E-biomed: A Proposal for Electronic Publications in the Biomedical Sciences (May 5, 1999 DRAFT)

Go to: [E-biomed comments index] [Link disclaimer]

June 7 - June 13, 1999

June 13, 1999

David B. Robinson, M.D., M.P.H., Alaska Psychiatric Concepts, June 13, 1999

Dear Colleagues:

I tend to agree with those who view the E-Biomed idea with skepticism. With only gross screening for submissions to the general repository, we may erode the standards of scientific proof and blur the public perception of what constitutes an "evidence-based" medical recommendation, at a time when identification of snake oil purveyors is difficult amid the "alternative medicine" response to inaccessibility of health insurance and allo- and osteopathic care. I'm all for breaking down the barriers of academic elitism, nepotism, and expense but not at the expense of methodological refinement, research accuracy, and the epistemological heirarchy which gives us confidence in our accumulating evidence.

David B. Robinson, M.D., M.P.H.

June 12, 1999

Stevan Harnad, University of Southampton, June 12, 1999

On Thu, 10 Jun 1999, D. R. Forsdyke wrote:

>sh> Archives are archives, a reliable, permanent place where all authors
>sh> can self-archive their journal articles on-line for free for all.
> The key here is "permanent". Files can be deleted and interfered with.
> Could you spell out what you mean by permanent?

Los Alamos is good enough for now. When the world's authors' eggs are all in the same virtual basket, collective interests will ensure that they are reliably upgraded in perpetuo. There is NOTHING here to detain us.

>sh> the only function left for the journals to perform will be
>sh> quality control and certification.
> You might point out two kinds of for potential
> readers who are unable for themselves to sort out the "good"
> to help those who judge the authors as meriting appointment, tenure,
> promotion and research funding. In the former case, I think better
> search-engines (see latest Scientific American) and a new class of
> professional sifters/reviewers) reporting directly to the internet, will
> replace journals. In the latter case, merit assessors ("peers") will
> really have to do their the author's application, not
> just skim through his/her publications and tick off how many are in
> Nature or Cell. (see

Nothing at all to point out along those lines:

Both peer review and tenure review could certainly do with some reform. But that has nothing WHATSOEVER to do with my one specific goal, which is to FREE the peer-reviewed journal literature, such as it is, not to "FIX" it.

The main point of my recommendations about self-archiving has been that these two agendas should not be conflated. Free the peer-reviewed literature (through author self-archiving via E-biomed) OR fix peer review, but do not hold the fate of one hostage to the other, especially given that the benefits of freeing the literature are already dramatically demonstrated whereas the reform schemes are all still untested pigs-in-pokes.

Stevan Harnad
Professor of Cognitive Science
Department of Electronics and Computer Science
University of Southampton
Highfield, Southampton

June 11, 1999

American Society of Biochemistry and Molecular Biology, June 11, 1999

Dear Dr. Varmus:

The Council of the American Society of Biochemistry and Molecular Biology has carefully considered your recently released proposal for E-Biomed. We definitely believe we should be alert to exciting developments in the use of the Internet for scientific communication. For this reason, we were particularly interested in the new ideas presented in the E-Biomed proposal. However, after extensive deliberation, we have concluded that the proposal in its present form should not be accepted. We recommend that further study be given to other ways to reach the same objectives, namely, maximum use of the Internet to further scientific communication.

By way of background, the Journal of Biological Chemistry, published by the ASBMB, was the first journal to make the entire text of its journal, including figures, available on-line. With the cooperation of David Lipman of the NLM, the JBC is now "hyperlinked" to Medline, as well as to the full text of approximately 90 other journals. The past few years have seen a dramatic change in publication techniques and a marked increase in the accessibility of journals around the world: since the inception of electronic publishing by the JBC in l995, practically all journals have an on-line version. The JBC has also initiated a simple single subscription system that permits easy access and involves modest charges (all scientists and staff at the NIH can access the JBC on-line for a subscription charge of only $1100). We anticipate that usage of our on-line version will increase rapidly and eventually lead to paperless journals. It is also clear that through the cooperation of many journals we have created a virtual barrier free library of the biosciences.

More recently, we have instituted on-line submission of manuscripts (via the Web), followed by on-line transmissions to the editors and referees and subsequent on-line submission and distribution of reviews. We predict that in the very near future this system will eliminate the need to exchange paper by mail, resulting in very considerable savings in time and cost due to mail or express charges. In addition, this use of the Internet permits easy access to reviewers and editors from all countries, facilitating and enhancing the international nature of science.

Thus, we, and now many other journals, have already initiated many of the publication practices that have been proposed in E-Biomed: electronic submission, review and distribution, publication of large data sets and other digital data like videos which are not suitable for the print journal, very low cost distribution and seamless links among many different journals. These recent accomplishments overlap significantly with many of the changes postulated under item (i) of the E-Biomed proposal titled "Submission to E-Biomed through editorial boards". While individual journals are not currently linked with regard to the handling of submitted articles, the exclusive use of the Web for submission, review and publication of manuscripts is rapidly becoming a reality.

With regard to the specifics of item (i) of the E-biomed proposal, we are concerned that the implementation of a monolithic organization responsible for all of scientific communication could become unwieldy. The establishment of a Governing Board may have a consequence of impeding the healthy diversity of existing journals. Also, the incentive for scientists to serve as members of editorial boards may be diminished with the depersonalizing impact of a central administration via a Governing Board. The advantage of scientific societies controlling the publication of some of the most prestigious journals should not be dismissed without careful thought. In the case of the ASBMB, the finances of the Society and the journal (JBC) are now separate, and the views developed in this letter are rooted in philosophical (not financial) principles.

Concern also arose regarding item (ii), "Submission to E-biomed through a general repository". All of the opinions that have been received from the Council, JBC Editorial Board members and members of ASBMB stress the enormous importance of peer review. Respondents emphasized that peer review gives some assurance of quality and, very importantly, that funding decisions, recruitment and tenure decisions are highly dependent on reviewed publications. While the experience of the Physics journals is often quoted as an example of the success of this type of publishing, the number of articles published in the area of biological sciences exceeds those in physics by a ratio of 25:1 and this discrepancy is likely to grow substantially higher in the absence of peer review. Researchers in the biomedical sciences already complain that there is far more material than they can possibly read. This would become immeasurably more problematic with the publishing of a huge amount of unreviewed material. There is also the question of who would be allowed to deposit information. Would, for example, high school and college students or lay people be able to submit their data, thoughts or hypotheses to the repository? Since this system would include medically relevant material, the availability (and implied approval by the NIH) of preliminary therapies could be undesirable and even dangerous. Finally, there is the issue of cost. Even if the NIH were to find outside sources to fund this initiative, it would still be placing its enormous stature behind the data contained within a non-peer reviewed repository.

In closing, we want to emphasize the commitment of the ASBMB to the use of the Internet to enhance scientific communication. We hope the above comments will be useful to you and taken as the basis for further communication. In particular, we offer the experience of the JBC in electronic publishing as a resource in devising a future plan. Representatives of the ASBMB would welcome the opportunity to meet with you for further discussion.

Judith P. Klinman
President of ASBMB

Barry Markovitz, M.D., St. Louis Children's Hospital, June 11, 1999

Dr. Varmus:

Your proposal for E-Biomed could not have come at a better time, or stated in a better way. The publishing industry has served us well when we had no other means of distributing information; now that we can make information available at a fraction of the cost to a far wider audience, their role must necessarily diminish, as should their consumption of the taxpayers resources. Peer review can and should remain as the best filter we can muster at present, but the bulk of this work is performed as a public service by our peers and can continue no matter what the "publishing" medium.

For those of us who have written, lectured, and given a great deal of thought to how the information revolution created by the popularization of the Internet could be used to disseminate biomedical research (and education) to the widest possible audience with the fewest barriers to access as possible, E-Biomed offers a "virtual" Holy Grail.

Thank you!

Barry P. Markovitz, M.D.
Assistant Professor of Anesthesiology and Pediatrics
Washington University School of Medicine
St. Louis Children's Hospital

Marcia R. Weis, Director, Administrative Affairs, UAB Department of Pathology, June 11, 1999

I read, with great interest, the suggestions about development and implementation of e-biomed. I believe that it is a very interesting concept that has great potential for communicating new scientific knowlege quickly and freely. A few concerns that I have about implementation are:

As it relates to patents and first discovery, there would need to be a mechanism for identifying "first" disclosures -- and it must be accepted by both the legal and scientific communities.

The peer-review mechanism needs to be fleshed out more. Many things in the academic environment are tied to the peer-review process, not the least of which is promotion and tenure. Not only does publication in well-respected peer reviewed journals help determine an investigator's productivity, but membership on respected editorial boards, also, reflect a person's level of respect in the scientific community.

There will still be people in underserved institutions that may not have the same access to online publications as do those people in more mainstream institutions and areas of the world. Would it then follow that those with greater electronic resources will continue to have a leg up on those that don't. In the interest of trying to make things more accessible, are we cutting out an underserved portion of the users?

The internet does not act as a clearinghouse to make sure that the users only get first-class, accurate data. It is very much a "buyer-beware" kind of enterprise. Will making internet publication the norm instead of an adjunct to the print media, will the amount of inaccurate data floating around taint that which is of high quality?

Culture shifts are difficult, changing "the way we have always done it" is a major challenge (especially if "the way we have always done it" still works fairly well). In order for this type of publication norm to be established, it will first be imperative to show the scientific community what they stand to gain and what (if anything) they stand to lose.

Marcia R. Weis
Director, Administrative Affairs
UAB Department of Pathology

Leslie Biesecker, M.D., Laboratory for Genetic Disease Research, National Human Genome Research Institute, June 11, 1999

Dr. Varmus,

The E-biomed proposal has intriguing implications for the study of rare human malformations and syndromes. This area of medicine involves the clinical delineation of malformations and the scientific analysis of the mechanism of abnormal and normal development.

On the positive side, E-biomed could be very useful as a archive that would allow comprehensive indexing and retrieval. Cross-referencing of reports of these rare disorders/findings could facilitate identification of malformation patterns that are sufficiently rare to preclude identification by current practices. In addition, the comment feature of E-biomed would allow many parties to contribute to these reports, enhancing the utility of the report beyond that achievable by reviewers and letters to the editor. These reports are a good example of a situation where the data may be of only minimal interest to the broad scientific community and therefore unpublished or published in obscure journals. However, they may be useful to the clinicians caring for the patient or scientists who study that, or a closely related, disorder.

On the negative side I have concerns about making such data freely available. There are more than a few people out there who harbor morbid fascinations of such afflicted persons and I would be deeply troubled by the abuse of this material. The ability to pick up images, repost them in non-scientific web sites, print them, etc. would not be a good thing, even if it is uncommon. Taking a step back from that concern, I wonder how willing patients or their parents may be to allow publication of such images if there is a potential for abuse. Most individuals affected by such conditions highly value the sharing of personal data among professionals when it might benefit others or their own care. The very problem we are trying to solve (clinical data buried in obscure journals) is a form of protection from unwanted intrusion.

A more general concern about the E-biomed proposal are protections for human subjects. We would need to insure that human subjects protections are being observed in every published report. This is a non-trivial issue when one considers that E-biomed would include data from US institutions/entitites that are not subject to US federal regulations (45CFR46) and international sites that may have very different standards of human subjects protections. Currently, all efforts funded by NIH that involve human subjects must meet standards of 45CFR46 or be individually judged by OPRR (Office of Protection from Research Risks) to meet an equivalent code of human subjects protection. The upside here is that the issue of variability of human subjects protections may finally have to be addressed in order to implement E-biomed for human studies. If that can be accomplished, we can improve human subjects protections, patient care, and the scientific investigation of rare disorders, an exciting possibility.


