On May 8Th, 2008 Human Brain Mapping (HBM) Published a Paper by A

Date: 16 June 2008

Dear Editors and Colleagues:

On May 8th, 2008 Human Brain Mapping (HBM) published a paper by A. Shmuel and D. Leopold entitled “Neuronal correlates of spontaneous fluctuations in fMRI signals in monkey visual cortex: Implications for functional connectivity at rest” (DOI: 10.1002/hbm.20580). The authors are former postdoctoral fellows of mine and the data were collected in my laboratory, so it is with some unhappiness that I write to you to disavow the paper’s contents and conclusions, and to let you know that my colleagues and I are preparing a rebuttal for publication. My objections to the paper are both technical and ethical.

The technical problem is that most of the measurements reported in the paper to be “at rest” were taken from the prestimulus interval of long observation periods with multiple trials during which visual stimuli were presented. The data came from another study of my laboratory (Shmuel, Augath, Oeltermann and Logothetis, 2006). In fact, a grey screen with a significant flicker at 30 and 60 Hz evoked electrical cortical activity, making it wrong to suppose that the comparison of neural with hemodynamic signals was within a rest period during which spontaneous activity alone prevailed (details in the rebuttal). This renders the conclusion in the paper, and thus the paper itself, invalid. These points were pointed out to the authors and to the editors as well, but the editors chose to accept it even in spite of the questionable ethical conduct of the authors, which the editors were also aware of, or should have been aware of (see below). In short, the report does not pertain to “spontaneous activity” and the findings are trivial, already having been demonstrated in our previous publications.

The ethical problem is that Shmuel and Leopold submitted this work without my knowledge or permission, after I had pointed out the problems with the data they were using, and after Shmuel agreed that the interstimulus data were inappropriate for this analysis and that new data would have to be collected before the work could be presented.

Shmuel did not make me aware of this publication until 6 weeks (19 April 2008) after the paper was accepted for publication (6 March 2008) and a few days before the publication-deadline, when he invited me to become a co-author. It is worth pointing out that he had invited Leopold to act as a coauthor in this paper but excluded from authorship the collaborators who did the actual data acquisition for him. When I tried at that late stage to intervene to prevent publication of this flawed work, I was frustrated both by Shmuel and Leopold and by the editors of HBM. The editor-in-chief and the two associate editors, who were responsible for the special HBM issue, entirely ignored my comments and the preliminary data analysis I had sent them to point out the serious flaws of the study. They also ignored the request of the Max Planck Society (MPS) to remove the name of the society from the funding sources of this type of work in the HTML version of the document; at least until today (16 June 2008). MPS has explicitly indicated that the accreditation is wrong, and that it does not want to be associated in any way with this particular publication.

I would be happy to give you further details of this unfortunate business, including my correspondence both with the authors and with the editors of HBM, should you be interested. But these details are not central to the point of this message: the published paper is profoundly flawed and does not represent the quality of work that I hope you have all come to expect from my laboratory. It has also been published in a way that many in the scientific community would regard as outrageous. I am taking steps to make this matter public and to expose the flaws in the paper, but in the meanwhile I hope you will accept this simple disavowal and not heed or cite the work reported.

With kind regards,

Prof. Nikos K. Logothetis MPI for Biological Cybernetics Spemannstr. 38, 72076 Tuebingen, Germany Tel: +49 7071 601-650 Tel: +49 7071 601-651 Secretary Fax: +49 7071 601 652 Vol 454|3 July 2008 NEWS Neuroscientist: my data published without authorization are ‘misleading’ The director of a top laboratory in Germany has permission to use data generated there. charged that two of his former research students Although he agreed at first, Logothetis with- took data from his laboratory without his per- drew his permission when he realized that the mission and published scientifically incorrect data — from functional magnetic resonance interpretations of them against his advice. imaging studies on monkey brains — were Neuroscientist Nikos Logothetis (pictured), being used to support a theory about sponta- of the Max Planck Institute for Biological neous brain activity. The data had been col- Cybernetics in Tübingen, further claims that lected when monkeys were looking at a grey the journal involved, Human Brain Mapping, but flickering LED screen. “The protocol was acted incorrectly by publishing the paper after just inappropriate for analysis of spontaneous he told them the data were inappropriate. He brain activity,” says Logothetis. says the journal has denied him the right to a Several months later, he says, he was timely reply. surprised to receive an e-mail from Shmuel One of the two editors-in-chief of Human containing a complete paper using the same Brain Mapping, Peter Fox of data, co-authored with another the University of Texas Health “The journal used the former research student, David Science Center in San Antonio, Max Planck Society Leopold, who worked in the told Nature that the paper was to excuse their own labs between 1992 and 2003, correctly refereed, but declined where he collected some of to add details. mismanagement of the data himself. Leopold is Logothetis is furious about the case.” now at the National Institute the publication of data, which of Mental Health in Bethesda, he believes will mislead the field, and about Maryland. Shmuel invited Logothetis to join the fact that the authors of the paper allege that as third author, telling him that the paper had wrote directly to the Max Planck Society he tried to stop them publishing the data for already been accepted for publication and (MPS), which runs 80 research institutes in personal reasons. would appear online in a few days. It had been Germany, claiming that Logothetis was trying The affair began in the spring, when Amir accepted six weeks earlier. to prevent him and Shmuel from publishing Shmuel, who worked in Logothetis’s labora- Matters escalated. “I told him that the data data for personal reasons. tories from 2002 to 2007 and is now at the were not publishable,” says Logothetis, who After consultation with Logothetis, MPS Montreal Neurological Institute of McGill also wrote to Fox proposing that the paper vice-president Herbert Jäckle wrote to the University in Canada, asked Logothetis for should not be published. But Leopold then authors giving approval for the use of the data, Turkish politics blamed for board block A prominent Turkish geologist is being denied a Istanbul Technical University, says that he has Akaydin, who heads the inter-university board. top spot in the nation’s higher-education system been blocked from joining Turkey’s council “He’s a good guy, a very well-known intellectual because, he says, his political views are out of of higher education (YÖK) and subjected to in Turkey.” step with those of the current government. spurious ethics investigations because he Nominations typically pass through a Celâl S¸engör (pictured), a professor at has spoken out against the government. straightforward approval process, but S¸engör “There is an atmosphere of terror,” says that YÖK’s leadership is holding up his he says. “It’s unbelievable what’s candidacy. This spring, he says, the leadership going on.” opened an investigation into a previously YÖK is a 21-member council that dismissed ethics charge concerning S¸engör’s oversees Turkey’s universities. One- ties to a family business. Then, just last week, third of the council is nominated S¸engör says he was informed of a second by the inter-university board investigation into his having allegedly — a group of university rectors and travelled abroad without university approval. other academic representatives. S¸engör believes that he is being persecuted In January, the board advanced because he is an outspoken critic of Turkey’s S¸engör’s name for a position on current, Islamist-rooted AKP ruling party. “They YÖK. The reason for the nomination want to get rid of anybody who is against them,” was straightforward, says Mustafa he says.

Please Login or Register Search

NewsBlog: >> July 2008 >> June 2008 Posted by Andrea Gawrylewski >> May 2008 [Entry posted at 2nd July 2008 09:51 PM GMT] >> April 2008 View comments(6) | Comment on this blog >> March 2008 >> February 2008 A prominent neuroscientist is accusing two former >> January 2008 researchers in his lab of taking data without his permission and publishing misleading interpretations of >> December 2007 them against his wishes. >> November 2007 >> October 2007 Nikos Logothetis, director of the Max Planck Institute for Biological Cybernetics in Tubingen, Germany, says >> September 2007 that two former researchers working in his lab took fMRI >> August 2007 data from monkey brain scans without his permission >> July 2007 and made misleading interpretations in a paper published this month in the journal Human Brain >> June 2007 Mapping. In addition, the journal has not guaranteed >> May 2007 him an opportunity to publish a response to the >> April 2007 findings, he told Nature. >> March 2007 The paper, authored by Amir Shmuel -- now at Montreal >> February 2007 Neurological Institute of McGill University in Canada -- >> January 2007 and David Leopold -- now at the National Institute of Mental Health in Bethesda, Maryland -- suggests that >> December 2006 the fMRI data is useful for studying spontaneous, or >> November 2006 random, activity in the resting brain. "The protocol [in their paper] was just inappropriate for analysis of >> October 2006 spontaneous brain activity," Logothetis told Nature. >> September 2006 >> August 2006 Peter Fox, editor-in-chief of the journal, refused to retract the paper and told Nature that the paper was >> July 2006 Survey Series correctly refereed. >> June 2006 Best Places to Work >> May 2006 >> April 2006 $alary $urvey >> March 2006 Lab Web Site and >> February 2006 Video Awards Rate this article >> January 2006 >> December 2005 >> November 2005 >> October 2005 Rating: 3.40/5 (15 votes ) >> September 2005 >> August 2005 Comment on this blog post >> July 2005 The Scientist Daily >> June 2005 Science headlines delivered daily. Register today. Return to Top comment: Who owns the data? by anonymous poster

[Comment posted 2008-07-09 07:58:50]

It is easy for a postdoc, who is relieved of all responsibilities for obtaining funds and running the lab, to believe that they are the only one who knows anything. Those individuals had no right to remove the data from the laboratory. They should have resolved their interpretation of the data with their mentor, and the paper should not have been published without his For Advertisers permission. They had no right to exclude him from authorship. Advertise with Us The rebuttal should be published in the best interests of science. Contact Ad Team

2008 Media Kit

Return to Top comment: This time the opposite happened by anonymous poster

[Comment posted 2008-07-08 13:53:58]

In my opinion, this dispute over the "stolen data" is rather a problem of an authorship.

Though the story is presented only from one side, I have an assumption what could be the reasons of parties on the other side. Typically, if a person (usually a post doc) leaves a lab under disgrace, the PI (ab)normally feels free to use their experimental data in his later publications and frequently without giving any credit (authorship) to these people. Although this kind of stealing is very frequent (I have both seen it and suffered it), it is usually considered to be accepted by the scientific community.

This time the opposite happened. The persons leaving the lab (probably under disgrace), ensured their authorship in their own work by publishing their own results before the PI could do it. Practically the two post docs did not steal anything, since the data were produced by them. Authors should have made substantial intellectual contributions to a study and providing just funding (or oversight) ? what a PI usually does - is not enough for an authorship, as is stated by the guidelines formulated by the International Committee of Medical Journal Editors (see: How to resolve authorship disputes; www.the- scientist.com/news/display/53485).

The "misinterpretation" argument of the story seems just to be a handy, additional casus belli.

Return to Top comment: Ridiculous Gossip Column by anonymous poster

[Comment posted 2008-07-07 22:54:29]

I find it ridiculous Nature has decided to post this one sided gossip piece. As a pillar in scientific advancement Nature should have known better that there are better avenue's that statements like these should be addressed.

Return to Top comment: Comment on one item in the comment by Ellen Hunt by Ruth Rosin [Comment posted 2008-07-04 11:03:20]

I want to comment here only on the following item in the comment by Ellen Hunt:

"However, I also think that any reputable journal needs to give space to properly stated criticism of any paper, whether it is a situation like this, or any other. In today's electronic world, publication of critiques should be a trivial matter. Additionally, such publications should be highly desired by the science community, because science is all about getting at the truth."

Her idea is a very good idea, but even the editors of a reputable journal like Nature, do not provide sufficient space for a proper, online critique of articles published in their own journal.

I've been trying for over a year to submit a proper critique of the report by Riley et al. (2005), on their honeybee radar- tracking study, where the authors express the hope that their results would be accepted as a vindication of v. Frisch's famous honeybee "dance language" (DL) hypothesis.

Von Frisch's sensational DL hypothesis, first published in a scientific journal in 1946, as presumably already fully properly experimentally confirmed, was openly criticized for the first time by Adrian Wenner & his team, in 1967.This gave rise to the well-known DL controversy. The controversy, which has by now been going on for over 40 years, has throughout the years become very complex and convoluted.

The study by Riley et al. (2005), cannot, however, provide any support for that DL hypothesis, because the hypothesis was stillborn, (more than 20 years before its inception), when v. Frisch fully justifiably concluded, on the basis of his first study on honeybee-recruitment, (published in an extensive summary in 1923), that honeybee-recruits use only odor, and no information about the location of any food. Moreover, the results he then obtained, already grossly contradicted the expectations from his 1946 DL hypothesis.

Nature, however, allows critiques on material published in the journal, only in the form of "Brief Communication Arising", that are limited to not more than 700 words, plus 10 references, and must be understood even by non-professionals. It is utterly impossible to deal with any aspect of the very complex DL controversy, let alone with the recent study by Riley et al. (2005), within such narrow constraints!

As a result, readers of Nature will continue to be misguided into believing that the study by Riley et al. (2005) can be accepted as a vindication of v. Frisch's stillborn hypothesis.

Return to Top comment: Very well said Ellen Hunt! by David Bird

[Comment posted 2008-07-04 00:45:12]

While this particular story may not fit the description, your thoughts and comments are spot on I think!

The comparison to Russian Mafia couldn't be better said. Implementing such a policy could be tricky, but I agree, all data, findings, etc. generated from public funding should (eventually) be released to the public trust.

Return to Top comment: Complicated by Ellen Hunt

[Comment posted 2008-07-03 11:59:26]

If it's a bad paper, that wouldn't be the first time one was published in a Nature publication. But I am not sure it is.

Over the years I have become less and less sympathetic to the idea that the P.I. who gets the grant and runs a lab should have control over what gets publshed and what does not. This idea is based on the premise that the P.I. is wiser, better able to interpret the data, etcetera. But if this ever was true, it certainly is not true today. In today's world I think that the most creative and interesting science is more often done by graduate students and post-docs.

I know of quite a few situations where a lab has amassed huge troves of data and the people in the lab and/or the P.I. are barely scratching the surface. The taxpayers paid for all that, and yet there is this unwritten law that nobody else, inside or outside the lab is allowed to publish based on that data or allowed to have access to it. It "belongs" in effect, to the grantee.

Frankly, it reminds me of nothing quite so much as the gangsters in the former USSR that I happened to meet in the 1990s who spoke of "their hard-won holdings" which was from money given by aid organizations. These men also had this idea in their head that the millions given in grants to buy things like generators or stoves or blankets meant that all those generators, stoves or blankets were theirs.

It was after that experience that, back in the States, I found myself seeing that it was exactly the same sort of mind at work in science. I could not stop myself from drawing the parallel.

Now I think this should be formally changed by law. I think that the law should say that after 12 months, or 18 months or some reasonable time period, P.I.s should be required to publish what data they have, and make it available to other scientists. Receiving a grant is a public trust. To me this is a matter of scientific ethics.

I think it would be good, but I don't see how it can be required, that as much as possible, privately funded research have the same ethic.

However, I also think that any reputable journal needs to give space to properly stated criticism of any paper, whether it is a situation like this, or any other. In today's electronic world, publication of critiques should be a trivial matter. Additionally, such publications should be highly desired by the science community, because science is all about getting at the truth.

Comment on this blog post

EDITORIAL

Protecting Peer Review: Correspondence Chronology and Ethical Analysis Regarding Logothetis vs. Shmuel and Leopold

1 2 3 1 Peter T. Fox, * Ed Bullmore, Peter A. Bandettini, and Jack L. Lancaster

1Research Imaging Center, University of Texas Health Science Center, San Antonio, Texas 2Department of Psychiatry, Cambridge University, Cambridge, United Kingdom 3Functional MRI Facility, National Institutes of Health

Abstract: Editors of scientiﬁc journals are ethically bound to provide a fair and impartial peer-review process and to protect the rights of contributing authors to publish research results. If, however, a dispute arises among investigators regarding data ownership and the right to publish, the ethical responsibilities of journal editors become more complex. The editors of Human Brain Mapping recently had the unusual experience of learning of an ongoing dispute regarding data-access rights pertaining to a manuscript already accepted for publication. Herein the editors describe the nature of the dispute, the steps taken to explore and resolve the conﬂict, and discuss the ethical principles that govern such circumstances. Drawing on this experience and with the goal of avoiding future controversies, the editors have formulated a Data Rights Policy and a Data Rights Procedure for Human Brain Mapping. Human Brain Mapping adopts this policy effective immediately and respectfully suggests that other journals consider adopting this or similar policies. Hum Brain Mapp 30:347–354, 2009. VC 2008 Wiley-Liss, Inc.

Key words: ethics; research ethics; responsible conduct of research; peer-review; data ownership

INTRODUCTION scientific disagreements as well as inter-personal conflicts inevitably arise. Investigators may disagree, for example, As the complexity of scientific investigation has on the adequacy of the data for publication, the most advanced, bio-medical research has progressively adopted appropriate analyses to be performed, or the appropriate a team-based approach to research. In the life sciences, conclusions to be drawn from the accumulated experi- brain imaging is one of the most technically advanced and ments. In the context of such disagreements, more funda- integrative disciplines. In this collaborative environment, mental disputes often arise, including the right of individual investigators to publish data acquired cooperatively. When efforts are made to publish disputed data, journal *Correspondence to: Dr. Peter Fox, University of Texas Health Sci- editors necessarily become involved. ence Center, Research Imaging Center, 7703 Floyd Curl Drive, San Antonio, Texas 78284, USA. E-mail: [email protected] The editors of Human Brain Mapping were cocorrespond- Received for publication 14 July 2008; Revised 12 September 2008; ents in a widely publicized [Abbott 2008; Gawrylewski, Accepted 16 September 2008 2008] data-rights dispute over a manuscript published in DOI: 10.1002/hbm.20682 May 2008 [Shmuel et al., 2008]. The core conflict was Published online 9 December 2008 in Wiley InterScience (www. between Dr. Nikos K. Logothetis, on the one hand, and interscience.wiley.com). Drs. Amir Shmuel and David Leopold, on the other. At

VC 2008 Wiley-Liss, Inc. r Fox et al. r the time the research was conducted and the disputed cal guidelines for journal editors, including the World data were acquired, Shmuel and Leopold were both Association of Medical Editors (WAME; http://wame.org), Research Scientists under the supervision of Logothetis, a the International Committee of Medical Journal Editors Department Director of the Max Planck Institute (MPI) fur (www.icmje.org), the Council of Science Editors (http:// biologische Kybernetic, in Tu¨ bingen, Germany. At the time www.councilofscienceedtitors.org), the American Associa- of manuscript submission, Shmuel was an Assistant Pro- tion for the Advancement of Science (http://www.aaas. fessor at the Montreal Neurological Institute and McGill org), and the Committee on Publication Ethics (http:// University, Montreal, Canada; Leopold was a Unit Chief in www.publicationethic.org.uk). the Laboratory of Neuropsychology of the National Insti- tute of Mental Health, Bethesda, Maryland. In brief, Logo- RCR and Data-Management Ethics thetis made two serious allegations against Shmuel and Leopold. First, Logothetis contended that data were used Permission to acquire data in living subjects, either without his permission, thus being unethical. Second, he human or animal, can be granted only by institutional contended that the data were misrepresented in the manu- ethics committees, such as the Institutional Animal Care script submitted to Human Brain Mapping, and the inter- and Use Committee (IACUC) or the Institutional Review pretations were misleading. For these reasons, Logothetis Board (IRB) for the protection of human subjects, which contacted the editors seeking to block publication. Concur- receive their authority from the national government. rently, Logothetis pressured Shmuel and Leopold through Ethics committee approval is granted on a case-by-case other channels to withdraw the manuscript. When these (protocol-by-protocol) basis, with an individual investiga- efforts to block the manuscript failed, Logothetis made tor being named as Principal Investigator (PI). Because vio- public statements discrediting the authors and the editors. lations of Ethics Committee policies, regulations and pro- In our opinion, the accusations against both the authors cedures can negatively impact an entire laboratory and, and the editors are not supported by the facts of the case; occasionally, an entire institution, it is not uncommon for rather, Dr Logothetis’ conduct in this episode clearly fell laboratory directors to have senior investigators named as short of widely accepted ethical standards. PI on protocols designed and carried out chiefly by more In what follows, the generally accepted ethical standards junior researchers, including post-doctoral fellows and that we consider applicable in this case are rehearsed. This students. In the absence of a policy or other adjudication is followed by a chronological account of the episode, to the contrary, authority for data management belongs to based on the relevant correspondence. An analysis of the the PI. With this authority come a number of responsibil- actions of the various parties and institutions, relative to ities. For example, it is the responsibility of the PI to make the described ethical standards, is then offered. Finally, sure that data acquisition procedures are well docu- recommendations for the prevention of similar episodes in mented, that all Ethics Committee stipulations and other future, including explicit statements of a new Data Rights applicable rules and regulations (e.g., regarding use of ra- Policy and Procedure for Human Brain Mapping, are dioactive materials) are followed, that subject safety and provided. comfort are maintained throughout, and that privileged information (e.g., protected health information for human subjects) remains confidential. Since the acquisition of data ETHICAL STANDARDS in academic institutions is typically funded by research grants from public sources, it is also a responsibility of the The case at hand involved two distinct areas of research PI to ensure that the data achieve the intended public ethics: responsible conduct of research (RCR), and the good, either through publication, or through data sharing, ethics of peer-review. RCR is a rapidly evolving area or through both. of ethics, the development of which is being driven by the When disagreements regarding data management and involvement of national governments in research funding. data publication arise between the PI and co-investigators, In the United States, the Office of Research Integrity (ORI) adjudication can be requested. Insofar as the rights to ac- of the Department of Health and Human Services (http:// quire and manage the data are granted by the institution ori.dhhs.gov) is responsible for ensuring the integrity of and, ultimately, are the responsibility of the institution, the the research practices in the 4,000 institutions world-wide which are funded by the US Public Health Service. The right to reassign any component of this authority resides ORI promotes research integrity through regulatory, inves- with the institution. In ideal circumstances, academic insti- tigational, preventative, and educational activities. Data- tutions have policies in place to govern the management access and right-to-publish policies are important domains of research data, including procedures for dispute resolu- within RCR for which the ORI and other institutions pro- tion. For example, the data-management policies of the vide guidelines. The ethics of peer-review for scientific University of Pittsburgh (http://www.pitt.edu/provost/ publication, on the other hand, have evolved chiefly retention.html) and Stanford University (http://www. through the efforts of journal editors. Various associations stanford.edu/dept/DoR/rph/2-10.html) are considered have evolved to promote the development of formal ethi- models in the field of RCR. In the event of a dispute, the

r 348 r r Protecting Peer Review r typical chain of adjudication rises through the institutional For journal editors, protecting the integrity of the peer- hierarchy, from PI, to Laboratory Director, to Department review process is a public trust and a responsibility both Chair, to Dean. Alternatively, the IACUC or IRB or an to the authors and to the members of the scientific commu- independent Ethics Committee may be appointed to adju- nity at large. Journal editors bear a particular responsibil- dicate such disputes. ity to the authors because peer-reviewed publication is the Dissolution of a research team, relocation of an investi- building block of scientific careers. Publication in peer- gator (principal or nonprincipal), and the participation of reviewed journals is the primary metric by which aca- trainees (e.g. post-doctoral fellows and students) in demic careers are judged and upon which employment, research, are common occurrences, and should be explic- promotion, tenure, funding, and status within the scientific itly addressed in any institution’s data management and community depend. Journal editors are responsible to the RCR policies. The most common policy is that all persons scientific community because the peer-reviewed literature involved in data acquisition retain rights to the data, constitutes the largest component of the collective corpus regardless of whether the team is intact or dissolved, of scientific productivity. whether the investigators are still on-site or have relocated, This is not to suggest that peer-review is able to guaran- and whether the team member is the PI, a co-investigator, tee the validity of each and every published finding. The or a trainee. For example: process of scientific discovery is not without error. Even the most rigorous, well-trained and well-meaning scientists ‘‘When a collaborative team is dissolved, University of can produce findings that cannot be replicated and Pittsburgh policy states that each member of the team should advance theories that are later overturned. Reviewers can have continuing access to the data and materials with which also miss fundamental flaws in submitted manuscripts. he/she had been working, unless some other agreement was Hence, the value of the well-worn research adage, ‘‘Repli- established at the outset.’’ cation is the best statistic.’’ The scientific enterprise is enor- and, mously powerful specifically because it is a collective undertaking. Errors eventually will be corrected. Thus, ‘‘Trainees and students who are an integral part of the journal editors and reviewers should carefully assess each research project should be allowed continued access to all paper, but not discourage novel findings and new theories records and data which pertain to their part of that pro- being presented to the community, trusting that replication ~ ject. (http://www.pitt.edu/provost/retention.html)’’ will be the final judge of the validity of a finding or Similarly, the ORI in its document entitled Data Manage- theory. Accordingly, scientists who disagree with a partic- ment and Lab Practices recommends that: ular finding, details of a methodology, or a theory should convey their arguments in a formal, peer-reviewed forum, ‘‘In general, each member of the team should have con- rather than attempting to impede publication. tinued access to the data/materials (unless a prior agreement was negotiated). (http://ori.dhhs.gov)’’ CORRESPONDENCE CHRONOLOGY

The disputed data first came to the attention of the edi- Peer-Review and Publication Ethics tors of Human Brain Mapping in an oral presentation, Peer-review of scientific manuscripts has been described ‘‘Spontaneous fluctuations in functional MRI signal reflect as a ‘‘cornerstone of modern science and medicine’’ [Rock- fluctuations in the underlying local neuronal activity,’’ by well, 2006]. The International Committee of Medical Jour- Dr. Amir Shmuel, Mr. Mark Augath, Mr. Axel Oeltermann, nal Editors (www.icmje.org) calls peer review ‘‘an impor- and Dr. Nikos Logothetis [Shmuel et al., 2007] presented at tant extension of the scientific process.’’ The most funda- the 2007 meeting of the Organization for Human Brain mental premise of peer-review is that it is based solely on Mapping (OHBM). Oral presentations at OHBM are very scientific merit. That is, peer review must be free of any high in quality, being few in number and awarded based ideological, political, financial, or personal biases or other on rigorous peer review. Shmuel’s presentation was of conflicts of interest. Consequently, a fundamental respon- particular interest to the editors of HBM because of its sibility of journal editors and reviewers is to avoid con- relevance to the topic of the annual HBM Special Issue for flicts of interest that could compromise the impartiality of 2008: Endogenous Brain Oscillations and Networks in the review process. For this reason, as with grants, it is Functional MRI [Bandettini and Bullmore, 2008]. On the commonpracticeinthepeer-reviewofmanuscriptsto basis of the quality of his presentation and its relevance to allow authors to identify persons whose reviews should the upcoming HBM Special Issue, Shmuel was invited to not be solicited because of real, suspected, or potential submit this work for inclusion in the Special Issue. Dr. conflicts of interest. However, it is equally the ethical David Leopold was also separately invited to submit an responsibility of the reviewer to excuse himself or herself article to this same issue. Subsequently, Shmuel and Leo- from the review process should a conflict of interest exist pold chose to co-author a single publication. It should be [Rockwell, 2006]. noted that invitation to submit did not guarantee publica-

