Provosts Push a Radical Plan to Change the Way Faculty Research Is Evaluated

By ROBIN WILSON

Peer review and publication of scholarly articles would be radically altered under a new plan being promoted by a small but influential group of academics. The group hopes the change would create new outlets for professors' work, both in print and on line.

The motive is primarily financial. Supporters of the plan want to create alternatives to high-priced journals, particularly in science and technology, for which annual subscriptions can cost as much as $15,000 apiece. University leaders backing the plan want to encourage scholars to publish less often in such journals, and instead to post their scholarly work more frequently on electronic sites. And they want universities to give up requiring that professors actually publish their work as a key step toward tenure or promotion.

"This is like the evolution to the automobile from the horse and buggy," says Charles E. Phelps, provost of the University of Rochester and a major proponent of the new plan.

The proposal, which has only recently been formulated and isn't yet widely known on campuses, already has some detractors, who view it as unrealistic. Among the skeptics are leaders of two of the nation's largest scientific groups, the American Physical Society and the American Chemical Society.

The plan also has important allies, though, in the Association of American Universities and the Association of Research Libraries. "We are a captive market being gouged by commercial firms that see us as having to buy the journals at whatever price they wish to charge," says John C. Vaughn, executive vice-president of the A.A.U.

Those two associations, which co-sponsored a meeting of scholars, librarians, and provosts last year with the Pew Charitable Trusts, followed up in March with a paper that endorses exploration of a new form of peer review. The point is not to change the goals of peer review -- identifying the best ideas and getting them out to the field -- but to create a new way to meet those goals.

"We are calling for neither a lessening of the importance of research in the criteria for promotion and tenure, nor a turning away from peer review," the paper says. "What we seek is an alternate means of achieving those ends."

Mr. Phelps and the provosts of Columbia University and the California Institute of Technology, among other academics, have been pitching the plan to learned societies.

In the current system of peer review, professors who want to get their ideas out must publish them in a scholarly journal, either non-profit or commercial. They submit their papers to a journal, which solicits reviews from two or three experts in the field. The experts, who usually remain anonymous to the authors, decide whether an article should be published.

The new proposal has been described as "decoupling," because it would separate peer review from publishing. Instead of sending an article to a journal to be reviewed, professors would send their work to a "certification panel" in their field.

The panels, which would be established by scholarly groups, would use the same kinds of experts that journals now use. But the new panels would give each article a grade or a stamp of approval. A professor could then do one of several things: try to get the article published in a major journal, take the article to one of several non-profit journals that would accept a certification panel's stamp of approval, or post the article on a World-Wide Web site and skip print publication altogether.

Publishing an article in a prestigious journal is a crucial part of earning tenure or a promotion. Under the system that Mr. Phelps envisions, universities would accept a certification panel's stamp of approval and not require that professors actually publish their work.

"It would be 'certify or perish' rather than 'publish or perish,'" he says.

If fewer professors seek paper publication of their articles, supporters of the plan say, some specialized journals might go out of business, and others would lower their prices in the face of competition from cheaper, non-profit publications, including electronic journals.

To help spur that kind of competition, the Association of Research Libraries has started a coalition to persuade non-profit scholarly societies to start new journals. Those publications would compete directly with journals from commercial publishers but would cost less, because the societies do not have to earn a profit on them.

The commercial journals that have come in for the most criticism are published by Elsevier Science, a company based in Amsterdam. They include such titles as Brain Research and Nuclear Physics B, which cost subscribers $15,428 and $11,453, respectively, this year.

John Tagler, head of corporate communications for Elsevier, says there are good reasons for the pricey subscriptions. It takes time and money to establish editorial boards and to hire reviewers to sift through the thousands of articles that are submitted to a journal, he says. And some of Elsevier's journals are published more than 100 times a year. Brain Research, for example, prints 125 issues a year. "Running an international journal with peer review and archiving are all very complex issues. They aren't simple or cheap," he says.

The idea of creating a new system of peer review is not outlandish, Mr. Phelps contends. In fact, it is based in part on what is already happening in physics.

Seven years ago, Paul Ginsparg, a researcher at the Los Alamos National Laboratory, created a data base of physics papers on his home computer. Its popularity has changed the way scholarly ideas in physics are shared. Professors post their work to Mr. Ginsparg's electronic archive before the papers are even sent off for peer review. They often revise their articles after receiving comments from colleagues who have read them on line. Some of the papers are eventually peer reviewed and published in journals, but others never see print. About 25,000 new articles are posted to the archive each year.

"Nobody is hurt by instant dissemination," says Mr. Ginsparg. "Authors are helped by getting instant feedback from large numbers of interested readers."

He adds: "We're hoping that researchers in other areas will wake up and realize that their entire mode of communicating their research is obsolete."

He may be hoping for too much. So far, the proposed changes in peer review and publication have received only a lukewarm response. Mr. Phelps wants learned societies to consider setting up the first "certification panels." But in a meeting earlier this month, representatives of the American Physical Society and the American Chemical Society said they didn't think the new plan would fly.

"They have the wrong answer to the right problem," says Robert D. Bovenschulte, director of publications for the chemical society.

Something should be done to reduce the amount of money that libraries spend on scholarly journals, but it's not clear that the entire system should be changed to accomplish that, he argues. "Today, if it's not published in print, it's not valuable," says Mr. Bovenschulte. Convincing professors that articles could be reviewed but not published "would be a change of values by fiat."

H. Craig Heller, chairman of biological sciences at Stanford University, agrees. "The system isn't broken, so why try to fix it?" he asks. He suspects that most professors would be dissatisfied simply posting their papers on the World-Wide Web.

"What if a paper gets a low grade from one of these review boards?" Mr. Heller adds. "People are not just going to throw the paper in the trash can. They are still going to try to get it published."

Other skeptics express doubt about the proposed certification panels. If the panels were controlled by the learned societies, those professors who felt that the societies did not represent their views would be reluctant to seek their stamp of approval.

"I worry about a lack of competition for ideas," says Peter Givler, executive director of the Association of American University Presses. "The current system is inefficient, but its virtue is that it guarantees that no single body is going to be able to stifle work or steer people away from work that it regards as unimportant."

Mr. Phelps realizes that it will be difficult to change something as fundamental as peer review. "This will either be happening within the next year or two," he says, "or it will die because nobody wants to go forward with it."