Leslie Biesecker, M.D.
National Institutes of Health
National Human Genome Research Institute
Laboratory for Genetic Disease Research
Human Development Unit
Bethesda, MD

Nancy Taylor, Ph.D., Greenville Hospital System, June 11, 1999

Dr. Varmus-

Here are my responses to your E-biomed "proposal":

Good things--and true--in this report:

Silly (or--like the first one--stupid) things in this report:

  • that journals have served us well because they have attractive formats and are bound between colorful covers, etc., that they conveniently fit into brief cases (so do computers), and that the pleasure-filled anticipation of their arrival is part of their attraction
  • ". . . liabilities [of the current system] . . . can be addressed by an evolutionary approach that need not threaten most of the benefits attributable to the print-based publication system"
  • the statement--in contrast to an earlier statement of purpose--that this proposal is "an effort to accelerate much-needed public discussion of electronic publication in the United States"
  • "Initially, some authors might hesitate to try this route or might use it only to report information perceived to be difficult to publish in current journals"
  • the distinction between "versions of reports containing supplementary information" and "a 1.0 file for the historical record" as a legitimate way to allow publication of inferior material
  • the whole concept of the Board of Governors as something different from present-day editors and editorial boards, and your naïve belief that such a board could ever be assembled or agreed upon--or that they would know what to do once assembled
  • that E-biomed will be a democratizing force any more than the present system and
  • the calling of this document a "proposal" when it is obviously a proposal only in the sense that we are encouraged to respond to it

Ideas and statements in this proposal that need to be more clearly defined:

  • the claim that electronic publication will "enrich the reading experience"
  • the "third outcome to the review process"--additional listings of unacceptable articles (ie, does this mean that nothing will ever be rejected? Does it mean that editorial boards will give limited "endorsement" to such articles, and how is that possible?)
  • what is meant by the statement that "each report would need to be approved by two individuals with appropriate credentials" (italics mine)
  • how credentials can be "broad enough to include several thousands of scientists" and, at the same time, "stringent enough to provide protection of the database": who will make these determinations?
  • how criteria for approval of reports can be "sufficiently firm to guard against gross abuse" and, at the same time, "sufficiently flexible to permit rapid posting": ibid
  • the meaning of the phrase "virtually any legitimate work" (p. 4)
  • "citation counts"--are these hits or actual citations as we now know them?
  • "the opportunity to supplement the [E-biomed] system with further improvements in the near future"
  • the "new magazines created as guides to E-biomed" ("real" journals or e-journals?)
  • reviews as "part of the vetting process that awards authors" (as in for promotion and tenure?)
  • the plan to "engage the editorial boards and publishers of existing journals, members of scientific societies, and the entire scientific community [talk about overstatement] in a vigorous international discussion over the next few months"--will this engaging be done electronically only?

Questions that need to be considered:

  1. Is physics, as an academic discipline, different from the academic disciplines of medicine and clinical research? Ie, if physicists allow everyone access to their preprints, does it follow without question that preprints of medical and clinical research projects should also be made available, or are there significant differences between the two? Another important question: were the physicists' preprints pre-peer reviewed before they were made into preprints?
  2. Is there a difference between the usefulness of universal access to data repositories and universal access to flawed interpretation of data of whatever kind?
  3. Would anyone--anyone--take the time to read, online, all the "appended commentaries" to the reports submitted through the general repository? (I won't even scroll through all the responses to your proposal, and I'm adding to them.) Even if we did, would skimming such commentaries suffice for understanding the results of clinical reports? How, then, will we know--any better than the present system tells us, and it's not totally dependable--whether the report in the general repository is bunk?
  4. Is it true that the only bad things the authors are concerned might result from either method of submission are "extraneous or outrageous material" and "gross abuse of the E-biomed repository"? What about the death of patients?
  5. Are the "novel peer review mechanisms, appended commentaries, citation counts, and accession data" necessary to "enhance the status and prominence of a report"? Wouldn't we rather the article had innate worth and not need enhancing?
  6. Will electronic review really be faster than regular peer review? Just because it's on your computer rather than on your desk, will you get to it any sooner, especially since you'll now have all this extra information at your fingertips that you need to read?
  7. Am I correct in assuming that you believe peer review will be speeded up in part because fewer people will choose to submit under method (i) and therefore there will be fewer articles to review? Yes, you are correct that excellent peer review (bad peer review, too) does indeed take time, and since time is of such importance to those wanting to get out their material, fewer of them will choose to go through peer review.
  8. How will electronic "editorial boards that develop the most efficient and most accessible review processes" be any more successful than or any different (except in time to "print") from current boards in competing for the best reports?
  9. What human being would like to live without libraries? We already knew scientists had access to the Internet. We also know that many equally promising individuals do not. They--and some of the rest of us--might need and even like to have libraries.
  10. Can those of us at "less prestigious institutions" (that's us!) really believe someone is going to give our papers equal consideration, even online?
  11. Do I understand you correctly to be saying that especially we "remotely located and poorly known" individuals will be submitting our reports only to the general repository? Is that because you suspect we'd never get electronically or paperly published through (i)?

The two most serious questions I have about E-biomed are ones you have not even addressed:

  1. Why, in the general repository, should we democratically make available huge amounts of information that may be unreliable?
  2. Whether the peer review system we now have always leads to publication of reliable information is not a given. But without peer review, especially in papers focusing on medicine and clinical research, how will physicians know what to believe, what is valid information for them to use in their own work? How will they know what information they can use when they treat us, their patients?


Nancy Dew Taylor, Ph.D.
Medical Research Writer
Department of Research
Greenville Hospital System
Greenville, SC

J.W. Suttie, Editor, The Journal of Nutrition, June 11, 1999

Dear Harold:

I appreciated the opportunity to sit in on your discussion of the E-Biomed Initiative with the FASEB Journal Publications Committee last week. Few would deny that rapid, easy, and inexpensive (or free) access to current biomedical research publications is an admirable goal that would be of value to all of us. If such a new effort would positively impact on the escalating costs of the journals (mainly the commercial publication) at the same time, it would be even more advantageous. It was apparent from the discussion that many see a move in this direction as a threat to the societal journals. I personally feel that the existing peer reviewed journals with their infrastructure in place will be needed for any plan that might be developed, and am not as concerned. I would hope, however, that as discussions proceed, the societal journals' editors would be involved. I have suggested to David Kaufman, the current FASEB President, that FASEB might take the lead in bringing this broad group of societal journal editors together for a serious discussion of the current and future roles of this important component of the publicatioin process.

From the perspective of a current journal editor, I would make these comments on the general plan. First, I would like to see a real effort made to utilize the current commercial distributors of on-line journals in some manner rather than to invent a new system. A significant portion of the current biomedical literature, and certainly most of the more prestigious journals, are now available through these sites, and some way of utilizing this in-place mechanism must be possible. Second, two aspects of the general proposal seem very positive: a mechanism for electronic sharing of large data bases that do not fit the current publishing mode and the scanning in of some portion of the past literature. Finally, an aspect of the general proposal that I do not find so positive is the posting of unreviewed manuscripts that will be archival and will be cited. I am not convinced that the potential good would outweigh the trivia and premature publications that could result.

The issues that you have raised are important and will obviously continue to demand our attention. I do hope that as discussions move forward there will be an increased amount of both formal and informal input from the large number of biomedical scientists who have spent a great deal of time working for their societal publications and have much to contribute.

Sincerely yours,

J. W. Suttie
Editor, The Journal of Nutrition

Michael Jacobson, M.D., M.P.H., FACP, Journal Club on the Web, June 11, 1999

Dear Dr. Varmus,

I have been following with great interest the debate concerning E-biomed. Whatever the outcome, the NIH is performing a tremendous service in bringing this topic to widespread public attention.

As evident from the discussion published on the website, most of the arguments about E-biomed center on two basic areas: quality control / peer review and copyright/publishing issues.

The most prolific author on the copyright question has been Prof. Stevan Harnad, with relatively few direct responses to his remarks. I would like to address his comments. I am a practicing internist, with an independent medical website (Journal Club on the Web and am particularly interested in the topic of biomedical publishing and intellectual property rights.

Dr. Harnad's basic premise is that although biomedical journals are well-suited to perform peer-review, they no longer have the legitimacy to usurp article authors' right to distribute their work. The reason authors ceded copyright protection for their work to journal publishers in the first place was because they had no other way to distribute their research results. According to Harnad, scientific authors were forced to strike a Faustian bargain. In his words:

> "The Faustian Bargain in the past was that all authors
> transferred copyright to their publishers because that
> was the only way they could gain the immortality of PUBLICation.
> This was fine for the trade literature (all books and magazines),
> because those authors contributed their texts for fee or royalty,
> and shared in the take from the toll-gate receipts. But this
> was never true of the refereed journal authors, who wanted
> only to reach the eyes and minds of their fellow-researchers
> with the reports of their research findings, so their work
> could have its full potential impact, and be built upon as
> broadly as possible. Yet they too had to transfer copyright,
> because there was no other way to cover the real expenses
> of paper dissemination. The access-restrictions imposed by
> the toll-barriers were against their interests, but the
> only alternative was even worse, namely, no access at all."

Now, with the widespread availability of the Internet and the ease with which one can distribute intellectual content online, authors can distribute their work worldwide, without needing to use the mechanism of paper-based journals, and should thus no longer be forced to give up their property rights. Of course, according to Harnad, journal imprimatur will still be needed and useful for vetting the quality of work. But authors should be free to publish online copies of their work before it is accepted by journals ("preprints") and after it has been accepted ("reprints"). This online publishing by authors is what Harnad calls "auto-archiving", and already is the standard in the world of physics. Why should biomedical journals allow this to happen? Because they have lost their power over authors, since researchers are no longer dependent on journals to distribute their work.

The problem with this analysis is that it attributes the stranglehold of publishers over authors solely to the ability of publishers to distribute scientific work. In fact, the reason scientific authors desire publication in the most prestigious journals is the same motivation that drives authors in other fields of endeavor: recognition and career advancement, or just plain fame and fortune. Obtaining the widest possible audience for their work is part of this, is both a prerequisite for and a consequence of recognition, but is not the entire goal in itself; the goal is also recognition and career. Much as I would like to believe that researchers want only "to reach the eyes and minds of their fellow-researchers with the reports of their research findings", I fear that the motivations of most of them are somewhat more complex.

As long as the leading medical publishers can dole out career advancement by rewarding authors with publication, they will be able to do so on their own terms and can continue to demand ownership of intellectual property rights. Just as the author of a detective novel will sell her copyright to a publishing house in return for distribution and financial reward, the author of a scientific paper will sell her copyright to the journal, in return for distribution and career advancement.

The ability to self-publish and self-distribute, or auto-archive, on the Internet in no way lessens the ability of biomedical publishers to influence the careers of researchers, and thus does very little to lessen their overall power over scientists.

Why has the power of publishers apparently not succeeded in resisting the power of the Internet in the field of physics? Although I am not intimately familiar with the situation in physics, I would assume that the amount of money at stake for publishers of physics papers and their power over career advancement are not sufficient to win the battle. In medicine, vast sums of money are at stake: the health care sector comprises some 15% of our economy. Pharmaceutical companies have fortunes to spend on advertising, which goes straight into the pockets of journal publishers. And the careers of researchers rise and fall on their publication in the most prestigious journals. Furthermore, many of the most prominent researchers are on the editorial boards of journals and have a vested interest in the continuation of this system.

Thus, unless the "unpublished researchers of the world unite", and overthrow the industrial-editorial complex (a rather unlikely scenario), the current status will not change greatly, at least as far as intellectual property rights are concerned. The New England Journal of Medicine will be able to enforce its Ingelfinger rule, if it so chooses, and will be able to interpret and enforce its requirements on authors.

Of course, journals are adapting to online publication and will continue to do so. They will surely collaborate with the NIH in order to allow more widespread distribution of their content, perhaps at reduced cost. But this will occur in a negotiated fashion, and is not likely to entail eliminating the toll-gate system that Harnad so deplores. The NIH would be well advised to consider the copyright system currently in place and its evolution (or lack thereof) in its laudable plans to make biomedical information more accessible to all.

Michael Jacobson, M.D., M.P.H., FACP
New York, NY

Shelby L. Berger, Section on Genes and Gene Products, National Cancer Institute, June 11, 1999


Shelby L. Berger
Section on Genes and Gene Products
National Cancer Institute

The advent of the internet has begun to change not only the way we communicate, but also the frequency with which we do it, the length of our remarks, the formality with which we write, and the speed at which we exchange information. These transformations are also being accomplished with a tremendous reduction in cost, the extent of which we can probably not yet imagine. With the internet poised to change business as we know it, we might take a few moments to consider how it will change our business-science-and how we should harness the electronic media to further the conduct of research. Let it be understood that the internet comes with no guarantees. Indeed, in the news business the speed of communication perpetrates truth and falsehood with equal efficiency, just as it can in our business. Furthermore, as we deploy the new technology, we should be aware that the decisions made now might affect aspects of science far removed from communication per se.

