IuK Workshop "Wege in die Zukunft – Elektronische Zeitschriften II – International Symposium on Electronic Journals" Berlin 16-17 February 1998

Returning Responsibility for Scholarly Communication to the Academy

Michael A. Keller, University Librarian, Director of Academic Information Resources, and Publisher of HighWire Press, Stanford University

May I start by thanking the organizing committee and especially Diann Rusch-Feja for inviting me and my colleagues to join this conference.

I wish also to thank my colleague and friend, Friedemann Weigel of the Otto Harrassowitz Buchhandlung, for operating the computer which is displaying the URLs for this lecture. Harrassowitz has become the European service bureau of HighWire Press. Thanks are due as well to the excellent technicians of the Humboldt University for the support they have provided to me and other speakers at this conference.

We are delighted to be with you and to continue in this place the vigorous debate on the ways and means to make the results of science available to other scientists as well as to the public who both benefit from and support the practice of scientific research. This paper is in four parts: a perspective on how we have gotten to where we are in scientific journal publishing, a demonstration of the Internet editions of high impact scientific journals published by a group of scientific society publishers working with HighWire Press, a section describing HighWire Press, and some concluding speculations on what might be salutary next steps in the return of responsibility for scholarly communication to the Academy.


During the decades following Sputnik, national investments in basic and applied research at universities created, among other products, an unprecedented out-pouring of scientific, technical, and medical articles. During these last forty years, we have seen the rapid growth, expansion, cross-fertilization, and re-definition of numerous subjects of research which processes have in turn led to new national and global communities of academicians working in related topics and thus needing outlets for the reporting of the results of their research. Until the time of Sputnik, I believe that it is accurate to say that the predominant means for disseminating the results of scholarly research was primarily under the direct control of scholarly societies, acting as publishers. Following Sputnik, the need for new outlets for new scientific topics outstripped the desire and perhaps interest of some of the scholarly societies to provide space in the existing journals. Thus the new topics, many of them, found new outlets, some of them provided by prescient for-profit publishers. Such topics were, and many still are, niches for fairly small communities of researchers. Others began small and are now gargantuan. One thinks, for example, of the field of tetrahedron chemistry which has generated beyond one journal started by Pergamon, three other journals. Over the past forty years, for-profit publishing houses have assembled lists of journals, by acquisition from smaller, niche publishers and by making shrewd investments in developing sub-disciplines. For many such publishers, handsome profits have been realized. In the same time, some scholarly societies acting as publishers have grown as their fields of study and their membership expanded.

All of this sounds quite benign as far as this description goes.

As researchers in the post-Sputnik days of seemingly inexhaustible national funding for basic research tackled more topics and the disciplines twigged and cloned, some scientific societies were convinced to contract with for-profit publishers for the publication of their journals. These decisions may have seemed wise back when taken, but some unintended consequences have resulted. In the past twenty years, some for-profit publishers have realized that some information, scientific, technical, and medical articles in particular, can be treated as a commodity. That is, the rapid publication of lots of articles leads to information commodities which can be priced on the basis of their real or perceived value, rather than on the actual costs necessary to acquire and publish them. What we have seen in these years is the pricing of journal subscriptions to individuals more or less at the marginal cost of producing each copy, but the pricing of subscriptions to institutions, particularly university libraries, at prices which initially reflected the community usage of such information, soon came to reflect value pricing in strategies of achieving large profit margins year after year. University libraries such as the ones at which I have been privileged to work, Cornell, Berkeley, Yale, and Stanford, have experienced outrageous price increases each year for certain publishers' journals. Such increases have not been and could not be matched by similar budget increases. So, research libraries have responded to the price increases in a couple of ways simultaneously: cutting books and journals from their collections and moving money to the sciences from other disciplines. The result in our universities of these responses has been dramatically reduced representations of the literatures of virtually every discipline formerly collected. The result of these responses among the publishers involved has been the increase of prices to account for the reduced number of subscribers. One publisher, Elsevier, has been quoted in Fortune magazine to the effect that because institutional subscribers had to purchase their publications, the company could increase prices as they wished. In short, the university libraries are felt to be tethered goats.

Scholars in the sciences, technologies, and medicine, were for many years oblivious to these situations. They could subscribe at individual rates and regarded library subscriptions as necessary for use by students and as the means to build reliable archives for information gathering when they moved beyond their immediate areas of research. Besides, the libraries' collection budgets were in part supported by income from their research grants. And furthermore, payment of enormous institutional subscription prices was another way of institutions subsidizing big science, in effect supporting the required publication of results.

