Pitfalls of editorial miscommunicationBMJ 2003; 326 doi: http://dx.doi.org/10.1136/bmj.326.7401.1262 (Published 05 June 2003) Cite this as: BMJ 2003;326:1262
- Karen Shashok, translator and editorial consultant ()1
When Human Immunology retracted an article from a special issue, it blamed the guest editor, who was also one of the authors. But journals should not hand responsibility to someone unfamiliar with that journal's editorial procedures without written guidance or oversight
In October 2001, the publishers of Human Immunology retracted an immunogenetics paper that some readers felt contained inappropriate political content. They also deleted it from the online edition of the journal and asked librarians to physically remove the pages the article was printed on. The first author of the controversial article was also guest editor of the special issue the paper appeared in. The case triggered much debate in editorial organisations and internet discussion groups, and the guest editor, editor in chief, sponsoring society, and publisher were all criticised for their roles in the affair. This article examines the claims and counterclaims.
Encounter between science and politics
Human Immunology, a journal published by Elsevier Science and sponsored by the American Society of Histocompatibility and Immunogenetics, produced a special issue on anthropology and genetic markers in September 2001. The guest editor was Dr Antonio Arnaiz-Villena, head of the immunology department at a large public hospital in Madrid and professor of immunology and cell biology at Madrid's Complutense University. Dr Arnaiz-Villena was recruited by the journal's editor in chief, Dr Nicole Suciu-Foca, because of his expertise in the new, interdisciplinary research front that he called historic genomics.1 He was expected to contribute a keynote paper written specifically for the special issue.2
The paper attracted attention not only for its scientific content but also for parts of the text some readers felt reflected political bias. The authors reported that their analysis of human leucocyte antigen gene variability and haplotypes showed that “Jews and Palestinians came from ancient Canaanites, who extensively mixed with Egyptians, Mesopotamian, and Anatolian peoples in ancient times.” The abstract concluded that “Palestinian-Jewish rivalry is based on cultural and religious, but not genetic, differences.”2
Immediate attacks by some readers and members of the journal's editorial board led Dr Suciu-Foca and the society to contact Elsevier Science and ask it to retract the article from Medline and delete it from the online edition of the journal. Paul W Taylor, a senior publishing editor at Elsevier Science, asked all subscribers to “ignore the article in question” or “preferably, to physically remove the pages.”3 In an editorial published in the October issue of the journal, Dr Suciu-Foca and Dr Robert Lewis (of the American Society of Histocompatibility and Immunogenetics publications committee) expressed regret over publication of the article and apologised to readers.4 They apologised that the authors had “confounded the elegant analysis of the historic basis of the people of the Mediterranean Basin with a political viewpoint representing only one side of a complex political and historical issue… The Editors deplore the inappropriate use of a scientific journal for a political agenda.” The same issue contained a statement signed by the society's then president, Dr Dolly B Tyan, that condemned “the use of a scientific forum to advance any bias.”5 Dr Arnaiz-Villena was dismissed from the editorial board.
A paper in Human Immunology was retracted and removed from the online journal because of alleged political bias
The first author of the paper was the guest editor of the special issue in which it appeared
The ethics of withdrawal of an article that is considered offensive or politically biased remain unclear
The actions of the guest editor, journal editor, publisher, and the society sponsoring the journal have all been criticised
Journals need to have clear written polices for guest editors of special issues, including details of responsibility for peer review and language editing
The retraction triggered a second wave of attention. Although the editor in chief tried to blame the guest editor for the negative reactions to the article, she, the society, and the publisher also received some harsh criticism. Suppression of the article was widely reported in the press and called unprecedented,6 and the Observer described retraction of the paper as a “drastic act of self-censorship.”7 Sheldon Krimsky, writing in Nature Genetics, noted that “the decision by the journal's editorial board and publisher to expunge the record of an existing published paper raises issues for the ethics of journal publication.”8 In reply, Dr Suciu-Foca stated that withdrawal was justified because the paper “was an inflammatory political manifesto poorly disguised as a scientific work” and further charged that the guest editor had failed to “uphold the journal's integrity, quality, and impartiality.”9
In the electronic edition of the BMJ, Sam Richmond challenged the journal's sponsoring society, Dr Suciu-Foca, and Mr Taylor to “justify their unprecedented censorship of the article” and observed that an explanation for their action was needed “regardless of whether or not it is related in any way to the origin of the charges against Dr Arnaiz-Villena.”10 Another correspondent, J K Anand, noted that if the paper did not contain fake data, it was hard to understand why the journal chose to “wipe it off the records.”11
Ethics of retraction
Scott Plutchak, editor of the Journal of the Medical Library Association, criticised the decision to withdraw the paper as an attempt to tamper with the published record. He pointed out that “there was no claim of scientific inaccuracy” and questioned the copyright holder's right to change the electronic record simply because it is technically possible to erase a document.12 Erasure meant that many people who wished to study the article were no longer able to obtain it except through the grapevine.
The article was published at a time when readers' emotions were heightened by the terrorist attacks in New York and Washington, DC. It may be that Dr Suciu-Foca's decision was partly informed by these events. She was apparently under strong pressure to take expedient measures, stating that immediate action was needed to prevent “mass resignations from the editorial board and destruction of the journal.”6
What went wrong?
Translation and editing lapses
Dr Arnaiz-Villena told me he was not given specific guidance or instructions on language editing or copyediting. This is regrettable, as English was not the first language of most of the authors in the issue. Because the editor in chief never mentioned this point in her correspondence with Dr Arnaiz-Villena, he assumed that he was not responsible for language editing. It is usual for accepted manuscripts to be copyedited by the journal's in-house or freelance editors.13 Dr Suciu-Foca has not replied to inquiries about what guidance the journal gave its guest editor.
