Blog Archive

Sunday, April 6, 2014

Stephan Lewandowsky: Revisiting a Retraction

by Stephan LewandowskyProfessor, School of Experimental Psychology, University of Bristol, Shaping Tomorrow's World7 April 2014

The journal Frontiers retracted our “Recursive Fury” paper on 21 March. Frontiers withdrew Recursive Fury due to legal fears, not academic or ethical reasons. The paper—probably the most widely-read article ever published by Frontiers—can now be found at

The retraction was accompanied by the following statement:
“In the light of a small number of complaints received following publication of the original research article cited above, Frontiers carried out a detailed investigation of the academic, ethical and legal aspects of the work. 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. The authors understand this decision, while they stand by their article and regret the limitations on academic freedom which can be caused by legal factors.”
This statement was the result of negotiations between the lawyer for Frontiers and a legal representative of the authors in the U.K., and it formed part of a formal retraction agreement signed by both parties. Although we disagreed with the journal’s decision, we were provided with sufficient information to understand it. Our position on the decision was shared by officers of the Australian Psychological Society and other organizations, such as the Union of Concerned Scientists.

Although there has been considerable media attention, the authors have made few public comments since the paper was retracted. I have continued to serve as a co-editor of a forthcoming special issue of Frontiers, I accepted a reviewing assignment for that journal, and I currently have another paper in press with Frontiers. After the retraction, I was approached by several Frontiers editors and authors who were dismayed at the journal’s decision. In all instances I pointed out that I continued to serve as author, reviewer, and co-editor for Frontiers.

A few days ago, the journal released another statement about the retraction on their website. This recent statement raised several points that were new to us and that can be interpreted as a departure from the earlier, contractually-agreed retraction statement. Because of the public interest in this issue I draw attention to three issues that are most in need of disambiguation. (I defer other issues that deserve correction to future posts):

First, in its most recent statement, the journal seemed to imply that the paper was retracted because it “did not sufficiently protect the rights of the studied subjects.” This stands in contrast to the contractually-agreed retraction statement, signed by legal representatives of both parties, that Frontiers “…did not identify any issues with the academic and ethical aspects of the study.” It also sits uneasily with public statements by Frontiers’ lawyer, such as“Frontiers is concerned about solid science and it's obviously a regret when you have to retract an article that is scientifically and ethically sound…

Second, in its recent statement Frontiers also stated that it had received no (presumably legal) threats. This claim sits uneasily with the public statement of at least one individual who explicitly stated that he had threatened the journal. Moreover, another complainant publicly alleged defamation, and asserted that the journal's apparent concern with “defamation liability” was justified: Details were provided by Graham Readfearn earlier. The journal’s recent claim also sits uneasily with the contractually-agreed retraction statement, which ascribed the retraction to an “insufficiently clear” legal context. I pointed out earlier that this legal context involved English libel laws in force prior to 2014. Those laws were sufficiently notorious for their chilling effect on inconvenient speech for President Obama to sign a lawthat makes U.K. libel judgments unenforceable in the U.S.

Third, the journal revealed the existence of a new paper that we submitted in January 2014 and that according to their latest statement “did not deal adequately with the issues raised by Frontiers.”

To resolve those discrepancies between retraction-related statements requires a brief summary of events.

During a Skype conversation on 14 June 2013, representatives of Frontiers informed me that they had decided that there were no academic or ethical grounds for a retraction of Recursive Fury, but that changes might have to be made to the paper to safeguard against the legal risk of defamation. I agreed that I would “… work towards a constructive solution with you [Frontiers] to get the paper re-posted when it is ready,” even though no such risk of defamation had been identified by the relevant officers of my host institution at the time, the University of Western Australia.

On 28 August 2013 I was informed by Frontiers that their analysis of the defamation risk—under English libel laws—had found the risk to be too great for the journal to carry the article, and that it would have to be retracted. This decision was accompanied by an invitation to submit a replacement article that dealt with the issues identified in the various reviews and assessments.

We submitted a replacement article on 1/1/14, by which time English libel laws had changed significantly. It is worth considering this replacement article in some detail because it went beyond the initial Recursive Fury in the following ways:
  • Our narrative analysis was independently verified and further refined by a philosopher and a historian of science.
  • We conducted two behavioral studies with naïve and blind subjects who were not aware of the background or purpose of the study, and who responded to anonymized web content. Those studies (a) confirmed the classification of hypotheses reported in Recursive Fury and (b) showed that naïve observers rated the web content extremely high (i.e., modal response was the top end of the scale) on dimensions related to conspiracist thinking but not on an attribute relating to the quality of scholarly critique.
  • Our narrative analysis was anonymized (by paraphrasing verbatim public statements until they no longer yielded hits in Google) to prevent identification of individuals while retaining the integrity of the study.
Frontiers rejected this replacement paper on 12 February, claiming that it failed to deal adequately with the defamation issue. Our (English) legal advice clarifies that defamation cannot arise if individuals  cannot be identified in the minds of a “reasonable reader.” It must also be noted that the laws in England changed significantly on 1/1/14 to now include explicit provision for the protection of peer-reviewed science.

To sum up:

Throughout the entire period, from March 2013 until February 2014, the only concern voiced by Frontiers related to the presumed defamation risk under English libel laws. While the University of Western Australia offered to host the retracted paper at because it did not share those legal concerns, Frontiers rejected an anonymized replacement paper on the basis that non-identifiable parties might feel defamed.

No other cause was ever offered or discussed by Frontiers to justify the retraction of Recursive Fury. We are not aware of a single mention of the claim that our study “did not sufficiently protect the rights of the studied subjects” by Frontiers throughout the past year, although we are aware of their repeated explicit statements, in private and public, that the study was ethically sound.

This brings into focus several possibilities for the reconciliation of Frontier’s contradictory statements concerning the retraction:

First, one could generously propose that the phrase “did not sufficiently protect the rights of the studied subjects” is simply a synonym for “defamation risk” and that the updated statement therefore supports the contractually-agreed statement. This is possible but it puts a considerable strain on the meaning of “synonym.”

Second, one could take the most recent statement by Frontiers at face value. This has two uncomfortable implications: It would imply that the true reason for the retraction was withheld from the authors for a year. It would also imply that the journal entered into a contractual agreement about the retraction statement that misrepresented its actual position.

Third, perhaps the journal only thought of this new angle now and in its haste did not consider that it violates their contractually-agreed position.

Or there are other possibilities that we have not been able to identify.

No comments: