Cyberjournals offer
faster, cheaper and
fuller research news

Stephen Strauss*
Globe and Mail, 6 April 1996, p. D8

ONE of a science writer's most entertaining pastimes is attending conferences at which researchers debate what the "literature" has to say about the issues of the day.

These litterateurs aren't exchanging bon mots about works by Kafka or Jane Austin. Rather, they are analyzing such classics as Co-ordination of three signalling enzymes by AKAP79, a mammalian scaffold protein. You might say that scientific literature is to literature as potatoes are to Einstein. The analogy is unfair, though: The purpose of scientific communication is to tell colleagues about your work, not dazzle them with your style.

Science produces a blunt literature. And today's scientists are being just as blunt when it comes to choosing between the printed word and the electronic word. Indeed, some scientists say print publishing is not about communication so much as slowing down communication -- and grabbing huge profits.

Therefore: Death to print.

Such hostility to the printed word is largely prompted by declining library budgets. To save money, the University of Calgary cut 3,000 journals to which it subscribes over a five-year period, similarly, between 1984 and 1995 the University of Toronto cut 4,000 magazines from its subscription list.

Why? Because scientific journals can be very, very expensive, even though scientists often have to pay several dollars a page to have an article published.

In a field so highly specialized that competition is often only a fiction, publishers can charge whatever the market will bear. According to Forbes magazine, scientific journals published by Amsterdam-based Reed Elsevier earned a pretax profit of nearly 40 percent -- $225 million (U.S.) on sales of S600 million (U.S.) -- in 1994. One of its journals, Nuclear Physics B, costs $9,909 (U.S.) a year, for 75 issues.

But it isn't just the cost of print that has turned scientists toward electronic publishing. Print is slow, too. It is often six months to a year before a submission to a print journal appears. This is called the molasses effect.

Which is why some scientists are deciding that the way of the future is a return to a computerized version of the 17th century.

Four hundred years ago, scientific publishing took place when researchers wrote one another describing their recent work. (Some historians have argued one of the most important elements in the rise at a scientific culture was the emergence of a quick and reliable mail service.)

Now comes Paul Ginsparg, a theoretical physicist at the Los Alamos National Laboratory, who has set up a 21st-century post office. Thanks to Dr. Ginsparg, physicists, mathematicians and others can get electronic previews of yet-to-be-published journal articles at an Internet Web site at LANL.

In the past, such preprints were circulated among select members of the scientific community. Now everybody is a select member: Each week, the web site (http:/// receives more than 200,000 requests for E-prints from tens of thousands of scientists worldwide.

E-prints are a very economical way of communicating. Dr. Ginsparg estimates he can store about 25,000 papers on a hard disk, at a cost of less than 2 cents apiece. A researcher can get an electronic copy of one of them for a few cents' worth of telephone time.

Sounding like the Karl Marx of the information age, Dr. Ginsparg describes his electronic repository as the death blow to an exploitive system in which publishers interpose themselves between the best interests of their contributors and their readers. "They [the publishers] get high-quality content for free and then sell it at a high price back to those who supply it," he says.

Electronic journals could be produced by scientific societies for a fraction of the cost of putting out a journal, he adds. In fact, an Industry Canada study of some Canadian journals shows producing and distributing what Dr. Ginsparg disparages as "chemicals adsorbed onto sliced processed dead trees" -- that is, ink on paper -- account for 35 to 60 per cent of their costs.

But times are changing. One indication: A 1995 directory of electronic journals contains 700 entries, up 66 per cent from 1994. One concern is that the software being used to produce electronic journals will be obsolete in 50 years and nobody will be able to read them. But this problem may provide its own solution, stimulating development of programs that will quickly translate documents written with obsolete software.

There's little doubt that electronic publishing is the medium of the future. Indeed, some magazines have begun using the Internet to complement their print offerings because journals can no longer print all the basic data that underlie most articles. For example, Nature magazine recently told its potential authors that they must make their papers on the molecular structure of genetic material available in a public electronic database or Web site where all readers could access them.

The magazine recently published a four-page brief description of the most complete mapping so far of human DNA. Publishing the complete research would have required more than 500 pages; instead, scientists were referred to a web site where all the information was stored. The four-page description in Nature was less a report and more of a signpost pointing to the data.

If only Marshall McLuhan were alive to turn all this change into an epigram. Maybe something like this: Fresh-baked print is stale science.

Stephen Strauss is The Globe's science reporter.

References and links about Electronic Journals
Writings page