Highlights from Syllabus Magazine

October 1998
Volume 12, No. 3

Science, Scholarship, and Internet Publishing:
The HighWire Story

By Michael A. Keller, Stanford University


Scholarly journals have found an effective and efficient outlet on the World Wide Web. This article details the history of Stanford's HighWire Press, and the advantages of this publishing model.

Scientists, scientific editors and publishers, scholarly society officers, and an enterprise unit of the Stanford University Libraries named HighWire Press have worked together over the past three and a half years to publish Internet editions of 70 influential scientific journals. Three significant accomplishments have resulted. First, there has evolved a mode of scholarly communication which serves readers, and facilitates research as much as it supports the clarity and validity of scientific discourse; this model has become a standard in Internet scholarly publishing. Second, an active community of scholarly editors and publishers has intensified the benefits of online scholarly publishing to the scientific, medical and technical communities at large. Third, the products of life sciences research in the advanced economies of Europe and North America are now more widely available than ever before, stimulating scientific and other cultural developments in other parts of the world.

HighWire Press was established in early 1995 in an agreement between the Stanford University Libraries and the American Society for Biochemistry and Molecular Biology (ASBMB). The purpose of the agreement was the joint development of an Internet edition of the Journal of Biological Chemistry (JBC), now the most frequently cited scientific journal in the world. At the same time discussions began with the editorial leaders of Science magazine, and in the fall of 1995, the first online issues of Science were released. Simultaneously, discussions with the publishers of the Proceedings of the National Academy of Sciences and the Journal of Neuroscience led to Internet editions of these highly significant journals.

Publishers working with HighWire pay for services provided, but we encourage these colleagues to treat HighWire as though it were part of their own information technology group. Publishers determine independent of one another and of HighWire their own business models and subscription rates. We have begun a marketing group for some of the publishers to supplement, on a global basis, their own marketing efforts. HighWire Press is a self-supporting service bureau.

"Church" and "State" Missions
The practical and technical underpinning of the HighWire Press services to scientific, technical, and medical (STM) publishers, mainly of non-profit society journals, are driven by two strategic missions: the enhancement of scholarly communication; and the improvement of the competitive position of such journals in the marketplace for such information. The first we call the "church" mission and the second has been dubbed the "state" mission. This dual mission derives from the fundamental roles of research universities, namely the creation and dissemination of knowledge and wisdom through research and teaching. For all scholars in all disciplines, reporting on the results of research in refereed articles and books is an essential aspect of the practice of scholarship. A key notion of scholarly communications is that of validation before publication; refereeing in scholarly publishing is a vital function because it guarantees adherence to principles of scientific method, establishes trust in the information and methods presented, and thus encourages the dialectic among the readers of the communication.

For the past two decades, treatment of STM information as commodities by certain publishers has created a crisis in research library budgets. Increases in STM journal prices have been responsible for fueling the 147 percent increase in STM journals from 1986 to 1996. During the same decade, as a result of library budgets not keeping up with such increases,

7 percent fewer journals and 21 percent fewer monographs were acquired on average across North America's 100 largest research libraries. Thus in a time when vast numbers of new articles are written and printed (and some would argue that too few articles are rejected), in most research universities, less of the growing literature is available. (See "University Libraries and Scholarly Communication" for an extensive description of the situation). Among Stanford's responses to the crisis in STM serials' pricing is the creation of HighWire Press, which in turn works with responsible STM publishers to improve their own competitive positions with regard to recruiting authors and articles, serving readers, adopting and promulgating subscription and business models, and returning value to the research community. A paper presented by this author at a conference of German scholars, librarians, and publishers provides some details in support of these notions (See "Returning Responsibility for Scholarly Communication to the Academy").