r 349 r r Fox et al. r tion. HBM Special Issues are subject to peer-review of biologische Kybernetik, Tubingen, Germany). The experi- rigor equivalent to that of noninvited submissions for reg- mental part of this work was supported by NKL through ular issues. Production of HBM Special Issues is timed so personnel and funds of the Max Planck Society.’’’ that issues are released at the OHBM annual meeting, held in June of each year. Inclusion required a submission no The second stipulation related to authorship rights of the other investigators: later than November 1. The Shmuel and Leopold manuscript was submitted on October 17, 2007, went through ‘‘Mr Augath and Oeltermann informed me that two rounds of peer review, and was accepted for publica- they...decided not to join the list of authors and that tion on March 6, 2008. there is no mentioning of their names (including the On April 25th, the editors of HBM were notified for the acknowledgement section). They agree however, that their first time by Shmuel and Leopold of an unresolved data- data can be used under the aforementioned conditions.’’ ownership conflict that the authors were concerned might prevent publication. Unknown to the editors, this conflict These two stipulations were viewed as binding both by had arisen prior to the submission of the manuscript and the HBM editors and by the authors. Both were carried had remained unresolved throughout the review process. out exactly as requested. Logothetis had been a willing co-author of the original The third stipulation requested that scientific criticisms presentation of the data at OHBM [Shmuel et al., 2007]. made by Logothetis (and contained in Ja¨ckle’s letter) be But, once Shmuel and Leopold had left his laboratory, addressed by the authors. This stipulation concluded with Logothetis was unwilling to allow the data to be published the caveat: in a full-length manuscript. In an earlier e-mail correspon- ‘‘However, if you in spite of this criticism feel that all dence with Shmuel, Logothetis acknowledged Shmuel’s the above points are scientifically unjustified, then stand- legal right to publish the data. On the basis of the initial ard scientific practice would still require that you address permission, Shmuel and Leopold submitted the manu- these points, for example, in the discussion section of your script to Human Brain Mapping. Logothetis subsequently manuscript.’’ denied Shmuel and Leopold’s right to publish the data. After the HBM editors accepted the manuscript, Shmuel and Leopold requested adjudication of the dispute from The HBM editors reviewed the scientific criticisms to Dr. Peter Gruss, President of the Max Planck Society (MPS; determine whether they revealed new, scientifically crucial http://www.mpg.de/english/). The MPS oversees more information that might have invalidated the prior review than 80 research institutes and research facilities, of which process; in our opinion, they did not. The only scientifically the Max Planck Institute (MPI) fur biologische Kybernetik substantive issue was whether a uniform grey monitor is one. While the more typical approach for adjudicating a being present during data acquisition should prevent char- disagreement with a PI would be to appeal to the acterization of observed effects as ‘‘spontaneous fluctua- Laboratory Director or the Department Director, in this tions’’. However, the stimulus conditions were explicitly case, Logothetis held both of these senior posts. In making stated in the manuscript and were not felt by reviewers or the appeal for adjudication, Shmuel and Leopold provided editors to invalidate the study. Furthermore, resting state is evidence that Logothetis had already demonstrated, a condition allowing a wide range of definitions, easily in their opinion, a serious but not scientifically based including a uniform grey monitor presented to an anesthe- professional conflict of interest, which must exclude him tized animal. Consequently, the editors concluded that from participation in the adjudication process. Conse- there were no valid grounds for initiating a scientific re- quently, Gruss appointed Dr. Herbert Ja¨ckle, a Vice review. President of the MPS, to adjudicate. While awaiting Another scientific criticism by Logothetis (contained in Ja¨ckle’s decision, production of the manuscript was halted Ja¨ckle’s letter) was that Shmuel and Leopold [2008] made by HBM, per request of the authors and decision of the no fundamental scientific advance over a prior article by editors. Logothetis [Logothetis et al., 2001]. This view was not On May 2, 2008, the HBM editors received a letter from shared by either of the reviewers, nor by any of the HBM Ja¨ckle addressed to Shmuel and Leopold, confirming that editors. Most obviously, the earlier article did not seek to the conflict had been resolved in the authors’ favor and address multiregional network properties using resting- granting permission to proceed with publication. In the state acquisitions. Even if the new work merely duplicated context of granting permission to release the paper for the prior work, replication of such difficult and unusual publication, Ja¨ckle’s letter contained four stipulations. experiments is of fundamental importance in science and The first stipulation addressed support acknowledge- would have justified publication. Further, we found it dif- ment: ficult to reconcile this criticism with the fact that Logothe- tis was the senior author on the abstract originally present- ‘‘Support of the MPI should be adequately mentioned ing these data [Shmuel et al. 2007]. in the Acknowledgement: ‘The data analyzed in this study The editors noted that the third stipulation did not spec- were obtained in the Lab of Nikos K. Logothetis (MPI fur ify that Logothetis had the right to approve any changes

r 350 r r Protecting Peer Review r made to the manuscript in response to this criticism. The ‘‘...there are two aspects to this decision: 1) scientific authors also viewed this third stipulation as nonbinding. ethics and data management/ownership legalities; and 2) This interpretation was supported by the fourth stipula- publication peer review.’’ tion, which gave Logothetis the right to respond in a subsequent submission. ‘‘As regards the first, the critical issue is who has the right to own/mange the data that were reported.’’ ‘‘In the event that the scientific flaws (I hereby refer to Dr. Logthetis’ comments above) are not adequately ‘‘I understand that the data were acquired in your labo- addressed in the published version, Dr. Logothetis reserves ratory, but this is not necessarily sufficient to establish his right to address them in a scientific correspondence/ ownership....’’ paper.’’ The reader should note that each of these stipulations ‘‘If questions or disputes arise, they are most properly specifically relates to the imminent publication of the al- settled at the institutional level. In the present case, your ready accepted manuscript. Stipulations 1 and 2 relate to institution has already reviewed the matter and determined authorship and acknowledgements. Stipulation 3 recom- that Shmuel and Leopold are entitled to use these data....’’ mends editing the discussion section. Stipulation 4 provides a contingency should the ‘‘published version’’ not ‘‘As regards the second, a standard peer-review process satisfy Logothetis. This is relevant in that some subsequent was done prior to manuscript acceptance.’’ statements attempted to claim that permission to publish was not intended. ‘‘...the peer-review process is complete and I do not In closing, Ja¨ckle extended the permission to publish the believe it is appropriate to over-rule this process, based on present manuscript to future manuscripts using additional input from an investigator having a dispute with the data in alert monkeys. authors.’’ ‘‘Finally, it is my understanding that additional papers After further correspondence, the HBM editors invited from your time in Dr. Logothetis’ lab are in the Logothetis to submit a commentary explaining his scien- ‘‘pipeline’’. Dr. Logothetis hereby authorizes the usage of tific concerns, including data re-analysis, if appropriate. data you obtained from alert monkey experiments. They The editors also requested that the MPS inform the editors can be used - only by you - in the same way as the data of their position regarding republication of the data by were used for the present paper by simply acknowledging Logothetis. Ja¨ckle responded to the editors on behalf of the where the work has been done (see above for the present MPS (May 27, 2008); in doing so, he made several relevant paper).’’ points, including:

Having received confirmation from the MPS of the ‘‘According to the international standard, the data authors’ right to publish and having accommodated the belong to the institute where they were obtained and stipulations regarding authorship and acknowledgements, funded. In this particular case, the corresponding repre- the HBM editors released the manuscript for publication. sentative is Dr. Logothetis (Director and PI).’’ Following the decision to release the manuscript for publication, the HBM editors received numerous commu- ‘‘The Max Planck Society gave the permission to use nications from Logothetis. In all, he challenged the publi- the data (to Shmuel and Leopold) because the authors cation on both ethical and scientific grounds. In some, he insisted that Dr. Logothetis holds the data because of per- argued that because the paper had not been modified suf- sonal, non-scientific reasons.’’ ficiently to address his concerns, HBM editors were wrong to release the manuscript and that it simply ‘‘Dr. Logothetis agreed to the use of data (by Shmuel ‘‘... must be retracted. (May 20, 2008).’’ and Leopold), although he had serious scientific concerns....’’ In other communications, he requested retraction but if this could not be agreed to by the HBM editors, he ‘‘The Max-Planck-Society made the data available to be requested used by Drs Shmuel and Leopold as requested by them. ‘‘... that you publish a commentary that my colleagues No strings and no bias (were) attached.’’ and I will submit immediately after the article goes to print. (May 12, 2008).’’ ‘‘I would like to mention that re-analysis of data is a process inherent to science, and it is my understanding In reply, the editor-in-chief (Fox) sent Logothetis an e- that not only Dr. Logothetis would have the right to re- mail (May 22, 2008) explaining the rationale for the deci- analyze the data: every person in the field, who may have sion to proceed with publication of the manuscript. The scientific doubts must have this right as a matter of course key points were, in order to allow the system to fix possible shortcomings.’’