Scientists are encouraged, perhaps mandated, to publish results in a timely fashion. The purpose of publishing is to convey reliable information so that all others may benefit either by furthering their own investigations based on the new data or by enhancing their understanding of the subject at hand. However, scientific publications at the close of the millennium are being asked to serve additional functions: they are the fundamental means by which achievement is measured and rewarded. We probably agree that those who have not succeeded in obtaining new information from their endeavors have nothing to publish. But if one does publish prolifically is that an indication of prodigious achievement? Should a system of publish or perish be propagated?

What we publish and why we publish it are no longer tightly connected to information transfer. What do we publish? Everything that we can. There is assuredly more "information" than is fit to print and far more than moves a field forward. For many, the least publishable unit is the ideal. I sat on a promotion and tenure committee at one time when a fellow member asked of the candidate's oeuvre, "If all of this disappeared beneath the waves, would the fieldhave suffered? Did all of those publications make a difference? Did any of them?" There are also examples of brevity, but not commendable ones. Fueled by the space limitations of the journals, the authors deliberately telescope the descriptions of methods and ignore vague passages.

Why do we publish? There are several answers. (i) We generate publications to obtain funding for future investigations. Even though the grants procedure tries to stress quality over quantity by limiting space for lists of publications on applications, deep down, we all behave as though "more is better." Quality is hard to measure, time consuming to determine, and subjective, whereas quantity is easy, quick, and objective. (ii) The published record serves as the infrastructure on which recognition is based. For some, travel and invited talks, together with the ego gratification they provide, are the essence of science. It's fun to be in the public eye. Papers aren't just meal tickets, they're airline tickets, too. (iii) Publications are required for promotion, as are letters of recommendation, which may or may not reflect the strengths and the weaknesses of the included bibliography. Often, one wonders how much of that list committees and referees have read before pronouncing judgment. One wonders, also, whether publishing manuscripts of inferior quality is a negative factor or whether poor work is disregarded when accompanied by noteworthy contributions. (iv) And finally, we publish to disseminate information to our peers. Our publications showcase our creativity, our perceptiveness, our analytical capabilities, and our grasp of detail, as well as serving as the ultimate record of our discoveries. But because we crave credit, we sometimes ignore the fact that publications are reflections of self. We compromise; we publish fragmentary material to protect our interests and to establish our priority; we try to obtain recognition for work slightly sooner than we have, in fairness, accomplished it; we don't calculate the cost both to our own reputations and to a vocation which depends doggedly on truth. With so many competing reasons for rushing into print, is it surprising that the prime reason-information transfer-isn't always foremost?

Will use of the internet for publishing return the focus to meritorious work, well written to aid in its dispersal among a broad spectrum of scientists? I believe the potential exists but is by no means a certainty. In order to explore the possibilities, it will be necessary to understand the platform which might be erected to support papers or abstracts and to establish the rules that will determine how manuscripts will be posted, revised, and maintained. It will be essential to determine the cost and to ascertain who will pay it. Finally, one must evaluate publication on the internet from the point of view of the authors, the readers, and the many administrators and committees who use the published record to discern the fitness of the author.

On the internet, space does not limit what is published or posted. Nor is cost the major consideration. With the dual constraints of page and price removed, it will be possible to restructure the publication process. If one thinks of the internet as a journal on a computer screen, one has failed to comprehend the revolution that is under way.

In the internet world, it will be possible to "publish" everything. If a particular web site restricts access in any manner, other web sites will arise to take up the slack. If an investigator becomes frustrated with the regulations at a given location, publication is available on one's own home page or with friends. Let us assume, therefore, that scientists will no longer publish everything-they-can-get-away-with, they will publish everything, period. If one sees this as the beginning of chaos, the descent into hell, or the end of quality as we know it, please read on.

Instead of decrying the apparent loss of standards, we should embrace the new-found freedom. We will be able to say what we please whenever it pleases us to do so. We will no longer revise a manuscript according to the wishes of a reviewer if we do not agree with the recommendations. Did the Impressionists repaint their canvases because the Salon rejected them? The Establishment will not be in a position to dictate terms. Our voices will be heard notonly as authors of manuscripts but as writers of letters to the webmaster (there may be no editors), as reviewers, and as critics. We will be able to appraise the work of others online as well as rebut analyses of our own contributions. There could be an end to the anonymous referee. Is there a downside to such an all inclusive conversation? You bet! In such a permissive environment, the internet may collect the largest pile of drivel seen by mankind. But just as some obey the law whether or not a policeman enforces it, some will choose to do quality work because they aspire to excellence.

I foresee several to many very lenient web sites. Theoretically, a lone site would suffice. However, a single site would stifle innovation at every level from the design and arrangement of the posted (published) material to the format for discussion. Since work can be continuously revised, rules will evolve to determine how and whether the original work will be maintained, how and when revisions can be effected, and how alterations, corrections, and additions and subtractions will be cited in the literature. No doubt, web sites with different points of view will appear. Some might focus on a subject, or emphasize discussion, or assist the reader with summaries of key papers. Sites or sections within sites devoted to science policy, government, and administrative events will arise. Other sites might attempt to re-create an establishment, with reviewers and editors poised to stifle expression in the pursuit of quality. Human nature being immutable, there will be in-groups, who will refuse to post your paper, and outlaws, who occupy the lunatic fringe. Regardless of the orientation, a web site will have to woo readers, just as journals do now, but with a difference. Journals, regardless of the number of people reading them, are supported primarily by subscriptions and page charges; few attract sufficient advertising to pay the bills. Web sites, in contrast, want only readers to be successful. There need not be subscribers because a modest fee for posting might suffice. Furthermore, since libraries rather than individuals purchase most printed material, journals are no longer devoted solely to their readers. Web sites, without the enormous overheadinvolved in type setting, printing, binding, and mailing, can easily be reader-driven.

It should now be evident, that the act of publishing, as such, will become irrelevant and that the reader will reign supreme. To attract a clientele, it will be necessary to provide the end-user with quality, certainly with a useful commodity or an interesting one. From the authors' point of view, the focus will shift significantly from getting published to getting read. Rewards that now depend on the number of papers and the journals that publish them will be meted out on the basis of readers. It will be impossible to ignore the comments, praise, and criticism that our readers append online. Work will be reviewed not by an anonymous peer or two, but by everyone who cares. Since it is easy to count and identify readers, the impact of a specific article can be made available and assessed by all. I believe that the emphasis on reading will be salutary. One also hopes that if no one must peruse a paper-not investigators, not reviewers, and not editors-there will be greater stress on courting the reader by posting only completed work, written well. Clearly, improvements in content, clarity, and style are long overdue. I am reminded of my college English instructor who opined that he read my essays because he was paid to do so. "But when you leave school," he said, " Nobody will be required to read what you write and, unless you markedly improve, nobody will." Amen!

The internet will destroy or change the journal as we know it. In order to justify the extraordinary cost of a journal relative to that of internet publishing, the journal will have to add value or metamorphose into a web site. The paraphernalia of acquiring, storing, cataloging, refurbishing, lending, and retrieving journals-libraries-will also cease to play a dominant role in scientific communications. With revenues normally disbursed for subscriptions, page charges, and other publication costs unspent, additional funds will become available for research and related activities. Furthermore, in the absence of journals, the internet will become the vehicle of choice for advertising and resolve the issue of funding. Since advertisers will also be able to ascertain the impact of a particular paper, they will have the capability ofdirecting their messages toward those most likely to purchase their products. Even a small, boutique web site, serving a group that does not frequently browse elsewhere, should be able to attract its fair share of the lucre.

The massive number of dispatches disgorged by your computer will commingle information with disinformation. The responsibility for sorting through the melange will devolve onto each of us acting in our own best interest. However, I foresee that a new class of individuals will be introduced into the world of science to aid us. For lack of a better term, let me call them the critic class-an extension of the scientific fourth estate. These individuals will not be investigators, although they may be drawn from the investigator class and train with it. They will be well organized, knowledgeable, perceptive, critical, analytical, extraordinarily fair, even handed, and literate. Each will choose a beat and organize the submitted manuscripts and commentary within it for assimilation by the investigator class. Far from producing annotated bibliographies, the critics will exult in good papers and disparage inferior ones. No fuzzy bears, these; it will be their job to point out that had Dr. Smith not ignored the possibility of an artifact and confronted it with an experiment, his opus might have been proof rather than provocation. Their column, "This Week in the Nuclear Matrix", for example, will review what's good, highlight what's controversial, ignore what's mediocre, and consign the dogs to the dog house. Authors! Your publication woes are over! Your greatest fear may be in not getting reviewed. In this conformation, science will draw closer to art, where post-publication review by individuals, who are not practitioners, is the norm. If the critic class seems authoritarian, compare it with what we have now, a censor class with near despotic powers. Unlike editors, critics will be unable to control what is published; unlike editors and reviewers, their work will be signed and subjected to review by us. On occasions, the investigative class may usurp the powers of the critic class to set the record straight.

Since the critics will not be practicing scientists receiving funds from universities,institutions, and grants, it will be necessary to devise schemes to remunerate them. Clearly, with the decline of the journal, sufficient revenue to support a robust class of critics should be available. Such individuals might be free-lance essayists, syndicated columnists, or employees of web sites, universities, companies, foundations or other eleemosynary institutions. Editors who have glimpsed the internet and recoiled might consider joining their ranks. For those who already possess the necessary skills, entry into the new medium is as simple as a shift from backstage to center stage.

Is internet publishing inevitable? I believe so. Driven by cost effectiveness, it will eventually displace more expensive methods of communication. Were there no other benefits, and there are many, the internet would be here to stay. Investigators might presently be loathe to risk their grants and promotions by selecting the internet over a journal, but that choice will soon become routine practice. Once the switch has been made, the contempt of the critics for rubbish might encourage quality to triumph over quantity. If a major institution chose to lend its good offices to such endeavors, the new era could commence now.

June 10, 1999

James Ankenbrandt, Systems Engineer, June 10, 1999

I saw the comments about your proposal on the page and came to look for myself. What you propose is evidence that there ARE people in government that understand current events and are willing to look beyond their fiefdoms to help form a future that is better then today.

Jim Ankenbrandt
Systems Engineer

Lee Giles, NEC Research Institute, June 10, 1999

Dear Dr. Varmus:

After our talk in CIS at Drexel yesterday on our autonomous citation indexing project, we were told that you might be interested in possibly using it for scientific indexing at NIH. A brief description of it can be found at and a demo can be found at

Keep in mind, it has been set up for indexing primarily computer science papers found on the web and has about 200K documents and 3M citations but would easily extend to any papers in a scientific/technical format.

One goal of this project is to give away the software for noncommercial use. If you like what you see, we can come down for a seminar and demo.

Best regards
Lee Giles
Steve Lawrence
Computer Science
NEC Research Institute

PS: You might find our demo metasearch tool of use. It can be found at

Rares Marian, June 10, 1999

As a netizen (whatever that is) I applaud your efforts to open up the information. I cannot for the life of me understand why bureaucrats, luddites, and the just too-dense can't simply burn an archive of resilient and durable CD's that include the viewing software on them. I'm shocked at how we still think of the Internet as a miracle.

Here's the link to the discussion on

Rares Marian

Rob Bos, Simon Fraser University Computing Science, June 10, 1999

Sir, I have to say that I really do respect your incredible work in expediting the flow of information in the medical community -- you are absolutely correct in realising that it could revolutionise research in the medical sciences world-wide to allow collaborative research on a hereto unheard of magnitude. It is foolish to think otherwise -- in many ways, advances in science strictly parallel advances in communication.

However, I feel that your approach to implementing the structure of publishing articles is fundamentally flawed in that you pay no attention to "heirarchies" of usefulness -- that is, some articles are more valid than others, some have better evidence than others. Exclusion of some topics of conversation would inevitably result because of this flattening out. People could not judge which is most important to them, or decide how deeply they wish to research a given subject.

The adoption of a more traditional, proven, collaborative style of online publishing may be more prudent -- the methods of publishing that apply to printed journals simply do not work online.

In short, I think a far more anarchistic model of publishing should be used, with the principle of peer review guiding which articles are truly relevant. Any one person, regardless of such petty things as "qualifications" should be able to submit their research to a central repository, and all readers should be able to page through it. That's what peer review /is/. Comments on that research can be added, and a system by which registered, qualified members of that site can decide which items go up to the next heirarchy, say, to a weekly highlights and updates page. The archives would maintain all submitted articles for date, with an appropriate search engine.

The print version could take the most compelling of these, and publish them along standard channels.

What Will Not guarantee the success of such a system is the use of an "editorial board" to censor and determine the final content of the overall site -- for even if an article is factually unsound, it may have ideas that could be of use at some time in the future, and ideas are key to the information age.

Such a system does not historically scale well -- that is, as submissions increase, the editorial board gets overworked and it is more and more difficult to filter the content.

In other words, the Internet is a bottom-up medium, where content is selected by peer review for gems. The higher up the heirarchy it is, the harder the scrutiny. What you are proposing is a top-down system, where the editorial board can reject(!) content, reject information! Requiring any one to have the approval of two or more "approved" people to submit to a public database is nothing short, sir, of criminal, and has the potentiality of covert censorship.

Examples of the principle at work are evident wherever you look on the Internet. Software, such as the Linux kernel, is developed using heirarchies of importance -- any one can submit a patch, which is reviewed by the people in charge of certain areas of code, and the final say is in the project leaders' hand.

I congratulate you again on your work in fighting against the somewhat ossified system of publication currently in place, and helping to create a more fluid, responsive, adaptive, and stronger research community.

Oh.. yeah. "E-Biomed" is a bad name. Sounds very commercial, and, well, kinda tacky.

Rob Bos, undergraduate, citizen of the world.
Simon Fraser University Computing Science.
(and no jokes about how anything that calls itself a science isn't).

John Lynch, University of Pittsburgh, June 10, 1999

To whomever it may concern,

I wish to praise the ideas behind e-biomed. Making scientific, especially biomedical, knowledge available cheaply and readily to the public should be a top priority for the NIH. Not only does it have the potential to increase medical literacy amonst the population at large, but the possibility of a larger audience to review and respond to research endeavors should be a clear reason for the biomedical community to sign onto this project.

Will there be problems? Most likely. Will a few quacks get through and gain a forum? Perhaps. How is this different from how things happen today, when quacks get the ear of the public anyways and often look credible (sometimes even more so than legitimate scientists)? It isn't.

Please continue democratizing science.

John Lynch
University of Pittsburgh

Tim, June 10, 1999

Dear Dr. Varmus,

Please ignore the criticism from bitter cranks like Dr. Arnold Relman and push through your idea for a peer-reviewed public website for scientific research reports. The journals that deserve to survive will do so, and in all cases it should force the publishers to electronically distribute the information for local hardcopy preparation. Surely it is not too difficult to obtain the (free) Linux OS, setup the (free) tetex typesetting package, and teach librarians how to print hardcopies using dvi2ps or similar. Don Knuth knew exactly what he was doing when he came up with TeX and DVI, and they are as useful today as 30 years ago, since they are not hampered by allegiance to any specific device or software.

[some of the following points echo those in your on-line draft of the proposal]

In point of fact, the process could easily be automated and the cost of publication vastly reduced for the publishers of said minor technical journals, whereas a subscription to Science or Nature is cheap enough to be worth buying. So the question of "who will be able to publish marginally relevant results in fringe disciplines" should not be a factor.

More importantly, with well-chosen reviewers earmarked to referee the papers, the review process ought to be more convenient and efficient for all parties involved. Proper selection of reviewers and distribution of notices to referee a paper could be a challenge, but should be trivial compared to some of the problems solved daily at NIH. And there is no reason to suppose that prestigious journals would not remain, if only because their cachet (when they appear on a CV) is such a boost to a scientist's career.

Relman's scare tactics are ridiculous -- the web is already a hotbed for fraud, chicanery, and dangerously flawed pseudoscience. A filtered stream of scientific papers, made accessible by a respected entity like NIH, would greatly improve the situation, as NIH has /earned/ the trust of the public. There is no reason why publication to the website should be "hastily published" or "without accompanying expert commentary" any more than it is necessary for the NEJM or Journal of the ACS to publish any old crap that arrives at their doorstep. Dr. Relman's comments represent the Good Old Boys' Club side of medicine at its absolute worst, and his self-serving comments about "the livelihood of traditional publishers of scientific research" being endangered are nostalgic at best. The "few weeks saved between acceptance and print publication" are not what most researchers care about. The few thousand dollars saved by libraries (per journal, in some cases) and the few million hours of gofering in university libraries for current results saved by practicing researchers and their students are what matters.

Basically, the name and expected date of delivery for e-biowhatever is irrelevent. It is an idea whose time has come, and one that has already seen parallels in several successful "communities" like Slashdot and photoNet, where a process not unlike the imperfect refereeing of research papers goes on for individual comments. These mass-moderated forums tend to be most successful at weighting objective information, which bodes well for your proposal's ultimate success. As mentioned above, the hurdle for a respectable scientific "community" site would be determining whose opinions are relevant to each paper reviewed in a timely fashion, so that the ease-of-use advantage of electronic distribution is extended to the reviewers as well. ( in particular has some discussion of how to engineer a successful, usable community which properly moderates itself given minimal supervision)

As a personal aside, I built a trivial version of what you are proposing myself as a student at Cornell back in 1994, long before I knew how to do things the right way, and can attest to the importance of properly weighting reviewers so as to avoid corporate influences.

I sincerely hope you are able to push this through, as it would improve the public perception of NIH and vastly increase the efficient dissemination of experimental results. My father worked at NIH for 14 years because he loved the caliber of research being carried out on the Bethesda campus and funded elsewhere. He left in part due to larger salaries in the private sector, but his departure was eased by people who acted like Relman, even at NIH. My sister is working at NIH right now and sees the same paradox. I hope you ignore the old cranks and place the needs of future researchers like her above their grumbling.

"When it is not necessary to make a decision, it is necessary not to make a decision." --Lord Falkland

Albert Henderson, Editor, Publishing Research Quarterly, June 10, 1999

Quite a number of larger publishers have already put their journals into the web, as can be seen from many comments already submitted. These initiatives have the advantage of being specialized, presumably tuned to the interests of their authors and readers. As such they will offer a diversity of format, review, editorial coverage, and editorial supplements not likely under a government program that is designed from an administrative point of view.

There are also a number of publishers who, by virtue of their smaller size, lack resources needed to employ this technology.

NIH would make better use of its resources by helping individual publishers make appropriate choices in upgrading their distribution technology. One of the challenges to updating is the present need for each publisher to invest in programming and rearrangement of work flow. Helping publishers to learn from each other by means of conferences and surveys would be helpful. Providing grants and technical assistance would also be welcome.

What can be done to assist the existing indexes and abstract services, including those run by the National Library of Medicine, to extend their coverage? These services were the first electronic publishers, thanks to government support in the 1960s. They should, at least, keep up with the growth of research output that they are expected to cover.

What can be done to synthesize and refine research findings, to bring coherence and quality? The answer provided by many editors and information scientists requires the interference of judgment. NIH would do well to allocate more resources to comprehensive reviews of the published literature. We need more and more peer review, not less.

Albert Henderson
Editor, Publishing Research Quarterly

Lee Giles, NEC Research Institute, June 10, 1999

We have autonomous citation indexing tool which we will be making freely available which may be of use in your E-biomed project. Right now the database is for primarily computer science papers we obtained from crawling the web. However, any paper in a scientific format can be used.

For a demo, please go to

Best regards
C. Lee Giles
Computer Science
NEC Research Institute

PS: Please try our metasearch tool at

Susan M. Sparks, R.N., Ph.D., FAAN, Division of Extramural Programs, National Library of Medicine, June 10, 1999

The following is the URL of an article published electronically and in hardcopy in Nursing Research in January/February 1999, entitled Electronic publishing and nursing research. Please be sure to take a look at the Electronic Publishing Preferences polling for some interesting data!

It is very similar in opinion and perspective to the draft at

Susan M. Sparks, R.N., Ph.D., FAAN
Division of Extramural Programs
National Library of Medicine

Ajit Varki, Professor of Medicine, UC San Diego, June 10, 1999

Please consider this listing, which is based on 5 years experience as the Editor of the JCI, and 25 years as an author.

Pros and Cons of the Current Peer Review System for Biomedical Research

Can help the author to improve the quality of the work
Can prevent or modify the publication of erroneous information
Can keep reviewers abreast of the latest advances
Can set standards of quality that all scientists can aim for
Can help scientists to decide what to read first
Can help academic institutions and granting agencies evaluate quality

Can delay the time to initial publication
Can give a technical advantage to unscrupulous reviewers
Can allow unscrupulous reviewers to delay the time to acceptance
Can allow unscrupulous reviewers to block acceptance
Can allow Editors to exert undue influence over dissemination of information
Can result in some unfair decisions

Ajit Varki
Professor of Medicine
UC San Diego

Peter Ukpeh, M.D., FACOG, June 10, 1999

Excellent idea that's long overdue!!!

Peter Ukpeh, MD., FACOG

Dale Armstrong, Kansas City, Missouri, June 10, 1999

Dear Dr. Varmus:

As an outreach professional in the public sector, I fully support your proposal to post research online.

Objections to doing so range from the inane to the conspiratorial. AP quotes the New England Journal of Medicine as saying, "It would make the journal - the paper journal, and also the journal Web site - merely archival. Redundant. Insofar as that happened, it would weaken the journals and maybe even destroy them." JAMA jumps on the same bandwagon, saying "... journal publication would be reduced to an archival function."

What nonsense! The journals can make what they want of themselves. They will only become archival and redundant if that's what they want to be. Why do they sound like the priests of not so long ago forbidding commoners to own or read bibles because they were not properly trained to understand them? Could it be that the high priests of medicine have a vested interest in holding close the secrets of the priesthood?

Those who seek knowledge will read and strive to understand. Those who don't won't waste their time. In either event, I assure you from many years of experience that the public is far better qualified to judge for itself what it wants to learn than any high priest of any profession.

If the journals fear irrelevancy, perhaps they should consider constructive criticisms of research published at E-biomed. But they have no business being the public's gatekeeper. Heaven save us from those who would protect us from ourselves.

Dale Armstrong, Kansas City, Missouri

June 9, 1999

Ray Esper, June 9, 1999

I am writing in regard to the proposed publishing of research materials on the Internet (E-Biomed). As a biomedical researcher, I applaud the attempts by the NIH to quickly distribute scientific findings. Some have argued that releasing this information before the process of peer review and scholarly publication will pose a threat to uninformed patients and consumers. I disagree.

Those of us in research laboratories must regularly confer with our colleagues to share research findings and theories. If the NIH publishes research reports through an easily accessible medium, such as the Internet, it will only serve to aid researchers in our various projects by allowing us to share information more quickly, consistently, and reliably.

Any small risk of potential harm to uninformed patients (who generally couldn't utilize the research without their physicians anyway), is greatly outweighed by the tremendous benefits to those same patients when researchers are better able to develop new medicines, vaccines, procedures, etc.

Thank you very much for your time.
Ray Esper

D. Han, June 9, 1999

I fully agree that E-Biomed should be started.

I think E-Biomed should be organized into systems biology - such as immune system, neurobiology, cardiovascular biolgy, cancer biology, genomics, infectious diseases, etc.

I think there is a lot of unpublished results that would be very useful to many scientists if available via E-Biomed. I think to establish the credibility of the submitted results as well as for the accountability of the reviewers, the reviewer's comments should be included in every publication. The authors should be given a chance to respond to these comments and if the reviewer find it satisfactory, a specific reviewer's comment could be removed or indicated that the specific point has been satisfactorily addressed.

Since there is a lot of time and money associated with submitting 3-5 copies of printed manuscripts and figures to Journals, which can be avoided by submitting to the E-Biomed, I think E-Biomed publication will quickly attract a large number of submissions.

Furthermore, the main focus of E-Biomed should be open distribution of scientific information, publication of reviewers comments, and accountability of the reviewers.

One could potentially make the reviewer accountable by making all the reviews from a particular reviewer linked; for example one could search for all the papers that has been reviewed by a single reviewer and identify possible pattern of abuse.

I will be in support of this change and will welcome the beginning of E-Biomed.


Professor G. Hellekant, June 9, 1999

Dear Dr. Varmus,

With great sympathy I have read about the NIH plan for online publishing which I strongly support for the following reasons.