From another perspective, that of institutional librarians, part of whose mission is to make capital investments in information for the future of scholarship, the astounding rates of increase annually in many stm journal prices resulted in diminished local information resources for scholarship and teaching for the present generation as well as ones to come. Profit taking by some publishers has been impeding a couple of the social roles of librarians: we can not and could not collect as widely for current students and faculty; and we have been unable to collect in depth to provide meaningful collections for future generations.

The economic crisis in stm journal publishing has been at the boil for some years. There are other limiting aspects of traditional journal publishing as well. We are all aware of the long lead time necessary for some journals to take a manuscript submitted and turn it into a published piece. And many academicians, scholars and librarians alike, have been annoyed by the seeming lack of selectivity in decisions on which articles to publish. Most of the most highly cited journals must limit the length of articles as well; often authors and editors reduce articles to only the most essential elements. Many authors and editors have wished to publish many more graphics and images to illuminate the texts. Distribution of journals has been for years dependent upon the vagaries of the various national and commercial postal systems; and distance from the place of publication has made staying current with the literature of any discipline difficult for those in distant places. Finally, as scientific research has benefited from information technology, new forms of information archives, image and data bases, are not easily reproduced in printed media.

So great has been the pressure on library budgets, so onerous the dilution of the ability of many universities to provide for the information needs of local scholars, that some organizations, the Association of American Universities for one, are calling for an extensive overhaul of the system of scholarly communication in order to protect the interests of the academy.

For all of these reasons, recourse to network-based journal publishing has seemed an obvious opportunity and many publishers, both scholarly societies and for-profits, have made serious progress in developing Internet editions, often with more information and services than the print equivalents.

Are we in the academy at a crisis point? Does it matter that there are measurably fewer books and journals coming to our libraries each year? Do scholars recognize that their own work is disadvantaged by the profit-taking of some publishers in some disciplines? The answer from my perspective as a university officer at Stanford and as a university librarian is yes unequivocally to each and every one of the preceding questions.

After years of giving away their articles to publishers, scholars in America are increasingly asking how to manage their copyrights to benefit themselves and their communities, including their universities. And a number of scholarly societies which had contracted with for-profit publishers for publication services are now asking their lawyers how to get out of the contracts. After years of library book budgets increasing at one rate, but receiving invoices for stm journals increasing at a multiple of that rate, faculty are realizing the consequences of the current arrangement. And the academy is beginning to stir itself into action.

One sees the AAU calling for separation of the validation and printing/distribution steps in the publishing process. We see talk of purchasing cooperatives, not just for electronic publications, but for print as well. We know of investigations in the U.S. Department of Justice to determine whether aggressive acquisition of once independent publishers by Elsevier constitutes a violation of American anti-trust regulations. We understand that there are parallel investigations in the E.C. as well. These investigations were instigated by librarians and by scholars.

Another reaction to the situation among an increasing number of collection development and acquisitions librarians is to question why an institution would pay more money to make available digital versions of low impact and high cost of stm print journals. We have seen in these early years of networked information resources a tendency to buy and make accessible for one's institution whatever is available in order to seem hip, to be part of the revolution and not a stick-in-the-mud traditionalist. Informed and demanding consumerism is on the rise among librarians as I suspect Hazel Woodward's paper at this conference will indicate.

And most especially, we see some efforts among scholarly societies and universities to improve their own scholarly communications through the use of network publishing to become much more competitive, to attract more authors and readers, and to take back responsibility for publishing the results of their own endeavors.

Quite in parallel, for-profit publishers of their fraction of the stm literature have created Internet editions of their journals. However, the business and access models are coercive and as a group the for-profit publishers are putting the building blocks in place to track each and every use of each and every article they put on the Web. The tracking mechanism begins with the development and marketing of the digital object identifier and will proceed with digital watermarks and fingerprints. The reason for the compulsive tracking of use is more profits, of course. For-profit publishers have for years complained that fair use rules in the U.S. and the rough equivalents in the E.C. have prevented them from making even larger profits than they already have! Another aspect of this trend is that of trying to create copyright laws and intellectual property right regulations which trample the rights of the citizenry to information, done under the guise of protecting the income-producing potential of authors. We should note well how many authors of articles in stm journals have become wealthy or even moderately well-to-do from their published articles.