r 351 r r Fox et al. r

Despite this clear confirmation of Shmuel and Leo- DISCUSSION pold’s right to publish the data, and despite the explanation of the HBM editors’ conclusions regarding the ethics The case of Logothetis vs. Shmuel and Leopold is com- and legalities of the matter sent (May 22) to Logothetis, plex and challenging. Nevertheless, it can be effectively an- and despite the editorial invitation to Logothetis to submit alyzed using the basic ethical principles presented in the a scientific commentary on the Shmuel and Leopold article, Introduction. Logothetis did not accept the decision of the MPS or submit a commentary. Rather, Logothetis went outside of the Analysis of The Relocated Investigators’ Actions accepted peer-review and adjudication processes. For example, Logothetis corresponded with the employers and The present controversy arose, by their own account, colleagues of the authors at their newly adopted research when the Logothetis retracted his permission to publish institutions, accusing the authors of unethical behavior. and the authors did not accept Logothetis’ authority to do Logothetis also continued to pressure HBM editors (by e- so. In our opinion, Shmuel and Leopold were justified in mail) to retract the manuscript and accused them repeat- submitting these data for publication for at least three rea- edly of unethical behavior. When a retraction was not sons. First, Logothetis had acknowledged their right to pub- forthcoming, Logothetis made public statements which dis- lish the data by appearing as a co-author on the original credited the authors and the HBM editors. These public publication [Shmuel et al., 2007]. Second, Logothetis statements began with a widely distributed e-mail (June acknowledged Shmuel’s right to publish the data after their 16) accusing Shmuel and Leopold of unethical behavior departure from his laboratory. Third, they were collabora- and accusing the HBM editors of ignoring his wishes and tors during protocol design and execution, giving them those of the MPS. This e-mail was followed up with a enduring rights to the data, in accordance with widely news article in Nature [Abbott, 2008] about the controversy, respected, published guidelines. Their right to use the data, which included interviews with Logothetis and Ja¨ckle and including publication, was subsequently affirmed by Ja¨ckle which portrayed Logothetis as the injured party. This story on behalf of the MPS. was subsequently picked up by The Scientist [Gawrylewski, On the other hand, the manuscript was submitted with- 2008] and republished online. out informing the HBM editors of the unresolved conflict. The public statements by Logothetis and Ja¨ckle did not This prevented the HBM editors from taking steps to provide an entirely accurate or consistent account of vari- resolve the conflict prior to manuscript review. Prior to ous facts of the dispute. One key aspect of the dispute that manuscript publication, however, Shmuel and Leopold did was repeatedly stated incorrectly was the outcome and appeal to the institution for adjudication, which was termsoftheMPS’adjudication.Asisdocumentedabove, appropriate in our view. Shmuel and Leopold also fol- the MPS had adjudicated the matter and ruled in favor of lowed the stipulations regarding authorship and MPS the authors (Shmuel and Leopold). Further, the MPS had acknowledgements imposed by the MPS adjudication, specified that MPS support should be acknowledged and which was also appropriate in our view. gavetheexactwordingfortheacknowledgement.Yet,in his public e-mail (June 16), Logothetis asserted, Analysis of The Laboratory Director’s Actions ‘‘MPS has explicitly indicated that the accreditation is wrong, and that it does not want to be associated in any In our opinion, Logothetis’ conduct is difficult to recon- way with this particular publication.’’ cile with the RCR and peer-review standards described above in four ways. First, it did not accede to the basic The MPS’ endorsement of the authors’ right to publish RCR principle that authority over data ownership belongs was also incorrectly stated by Logothetis and Ja¨ckle, to the institution, in that he repeatedly advocated against claiming that the MPS only gave authority to ‘‘use the publication by the authors after the MPS had granted data,’’ as distinct from publishing [Ja¨ckle quoted in them the right to publish. Second, it did not conform to Abbott, 2008]. the basic precept that peer review be free of conflict of in- A third error regarded manuscript authorship. Ja¨ckle terest. The authors identified Logothetis as having a con- had stipulated that two engineers associated with the ex- flict of interest to the HBM editors and the MPS, request- perimental work, Mr Augath and Mr Oeltermann, were ing that he not be a party to any decision regarding their not to be included as authors. Yet, in his public e-mail, access to the data or its publication. The HBM editors noti- Logothetis represents this as solely Shmuel’s decision, fied Logothetis that they regarded him as having a conflict of interest. Despite this, Logothetis endeavored to influ- ‘‘It is worth pointing out that he (Shmuel) had invited ence the peer review process, repeatedly contacting HBM Leopold to act as a coauthor in this paper but excluded editors to criticize the manuscript, make accusations of from authorship the collaborators who did the actual data unethical behavior and demands for retraction. Third, Log- acquisition for him.’’ othetis’ conduct did not properly correspond with the dif- To date, Logothetis has not submitted a commentary to ferent domains of authority over the data-management Human Brain Mapping. process (the domain of the institution) and the data-publi-

r 352 r r Protecting Peer Review r cation process (the domain of the journal). In correspon- ‘‘Editors should not reverse decisions when authors have dence with the journal editors, Logothetis argued his case been told that their papers will be published unless serious on both ethical (RCR) and scientific grounds. Subse- problems are identified with the papers.’’ quently, he accused the HBM editors of unethical behavior by complicity with the (perceived) ethical lapse of Shmuel Analysis of the Home Institution’s Actions and Leopold, simply by virtue of allowing the paper to be published. As the HBM editors clearly expressed in writ- The Procedures and Regulations of the Max Planck Soci- ing to Logothetis, the RCR concerns (i.e., issues of data ety are openly accessible. These contain the MPS Rules of ownership and access) are not the domain of the journal. Good Scientific Practice (http://www.mpg.de/pdf/rules Fourth, Logothetis moved this dispute from the proper ScientificPract.pdf). Of these rules, Rule 4, Securing and domains of discourse and adjudication, i.e., the institution Storing Primary Data, addresses data access rights. and the journal, into the public domain. Outside parties ‘‘The (individual) institute management is responsible (supervisors, colleagues, the community at large, the news for regulating and setting out in writing all further media) were appealed to repeatedly, none of whom had details and responsibilities, in particular for detailing any proper role in resolving this dispute. In our view, the proper reporting standards and access regulations for the only appropriate manner in which the ‘‘higher authority’’ use of data.’’ of community opinion should have been appealed to would have been by publishing a peer-reviewed article That is, the MPS locates the responsibility for develop- challenging the scientific status of the results published by ing data access policy at the level of individual institutes, Shmuel and Leopold [2008]. rather than promoting a single policy to cover the more than 80 institutes and research laboratories that the MPS Analysis of the Journal Editors’ Actions governs. Despite this mandate from the central authority, no such policy or procedure statements are readily evident As regards Human Brain Mapping, the RCR analysis is at the web site of the Max Planck Institut fu¨r biologische relatively straightforward. In general, RCR is the responsi- Kybernetik, Logothetis’ home institute. Further, the web- bility of the institution carrying out the research. Journal posted MPS policy fails to dictate specific provisions editors have a responsibility to ensure the integrity of data regarding data access for dissolved teams, relocated inves- management and to ensure that the authors have institu- tigators, or trainees, such as those provided by the policies tional approval to use the data they seek to publish. How- of the University of Pittsburgh. Given the lack of necessary ever, in our view, editors have no charge to independently details in the MPS policy and the apparent lack of local judge the ethical merits of a particular case reviewed by compliance with an MPS-wide directive, it is arguable that institutional authorities, having no direct access to the the MPS might have been better served by having a institutional approval process and documentation. Never- detailed policy applying to all its subsidiary institutions. theless, when the institution’s RCR safeguards fail, jour- In this instance, the MPS President assigned a single nals can become unwittingly involved. At the time this person to resolve the RCR-derived conflict, rather than case arose, HBM had no formal policy or procedures in passing the case to a standing ethics committee. In our place to protect the journal from an RCR issue arising de- opinion, some actions taken by this individual (Vice Presi- spite the due process of an academic institution, such as dent Ja¨ckle) on behalf of the MPS are also difficult to rec- the Max Planck Society. This omission is now corrected by oncile to the RCR principles stated above. For example, the Data Rights Policy and Procedure presented below. requiring Shmuel and Leopold to respond to Logothetis’ As regards peer-review ethics, the HBM editors carried criticisms inserted an RCR action (the institution’s domain) out a rigorous peer-review. Once this was complete, Logo- into the peer review process (the journal’s domain). Addi- thetis was viewed as a nonauthor named by the investiga- tionally, allowing Logothetis to publicly dissent from the tors as having a conflict of interest, and was not allowed MPS decision, once rendered, undermined the authority of to insert himself into the peer-review process. Logothetis’ the MPS as the sole and final adjudicator of RCR matters. criticisms did initiate an editorial review, to judge whether Other actions appear mutually contradictory. His first let- scientifically crucial information might have been withheld ter, from Ja¨ckle to the authors, grants Shmuel and Leopold which would have invalidated the prior review process. the right to publish with some stipulations (listed above). No evidence of this was found. Once the editors found no His second letter informs the HBM editor that the right to scientific grounds for withholding the manuscript and had publish was given ‘‘with no strings and no bias attached’’, confirmed RCR compliance with the MPS’ ruling, we were i.e., with no stipulations. This same letter, moreover, con- ethically bound to publish the manuscript. In this regard, cludes by attributing access to this data to ‘‘every person we followed the Code of Conduct for Editors of Biomedi- in the field’’. Yet later, Ja¨ckle is quoted as asserting that he cal Journals (http://www.publicationethics.org.uk/guide- granted Shmuel and Leopold only the right to use the lines/code) published by the Committee on Publication data, but not the right to publish [Abbott, 2008]. Ethics (http://www.publicationethics.org.uk/) that specifi- In general, the HBM editors respectfully suggest that cally recommends: RCR institutional adjudication of RCR issues should be