Over the years I have with great dismay and worry followed the successful transition of the rights of the researcher's results from the scientist to the commercial publisher. One result of this is that an increasing number of journals ask us to pay for publishing. Further, w e have soon no access to our own published data without pay, since in our attempt to publish to avoid to perish, we transfer more and more our rights as inventor or discoverer to the publisher. The additional consequence of this commercialization is that the number of journals available on library or departments is shrinking every year. It is apparent that you are aware of this.

The fact that the publishing industry can not exist without products of scientists' labor is the key solution, because we control the product; there is little to publish without our efforts. The fact is that the commercial publishers are completely dependent on the efforts by public institutions such as NIH and universities. They are aware of this situation, otherwise we would not have these extensive transfers of copyrights. These must be seen as attempts to keep control over the product.

It is time to realize that we don't need the commercial publishing houses any longer. They are hinders not promoters of scientific progress. Today we have the technology to deliver a basically ready product, spell checked, formatted, with reference lists to the customer, other researchers and the public. Even illustrations is not a problem with modern publishing programs. ( We draw, size and insert illustrations all the times in grant proposals.) What remains of the contribution from the publisher is only the critique, printing and dissemination.

It seems that the idea of a "filter" as mentioned in Science 1999, 283, 1611 or Editorial Board (May 1999) could serve replace the present review solution. After all, the intent of the critique in its present form is to protect the publisher, so that the subscription numbers don't decrease, not the author.

With regard to printing, all Departments I know have laser printer or color printer, and reprints can be printed as needed to supply the number. This can be done at a very low cost, much less than the $100-300/page + $5 for each copy I am paying now. Finally, it seems that your group offers a solution to dissemination.

I would also keep the rights to my work, without having to ask for permission if I want to reuse e.g., an illustration for a review. The time from manuscript to publication would be cut with 1/2 to 3/4 of a year. Such an approach would not have been possible 20 years ago, when we delivered typed, sometimes even handwritten, manuscripts filled with typographical incompleteness and errors.

I for one is thoroughly tired of the present situation, which, if it continues in the direction it is going, will turn more and more research results into intellectual property unavailable to the public without charge. It also presents the scientific community with all kinds of proprietary hinders. If the present trend continues, within 10 years most of us will not be able to publish data that should be available on the public arena according to the mandate of public universities.

I fully support your initiative.

G. Hellekant

Alex Baumgarten, M.D., Ph.D., June 9, 1999

This is an excellent proposal, but I would suggest a minimalistic review system. I am concerned that the current editorial/peer review system will again settle a dead hand over publications. I would argue that novel and contrary observations and ideas, or ones in seemingly obscure areas, are most severely adjudicated and most prone to rejection, with ultimately unfortunate consequences. I am especially troubled by the criteria for selecting reviewers, such as those with current grants: is this not a perpetuation of the shoe that fits the baby now and rejecting the one which may fit when the baby grows? This proposal offers an unparalleled opportunity to remedy the situation. Someone should be responsible for ensuring that rejection occurs only because the material is intrinsically clearly inconsistent or unsupportable, or is presented in a manner that cannot be understood with reasonable effort. Emphasis should be on exhibiting the dog, not the speck on the tip of its tail. Unless someone "watches the watchers" and ensures that there are very valid reasons for rejection of material, this proposal with potentially revolutionary effects will become simply an asterisk to the existing system. Principal criticism of all material should come from the readers, with supporting or contradictory observations, differing ideas for interpreting the data, or offering novel suggestions for verification or extension. These comments, if not independent papers, should be attached to the eventual archive by an article-defining identifying number. Archived articles should be accessible through an initial, short, author-provided summary and/or key-words. I trust the vision of this proposal is implemented in a manner that truly re-creates the way that original material can be brought to the notice of the interested community, current and future, and does not become just another hill for Sisyphus.

Alex Baumgarten, M.D, PhD.

samrain, June 9, 1999

dear dr. v.,

ebiomed is a GREAT GREAT idea! don't let the pious hypocrites such as the oh so pompous kassirer and angell of nejm divert you from this VERY IMPORTANT work. these two and others of their ilk make it a practice to puncture the pretension of those in medicine who have trouble disguising their financial interests. how delicious to see them hoist by their own petard. the one downside of your ultimate victory will be the silencing of their hilarious rococo explanations. i will also miss the very slightly less entertaining statements of the assorted nutjobs, without a vested interest, who have also found wild and wonderful reasons why your SUPER idea is bad bad bad.

..................... (i am a 61 yo MD.)

Eric Halgren, Ph.D., Professor, University of Utah, June 9, 1999

I think this is a great proposal, and should be implemented ASAP. It would greatly accelerate biomedical research. Despite our visceral affection for print journals, it is clear that their time has past. The electronic archive should be entirely funded by NIH as a public service.

The most difficult problem is that of peer-review. The current print journal system is highly imperfect, and the flexibility of electronic publication should lead to its improvement, by making available several alternative mechanisms of peer-review, at the discrection of the submitting authors:

Peer-Review option 1: none
Peer-Review option 2: acceptance by a traditional journal
Peer-Review option 3: acceptance by an electronic editorial board

These would be spontaneous, self-organizing groups who would take it upon themselves to review articles. Presumably they would be established mainly by professional societies. The membership of the boards would be available at the NIH site, but NIH would not approve the boards. Peer-Review option 4: self-organized review Peer-reviewers would be selected by the submitting authors and the interactions between the authors and reviewers would be direct (i.e., not through an editorial staff). The reviewers and their comments on the final (i.e., epublished) article would be posted with the article.

Eric Halgren, Ph.D.
Professor, University of Utah

Stevan Harnad, University of Southampton, June 9, 1999

On Wed, 9 Jun 1999, ALPSP wrote:

> Many thanks to Stevan Harnad for the unsolicited publicity! Anyone who is
> interested in actually reading the research study, so as to discover what it
> was attempting to do, what questions were actually asked, and what the
> conclusions were, is very welcome to order a copy (100 pounds or 200 dollars
> to non-members of ALPSP) - just contact me.

Note that no endorsement was intended; indeed, I was criticizing the ALPSP study, sight unseen, based on the little that was already said about it. The passage below evokes the same criticism:

> In brief, our primary aim was to understand the motivations and concerns of
> authors in the current journal publishing situation. It was clear that the
> standing of the journals in which they publish is of key importance - more
> so than price or absence of page charges; this would seem to indicate that,
> on its own, an 'author pays' model is unlikely to be attractive. What it
> would need, I feel (and I'm actually very attracted by the model) is for the
> publisher of a large, highly rated, successful journal to take the plunge -
> new journals will always struggle initially, regardless of their economic
> model.

What authors should be asked is not whether they are interested in paying page charges. That would be an absurd question, under current conditions, with access to the literature blocked by Subscription/Site-License/Pay-Per-View (S/L/P) tolls. What they should be asked is:

(1) IF IT WERE POSSIBLE, would you prefer (as authors) that everyone everywhere have constant, permanent access to your refereed, published articles for free (and would you, as readers, prefer to have such access to everyone else's refereed, published articles)?

Having elicited a clear and resounding "YES" to (1) (formulated without any false oppositions about a loss of the prestigious journals or of quality or of having to pay to get it), they could be asked the following question:
(2) Would you prefer that your institutions continue subsidizing access to those refereed articles on the reader-end, via S/L/P tolls, or would you prefer that they pay on the author-end, thereby bringing about (1)?

(3) If your preference in (2) is conditional on which alternative would cost more, what if author-end payment proved to cost considerably less?

(4) Would you contribute to making (1) come to pass by publicly self-archiving your refereed articles on the Web free for all (just as you distribute paper reprints to reprint-requesters) if this right were part of your journal copyright agreement?

At this point it would become abundantly clear to one and all, that any author reluctance about doing (4) could ONLY be based on perceived or actual attempts by the Journals to formulate copyright agreements that explicitly forbade public online self-archiving by authors.

It is then that this glaring conflict of interest will be seen clearly, and head-on, and that the new models for copyright agreements in the online era provided by the progressive publishers, such as the APS, and perhaps soon the BMJ, will be there for everyone to see as not only the possibilities, but the actualities that they represent, for one and all.

And only then will it be clear PRECISELY what is at stake in the question of whether there is any justification for continuing to hold the journal literature hostage to access-tolls now that they are no longer necessary.

> We did ask some more open questions at the end about authors' (a)
> expectations and (b) wishes for future journal publishing. Interestingly,
> very few predicted, or hoped for, a radically different model from the
> present one. So if we want it to happen, we clearly have some way to go
> with authors as well as publishers!

You are simply polling habit and ignorance, if you do not make the new possibilities crystal clear before asking about satisfaction and preferences.

Stevan Harnad
Professor of Cognitive Science
Department of Electronics and Computer Science
University of Southampton
Highfield, Southampton

Oliver Obst, Ph.D., June 9, 1999

If publishers would take their responsibility for the scientific communication process seriously, there were no need for the proposed E-Biomed web repository. But some publishers are more interested in their profit than in the free flow of information. Some - everybody know their names - increase journal prices beyond all measure so that libraries have to cancel dozens and even hundreds of serial subscriptions each year. If faculty members can no longer walk into the library, pick up an journal issue, and read their own articles, the access to information seems to be severly disturbed.


Dr. Oliver Obst, 2nd Chairman Association for Medical Librarianship, Germany

Yours sincerely
Oliver Obst, Ph.D. Medical Branch Library, ULB Muenster

Michael Phillips M.D., FACP, New York Medical College, June 9, 1999

Dear Dr. Varmus:

I cannot wait for E-Biomed to get started. Clearly it is an idea whose time has come. Criticism of established journals seldom gets published in established journals, which is hardly surprising. So there is little public discourse on the dirty little secret that editors of paper journals abuse authors and abuse the system. Their principal sins are slowness, carelessness, and arrogance. I say this as one smarting from the recent experience of submitting a paper to "a famous high-profile medical journal" and enduring their incompetence. This extended to their loss of the manuscript (more than once), failure to secure timely peer review, and loss (yet again) of the corrected manuscript. Time elapsed from submission to publication? Nearly one year!

Anything that could help dent the arrogance (and restrict the monopoly) of some of our most allegedly distinguished print journals has my enthusiastic support.

Yours sincerely,
Michael Phillips M.D., FACP
Professor of Clinical Medicine
New York Medical College

June 8, 1999

S. Robert Levine, M.D., Progressive Policy Institute, June 8, 1999

Dr. Varmus,

Allow me to offer my strong personal support for your E-Biomed proposal. If there is anyway that I can be of help to you re: generating support for its adoption, please let me know.

Sincerely yours,
S. Robert Levine, M.D.
Chairman, Health Priorities Project
Progressive Policy Institute

Brian Guenter, Microsoft Research, June 8, 1999

I just read your article about E-biomed. A great idea and none too soon. However there may be a more efficient way to separate the wheat from the chaff than by having a strict separation between peer reviewed papers and completely unreviewed papers. The basic idea would be this: people using the site would be self segregated into two groups: readers and reviewers. Readers just look at papers but reviewers read and evaluate papers. Each reviewer would be given a unique password and their name would be entered into a database. Each time a reviewer used the system to read a paper they would be asked to review a paper as well. In addition they would be asked to rank other reviewers in the database. The ranking of reviewers would be kept strictly confidential. Every paper on the site would be assigned a ranking which would be some weighted average of all the reviewer's ratings. Over time the authorities in the field will achieve a very high ranking because many people in the field will give them high rankings and their reviews will receive great weight. Crackpots and second tier reviewers will have no or lesser weight and correspondingly smaller effect on the ratings of papers. This system completely bypasses the normal process of choosing reviewers, harassing them to finish their reviews, etc., and it allows anybody to post whatever they want on the system. To separate quality papers from garbage a reader can choose to see only those papers with a rank above a threshold and thereby never even see crackpot papers.

Brian Guenter
Senior Researcher
Microsoft Research

Michael B. Helgerud , Stanford University, June 8, 1999

I just read the online NYTimes article on your recent proposal to publish an e-journal of medical research, free. I think it's a great idea. To address the criticism of keeping bad science off the site, I suggest the following. One person's bad science is the next generations break through idea. I think the best way to deal with the review process for this journal is to publicly post the reviews (anonymously if need be) and allow readers to post reviews. That way, the author's original work is presented as they desire and the communities reaction to it is also documented. I would also suggest that you allow authors to alter their papers if obvious errors are found by subsequent readers.

