Darwinism And The Internet

Stevan Harnad does not look like a revolutionary. But this distinguished researcher and director of the cognitive sciences center at Britain's University of Southampton admits he's urging the virtual overthrow of one of science's most cherished institutions: print journals. With Internet usage exploding, Harnad and a growing legion of scientists say the venerable scientific journals are just too slow. They are urging scientists to post research papers electronically first, and only then submit them to journals.

It's a radical notion that is being advanced at conferences and, of course, on the Net. If it succeeds, it could dramatically change--and perhaps weaken--the traditional process by which refereed periodicals vet research and shape entire fields of science.

ON GUARD. Not surprisingly, the publications already are pulling up the drawbridge. On June 22, an editorial in the New England Journal of Medicine will lay out its stance. NEJM will "apply the same rules to Internet that apply to publishing anywhere else," says editor Jerome P. Kassirer. Translation: It won't consider articles previously published, including on the Internet.

This is no small dispute. The development of electronic publishing is rocking the $4 billion technical publications industry, which includes roughly 50,000 peer-reviewed serials. Journals want to retain their prestige and franchises, yet economics--and agitation by scientists--are pushing them toward electronic formats. "Costs are up, postage is up, and ad revenues are down," notes Larry E. Joyce, president of publishing and multimedia for the American Medical Assn. "You can't grow enough new revenue sources. We've got to look at electronics as the future."

There already are 150 online-only scientific journals, and dozens of publications now exist in both print and electronic forms. In the coming weeks, some of the nation's best-known journals are going electronic. On June 23, Science will post its first Web page, offering a table of contents to the current issue, an editorial, and a discussion forum for scientists. Later, it hopes to add other features not found in the magazine. By August, the Journal of the American Medical Assn. will electronically publish abstracts of current articles and archived back issues. Britain's Royal Society of Chemistry is developing an electronic journal that will let readers rotate three-dimensional graphics of molecules on their computer screens--a valuable tool for chemists and biologists.

But the trend that's raising hackles is the spread of "preprint" Internet servers--there are about 70 up so far--offering scientific papers that have not yet been formally published. Pioneered by physicists, the preprints' speedy dissemination of ideas is being adopted by scientists in other fields. Patricia Kreitz, a librarian at the Stanford Linear Accelerator Center (SLAC) in Palo Alto, Calif., says "the influence of the Web on the physics community has been phenomenal," decreasing travel and increasing

collaboration. Monthly "hits," or requests for SLAC preprint data from physicists, have grown about sixfold since 1993 (chart).

BUM STEERS. What worries some scientists is the potential for the publication of flawed research if the vetting process used by the journals is bypassed. Nobel laureate and Rockefeller University geneticist Joshua Lederberg routinely posts works-in-progress on his own Web page, but he worries that no-holds-barred Net publishing could result in "low credibility, instant regurgitation." Journals such as Science and NEJM have intense and sometimes protracted vetting procedures that reject 90% of the articles submitted. Losing such procedures may not be so bad in some disciplines--an error in a physics string theory will eventually be found and corrected. But in medicine, wide and rapid dissemination of bad advice could literally be deadly.

Still, scientists such as Los Alamos physicist Paul Ginsparg contend that peer review can be just as rigorous and unbiased--and far more efficient--when done electronically. Indeed, Harnad's review, Psycoloquy, is both electronic and refereed. And many editors of traditional journals admit they'll have to adapt. "We've only begun to scratch the surface of how much more effectively we can communicate," says Floyd E. Bloom, editor of Science. The way scientific use of the Net is growing, he and other journal editors will soon find themselves scratching pretty hard.

Before it's here, it's on the Bloomberg Terminal.