Hyperlinks to Medline
From the start of the relationship with the ASBMB, an agreement to demonstrate how hyperlinking would benefit research was forged with the leadership of the National Center for Biotechnical Information at the National Library of Medicine. NCBI staff creates and publishes Medline, the single most important database of abstracts and other information about articles in the clinical and life sciences. The result of the agreement was the transmission in digital form of abstracts and bibliographical information about articles in JBC (and subsequently all the other journals published on-line with the assistance of HighWire Press); such transmission accelerated the intake of such information into Medline, a benefit for physicians and researchers in the life sciences world-wide. This agreement with NCBI created a model for linking cryptic citations of articles and other information sources in footnotes and references of newly published scientific reports to the abstracts in Medline.

Digital Services
Most of the journals' editors and publishers have used the technical possibilities of Internet publishing to intensify the benefits to their readers though such services as "toll free linking" and "cite tracking."

"Toll free linking"permits linking from the citations of articles to the articles themselves when such cited articles are available in digital form in the HighWire environment. The publishers participating in this program decided that subscribers (whether institutional or personal) to the citing journal would be permitted to read the cited article in another journal even if the reader did not have a subscription to the cited journal. The practical value of "toll free linking" to readers of journals in the HighWire environment is apparent when one realizes that over 50 percent of all journal articles in the life sciences in the top 500 most cited STM journals are provided by those publishers whose Internet editions are supported by HighWire Press. Publishers of journals not working with HighWire Press have recognized the value of this sort of function for researchers and readers, and have begun to do the same within their own communities of discourse.

A new function available to members of certain societies publishing Internet editions is that of "personal research alerting" or "cite tracking" by which members are alerted by e-mail whenever new content in a journal is published that matches criteria set by the Member. The criteria can be based on topics, authors, and particular articles. Or HighWire can announce to a member when a new issue has gone on-line along with the entire table of contents. For each of these services, the e-mail message includes live hyperlinks, which send the recipient directly to the cited article. The value of these services to researchers must be seen in the context of the number and significance of journals and thus the number and significance of articles available through the service. Together these 36 journals served by Highwire publish thousands of articles each week. All of the 36 journals involved in the two services are either in the 500 most cited journals in STM publishing or are sibling publications of one such; a list of participating journals is available at http://www.jbc.org/help/citetrack/journals.dtl.

HighWire Press has joined other parties to ask the NCBI to provide a service called PubRef which will provide what might be called the "biblio-technical" information necessary for one publisher to make links to another's journal articles. The key to making such links will be agreement between and among publishers that such links should be made. For some publishers, a powerful consideration will be those of HighWire's "state" mission, enhancing scholarly communication. For other publishers, considerations will be those of their own economic or business needs and models.

Globalizing Scientific Scolarship
Such has been the demand for effective access to these high impact journals that some innovations in distribution and in subscription models has been necessary so that the publishers could satisfy readers. The problem of reduced rates of transmission of data across busy time zones direly affected readers in Western Europe each weekday after about 2:00 p.m. European time. When people awoke in the Eastern time zone of the U.S., their use of the Internet for all sorts of purposes essentially pre-empted bandwidth in that sector, thus impeding communications between European readers and HighWire's West Coast servers at Stanford. After more than 18 months of considering alternatives and attempting to interest the major providers of Internet services, HighWire discovered a company which offered what amounts to private global Internet services. That company, Digital Island, has provided HighWire with distribution services avoiding the public Internet saturation effects in various time zones. The result is that, in many countries worldwide, readers of certain journals working with HighWire can now avoid transmission problems on the public Internet.

Another effect of Internet publishing of high impact STM journals has been the increased interest and awareness of scientific advances in marketplaces formerly served inadequately by traditional surface and airmail. What has been discovered, most often due to the initiative of leaders of national scientific and library or information service providers in such countries as Australia, the Peoples Republic of China, India, and Russia, as well as several South American countries, is that Internet editions of these journals are available to their researchers and teaching organizations instantly rather than after a wait for delivery of the physical issues. Internet editions thus provide a possibility for leveling the playing field in sciences. Some of the publishers working with HighWire Press have decided to offer special subscription rates for national or provincial consortia of universities and research organizations, as well as Third World subscription rates.