My 2 cents,
Michael B. Helgerud
Stanford University
Graduate Student
Geophysics Department

Richard B. Darlington, Cornell University, June 8, 1999

I just read about E-Biomed in the NY Times. This is to tell you that I approve strongly of NIH's plans in this regard. It's hard to imagine anything more important than the rapid, convenient, cheap dissemination of scientific information. Given that free dissemination is necessarily a money-losing enterprise, it's natural for this effort to be undertaken by government.

Richard B. Darlington
Fellow, AAAS

Department of Psychology
Cornell University

Richard H. Ebright, Howard Hughes Medical Institute, June 8, 1999

I strongly support the proposed creation of a centralized, free electronic repository for biomedical research reports.

The creation by NLM-NIH of a centralized, free electronic search tool for biomedical research, PubMed, has revolutionized information searching and has saved researchers countless hours and taxpayers countless dollars.

The one limitation of PubMed is that it furnishes access only to abstracts, not to full texts (and although linked to a number of journal-owned sites providing access to full text, is not linked to all, is not linked seamlessly to any, and, most important, relies on payment of subscription or article fees for access to full text). E-Biomed is the obvious next step: PubMed with full text.

I currently maintain more than thirty personal print and/or online subscriptions (Science, Nature, Nature Structl. Biol., Cell, Mol. Cell, PNAS, EMBO J., Genes Dev., J. Mol. Biol., J. Biol. Chem., Nucl Acids Res., Mol. Microbiol., Mol. Cell. Biol., J. Bacteriol., Proteins, Protein Science, Protein Engineering, J. Peptide Res., Biopolymers, J. Biomol. Struct. Dynamics, Biophys J., Biochem., J. Am. Chem. Soc., J. Org. Chem., J. Combinatorial Chem., Bioconj. Chem., Chem Biol., Curr. Opin. Chem. Biol., Curr. Opin. Structl. Biol., Curr. Biol., Trends Biochem. Sci., Trends Biotech., Trends Genet....). These numerous subscriptions are expensive. The chore of renewing and maintaining these numerous subscriptions, especially the online subscriptions with their different passwords and different interfaces, wastes time. The need to switch from site to site when accessing online journals, including some twenty to thirty additional online journals I access through institutional subscriptions, wastes still more time. And, most important, the fragmentation of information makes it difficult to locate needed information.

The current practice of preparing manuscripts for the most highly ranked journal and, if rejected, reformatting and submitting to the next most highly ranked journal, and so on--sometimes through four or five iterations wastes still more time.

And, finally, reviewing manuscripts (including re-submitted manuscripts on their second, third, and fourth tries) for more than thirty journals and editing manuscripts for one journal (J. Mol. Biol.) wastes still more time.

Please end this *insane* system. Ignore the self-serving naysayers, and implement E-Biomed. The sooner, the better.

Richard H. Ebright
Howard Hughes Medical Institute
Waksman Institute
Rutgers University
WWW 1:
WWW 2:

Prof. Michael V. L. Bennett, Albert Einstein College of Medicine, June 8, 1999

This may be a litte late, but here goes.

PubMed has made a tremendous difference in how one minds and mines the literature. A couple of years ago, it became obvious that one should write a very good abstract of 250 words (4096 characters), because that is all that most people will see. The abstract should be informative rather than seductive, because the probability of a trip to the library is low except for a few essential articles. Now for those with electronic access, primarily affiliates of academic institutions with deep enough pockets, and the necessity, to subscribe to numerous journals, the situation has changed. PubMed remains the search engine of choice for finding the significant publications on a given topic, but now one can call up and down load most articles of interest published in the last two years. These combined possibilities will make a tremendous difference in how reviews, introductions to research papers and grants are written. E-biomed should rationalize and democratize the dramatic new tools made possible by the Internet.

One can lament that the literature before PubMed has largely disappeared. One can anticipate that '97-'98 when on-line journals became common will develop into a similar event horizon. This horizon will recede as time passes, and the E-biomed proposal to "convert material already published on paper to digital text and image format, with hyper-linked citations" would cause it to recede even more.

One can anticipate a serious dislocation in the scientific publishing community, both for for-profit companies and for non-profit organizations that fund their activities through sale of their journals. Economic stresses are to be anticipated, with or without E-biomed, as electronic publishing costs drop, and Internet based communication through email, bulletin boards, and chat rooms develop more sophistication and acceptance. This writer has no clear scenario of the immediate future with or without E-biomed. Making the scientific literature freely available on the Internet, as PubMed, Genbank and other data bases are now, would appear to be a powerful force for innovation and progress.

Prof. Michael V.L. Bennett, D.Phil. (Oxon)
Department of Neuroscience
Albert Einstein College of Medicine
Bronx, NY 10461

Stevan Harnad, University of Southampton, June 8, 1999

From: Stevan Harnad
To: Declan Butler
Subject: Re: Latest on E-Biomed from Nature

Comments on:

Butler, D. NIH plan brings global electronic journal a step nearer reality. Nature, 1999 APR 29, V398 N6730:735.


Here is some feedback. You are doing fine if you wish to be a passive conduit of the opinions that are being voiced willy-nilly. But if you want to exercise some reflective judgment over the spectrum of reactions, you might consider some of the following:

> Axel Kahn, editor of Médecine et Sciences, the leading French-language
> biomedical journal, says the proposal challenges the 'naked emperor' of
> scientific publishing -- that "80 to 90 per cent of what is published is of
> little real interest". Publish or perish "rather than intrinsic merit" has
> become "the principal justification for much of the output," he says.
> Kahn claims that most journals are infrequently consulted, and that E-Biomed
> would "allow you to have access to the articles you want, without having to
> browse hundreds of journals". A shake-out of the journals system is long
> overdue, he says, adding that there is only a real need for the cream of
> journals, and in particular the best multidisciplinary journals.

Kahn's point about most of the literature being neither read nor cited is correct. It has been made by many others, including, prominently, Stephen Lock, former editor of the British Medical Journal:

Harnad, S. (1986) Policing the Paper Chase. (Review of S. Lock, A difficult balance: Peer review in biomedical publication.) Nature 322: 24 - 5.

But that point has absolutely nothing to do with online archiving or free archiving! An online archive accessible to all would be just as welcome if its contents were 80% chaff or if they were 100% wheat. Either way, the online access would make it more useful and navigable. If you want to consult only the "cream," set your browser to search only what is tagged "cream."

(Nor is the Ebiomed Proposal really claiming to solve the problem of the chaff/wheat ratio, which is a problem WITHIN-journals as much as it is between-, and there is no way for ANYONE to tag THAT for a browser. Moreover, human performance being what it is, we must expect a bell curve in every population.)

> One major concern is that the proposal could harm the best existing
> journals, without accelerating improvements that might gradually occur in
> any case as a result of market forces and more diffuse web efforts by
> not-for-profit science publishers. Several observers say it might create an
> unhealthy monopoly, erode the diversity of existing journals, and reduce
> competition between journals for the best papers.

Have you looked at the proposal? How is it to create a monopoly without the collaboration of the journals? By competing with them? But that would just be a big attempt to found new journals -- and it would fail. (And if it had succeeded it would not have been a monopoly, but a successful rival journal(s) bid!)

And in any case it has nothing to do with the heart of the proposal, which is to provide the journal literature online for free.

> The Varmus proposal notes that the current journal structure has served the
> biomedical community well for 300 years. "So the first question I ask is, if
> it has served us well for 300 years, why change?" says Martin Frank,
> executive director of the American Physiological Society, which publishes 14
> journals and 36,000 pages of articles each year.

Because this structure, which has served us well for 300 years, and will continue to serve us well for 300 more, would serve us all even better if it were available (unchanged) online for free.

That is obvious to any reader of the proposal: What is the advantage of repeating non sequiturs?

> "It [E-Biomed] is extremely cumbersome and is not going to be easily
> implemented," says Frank. "It is so unclear in terms of process that it's
> going fall under its own weight."

Based on its source, it is evident that this has a large dose of wishful thinking. For if one simply drops from the Proposal the needless parts I criticized in my comments, one is left with something as easily implemented and as sure of success as LANL.

> Frank and other non-profit publishers are irritated at what they claim has
> been their omission from early discussions of the proposal, even though it
> intimately affects them. Thirty non-profit publishers wrote to Varmus on 29
> March, as word of the proposal began to spread, asking him for a meeting to
> discuss the plan.Varmus points out that a series of meetings is being
> scheduled with organizations worldwide such as the European Molecular
> Biology Laboratory (see box opposite), but that these will take some time to
> arrange and conduct. He also intends to post the proposal on his NIH web
> page for comment.

Open dialogue is always preferable to subterfuge, although based on the depth of reflection in the feedback that you report here, a wider dialogue is going to generate far more chaff than wheat.

> Some observers note that the page charges collected by some publishers
> provide them with a cash cow -- and that in the United States the NIH is one
> of the largest that is milked. Under the existing system, page charges can
> be passed on to the biomedical agency by investigators that it supports. The
> potential loss of this lucrative system is alarming publishers -- especially
> non-profit organizations --which rely heavily on it.

This is based on such a convoluted misunderstanding that it takes one's breath away:

There are three current senses of "page charges," and this passage completely conflates them: One of these is (1) page charges charged to authors in paper journals today. There are few of these, and they are dying out. The other is (2) the charges for paper reprints of one's journal articles. These exist, but are not a cash cow either. The only real cash cow is the (3) charge for the pages of the journals themselves, paid through Subscription/Site-License/Pay-Per-View (S/L/P).

Then there are my own proposals for (4) page charges for quality control, once all papers are available online for free.

Now how do these map onto the message of the above paragraph? It is not the Reprint cash cow that is at risk, but the S/L/P cash cow!

> Proposals that E-Biomed should coordinate peer review of its contents are
> controversial. Noorman argues that centralization of peer review would
> threaten the diversity of schools of thought provided for by journals.

The peer review component of the E-biomed Proposal was inchoate and incoherent in that draft; but it was also inessential, and the Proposal will look just fine once it's dropped.

> This concern is shared by many scientists and learned societies, who feel
> that a centralized structure may obscure the well-defined hierarchy of best
> science provided by journals, and that scientists may be more reluctant to
> give their time and energy free to a central structure.

Correct, but it's not at all clear from the Proposal that that was what was intended. In the next draft that will no doubt be remedied. But the core idea of a free online journal literature as a centralized resource will remain, and that is what it is all about.

> Andrew Odlysko, a mathematician at the AT&T telecoms corporation and an
> expert on scholarly publishing, argues that it would be simpler to separate
> the distribution and peer-review functions of the repository, as is done at
> the Los Alamos physics e-print servers, where peer review is provided by
> journal 'overlays' to unrefereed papers.

Correct, and the first sensible remark adduced in this article so far.

(It was also the gist of my own critique of the Proposal, as you well know, and Andrew wrote what he wrote in response to my critique. I am not carping about being left out, by the way; just about the needlessly low wheat to chaff ratio.)

> Lynn Dobrunz, a postdoctoral neurobiologist at the Salk Institute in San
> Diego, asks: "Would E-biomed be in addition to the current system of
> journals, or instead of it?

Reasonable question, so far.

> If there was a consolidated site that published
> online versions of all the articles that are currently published... that
> would be fantastic.

Indeed, especially if it was not just consolidated online, but free.

Online (and consolidated, by a click-through monopoly) is already on the way from journal publishers anyway.

> If it's instead of, and especially if it has this
> non-peer-reviewed track to it, I think that is a much less good idea."
This is now a postdoc's uninformed opinion to the effect that:

(1) "I think preprints should not be archived." (Fine, don't archive yours; but 100,000 in LANL, for example, have a different opinion),

(2) "I think Online-Only will not be enough. Keep the paper corpus (and keep paying for it)." 5000 hard-pressed libraries may have another opinion; so might the 100,000 physicists who no longer use the paper version).

> The Varmus proposal suggests that scientific societies could be one source
> of peer review. But the societies are worried that E-Biomed may undermine
> the journal revenues on which many of their other activities, such as
> fellowships and meetings, depend.

True. But then the thoughtful question is: Are those other activities worth holding the journal literature hostage to? For how much longer?

> The head of one society says he is open to change, but would need guarantees
> that revenues would be preserved. Given such guarantees, societies might
> consider joining the initiative, he says. "E-Biomed will only fly if learned
> societies and their journals can be brought to the table," predicts Tony
> Delamothe, web editor of the British Medical Journal, and a supporter of
> E-Biomed's goals.