Certain words and phrases in the introduction section of the retracted paper were politically insensitive and potentially offensive to some readers. As Arnaiz-Villena readily admitted, the literal translation of cólonos as “colonists” rather than the less politically loaded word “settlers” was an error. A statement that seemed to blame Israel alone for the outbreak of war in 1948 may have been the result of a frequent (but none the less unpardonable) translation error that confuses the active and passive voice. The phrase “and started a war,” which occurred immediately after a mention of Israel, probably should have been translated as “and a war started.”
Because of the multidisciplinary nature of the study, some parts of the introduction might have seemed out of place or irrelevant in a medical journal. This, and a few unfortunate choices of word, may have led some readers to misinterpret the cultural and historical material as evidence of political bias. Kerans has published a detailed analysis of how the text may have prejudiced readers by not fulfilling their expectations of content and language.14
Although some readers were understandably upset, it is illogical to assume that Dr Arnaiz-Villena would risk harm to his own and his coauthors' reputation by intentionally offending them. Experts familiar with journal publishing and the difficulties of researchers whose first language is not English agreed that the problematic words and phrases were probably the result of inadequate translation, ineffective language revision, and inadequate (or non-existent) copyediting.14 15 Those who complained about the perceived bias in the text—including the journal's editor in chief—may have overlooked the fact that the data were collected, and the paper was written, many months before publication.
Responsibility for peer review of special issues
Peer review was a major point of disagreement.6 7 16 Dr Arnaiz-Villena told me that when he agreed to be guest editor of the special issue, he was not given specific instructions about how to select manuscripts or prepare them for publication. Although Dr Suciu-Foca has described the article as “non-peer-reviewed,” Dr Arnaiz-Villena has stated that all papers for the special issue, including his controversial article, were reviewed by two experts. 6 16 However, as he was given no explicit instructions or guidelines about the journal's usual procedures for peer review, there was potential for misunderstanding between the guest editor and the journal's regular staff.
The society has said that review policies at Human Immunology were changed immediately after the retraction to prevent the mishap from recurring. Dr Tyan was quoted as saying that the paper by Dr Arnaiz-Villena and colleagues went to press without peer review and editorial oversight because it was part of a special issue, and that “the society had changed its policies so that in the future the journal's editor-in-chief will review special issues.”16 Klarreich reported that “the incident has prompted the journal's editorial board to revise its policy so that in the future the Editor-in-Chief will supervise work by guest editors.”6 These comments imply that up until the controversy, the manuscripts accepted for special issues of Human Immunology were not always thoroughly read by the journal's in-house editors.
In March 2002, Dr Robert Bray, who succeeded Dr Tyan as the society's president, sent me a copy of the society's undated statement about the controversy (available from 17) along with a note stating that the society had no further comments on the issue. The statement admits that the historical content (but not the “social or political” commentary) in the article was appropriate but repeats the accusation that the controversial paper did not undergo “normal” peer review. According to this explanation, “Normal peer review of scientific papers prior to publication results in suggestions (sometimes strong) as to how a paper should be revised or further supported with data and recommendations by the reviewers whether to publish or reject the article. Generally an author must decide whether to accept the critiques of reviewers and appropriately edit the manuscript… or whether to submit the article elsewhere. In this particular instance this important process did not occur.” To my knowledge, this document has not been published. In the absence of further clarification from the editorial office, it is hard to avoid the conclusion that Dr Arnaiz-Villena was summarily punished by the journal with little regard for due process—such as it exists in journal publishing.
Consequences and conclusions
Journal publishing is a human process, vulnerable to human errors. When a mishap occurs the best policy is to admit to the lapse, offer apologies, and make the records available for scrutiny. Secrecy and stonewalling will only increase suspicions. Although I sent several emails to the society and the editor in chief asking them to explain their processes and offer their side of the story, neither provided any further information other than that already quoted.
All those involved in journal publishing can learn from this experience as there is no single model of peer review or editing that is appropriate for all journals. However, each editorial office should have written procedures for peer review and copyediting that are available to all editors, including guest editors. Dissemination of detailed information about editorial processes may help prevent possible errors from having serious consequences.
Dr Arnaiz-Villena continues to be an active and valued member of the international immunogenetics community. His expertise in historic genomics has not been questioned. His career seems to have survived the chain of mistaken assumptions and hasty decisions that led to the retraction of a peer reviewed article for reasons unrelated to science. But nobody has come through the controversy completely unscathed.
Chronology of Human Immunology retraction
9 September 2001
Publication of the special issue of Human Immunology guest edited by Dr Arnaiz-Villena
Editor in chief receives complaints over Dr Arnaiz-Villena's article “The origin of Palestinians and their genetic relatedness with other Mediterranean populations.”
11 September 2001
Terrorist attacks in New York City and Washington DC
Editor in chief dismisses Arnaiz-Villena from the editorial board of Human Immunology.
3 October 2001
Elsevier Science advises subscribers to ignore the article by Dr Arnaiz-Villena and colleagues.
11 October 2001
A retraction in the journal is flagged up in Medline, and the article is deleted from the online edition of the journal.
American Society of Histocompatibility and Immunogenetics composes a “statement” about the controversy, but does not make the statement public.
I thank Mary Ellen Kerans for encouragement with writing this article and Liz Wager, Stephen Lock, and Tessa Richards for helpful suggestions. Sam Richmond and Enrique Garcia Olivares provided copies of important articles.
Competing interests Competing interests: None declared.