Most scientists in the academy believe that in the near future, say within five to ten years, most of the reports of scientific, technical, and medical advances will be published on the World Wide Web. For the rest of this presentation, I will try to convince you that while the processes of scholarly communication are in transition, members of the academy whether associated with their universities or with their scholarly societies can innovate and improve the very nature of their communications as well as effect changes in the marketplace, in the economics of scholarly communication which would bring back to our libraries and our universities and our institutes broad and deep access to a much fuller range of information valuable for the scholarly processes and the general advancement of civilization.


In early 1995, we began the HighWire Press as a unit of the Stanford University Libraries to assist responsible publishers of high impact stm journals in creating Internet editions with advanced features. The American Society for Biochemistry and Molecular Biology and the American Association for the Advancement of Science were our first collaborators. From the first we have attempted to apply high technology and intelligent design to stimulate and then realize the specifications of authors, editors, librarians, and readers.

Stanford's HighWire Press has two over-riding objectives, "church and state" missions, if you will. The "church" mission is to improve scholarly communication. The "state" mission is to contribute to a long overdue correction in the stm journal publishing industry. Stanford is using this method on the publishing side to effect change and is working with the faculty to adjust their thinking and their practices regarding the placement of their articles on the author side of the system.

Today, three years later, HighWire provides services to 28 journals with dozens more in the schedule for design, development, and production in 1998. As the number of publishers working with HighWire has grown, a community of scholarly society and "responsible" journal publishers has grown up, working together to improve scholarly communications. We will have the third meeting of these publishers at Stanford at the end of February. In the life sciences and increasingly in medicine, the concentration of publishers working with HighWire has become a magnet for other journals in some interesting ways.

Rather than speak about the advanced modes of scholarly communication pioneered and now imitated to one degree or another, let me show you some articles to illustrate our approach.

First, note that as I indicated, HighWire Press is a unit of the Stanford University Libraries and Academic Information Resources (www-sul.stanford.edu and highwire.stanford.edu). Among the publishers working with HighWire, the American Association for the Advancement of Science has arguably been one of the most aggressively involved in exploiting the potential of the World Wide Web. One can see on the home page the number of other AAAS communications vehicles in addition to Science Magazine. AAAS presents: Science Now, a daily news service; Science's Next Wave, containing news, forums, columns on career development in science; Science Professional Network citing jobs, career resources, career fairs, meetings, & graduate programs; and the Electronic Marketplace featuring product information, advertiser directory, & reader service. Science Magazine is now a subscription publication available for individual members only, but soon AAAS will offer institutional subscriptions. Science's Next Wave will become a subscription item soon and can be obtained for institutions and I believe consortia or on national site licenses. In the current issue of Science [explain current issue page] (13 February 1998) (www.sciencemag.org/content/current/), there is an article by Gallagher et al. labeled "enhanced." After the authors worked with the editors of Science, librarians at Stanford added what we are calling "hypernotes" to the article to provide a larger context based on relevant Internet resources. Other enhanced perspective articles are listed through the link at the top of the Gallagher article; we may get back to them later. The first hypernote takes the reader to a site named "About Bio Tech" which is addressed to secondary school students and their teachers. In the general hypernote section there is a biotechnology and life sciences dictionary linked. Let us go to that site and search on the term "prion." Returning to the Gallagher article, one notes that the first regular footnote points the reader to a related article in the current edition of Science. Footnote six points to a Medline reference which we shall visit in order to demonstrate the notion of "related articles" in Medline. Additionally, there is a link to an article in an earlier issue of Science, one by Pavletich. The link takes us to the abstract, but can quickly go on to the full text of the article, noticing that there is a link to a section listing other articles which have cited Pavletich among the journals working with HighWire, one in the Journal of Biological Chemistry and another in the Proceedings of the National Academy of Sciences. Note as well that there is a section of "related articles" on subjects close to Pavletich's in Science itself.