Or, if all Biomedical authors self-archive all their refereed reprints (and their unrefereed preprints, if they wish) in it.

> Another broader threat, expressed by many scientists, is that NIH might come
> to dominate much of the biomedical literature, leading to homogenization or
> to discrimination against scientists from smaller countries. "Who would
> select the governing body?" asks an official at one European scientific
> society. "Who would select the editors and decide what is allowed to be
> published? Who will determine costs and access rights?"

Confusion caused by the current draft. This will no doubt be sorted out soon. Classical, pluralistic peer review will continue as before, and the Archive will simply be the free repository of its products (along with the pre-peer-reviewed drafts).

> Many are also uncomfortable with the prospect of public funding for
> scientific publishing, an activity currently dominated by for-profit and
> non-profit publishers in the private sector. At the same time, however,
> there is growing resentment among scientists and librarians at the
> spiralling inflation in journal subscriptions.

Journal publishing is currently subsidized completely by S/L/P funds from public and private institutional libraries. Networks and Archives (including LANL) are also subsidized by public and private institutions. The latter will simply be taking over the load from the former, saving a great deal of money, and providing the product for everyone for free.

> Competition between scientific publishers is less than in other industries
> because of distortions in the market, and profit margins as high as 40 per
> cent are not uncommon (see Nature 397, 195-200; 1999).

And inelastic demand from the libraries that have to keep subsidizing them for lack of an alternative. Ebiomed would provide the alternative.

> Graham Cameron, head of services at the European Bioinformatics Institute
> (EBI) in Cambridge, England -- an outstation of the European Molecular
> Biology Laboratory -- points out that public domain databases such as PubMed
> and GenBank and the EBI are widely considered to provide high levels of
> cost-effective service to the community.

True, but until further notice, publishers' journals are not public databases. (Authors' self-archives, in contrast, CAN be.)

> Many believe, however, that the wider and cheaper access promised by
> E-Biomed may happen anyway as a result of market forces. "Most scientific
> society publishers are already doing what Varmus is proposing," says Frank.
> "We are putting our journals on the web. We are linking our journals through
> PubMed to our sister journals on the web. We are developing interfaces for
> the submission and review of manuscripts on the web."

Indeed, and the objective is a click-through monopoly, with the online journal literature continuing to be held hostage to S/L/P till doomsday...

> Similarly, consortia of library and other users are increasingly negotiating
> electronic licences for journals for entire institutions and even countries.
> Scientists at such institutions can already access much of the literature
> online.

That's the "L" in S/L/P, in case you didn't notice, whereas the fundamental underlying issue here is FREE access.

> "My initial reaction to E-Biomed is, 'so what?'. Virtually every library has
> almost all major journals," says Heinz Steiner, a neuroscientist at the
> University of Tennessee College of Medicine in Memphis. What is the point of propagating this nonsense? He might as well have said "Let them eat cake!"

And show me an active, busy researcher -- even at the most prosperous university in the world, Harvard, which subscribes to them all -- who, every time he needs a paper, prefers to walk to the library or send a student to photo copy, and shuffle through piles of photocopied offprints rather than having the entire journal literature on his desk and at his fingertips at all times.

> Market forces are also driving a flurry of deals among publishers that may
> enable researchers to move rapidly and seamlessly from a citation to full
> text across journal boundaries.

Via a click-through monopoly involving L and P deals (out of the S/L/P troika), done by those institutions with enough money to make them; everyone else is out of luck, as before.

> Frank asserts, for example, that the web site of HighWire Press already
> accounts for a large proportion of the biomedical literature. This
> not-for-profit outfit was set up in 1995 by Stanford University Libraries
> and Academic Information Resources to help universities and societies to
> publish on the web at low cost. "So I don't know why we need to create
> E-Biomed," says Frank.

Ah me! Substitute for the 35,000 daily users of LANL (augmented all their potential biomedical counterparts worldwide) what the access levels would be if regulated by HighWire's "low" S/L/P costs.

> Indeed, the head of one scientific society argues that resentment over the
> huge costs of the current journals system is confusing the many complex
> issues involved in scholarly publishing. "If publishers are charging too
> much then we should attack this problem directly, but not attack the entire
> system. E-Biomed is a not very selective nuclear bomb."

Ebiomed would attack the system by providing an alternative, free route to exactly the same literature. You can't get more selective than that.

> Noorman, while admitting that Elsevier's profit margins "are higher than the
> average," says that the arrival of web publishing is putting pressure on
> commercial publishers. "Scholarly publishing will become a proper [not
> distorted] market," he predicts. "Elsevier is not in the world to keep that
> profit margin high. We are in the world to stay in the market. If the web
> causes us to have to agree to lower profit margins, then so be it."

But don't hold your breath.

Stevan Harnad
Professor of Cognitive Science
Department of Electronics and Computer Science
University of Southampton
Highfield, Southampton

Andrew B. Onderdonk, Ph.D., June 8, 1999

Dear Dr. Varmus,

After reading your proposal for the establishment of an electronic publication for the biomedical sciences, several important issues do not appear to have been addressed. These include whether it is desirable or appropriate for the agency that provides funding for the biomedical sciences to also assume responsibility for publishing the results of studies supported by NIH. Although the peer review process will be preserved for part of the e-biomed publication, it is not at all clear how editorial boards will be selected or monitored. For unreviewed publications, the probability for publication of false data and clinical trials appears to be a threat which has not been addressed in your proposal. Ethical considerations, authorship, disputes over publication rights and copyright issues do not appear to have been considered in a systematic and thoughtful manner. I believe that the community of biomedical scientists are well served by the current journals and society publications. These publications are already migrating to the electronic publication arena with many of the same features described in your proposal. I urge you to reconsider your proposal to impose yet another government sponsored and controlled program on the scientific community. Peer reviewed, independent scientific journals perform an important function and should be retained as the primary source for research information.

Andrew B. Onderdonk

Charles Pyott, June 8, 1999

I would like to express my support for E-biomed, as outlined by Harold Varmus. The controversy, I think, could be minimized if NIH collaborated with the major scientific journals, somehow, in the creation and maintenance of the proposed database. A complete and accessible database does not have to exclude the major journals, slash their revenues, or eliminate the peer review process. These should be combined, into a single system with elements of each. All interests can be served, if each party is willing to make a modest sacrifice.

David A. Orbits, June 8, 1999

I think this is a fantastic idea. No doubt problems will arise with bogus research but that will be self-correcting.
I am a software engineer and I would very much like access to this data that I helped pay for with my tax dollars.
Spreading knowledge in as "friction-free" a way as possible will only help to produce more knowledge. This is very exciting. You are to be congratulated.

David A. Orbits
Redmond, Wa

June 7, 1999

Thomas Sneider, Colorado State University, June 7, 1999

The on-line publishing proposal has good and not-so-good aspects. Just a few examples: good: enhancing the rapidity of communication and synergism amongst people working on similar problems; widening the circle of "those-in-the-know" (decreasing the influence of the old-boys networks). Not-so-good: potential circumvention of the quality assurance often (but not always) effected by peer review; the potential for a deluge of materials that might be viewed as place-holders to claim priority of discovery for patent and/or aggrandizement purposes; and, most importantly for university -based researchers, the difficulty that the system will impose on university procedures long in place to assess qualifications of individuals for promotion and tenure. For example, consider just one aspect: how will one track usage and impact factors associated with e-publication?

However, the e-publication makes a lot of sense both economically (page charges are outrageous) and democratically (the proposed system levels the playing field and widens the pool of cognoscenti).

Thomas Sneider
Associate Dean
College of Natural Sciences
Biochemistry & Molecular Biology
Colorado State University

John D. Leith, M.D., Ph.D., June 7, 1999

Dear Dr. Varmus:

I note Dr. Arnold Relman's editorial in the current NEJM (June 10, 1999, p. 1828-29) on your "E-biomed" proposal, and I have read your draft of it dated May 5, 1999. I am a pathologist, retired 1993 from the Brockton (Mass.) Hospital, where I served as chairman of its IRB for 8 or 10 years. The latter position gave me some additional insights into the complexities of clinical research and its interface with the public via so-called informed consent forms.

Dr. Relman puts forth numerous cogent reasons why the mechanisms as you propose them would or might be harmful to medical research or to the public at large. I support most of his positions.

Let me present here some of my own reactions.

The first is the enormous propensity of people to use the web to broadcast all sorts of fiction as "published fact". If you browse the web much (which I doubt you do!), you can see this in action, like the childhood game of "telephone", as rumors mutate into absolute truth through distortion and repetition, and then are spread like chicken pox all over the 'net. There is already too much wishful thinking by the public when it comes to easy medical cures, and full 'E-biomed' publication with a search engine would allow careless, ignorant or intentionally fraudulent extraction of phrases and ideas, then to be disseminated for mental or monetary gain by those with medical (includes mental) problems or seeking profit. We have plenty of alleged cancer cures already; we don't need to be inundated by more, put forth by every ignoramus cruising the biomedical literature armed with a search engine. And we don't need to clog up the internet this way, either.

A second problem relates to the publication process. The "Journal of Irreproducible Results" some years ago, as I remember, had an article on the "PPPP" phenomenon, the "Peculiar Propensity to Premature Publication". With the 'E-biomed' avenue open, we could expect to see more words published on fewer ideas: a further fragmentation of discovery into its components, separating into small papers what ought to be put together and presented as organic wholes. While a search engine might allow us to find isolated fragments even if published separately, we would tend to lose out on the broader synthesis of implications and conclusions that the thoughtful author should be encouraged to produce along with his data.

A third problem is what this plan would do to libraries and to journals. I believe it would over a decade or so kill off most biomedical journals, maybe 80% of them. (On the other hand, I also believe that might not be a bad idea! They have multiplied like rabbits, and now there are far too many for a conscientious reader to even scan the tables of contents, whether he is a specialist or a generalist. This is partly related to the fragmentation of research publication referred to in the prior paragraph.) Librarians would likely become clerks busy all day downloading ariticles for constituents, be they staff members, patients, or the visiting public. Or, that task might be in addition to their present tasks such as ferreting out articles related to some particular clinical problem. Surely the busy scientist or clinician will not do the downloading himself if he can avoid it. I suspect many medical libraries would cease to exist as such, a revolutionary change in my opinion that ought to be most carefully approached.

Your draft paper did not estimate costs and cost-shifting. That should be a factor in discussions. Dollars, and clogging of the Internet, and shifting of payment for medical publication from present customary sources to new sources all need discussion. Loss of journals and of publishing companies needs further consideration.

I question any real need to speed up further the rate of publication of scientific findings. Whose ends are being served? Who would profit most and least? I don't think the usual medical excuse of "saving lives" can be easily used in this discussion, as some lives would be lost from premature publication of wrong data, wrong conclusions, and misinterpretations that would be spread around the WWW. I can easily imagine 'E-biomed' as costing lives on that basis.

<<<<<<<<<<<<<<<<<< >>>>>>>>>>>>>>>>>>>

However, let us not throw out the baby with the bath water. How about amending the proposal to have NIH pay publishers, for either the publishers or NIH to publish papers electronically, say, 3 months _after_ the papers have appeared in print in a "normal, refereed journal". This would cut down some on journal subscriptions (which NIH subsidies could compensate for), but perhaps not too much, would ensure hopefully proper refereeing, and yet would also allow the vast body of poor researchers whom you seem to be concerned about to access the literature freely in a really pretty timely fashion, and would allow us to use search engines on this still-pretty-recent literature. I see no need to rush things into print, and a lot of problems and disadvantages!

A three-month delay would allow journals to publish comments on a paper, and would allow the popular media (such as they are) to think about what they were doing before putting out rumors. It might not cut down much on ignorant or fraudulent clogging of the web with snippets that misinform, mislead, and mistreat, unfortunately.

Sincerely yours,
John D. Leith, M.D., Ph.D.

Lisa A Perry, Minnesota State University, June 7, 1999

While I believe the goals of E-biomed are laudable, I disagree with the proposal to "publish" scientific research in an on-line format previous to or instead of peer review publication. As a humanist researcher, teacher, and coach of forensics (speech and debate, not pathology) I am worried about the repercussions of this form of "publication."