Technical Questions and Answers
The progress of HighWire Press is the result of the brilliant and focused efforts of John Sack, Director and Associate Publisher, working closely with the editors and publishers and with a talented and dedicated staff of engineers, designers, production specialists, network technicians, librarians, and managers at Stanford. Of special note are the specifications for the system of Internet publishing developed by HighWire which require as much automated processing of files and images from publishers as possible, and a robust technical environment involving an array of servers and supporting services in a highly evolved computer and network environment provided by Stanford's own Computing and Communications Services group.

Is it possible to copy HighWire so that other institutions could provide similar services? In principle, we think it could be done, but a copy is most likely to be successful in an environment like the one provided by Stanford, the Stanford University Libraries, and the Stanford computing and network infrastructure. What we are sure of is that HighWire's software applications and process cannot be shrink-wrapped and supported as an integrated product. There are too many rapidly changing elements to the software and networking environment to afford such an approach. Rather, we imagine similar developments at universities and scholarly societies with some degree of information sharing and sense of common causes among those sharing our church and state missions.

Another question often asked is whether more disciplines will be served by HighWire within the STM specialties or in the humanities and social sciences. We are interested in such expansion, but it is precisely because there is so much published in the sciences, technologies, and medicine that Internet editions such as the ones we support are affordable. We have seen our costs decline rather dramatically from our earliest efforts. For the kinds of enhanced scholarly communications we are now supporting to be most effective, a critical mass of articles and other information sources must be assembled. Without such a mass, the extensive benefits we are now able to provide would be prohibitively expensive.

Coming next will be more functions and information available only in Internet editions. Some publishers working with HighWire have pioneered a rapid publication of articles judged by editors and referees to have special significance; Science and Pediatrics come to mind as early adopters of such possibilities. The practice of including supplementary data, moving images and sounds, and complimentary hyperlinks to Internet sites with relevant information will increase as well. We are participating in the international effort to create reliable and acceptable ways of providing archival storage of Internet publications too.

The traverse to a much more vivid world of communications via the Internet and its successors is in its infancy. As research and development in computer and networking sciences advance and provide smart applications engineers, systems integrators, librarians, and editors with more possibilities, new functions and services will appear. At HighWire Press and Stanford at large, we seek to contribute to such advances through the fulfillment of our strategic missions.

Michael A. Keller is university librarian, director of academic information resources, and publisher of HighWire Press at Stanford University. His earlier career appointments were as lecturer and music librarian at Cornell, Berkeley, and Stanford. Before coming to Stanford in 1993, he was associate university librarian and director of collection development at Yale University.


"Back to the future: at last librarians chart a new course in scholarly electronic publishing" by Frances C. Wilkinson, Nancy K. Dennis, and Barbara Rosen in Against the Grain; linking publishers, vendors and librarians, vol. 9, no. 5, November 1997, pp.80-85, 93 at http://www-sul.stanford.edu/staff/pubs/atg.html and http://highwire.stanford.edu/about.dtl.

University Libraries and Scholarly Communication. A Study Prepared for The Andrew W. Mellon Foundation by Anthony M. Cummings, Marcia L. Witte, William G. Bowen, Laura O. Lazarus, and Richard H. Ekman Published by The Association of Research Libraries for The Andrew W. Mellon Foundation in November 1992 at

"Returning Responsibility for Scholarly Communication to the Academy" a paper presented at the Information und Kommunikation Workshop "Wege in die Zukunft ñ Elektronische Zeitschriften II ñ International Symposium on Electronic Journals" Berlin 16-17 February 1998 at http://www-sul.stanford.edu/staff/pubs/makgerman.html.

A list of the participating journals is available at http://www.jbc.org/help/citetrack/journals.dtl

A press release by Digital Island is available at http://www.digisle.net/customer/stanford.html.

October highlights

Syllabus home

Copyright © 1998 by Syllabus Press, Inc. All rights reserved.