The so-called "Budapest Open Access Initiative" calls on scholars to post their work on the Internet and to create alternative, Web-based journals available for free to all researchers.
Backing from Soros's Open Society Institute will amount to $1 million a year for three years, Darius Cuplinskas, society spokesman, said. The money will be used to find new ways of publishing scholarly literature while maintaining its quality and making it freely available to all, he added.
"Having something like the Soros Foundation backing us says to the world this is real, this is not just a bunch of idealistic, naive scientists running around," said University of California genomics professor Michael Eisen.
Eisen is one of the founders of the Public Library of Science, a group trying to start Web-based, free journals dealing with the life sciences. Under their plan, the costs of reviewing, formatting and posting a paper would be borne by the author, as one of the costs of doing research. The cost per published paper would probably be roughly $300, Eisen said.
Currently, most research appears in printed journals, many privately owned and available only by subscription. The cost of such journals, Eisen said, creates "an impediment to the free and open exchange of ideas."
Universities and other research institutions currently can afford only a fraction of the roughly 20,000 scholarly journals published around the world every year, said philosopher Peter Suber, one of the architects of the Budapest plan and a professor at Earlham College in Richmond, Ind.
Unlike most writers, researchers are not paid for published work, Suber said, and neither are the academic reviewers -- dubbed "peer reviewers" -- who must approve manuscripts for publication.
"A lot of people are donating their labor and their intellectual property," he said, "but the readers aren't getting it for free."
Aside from creating Web-based journals, the Budapest plan calls for academics to "self-archive" their work, according to cognitive scientist Stevan Harnad of the University of Quebec in Montreal, who is the main proponent of the idea. Under that system, researchers would submit their work to journals as usual, but as well they would post the peer-reviewed final version on special university-run Websites. "That would give open access right away," Harnad said.
While some journals are relatively inexpensive, the cost and number have steadily been rising, according to Graham Bradshaw of the University of Toronto library. One of the most expensive, he said, is Brain Research, from the Dutch publishing house of Elsevier Science -- a 2002 subscription costs libraries $18,578.
Despite spending nearly $5 million a year on journals, "we haven't been able to acquire all of the things we should have," Bradshaw said.
It's possible for journals to keep costs down, said Jeffrey Drazen, editor of the prestigious New England Journal of Medicine, which costs $140 a year but is free to physicians and researchers in the Third World.
Drazen said his journal offers a valuable service to readers, sorting through thousands of research submissions, professional editors work with academic reviewers to make sure the published papers are accurate and valuable.
"We don't just take what the author sends us," he said. Subscribers "pay us to filter out what is important," Drazen said.
(Reported by Michael Smith in Toronto.)