diverse conspiracy theories (‘conspiracist ideation’).

After the publication of Lewandowsky et al (2012), a number of blog posts appeared rejecting the conclusions of that paper. In RF, the authors argue that many of these blog posts themselves displayed the kind of conspiracist ideation that the original paper had investigated. For example, they say:

Initial attention of the blogosphere also focused on the method reported by Lewandowsky et al (2012), which stated:“Links were posted on 8 blogs (with a pro-science stance but with a diverse audience); a further 5 ‘skeptic’ (or ‘skeptic’-leaning) blogs were approached but none posted the link.”

Speculation immediately focused on the identity of the 5 ‘skeptic’ bloggers. Within short order, 25 ‘skeptical’ bloggers had come publicly forward to state that they had not been approached by the researchers. Of those… five were by individuals who [in fact had been] invited to post links to the study… Two of these bloggers had engaged in correspondence with the research assistant for further clarification.

This apparent failure to locate the ‘skeptic’ bloggers led to allegations of research misconduct by Lewandowsky et al (2012) in blog posts and comments. Those suspicions were sometimes asserted with considerable confidence; “Lew made up the ‘five skeptical blogs’ bit. That much we know”…

And the rest of the article continues in the same vein.

Why was it retracted? According to the original Frontiers retraction notice of 27th March

Frontiers carried out a detailed investigation of the academic, ethical, and legal aspects of [RF]. This investigation did not identify any issues with the academic and ethical aspects of the study. It did, however, determine that the legal context is insufficiently clear and therefore Frontiers wishes to retract the published article…

However, in a later statement Frontiers contradicted this. On the 4th of April, they said that (emphasis mine)

Frontiers came to the conclusion that it could not continue to carry the paper, which does not sufficiently protect the rights of the studied subjects. Specifically, the article categorizes the behaviour of identifiable individuals within the context of psychopathological characteristics…

One of Frontiers’ founding principles is that of authors’ rights. We take this opportunity to reassure our editors, authors and supporters that Frontiers will continue to publish – and stand by – valid research. But we also must uphold the rights and privacy of the subjects included in a study or paper.

This second statement is the focus of the rest of this post. It seems to me to involve a fundamental misunderstanding of scientific ethics. My comments also apply to Frontiers’ third statement which was very similar.

Frontier’s second statement would make perfect sense if RF had been a conventional psychology paper reporting on the results of an experiment or survey – like, say, Lewandowsky et al (2012). In that kind of a paper, it would be deeply unethical for the researchers to publish any details that allowed the subjects who’d taken part to be identified.

In a conventional psychology study the researchers have an express duty of confidentiality and of privacy vis a vis the participants, or subjects. This duty stems from the fact that the subjects have entered into an explicit contract of trust with researchers.

They have granted the researchers access to personal information (e.g. about their opinions, or other personal matters), information that they might otherwise keep private. Volunteers do this because they trust the researchers to use the data only for research purposes and to publish it in a nonidentifiable way. Clearly, the researchers have an absolute duty to honour this trust. Every code of ethics for psychological research spells this out.

This is the ethical frame of reference that most research psychologists operate in. But to apply this frame of reference to Recursive Fury is blinkered. RF just wasn’t that kind of study.

No-one entered into a relationship of trust with the authors. No-one had a right to privacy: no private information was involved. No subjects volunteered to participate in an experiment: there was no experiment. In fact, RF includes no mention of the term ‘subjects’ except in reference to other papers.

What the authors did was to take already published statements and interpret them. This makes all the difference. Once someone has published something, no-one needs any further permission in order to read it, quote it, criticize it and interpret it – within the law. Publication is consent for lawful discussion – this is an axiom of public debate, not just in science but elsewhere.

The appropriate ethical frame of reference to apply to RF, then, is not the discourse of privacy but the rules of public discourse. Frontiers’ attempt to invoke the ‘rights and privacy of psychology subjects’ is simply not appropriate.

Rather, the standard by which RF should be judged is, broadly speaking, the same standard that we apply to book reviews, peer reviews, and works of literary criticism – because this is the genre it falls into, the genre of public debate and discussion.

*

An article of the genre of RF would be unethical, had the authors, let’s say, used made-up quotes in order to make people look bad. Or, let’s say, gone beyond analyzing people’s published works and criticized their private lives; or hacked someone’s personal emails and published them without consent.

But no-one has even suggested that they did anything of the kind. No-one, in fact, has ever pointed to an applicable set of ethical guidelines that the authors of RF contravened.

Given all of this, Frontiers statement that RS “categorizes the behaviour of identifiable individuals within the context of psychopathological characteristics” is beside the point. When the ‘behavior’ in question is someone’s published work, there is nothing wrong with doing that per se. Otherwise, every book review (say) in which an author is described as being “obsessed” (say) would be unethical – even if it were accurate. The intellectual landscape would be an arid place if this kind of political correctness took hold.

I don’t think RF does psychopathologize, anyway. The paper is about ‘conspiracist ideation’ which is not a recognized psychopathological term.  At no point did Lewandowsky et al use terms like ‘mental illness’, ‘disorder’, ‘obsession’, ‘delusion’, ‘(in)sanity’ etc. The absolute closest they came to this is when they note that, according to one study, ‘paranoid ideation’ is one of the traits correlated with conspiracy beliefs.

So the ethics of RF are sound. This isn’t to say that it’s a nice paper. The paper’s tone is rather contemptuous, and I can see why it is not to everyone’s taste. It’s un-PC. I am also not saying that RF is a good paper, scientifically speaking. That’s not for me to judge.

Frontiers in Psychology would have been well within their rights to decline to publish RF in the first place, had they felt that it wasn’t suitable for their journal on the grounds of tone. But they did publish it. That was a judgment of taste.

Yet the paper is no better or worse now than it was when it was first submitted. It was tasteful enough for Frontiers in Psychology then. Why did they change their minds?

Comment Policy: Comments that do not relate to my arguments in this post will be deleted. There are plenty of other venues for general discussions of climate change and Lewandowsky et al’s papers.

Lewandowsky S, Cook J, Oberauer K, Marriott M (2013). Recursive fury: conspiracist ideation in the blogosphere in response to research on conspiracist ideation. Frontiers in Psychology, 4 PMID: 23508808