This worry stems from the fact that while the primary audience may be other biomedical scientists and medical doctors, the fact remains that the secondary audience is vast. Lay persons invesitgating illnesses and treatments, students doing research for papers, students doing research for speeches, all would have access to this information that has not been reviewed. I can see 4 reasons why this proposal is not only unworkable, but also potentially harmful.

First, nothing prevents researchers from posting research on their own web pages at any time. In fact, this self-"publication" would be a better way to gain any possible benefit as readers would be able to evaluate the credibility of the source and information without the misleading connection to the NIH or other possible "official" sponsors.

Second, this poses problems for faculty members who are worried about tenure and promotion. Many journals will not publish research that has been previously published in any other format. For example, the "New England Journal of Medicine" states "Manuscripts containing original material are accepted for consideration if neither the article nor any part of its essential substance, tables, or figures has been or will be published or submitted elsewhere before appearing in the Journal." And the Uniform Requirements for Manuscripts Submitted to Biomedical Journals states "Most journals do not wish to receive papers on work that has already been reported in large part in a published article or is contained in another paper that has been submitted or accepted for publication elsewhere, in print or in electronic media ("Uniform Requirements for Manuscripts Submitted to Biomedical Journals" The New England Journal of Medicine. 336[4, 1997], While that is problematic for all, those faculty under the tenure gun face an additional problem as many universities and colleges will not consider electronic or non-peer reviewed publications for tenure or promotion, even in the humanities.

Third, while the proposal states that this would be a "democratizing force that makes distance and wealth nearly irrelevant," the fact remains that for international scientists to read the reports requires access to a sufficiently powerful computer, and more importantly, electricity and phone access. And for those that do have the access, some of this information could already be found at individual web sites which can be bookmarked or otherwise personalized.

Fourth, having the information be submitted to the "general repository" does not guarantee anything about the quality or accuracy of the information. And while the primary audience of biomedical researchers may be able to dismiss faulty research out of hand, this information will not (and should not) be limited to this audience. The press, individuals, students, anyone can access this and, because it is affiliated with an "official," even governmental body, ascribe additional (and even unwarranted) credibility and ethos to the information. We have seen it happen with the misuse and misapplication of the so-called "Mozart-effect" that made the rounds of the press in the last few years.

And fifth, the element of personalization and specialization can be a bad thing. Journals at least leave the possibility of someone reading "outside" their area and developing and encouraging a more organic and systemic view of biology, research, and methodology.

In short, while I think the development of a standardized and central way of accessing information on medical research is a good thing, I believe that the E-biomed proposal is not the solution.

Lisa A Perry
Assistant Professor and Assistant Director of Forensics
Minnesota State University, Mankato

Jean-Claude Guédon, Université de Montréal, June 7, 1999

Dear Dr. Varmus,

For what it is worth, I enthusiastically support the general thrust of the e-biomed project. Only projects like these and the Los Alamos archives devised by Paul Ginsparg can hope to counter-balance the effects of the huge collections of expensive journal collections organized by large commercial publishers.

This said, an issue seems to have been somewhat overlooked in most of the debate so far, and it is the issue of the technical dimensions of this archiving/publishing device.

After thinking long and hard about the topic, three e-publishing groups have independently hit upon the necessity of some stable archival format and the necessity of distinguishing the archival format from the potential display formats.

The three groups are:

The Virginia Tech initiative for theses (led by Ed Fox):

The International Consortium for Alternative Academic Publications, led by Mike Sosteric in Canada:

And the Erudit project accompanied by a thesis project at my own university (Presses de l'Université de Montréal, led by Guylaine Beaudry)

There are probably other groups that are moving in the same general direction.

In all three cases, an archival format has been at least considered, if not always implemented, and it generally rests on some variation of SGML/XML.

TeX could be an alternative and TeX is a fine format but, oriented toward math typography as it is, it concerns scientists in a small number of disciplines. Outside Physics, astronomy, math and some parts of chemistry, molecular biology, etc., it is simply not used. As a result, the use of TeX appears as an exception that splits the collections of texts being developed. The advent of MathML as an extension of XML will solve this when it is widely accepted. But that will require a few years.

More broadly, I believe that we need a kind of "electronic press". Such a "press", in my opinion, should be totally open to all and would evolve in a fashion reminiscent of the Linux movement. Mike Sosteric, named above, has moved a fair bit in this direction and he is ready to set up the usual tools to encourage open participation in the development of the needed software to create this electronic press. In fact, developing this electronic press is what is needed to provide the e-biomed project with a firm conceptual and technical basis.

An electronic press would also be simple enough to use to support the kind of self archiving of papers that Stevan Harnad supports. It would also provide for the possibility of efficient search engines to be developed on top of the distributed archives.

As I see it, the electronic press needs good, probably perl based, scripts to transform basic digital input into whatever format, word processing, ascii text, etc. into SGML/XML. This assumes that we can converge on a small number of DTD's. The e-biomed project could usefully support the design and support of an adequate DTD, or help use an already recognized DTD.

We also need a good translator between TeX and SGML/XML, again with the assumption that we can converge toward one or a few DTD's. Lyx might offer a very useful base as it is free source code and already outputs Latex. It is in fact a front end for Latex production. It could be extended further to become a front end for XML production and thus allow a translation back and forth between the two formats. Achieving this would lead to overcoming the present split between the TeX based communities and the non-TeX based ones.

Mike Sosteric has developed macros to move gracefully (and efficiently) from XML to HTML and that is an important step. His example should be looked at and, if needed, improved upon.

The outcome would be a set of tools that, when concatenated together, would amount to a simple, free, electronic press that would be easily used by any scholar that wants to publish electronically. Done in free source code and running in particular on top of Linux, it would also open the road to e-publishing to most of the Third World, another important issue.

In short, I am suggesting we treat the technical side of electronic publishing as an assembly line that would be organized on the basis of various pieces of software. I further contend that these different software parts should be developed in the spirit that energizes the Linux/Gnome/KDE/Gimp/Apache/Perl/etc. projects. This would add to the conceptual notions of self archiving, etc. that Stevan Harnad, Paul Ginsparg and many others advocate.
As a final recommendation, good lines of communication should be maintained with W3c, the WorldWide Web Consortium as they are responsible for the development of XML, MathML,etc.

I will end with one last suggestion. I have just been named program co-chair for Inet2000, the Internet Society annual meeting that will take place in Japan. Having a strong round table on this topic with your own presence would certainly create a lot of interest at our conference in June 2000. By then, many pieces of the plan should have appeared to make a progress report quite exciting.

Again, allow me to congratulate you on a fine initiative, one that spells good omens for the advent of a society based on distributed intelligence.

Jean-Claude Guédon, Ph. D.
Département de littérature comparée
Université de Montréal
Montréal, Canada

Join the Internet Society and help to make it so.
See you at INET'99, San Jose, June 22-25, 1999

Dr. John R. Skoyles, June 7, 1999

E-Biomed is a challenge to rewrite history -- imagine that scientific publishing had started off not with print but with the internet, what would it look like? Are there traditions about how we publish articles on paper that are compromises linked to the practical problems of using the print medium but that are not part of good scientific communication -- traditions that have become so in grained into the practice of science that we fall to see how they impair science and ideally should abandoned in the move to the electronic media?

One such tradition I suggests exists: if science had started around the internet all papers as a matter of course would contain links to archived raw and semi-analysed data. However, in spite of mentioning the advantages of data sharing and data archiving, the E-Biomed proposal fails to adequately to challenge this tradition -- whether raw and semi-analysed is posted with a paper is still left to its authors discretion. I shall argue below it should be both expected and compulsory -- we should not mistake our familiarity with its omission with being part of the scientific process -- its traditional omission lies not in science but to the limits of print as a medium.

Compulsory data archiving in summary would (1) make fraud easier to detect, (2) encourage scientific criticism and (3) aid the scientific process in general.

Experimental data only rarely needs to be read -- usually we are happy with their author's own statistical treatment. But not always and in such cases it is important. Researchers do not always fully analyse their data; sometimes editors restrict their publication space; and sometimes we have an idea we would like to try out on those data. It would be nice if the experimental data we read about were easy to access. Though there are several potential problems with compulsory archiving of published data, the benefits would, I believe, vastly outweigh them.

Here follows a case for the compulsory archiving of data. I also raise a few objections.

First, electronic data archiving should be easy to implement and will become increasingly so. Most researchers would have little trouble archiving their data upon publication. Most results sections are based upon computer analysed ASCII data files or other standard data formats. Most researchers should thus have their raw data stored in a form (i.e. file and subdirectory names) which makes it easy for other researchers to use. The commands and procedures for transferring it to a central data archive will be familiar to most and be no greater than archiving papers.

Second, the scientific ethic is to make error correction as easy as possible. Scientists are not always entirely competent or honest. Numerous cases of fraud and intellectual dishonesty have occurred in all areas of science. Researchers are subject to enormous pressures to publish but unfortunately this normally requires positive findings. This puts pressure on researchers to rerun analyses (changing criteria for categorizing data, excluding subjects, treating missing data, etc.) when only negative findings turn up. It is not clear how many researchers resist these pressures on the integrity of data analysis. At present, it is difficult to check. In a case reported in Science, two psychologists were only able to check the data analysis of another psychologist through the intervention of lawyers (Palca 1991).

There has been public disquiet in the US Congress (notably, on the part of Congressman John Dingell) concerning fraud and intellectual dishonesty in science. Research on published fraudulent papers has revealed many defects (Stewart & Feder 1987). It is likely that any archived data would contain even more accessible and noticeable defects (in their data distributions, treatment and analysis). Archiving data would thus make it easier to detect both fraud and intellectual dishonesty.

Third, much honestly obtained and analysed data is incompetently handled, yet many legitimate criticisms never arise because of difficulties accessing data. At present, if a scientist that suspects that another researcher's analysis gives only part of the story or is misleading, faces an involved process of contacting them for the original data (something inconvenient to all concerned). Archiving data would increase the opportunities for legitimate criticism of published work.

Fourth, researchers ask different questions. Sometimes a researcher may wish to reanalyse data to answer questions the original authors ignored. People carrying out meta-analyses will often want to check the quality of the work they are using. At present this is not possible.

Fifth, students could gain much by examining real research papers and then "playing around" with their data, seeing the affects of different data-analytic strategies. They might even find things overlooked by their authors.

Sixth, much data is accidentally lost (despite the requirement of most journals that authors retain their data for a number of years). An archive would make a convenient data backup.

Seventh, scientific papers are printed on paper -- this, not the nature of science, is the reason data are not normally made accessible at this time. Science is about open communication that maximally exposes ideas and arguments to criticism (one legitimate criticism of an idea is the way its data are handled). Printed paper is a convenient means for opening written ideas to criticism, but it is unsuitable for making data accessible to criticism (it limits the quantity which can be published and communicates in a form that is inconvenient for computer reanalysis). Print has until recently been the only means for disseminating scientific ideas and data. Hence the tradition has arisen of limiting the dissemination of data. We should recognise the opportunity that electronic archives provide for breaking with this.

There are some reasons against archiving:

Certain classes of data (e.g., clinical data) may have to be excluded to preserve the confidentiality and privacy of those from whom it is collected. This constraint does not apply to large portions of research which involves research on animals, reaction time studies on student subjects, or computer simulations.

Researchers certainly have the right to the "first go" at their data. However, the fact of publication, unless contrary notice is given, usually signifies that the data have already been substantially analysed, and frequently no further analysis is intended.

There is another objection that is however entirely invalid. Many researchers will be uncomfortable with their data being archived because none of us are perfect. If our data can be reanalysed we may be shown to have carried out, quite unintentionally, inappropriate or misleading analysis. To some extent the present state of affairs is quite convenient for hiding the fact that many researchers could be better statisticians and could keep better records.

Data archiving as a standard part of electronic paper publication of course would involve some cost and effort, perhaps even some inconvenience. However, with the public and congressional concern about whether scientists are maximally ensuring the integrity of their data, an archive would show a commitment from the scientific community to ensuring honesty in published research.


Palca, J. (1991). News and Comment: Get-the-lead-out guru challenged. Science 253: 842-844.

Stewart, W. W. & Feder, N. (1987). The integrity of the scientific literature. Nature 325: 207-214. Based upon FTP INTERNET DATA ARCHIVING: A Cousin for PSYCOLOQUY

Dr. John R. Skoyles
London, UK

Check out my Golden House-Sparrow award winning homepage