From Pavletich, we can link to the full text of the Forget article in the Journal of Biological Chemistry regardless of whether we have a subscription to the JBC, under the terms of an agreement among many of the publishers working with HighWire which we have dubbed "toll free links." Readers in one journal can read full-texts of cited articles in participating journals' pages regardless of their subscription status in the target journal. This is an example of the journal publishers assisting one another to make reading and research more rewarding among the HighWire journals. While we are on the first page of the Forget article, note that one can search for other articles published by Forget and his colleague Gallagher in JBC. Note as well the list of articles citing the Forget article, an example of reverse hypernavigation (forward in time rather than backward in time). Let me increase the resolution of figure 5, a gel of dna material, a couple of times to that you can see the quality of these images through a computer and projected through Humboldt's "beamer." Toward the bottom of the article, there is a list of additional links from this research to the genomic data bases at GenBank and EMBL Bank. One might notice as well in footnote 69 another hyperlink to a full-text article, one in the Journal of Clinical Investigation, published by the Rockefeller Press. Notice that footnote 70 contains a hyperlink to a Science Magazine article.

Hypernavigating to JCI, we find another Gallagher article published in January 1996. Let us go directly from the abstract to the pdf so that you can see what one might do to get passive page images versions of the articles. Before we download the pdf, let us link to the JCI server at Stanford through the German node of Digital Island, with whom HighWire has contracted to provide steady, reliable access to most of the HighWire journals for readers in Europe and elsewhere even when the East Coast of North America arrives at the office and soaks up all the available band-width of the Internet. Let's go to the pdf now; we have pre-loaded it to save time in the talk – it took about 90 seconds. With a pdf and an ordinary laser printer, one can make a paper copy which is more accurate than a photocopy of the article from the printer version. We need to go back to the html version to benefit from hypernavigation and additional information resources. Beyond the references is a section listing other articles citing this article in JCI, one in Blood, another in JBC, and one in JCI itself.

As long as we have mentioned the Digital Island dedicated bandwidth, I should mention that the very high speed backbone network service (VBNS) and Internet II will dramatically affect speed of the Internet among North American research institutions and eventually these techniques will be available to other regions of the world. Such is the growth of Internet usage, we predict that even as the common Internet nodes communicate with one another much more quickly, dedicated band-width such as the Digital Island service will still be necessary.

I happen to know that a recent article in JCI by an author known as Erzurum features an example of a link to a commercial supplier. Let me use this fragmentary information to generate a search. Note in the methods section of the JCI article the link to Sigma Corporation where one can search to locate the assay method. We expect somewhat later that some links will be provided directly to a supplier's catalog descriptions rather than just to a home page.

Let us now go to the home page of the Proceedings of the National Academy of Sciences. Notice that there is an enticement to register for receipt of tables of contents of PNAS issues as they are issued. Soon HighWire will be offering readers personal alerting services based on individual profiles of subjects of interest by which e-mail messages with live links to relevant articles will be sent as the new articles are published on the web in HighWire journals. Numerous other services will follow during the course of 1998.

You may remember the article a year or more ago on the subject of life on Mars. I cannot remember where the article appeared, but suspect it was in the PNAS. So, let us conduct a search there on the terms life AND Mars. The results show only three articles, not one of which is the one I am seeking. Therefore, let us resume the search in the next most likely journal, Science. We expect soon to be offering simultaneous searching across multiple journals. The results show numerous articles, but the obvious one is by Zare et al. Let my expand a couple of the images to show you once again how good the resolution can be. The first is of a slice of Mars rock recovered from Antarctica; the yellow bits are the target areas for investigation for fossilized bacteria-like structures. The second is indecipherable at the thumb-nail resolution. At the highest resolution, we can see precisely what a researcher can see on the screen of an electron microscope. Once again, at the bottom of this article is a list of related articles.

Since we have come full circle, let me show you the list of enhanced perspective articles and search for one which I know contains a link to a QuickTime movie to illustrate a point. (search on Zahs, download the movie and show).

Finally please note that HighWire, working with Oxford University Press is bringing an Internet edition of EMBO, the journal of the European Molecular Biology Organization. And later this year, an Internet edition of the British Medical Journal will be ready for readers. Working with HighWire's European Service Bureau at Harrassowitz in Wiesbaden, Germany, we are fully prepared to serve more learned societies in Europe.


What is HighWire Press:
1. It is Stanford's principal Internet publishing service;
2. It currently provides such services so that 28 different journals are up now and another 90 will follow in 1998;
3. Publishers working with HighWire are largely not-for-profit scholarly societies, but there are some commercial publishers involved (Cell, W.B.Saunders, Williams & Wilkins);
4. Most of the journals are in the areas of life and medical sciences;
5. Most of the journals focus on basic research, but there are many coming on board with a clinical focus;
6. Most publications are journals, but some are abstracting and access journals (Journal Watch), one is a news service (Science Now), and coming soon will be some major reviewing journals (Annual Reviews);
7. New, derivative information products and services are in the works;
8. A monograph publishing project is underway with the Stanford University Press.

