We have received an elegant letter from some US authors employed by a government agency who have explored the association between voting for George Bush and smoking cessation throughout the USA. This follows a research tradition such as in:

Shaw M, Dorling D, Smith GD. Mortality and political climate: how suicide rates have risen during periods of Conservative government, 1901-2000. J Epidemiol Community Health. 2002;56:723-725.

Page A, Morrell S, Taylor R. Suicide and political regime in New South Wales and Australia during the 20th century. J Epidemiol Community Health. 2002;56:766-772.

The authors wish to remain anonymous, fearing being disciplined by their employer. George Lundberg (http://bmj.bmjjournals.com/cgi/content/full/318/7180/394) would know a thing or two about that. What do other editors think about allowing the possibility of anonymous authorship, which I noticed occasionally occurs in serious newspapers when letter or op-ed writers have reason to fear repercussions. Does anyone have policy or guidelines on this?

Simon Chapman
Editor, Tobacco Control (www.tobaccocontrol.com)

___________________________

In the past, anonymous authorship was supposed to confer divine attributes upon the work. We no longer believe that, because we have come to realize that authorship without accountability is meaningless. Indeed, anonymity is the hallmark of those who write obscene and threatening messages.

While there might conceivably be a reason if the authors lived in a country where they would suffer death or imprisonment, but that is not the case in the US. These authors do not have to publish. Indeed, unless they, the authors, think their work is important, they should not. And if they do think it important, they should be prepared to stand behind it with their names. That's what we all have to do, and the reason is that it allows the reader to respect and believe what we say.

There's another reason why you should not allow their anonymity. There are notable examples of government scientists in the US taking principled stances that are unpopular with their bosses. Their heroic example would be diminished if the present authors crept into print under cover.

Later this week, the Turkish writer Orhan Pamuk, goes on trial in Istanbul for "publicly denigrating Turkish identity." That hasn't stopped him signing the first article appearing in this week's (December 19th) New Yorker. A stand-up sort of guy. I am impressed and moved by that fact.

Drummond Rennie

___________________________

Thanks, Drummond. I would expect nothing less from you than your eloquent response. However, can I put it in a different perspective? Celebrity old crusts like yourself are rather bullet-proof. If someone had the gall to sack you for political incorrectness of the sort we are talking about, you would be snapped up as a prestigious editor and doubtless receive the same sort of international valedictory shower & cries of "shame!" that George Lundberg received when he was marched. However, many authors are far more mortal and fear retribution—the outstanding reason given by younger, ground floor people refusing to sign their reviews.

The Economist, unless I am wrong (since I tend to only read it at the dentist), is entirely anonymous. Is it unaccountable? What about in camera evidence in courts and in parliamentary enquiries, witness protection schemes, and whistleblowing legislation—all noble recognitions of the reality of the perils of speaking out. Your assumption is that researchers should be such lionhearts that they are above such intimidation. In Australia, the Federal Minister for Education has taken to personally vetting research grant award lists, ostensibly to prune projects with ridicule potential ("Comparative Easter Bunny mythology: a post-structuralist exigesis" type stuff) but where might that lead in today's climate of more interventionist government is anyone's guess.

I broadly share your sentiments, sign all my reviews, and have had my picture on a few government dart boards. But I do feel for junior researchers holding politically volatile material in their trembling hands (not that this present paper entirely satisfies that description) and so am open to the idea of being more accommodating.

Simon

___________________________

Dr Chapman, I think you've indeed been "open to the idea of being more accomodating" by starting this discussion, and it was right to be open to the idea. I don't think that means you ought to publish the letter anonymously.

May I suggest 3 points to consider from my "non-celebrity" position-though my conclusions happen to weigh in with Dr. Drummond's. Possibly someone who has remained demonstrably "junior," like the authors, should speak up now.

1. I don't think you need to worry too much about ruining the authors' potentially brilliant careers by reaffirming your journal's usual practice of publishing names. Sure, political perils are real-I even signed a letter in 1981 that had concrete employment repercussions in 1990. But my overall pattern of speaking up has mattered more than any single act. And life in the US will be just fine for them on the bottom rungs, as it has been for me in Europe. Dr. Drummond is right.
2. At the risk of seeming pedantic, on scientific politics: we need more people who speak up to make real information comprehensible to decision makers and voters. We don't need more jokes, however cleverly framed as correlation studies in letters to the editor. Could it be that the authors don't sign because they don't believe in the importance of the letter themselves—otherwise they'd have done a full-blown study?
3. The argument from The Economist's practice doesn't really extend to scientific journals—which started their history as signed "letters." I'm not a reader of that magazine either, but if all its articles are anonymous, they're extending the "editorialist" position to all their writers. Are editorials still unsigned in the US? They are in the papers I read in Europe. What we're meant to understand by that very traditional practice is that it is "the Paper's" considered position we're reading, and "the Paper" stands by the integrity of what is written. The Economist extends that to its whole staff, giving the magazine an aura of great prestige, implying their entire staff is venerable and speaks with one voice. But I don't personally know what happens inside their magazine because they extend that status to so many! I bet there's plenty of folklore to be investigated!