r 353 r r Fox et al. r the purview of an ethics committee, rather than an indi- cation in the usual manner, accepting no provisions, stipu- vidual, and that the committee should be guided by a lations, conditions, or duties from any party. well-developed, fully-speciﬁed body of policies and procedures. Data-Rights Procedure

In all submitted manuscripts, it is required that the Recommendations to the Community cover letter state that the authors have the authority to In Poor Richard’s Almanac, Benjamin Franklin wrote, ‘‘An publish the data contained therein. If the PI is not an ounce of prevention is worth a pound of cure.’’ In our author on the submitted manuscript, a signed letter from view, the scientific community can derive the most benefit the PI giving the authors permission to publish must be from this unfortunate series of events by using them to provided. Alternatively, a written statement from a higher guide the development of policies and procedures institutional official can substitute for the letter from the designed to prevent similar occurrences in the future. PI. These higher institutional authorities may include labo- As regards the RCR, we suggest academic institutions, ratory chiefs, center directors, department chairs, or other at the very least, should develop policies similar in spirit officials senior to both the PI and coinvestigators. to those of the University of Pittsburgh and Stanford Uni- In all manuscripts, the usual statement appearing in the versity. Because schools, institutes, departments, centers, methods section regarding institutional authorization to and laboratories may have individually varying circum- perform the investigations must be extended to include: stances that require more detail than an institutional pol- naming the PI on the approved protocol, naming the icy, these organizational levels within the academic hierar- approving institution, and citing the protocol approval chy may also wish to consider extensions of their institu- number. As for example: tional policies to govern their specific situations. Finally, ‘‘The studies reported herein were approved by the Institu- all investigators (including students and post-doctoral fel- tional Review Board (IRB) of the University of Texas Health lows) entering into collaborations should consider imple- Science Center at San Antonio (UTHSCSA) in protocol menting the option indicated in the University of Pitts- #12345, granted to P. Fox as principal investigator.’’ burgh and ORI policies, i.e., a written agreement at the outset of any collaboration. As regards journals and the ethics of scientific publica- ACKNOWLEDGMENTS tion, the editors of Human Brain Mapping advocate that The authors thank Elizabeth Heitman, Ph.D., (Center for journals have written policies and procedures enforcing Biomedical Ethics and Society, Vanderbilt University Med- RCR compliance. To this end, the editors have formulated ical Center), Nancy M. P. King, JD (Program in Bioethics, the following statements of Policy and of Procedure, here- Health, and Society, Wake Forest University), Peter S. Can- after applicable to all manuscripts submitted to Human elias and Gary Egan (University of Melbourne) for helpful Brain Mapping. For additional guidelines on policy devel- discussions and manuscript critique. Ed Bullmore is opment, the ORI website provides a wealth of information employed half-time by GlaxoSmithKline plc and half-time regarding laws, rules, regulations, guidelines, and common by the University of Cambridge. practices, as well as links to other resources. REFERENCES Data-Rights Policy Abbott A (2008): Neuroscientist: My data published without For all manuscripts submitted to Human Brain Mapping, authorization are ‘‘misleading’’. Nature 454:6–7. the authors must have legal rights to publish the data Bandettini PA, Bullmore E (2008): Endogenous oscillations and established in advance by their institution. In all matters networks in functional magnetic resonance imaging. Hum Brain Mapp 29:737–739. regarding authorship and the right to publish, authority Gawrylewski A (2008): Neuroscientist claims stolen data. The Scientist: rests primarily with the PI in whose name the rights to ac- Available at:http://www.the-scientists.com/blog/display/54818. quire the data were granted by the relevant ethical review Logothetis NK, Pauls J, Augath M, Trinath T, Oeltermann A boards. PIs receiving their authority from the institution, (2001): Neurophysiological investigation of the basis of the this authority can be overruled or replaced by higher insti- fMRI signal. Nature 412:150–157. tutional officials. The editors and publisher of HBM do not Rockwell S (2006): Ethics of Peer Review: A Guide for Manuscript assume the authority to arbitrate, moderate, or intervene Reviewers. Available at: http://ori.dhhs.gov/education/prod- in any manner in disputes concerning authorship, data ucts/yale/prethics.pdf. ownership and publication privileges arising between Shmuel A, Augath M, Oeltermann A, Logothetis NK (2007): Sponta- neous fluctuations in functional MRI signal reflect fluctuations investigators. Nor do the editors or publisher of HBM in the underlying local neuronal activity. Neuroimage 36:S58. assume any authority to judge the ethics of any disputes, Shmuel A, Leopold D (2008): Neuronal correlates of spontaneous such judgements being the sole purview of the host aca- fluctuations in fMRI signals in monkey visual cortex: Impli- demic institution. Once the right to publish is assured, the cations for functional connectivity at rest. Hum Brain Mapp editors of HBM will proceed with peer review and publi- 751–761.

