This is an archived article.

May 19, 2005

Addressing the growing dilemmas of scholarly publishing

News and Information

Some people would describe the current system of scholarly publication as crazy. Taxpayer-financed research is published in journals (many part of for-profit enterprises), and to get access to these journals, libraries must seek additional public funding for the ever-increasing price of subscriptions.

Is there a way out of this dilemma?

A panel at the UW recently discussed the rapidly-changing and contentious world of scholarly publishing. Production of scholarly articles continues to increase, fueling an increasing number of journals, and libraries are faced with subscription costs that are rising at about three times the Consumer Price Index, said Joyce Ogburn, associate director of UW Libraries.

“If we cancel a subscription, we can lose access to all current articles, because in many cases all we have is a license. This means individuals would have to pay for their own access,” Ogburn said.

The Internet is creating new possibilities for sharing knowledge, including an idea known as open access, which Ogburn defined as “being able to share products of intellectual inquiry at any time and everywhere.” Congress recently considered legislation that would have required research funded by the National Institutes of Health to have its papers posted on a site with free public access within six months of publication. This proposal was scaled back to a voluntary process within a year after publication, when the proposal was greeted with a storm of protest from commercial publishers.

“The current system is broken,” said Carl Bergstrom, UW assistant professor of biology. He talked about his own analysis of journals, revealing a “staggering difference” in the cost per article between for-profit and nonprofit journals. His research shows that for-profit journals cost libraries three to six times as much per page as do nonprofit journals. “The cost is unrelated to quality,” he said. “In fact, the top journals, in terms of impact, are usually the nonprofit journals.” Detailed charts and figures are available from Bergstrom’s journal economics website:

Barbara Cohen, a senior editor at the Public Library of Science, talked about the stake that society has in maximizing access. “Society supports scholarship because it improves society, but only if the results are shared widely,” she said. The Library of Science’s model of open access ( has two key elements. First, information from articles it publishes is licensed for use in “any lawful way,” as long as the information is attributed to its source. In addition, a complete, electronic version of the article will be maintained in a public repository in perpetuity.

“We should have no financial barriers to information, and this is especially true for health-related information,” Cohen said. “Online publication makes the information available for many uses, including text- and data-mining. It potentially maximizes the benefits from the research, and it helps to promote public trust in scientific research.”

However, as Cohen acknowledged, the financial model for open access is not yet proven. The Library of Science has received a $9 million startup grant, which it “burned through pretty quickly,” Cohen said. Its initial publications have rejected about 90 percent of submitted articles and hired many professional editors; they also have contained a “magazine” section of news and comment that is costly to produce because many writers contribute to it. But the library is launching a series of journals that more closely resemble those published by professional societies, which will accept a higher fraction of submitted articles and will not contain a magazine section.

Such journals, Cohen said, can be sustainable at a cost of about $1,500 per published article. The Library of Science currently charges authors that much for an accepted paper, but will waive the charges if the author is unable to pay.

“We believe in the long run that open access will bring down publication costs,” she said. “We’re at a transition phase now, so it may not be apparent in the near term.”

There are alternatives to the Library’s model, including “open archives,” in which a publisher agrees to have the final submitted copy of an article published on an individual or institutional Web site. Such an approach raises other problems, including the standardization of formats so that the articles can be located using search engines. If the article is republished on individual sites, it may not be permanent. “And if this idea catches on, are publishers likely to allow it to continue, if it affects their profitability?” Cohen asked. Moreover, permission to place articles in open archives usually comes with copyright restrictions, so the article cannot be integrated into scholarly databases.

The ground rules that publishers try to set when an article is accepted for publication are not immutable, said Gail Dykstra, software technology manager with UW TechTransfer Digital Ventures. The Digital Ventures Web site contains information about how authors can modify the agreements, either by changing specific passages or adding an amendment. ( “Universities can help their researchers by developing models and examples, and also by making it clear that they support scholars in modifying agreements to meet their needs,” she said.

“We’re at a time of revolutionary change,” Dykstra said, “and it’s hard to make the transition.”

Ogburn said, “We need more competition and experimentation if we’re going to break down the old way of doing things, which to many people appears no longer sustainable.”

The panel discussion was presented by the Digital Media Working Group, the Institute for International Policy’s Center for Internet Studies and the Simpson Center for the Humanities.