I vote for sharing a good laugh with your Editorial Board about the clever little correlation study, refusing to consider publishing anonymously, asking if they would like to have it re-considered for publication signed, and saying you certainly enjoyed reading it. The authors will have gained as good an anecdote for telling over academic beer drinking as they would have if the letter had been published. That's their real intention.

Mary Ellen Kerans
Barcelona, Spain

___________________________

May I offer the view from different personal background and national/history perspective?

We may name in person great people who did great things, and stand heroically. Some of them survive. If the survival rate is good, people get accustomed to thinking that it is very natural to speak aloud about what you think.

If you live in the country where you may lose every thing only because you publish the wrong article, like in my country, where thousands were shot or imprisoned and died in camps (it is about survival rate—mortality in Siberian camps was like 400%/year) only because they defended their research field—genetics. If you live in such country, you will understand that nobody can ask another person to be a hero.

It is why I enjoy the Dr Rennie's arguments, but I am sympathetic to what Dr Chapman explained.

Vasiliy Vlassov
Russia

___________________________

To weigh in, I would vote against peer-reviewed journals publishing anonymous letters or articles, for several reasons. First, the authors' conflicts of interest cannot be assessed by the reader. Second, there is need for anonymity because there are political (or econo-political) issues at play. A peer-reviewed scientific journal cannot afford to be labeled as promoting a political agenda, or its objectivity for selecting articles to publish could be assailed (especially for those journals that reject the majority of the articles submitted to them).

Finally, and most importantly, peer-reviewed journals are not the best forum for anonymous "articles" making political points. (And the example here—about a relationship between a conservative government and smoking cessation programs—is clearly political in nature.) An Internet weblog ("blog") would be a better forum for this. The whole world has access to such an article (unlike many medical journals), it can be anonymous, the data can be presented, and the audience can make up its mind. If peer review is important, then invited commentaries/critiques can be added (as the Annals of Family Medicine routinely does on its Web-based journal).

William M. Tierney
Co-Editor-in-Chief, Journal of General Internal Medicine

___________________________

I generally agree that authorship should almost never be anonymous. The "almost" is because, in my view, there can be rare circumstances that justify anonymity. For what it's worth, we published only a single article anonymously (actually, pseudonymously) in my time as editor of Annals of Internal Medicine. The paper in question was a "wrenching" personal account by a physician in the state of Oregon of his participation in the assisted suicide of one of his young patients-a legislative measure that he had voted against twice in the prior Oregon referendums (assisted suicide is legal, under strictly regulated conditions, in Oregon). [Kade WJ. Death with dignity: a case study. Ann Intern Med. 2000;132:504-506.]

The special circumstances that led us to "agree to a pseudonym" were that the patient's family gave permission to publish the report, but specifically asked that the author's real name not be used, to protect the patient's and the family's privacy.

So, sometimes even the most zealously guarded editorial principles melt in the face of legitimate human concerns.

Frank Davidoff

___________________________

Dear Drummond,

It is clear from your advice that you believe "it cannot happen here." Indeed, it is happening here. Unfortunately, Sinclair Lewis' 1935 novel is relevant today.

If the work is worth publishing, I agree with the researchers that their identity should be protected. Indeed, I am pleased to see that Simon Chapman recognizes his own vulnerability, considering George Lundberg's experience after promptly publishing accurate and relevant research results that were deemed politically inconvenient to the medical association's directors at the time.

In principle, I also agree with you about honoring the sacrifices of other who speak loudly in spite of adversity. Nevertheless, society needs quiet heroes who have the courage to conduct research that they know would be squelched, as well as those who openly challenge attempts at silencing the truth. We must be grateful that there are editors who are willing to take risks for the benefit of a better informed society.

PS: I suppose that my experience growing up Buenos Aires during Juan Perón's dictatorship makes me acutely aware of the repercussions of taking freedom for granted. One must experience it to understand.