r 354 r r Human Brain Mapping 30:1935 (2009) r

Letter to the Editors

Herbert Ja¨ckle

Max-Planck-Gesellschaft zur Fo¨rderung der Wissenschaften e.V., Hofgartenstr. 8, 80539 Mu¨nchen, Germany

r r

In your recent editorial in Human Brain Mapping (Fox conflict of interest, financial or otherwise, related to the et al., HBM, Early View, published online on December 9, submitted work must be clearly indicated in the manu- 2008), you formulated a Data Rights Policy and a Data script, or in a cover letter accompanying the submission’’) Rights Procedure for Human Brain Mapping. The back- when the conflict of interest became evident prior to ground and, in fact, the motivation for this editorial, entitled publication; ‘‘Protecting peer review: Correspondence chronology and – No explanation is given for why the editors did not ethical analysis regarding Logothetis vs. Shmuel and Leo- take into account the serious scientific concerns of Dr. Log- pold,’’ is a conflict over a publication by Drs. Amir Shmuel othetis regarding the reliability of the data, concerns sub- and David Leopold in your journal. Fox et al. lay out what stantiated with a re-analysis of the data, plots and they present as facts on which they base allegations against description of the flawed methods and results; Dr. Logothetis, myself, and the Max Planck Society for the – Inaccurate statements were made about my role in the Advancement of Sciences which are not sustainable. process. I was never an ‘‘adjudicator’’ of the Max Planck I hereby respectfully request the following rectification Society, but instead acted as a mediator; of facts to be published by you as soon as possible. – Inaccurate statements were made concerning the adju- The presentation of the unpleasant episode in the above- dication/mediation procedures within the MPS relevant to mentioned editorial contains serious factual flaws and the ethical analysis in question. omissions as well as pure speculations about the intentions In addition to these formal corrections, there will be a sci- of the different parties which discredit the people and entific rejoinder by Logothetis, N.K., Murayama, Y., Augath institutions involved. Furthermore, quotations from the M., Steffen, T., Werner, J. and Oeltermann, A.: NeuroImage ‘‘relevant correspondence’’ are selective so as to imply con- 2009 (in press). clusions which are neither substantiated nor sustainable. In order to not interfere with the journal’s space for scien- Particularly flawed is the description of the role of the tific matters, the clarifications will be summarized as ‘‘The Max Planck Society, which has been accused of having other side of the coin’’ and will appear on the website of interfered with a peer review process. Just to highlight Dr. Nikos K. Logothetis’ laboratory (www.kyb.mpg.de/lo/ some of the major flaws: index.html). The clarification concludes with a summary of – The editorial does not acknowledge the fact that the the pertinent events in chronological order. The corrections paper in question was submitted without the knowledge, and comments show, in my view, that the authors of the edi- let alone the consent, of the Principal Investigator; torial failed to take the time and effort to properly investigate – No explanation is given for why the editors ignored and report an ‘‘episode’’ for which they provide what they the HBM regulation (which requires that ‘‘any possible call an ‘‘ethical analysis’’ in the above-mentioned editorial.

DOI: 10.1002/hbm.20798 Published online 12 May 2009 in Wiley InterScience (www. interscience.wiley.com).

VC 2009 Wiley-Liss, Inc. r Human Brain Mapping 30:1936–1937 (2009) r

Editorial Reply to Ja¨ckle

1 2 3 1 Peter T. Fox, Ed Bullmore, Peter A. Bandettini, and Jack L. Lancaster

r r

Dear Dr. Ja¨ckle: Regarding the scientific concerns of Dr. Logothetics, the The goal of our editorial [Fox et al., 2009] was to use editorial provided a full explanation of our rationale for this conflict to discuss the ethical principles, which govern not allowing a person identified by the authors as having responsible conduct of research (RCR) and peer review, a serious conflict of interest to preempt a completed peer- and thereby to develop guidelines, which might prevent review process. Dr. Logothetis was invited to provide a future occurrences of this nature. We note with regret that scientific reply, which he declined. We are pleased, how- your reply did not address issues of ethical principles but ever, that a reply will appear in NeuroImage. This pro- focused only on perceived ‘‘factual flaws and omissions’’ vides a proper forum for the scientific community to (Ja¨ckle, 2009a). However, we are heartened that we agree assess the scientific issues at hand, as recommended in completely on ‘‘the only point that is indisputable: our editorial. Schmuel and Leopold had the right to publish :::’’ (Ja¨ckle, Regarding whether your role in resolving this conflict 2009b, pg 11). In the following, we first reply to the per- was as ‘‘adjudicator’’ or ‘‘mediator,’’ this distinction is ceived errors and briefly restate the ethical principles, irrelevant to the ethical analysis provided. Respecting the which govern such matters. principle that the home institution has the right and Regarding selectivity in quotations from the correspon- responsibility to determine for its faculty who has the dence, the editorial quoted only the points of correspon- right to publish, the journal requested a ruling. The journal dence directly relevant to RCR and peer-review principles; was neither privy to the process by which this ruling was quotations were limited to correspondence, which achieved, nor did it need to be. The journal only required occurred after the dispute had been brought to the atten- a ruling, which you provided (quoted extensively in the tion of the Editors-in-Chief and after adjudication had editorial), in favor of the right of Shmuel and Leopold to been requested from the MPS. None of the earlier corre- publish. spondence from the editors, some of which was less well On the matter of the alleged ‘‘inaccurate statements’’ informed and less formal, was included, as it was not rele- regarding the internal procedures of the Max Plank vant to this goal. Society (MPS), we repeat that the Editors were not Regarding a failure to explicitly acknowledge that the privy to the process used by the Society. The editorial manuscript was submitted without the knowledge or con- does note, however, several mutually contradictory sent of the Principal Investigator, this level of discovery is statements made by yourself on behalf of the Society in not the role or responsibility of the journal. When the con- this matter. We respectfully suggest that the MPS may flict was brought to the attention of Editors, it was referred wish to consider formal review of the adjudication/ to the home institution for a ruling. We had (and still mediation process, which you oversaw, and determine have) no direct knowledge of the internal circumstances whether the Society’s interests were well served by the surrounding this conflict. process. Regarding the assertion that the Max Planck Society was ‘‘accused of having interfered with a peer-review process,’’ DOI: 10.1002/hbm.20808 no such allegation was made. Rather, the assertion was Published online 12 May 2009 in Wiley InterScience (www. made (and appropriately documented) that Dr. Logothetis interscience.wiley.com). repeatedly tried to interfere with the peer-review process,

VC 2009 Wiley-Liss, Inc. r Letter to the Editor r even after having been explicitly requested to cease doing were obligated to proceed with publication of an accepted so (e-mail, May 22, 2008). manuscript. In closing, we would emphasize that you do not dispute what we view as the crucial facts of this case: (1) that determinations of data ownership and right to publish rest with the institution, not the journal; and (2) that the MPS REFERENCES confirmed Shmuel and Leopold’s right to publish this data and formally communicated this decision to the journal. Fox PT, Bullmore E, Bandettini PA, Lancaster JL (2009): Protecting peer review: Correspondence chronology and ethical analysis As cited above (Ja¨ckle, 2009b, pg 11), you clearly agree on regarding Logothetis vs Shmuel and Leopold. Hum Brain this point. As Editors of Human Brain Mapping, we can Mapp 30:347–354. assure you and our readership that we would not have Ja¨ckle H (2009a): Letter to the Editor. Hum Brain Mapp 30:1935. proceeded with publication if we had not received formal Ja¨ckle H (2009b): The Other Side of the Coin. Max Planck Institute communication from the Society confirming the authors’ for Biological Cybernetics. (n.d.). HBM—MPG Conflict. http:// right to publish. Having received that confirmation, we www.kyb.mpg.de/lo/hbm/hbmstart.html

r 1937 r