How does HighWire serve publishers:
1. HighWire is an extension of each publisher's own staff;
2. HighWire is a kind of partner & service provider; so far it is not a publisher or a licensee;
3. HighWire develops well-engineered production software; this week 10,000 page equivalents will be added to HighWire's servers and over 1 million readers will access them;
4. HighWire's publishers are either already high-impact (i.e. highly cited) or rapidly becoming so;
5. These leading publishers drive HighWire's team of engineers, librarians, designers, systems administrators, network specialists, and business staff to innovate constantly;
6. HighWire shares best-of-breed solutions and innovations whenever possible; publishers working with HighWire routinely stand on one another's shoulders;
7. HighWire's publishers share costs whenever possible;
8. HighWire's architectures are strong and flexible; innovation capacity is built in; (see .ppt slides on basic architecture)
9. HighWire's publishers meet together with the strategic and senior staff to work together to explore common ground and to innovate together;
10. HighWire is based at one of the world's great information technology institutions; we have access to many researchers working ahead of the industry for advice and involvement.

What does HighWire offer to publishers?:
1. responsive and collegial working relationships;
2. strong connections to researchers, instructors, & clinicians;
3. a team of adrenaline addicts;
4. assistance in developing and testing business models;
5. simple, yet effective access controls which emphasize discourse in the academy;
6. a commitment to innovation;
7. a diverse, high quality partnership base;
8. freedom of choice – no required packages and no ties to particular typesetters and printers;
9. friendly, usable systems which stay out of the way of complex information;
10. a clear sense of publisher ownership and control;
11. a compelling environment for scholarly communication with high impact content, useful links, and numerous services, which together draw and hold authors and readers. (see .ppt slides on concentration in top 500 jrnls)


All this having been said, this talk is not meant merely to be a marketing presentation for HighWire's services. Rather it is a wake-up call on the state of affairs of scholarly communication in the scientific, technical, and medical disciplines. It is a call to encourage scientific societies and universities to reassert their historical and beneficial role in publishing the results of work by their own academicians. We in the academy need to do this in order for our investments of time and effort to become more generally accessible to the global academy than is now the case. We need to reduce redundant costs in the system so that more information is available to more academicians, as teachers, as researchers, as students. Peter Murray-Rust's paper on xml at this conference will present some tools for scholars which will make network publishing easier.

The Internet offers alternatives to the present situation by making available high quality, refereed scholarly journals to increasingly networked communities related by topic and discipline rather than being separated by geography and time. We are only a few years from the wide-spread reality of global communities working across time zones and great distances on common problems. The plethora of representations of print journals on the World Wide Web, for instance those distributed on the Web primarily in page images, as well as the smaller number of really effective Internet editions, along with the development of numerous meta-data services and versions on the Web are pointers to new forms of scholarly communication. Paul Ginsparg advocates the creation of new forms of scholarly communication principally involving pre-prints and validation after publication. Stevan Harnad advocates Internet-only scholarly publication with a form of page charges rather than charges to readers or their institutions. Both Harnad and Ginsparg argue that the current business models are doomed ultimately to failure. This may be true, but we at Stanford have decided to assist the society publishers in the transit from a known, but failing system to an unknown but functional one, focusing on the functions and on the tactics and strategies of change which work for each publisher's situation from year to year.

Twenty-seven distinguished scholarly societies have chosen to produce their Internet editions in collaboration with Stanford's HighWire Press. A number of other scholarly societies are either producing Internet editions through their own publishing groups, as in the case of the Institute of Physics, the American Chemical Society, and the American Institute of Physics. Still others are working with university presses, such as those at Johns Hopkins University, the University of Chicago Press, and the MIT Press.

The societies represented in this conference and the universities and research institutes at which their members perform their science are capable of re-capturing the rights to publish their own reports, especially in Internet editions. Whether you seek to work with HighWire or another outfit like it, to create a similar service based in a German university or institute, or to leap over the current state of affairs to a model based on Ginsparg or Harnad is, in a way, less important as method than such a move would be in principle. You and others like you can and should take back responsibility for publishing your own science.

Thank you for your attention.

[an error occurred while processing this directive]