___________________________

Though not from Buenos Aires, I have to agree with Dr Cann, at least as far as the danger is concerned. Negative repercussions can be subtle, yet still worth avoiding, so it may not be best to speak in absolute terms ("we will never consider such a thing"). The risks have been well articulated in the WAME discussion, yet presumably there is also a benefit to having this piece published, even if anonymously? If so, then the question is no longer "Is there a downside?"—we know that there is. Rather, the question concerns whether or not the benefits outweigh the risks. It would seem to me that this is a question that each editor is best suited to answer for her or his own journal, because the answer may not be the same from journal to journal.

Vance

___________________________

Permit me to add my bit to the discussion. And do pardon me for its length.

Anonymous writing can be for 3 main reasons:

1. To slander someone
2. To expose someone without suffering the consequences of the expose
3. If someone does not want to be identified but wants to excite discussion among readers, eg an editor, editorial board member, well known writer, someone who does not want his/her position/affiliation to influence readers' responses to the write-up.

While every well intentioned editor is justified in being wary of anonymous writing whose primary intention maybe to slander or spread malicious gossip about researchers, editors, journals etc, it is equally his duty to be aware of the nefarious goings-on in medical research. Now, it is all right to say that a whistle blower must have the courage to get recognized and adopt a public stand, and most do; but it is equally important to accept that not all can. The line dividing courage and bravado is thin, and not everyone would risk his/her position to become a messiah of change. That does not, however, mean what he/she has to say may not be based on evidence. The editor/editorial board's prime responsibility is to determine whether it appears prima facie evidence based, like it does for all other manuscripts, and is couched in parliamentary language. Having ensured this, there is no reason why it cannot go for peer review for which, in any case, it has to be anonymous. If the editorial board considers it of great import and time is of the essence, it may even consider it for publication if the journal policies permit such 'fast track' publishing. If possible, such publishing should be accompanied by a rejoinder from the authors/journal concerned. If the latter refuse to respond, the editor/editorial board could always mention below the article that they welcome a response/rejoinder/rebuttal even later, provided it is couched in parliamentary language and produces counter-evidence.

In any case, no editor/board/journal should engage in anonymous scandalous writing about another editor/board/journal, taking cover of anonymity. How to ensure this happens is one vexing issue. However, if the parameters are clear, the evidence is strong, the other journal/editor is refusing to publish because it stands to get exposed for its wrong doings, I see no other option but for the article/communication to be published in another journal. The readers will judge it for what it stands, and will definitely pull it down a peg or two for its anonymity, as well as lack of declaration of 'conflict of interest'. But even if pulled down so, it may still stand scrutiny, and become the start of many other skeletons tumbling out of other cupboards.

In the case of something published in one's own journal, anonymous rejoinders to published material may be easier to accept, with the rider of evidence and parliamentary language mentioned earlier.

I realize that accepting anonymous writing can open the floodgates of slander/libelous writing, which maybe difficult to lock later, but it is equally important that truth prevails. For those who are high and mighty in research, and who have a reputation of being ruthless towards their detractors, what method of expose is available to those who know about their wrong doings except to reveal the truth under cover of anonymity? In other words, if anonymous writing is not allowed, we allow them to continue till Kingdom Come. And help spawn many more of their ilk, for the smart alecks learn quickly the ropes to climb. Moreover, to expect every truth carrier to also have the strength to bare the consequences of truth telling is to expect each man to be a Christ or a Gandhi. Is it not a little too much?

In the light of the above, I wonder if we could come to the following conclusions:

1. Editors should be very wary of anonymous writing landing on their desks, and in general should be very suspicious of the motives of such writing. Something couched in slander and abuse should be rejected outright.
2. However, if an editor feels there is substantial evidence produced, it conforms to journal style and is in line with journal policies, a serious second/third read must be given. The editor may at this stage encourage the writer/s to remove the anonymity condition so that the writing stands complete, and the authors muster courage. Sometimes an editor's persuasion may work for the welfare of all concerned.
3. The editorial board then discusses the issues raised in the anonymous article threadbare. If the evidence produced appears irrefutable, it must be sent for peer review, and considered for publication according to the regular process. It must be sent to the concerned person/persons for a rejoinder, wherever possible.
4. The rejoinder, if forthcoming, and only if couched in parliamentary language producing counterevidence, or accepting totally or partially the claims of the anonymous article, whatever, should be similarly considered for publication.
5. In any case, anonymous writing should form only a miniscule part of the journal's writings, and its rejoinder and counter to the rejoinder should not be used to convert the journal into a tabloid.
6. The controversy should be swiftly resolved. The journal should never play "running with the hare and hunting with the hound" to extract mileage and increase circulation at the cost of researchers' reputation.
7. No journal should encourage people to write anonymously as a part of its policy, but accept an occasional work of this type in extraordinary circumstances, if and only if there is sufficient prima facie evidence for such an acceptance, and due process of submission and review are followed. Editorial discretion and judiciousness are of paramount importance here.
8. Proper statutory bodies, or judicial process should deal with any journal that indulges in anonymous writing to spread malice and slander about researchers and other journals.
9. Rarely should the editor/editorial board ever associate itself publicly with anonymous writing. The disclaimer about all other work published applies equally to anonymous writing, in fact more so. A clause that it is an anonymous writing and should be judged accordingly is to be added below, and is itself a declaration of conflict of interest of sorts. The editor/editorial board only considers it print worthy based on its content. It does not endorse its contents, however.

I am aware many editors who read this may be tempted to play it safe and avoid anonymous publishing altogether. And maybe they are doing the pragmatic thing after all. For don't we all know that every editor who survives learns to play it safe. But every editor who not only survives but makes a mark also makes bold to play it on the front foot at times.

Well, each one to decide his course, and his destiny.

Ajai

___________________________

What about a time capsule approach? When warranted, publish the article anonymously, but with the agreement that the identity of the author(s) be revealed at some stated time in the future, presumably when there is no further risk of retaliation. The nature of the fear of reprisals may be split into groups, such as specific to a given employer (so this ceases to be a problem once employment ends), ruining of a reputation (this would remain in effect longer, and possibly for the duration of a career or even a life), and those in which there is a fear that even the collaborators (or children) of the author(s) may find themselves on a black list, in which case perhaps this time capsule approach will not mitigate fears. But there should be some cases in which this delay in identifying the author(s) will work, and ultimately we (the research community) get what we want, namely a complete record of who said what and when. Thoughts?

Vance

___________________________

Brilliant. Didn't occur to me it could be done. In totalitarian regimes, and even setups (where overtly democratic norms are supposed to prevail, but may not), this is a worthy option to explore.

However, the caveat that anonymous writing should always be occasional, only when other options of an expose are truly closed, and only in the cause of promoting genuine research, must prevail.

Ajai

___________________________

Researcher member of WAME here.

Medical journals should not publish anonymous work, except perhaps under the rare circumstance described by Harold Sox. Editors may not be given the true, or whole, story; anonymity may be used legitimately, to protect privacy, but it may also be used by cowards, or as a tactic by those who seek gain, or revenge. Bill Tierney has already noted that conflicts of interest of anonymous authors cannot be examined. Anonymity is a dangerous precedent. The article in question sounds like a weak correlational study…but in any case the authors are US government employees, and subject to federal ethics rules. I used to be at NIH, and published articles and a book (in a different field than I was working in) while I was there. I had to report these in outside activity forms that were vetted by the office of general counsel to see that they didn't constitue conflicts with my work, but no one censored my work. The current US administration is extremely controlling (although nothing l ike some Eastern European and other WAME members have described). The fact remains, however, that the government, or any employer, is allowed to control publications of their employees, at least on subjects that fall into their scope of work. The real heroes are those willing to be named, like David Graham of the Food and Drug Administration. Laws protecting whistleblowers have been weakened in the US, but there are still some laws in place, and projects, such as the Government Accountability Project (a nonprofit law firm that represents Graham and others) that defends whistleblowers. See http://www.whistleblower.org.

Adriane Fugh-Berman
Associate Professor, Complementary and Alternative Medicine Master's Program, Department of Physiology and Biophysics School of Medicine, Georgetown University

___________________________

Just a point of clarification—what does the word "anonymous" mean? I am a native English speaker, and have a dictionary handy, but this does not tell me the intended meaning in this context, because I had been operating under the assumption that the anonymity applied only in the would-be publication itself, but not in the correspondence. The significance of this distinction is that if this is true, then the editors would, in fact, know the identity of the authors, and might be in a position to evaluate any conflicts, real or perceived. The editors might even be able to reveal these conflicts without revealing the identity of the authors. Does this in any way modify your argument?

Vance

___________________________

Agreed that the assumption is that the editors know who the authors are. Editors do try to identify conflicts, but authors who wish to remain nameless may have other secrets as well. Readers may be able to identify undisclosed conflicts.

Adriane Fugh-Berman

___________________________