Bowing to pressure from commercial publishers and the Software and Information Industry Association (SIIA), the US Department of Energy on November 4 closed the PubScience database. Patterned after the National Library of Medicine's PubMed, PubScience had provided free access to research publications on the physical sciences, creating unfair competition for commercial databases, publishers argued.

"The introduction of a free product, even an inferior one, runs significant risk of driving out cost-based products and therefore eliminating competition, resulting in a lack of choice for users," according to David LeDuc, public policy director of the SIIA. LeDuc told The Scientist, "cost pressures were already there" for SIIA members because of PubScience.

SIIA says that PubMed is safe from their efforts, which are now focused on other federal databases in the fields of law and agriculture. The National Agricultural Library's AGRICOLA database is most likely next to be targeted, LeDuc said.

"PubMed is more statutorily sound," he explained, because PubMed was created before more recent Federal regulations that require the government to back off if proposed projects compete with private sector offerings.

"Increasingly, there's a blurring line about what the government's role is," LeDuc added. "It's not always in the best interests of the people for the government to provide everything the people can't afford."

Librarians, not surprisingly, disagreed and the American Library Association mounted an unsuccessful campaign to save PubScience. In a 2001 review, Peter Jasco, associate professor and chair of the Library and Information Science Program at the University of Hawaii, offered a scathing commentary on one of the proprietary alternatives to PubScience, Elsevier's Scirus. Though the Scirus database advertises "scientific information only," Jasco's review described tests demonstrating how often Scirus returned pornography, personal Web sites and other "useless trivia."

But PubScience was also far from perfect, according to George Porter, librarian at Caltech's Sherman Fairchild Library of Engineering & Applied Science. "It was not a real research-caliber database. The financial impact on research libraries will be zero. No research library, deserving of the "research" designation, was dependent upon PubScience for subject access to the literature. The database didn't have the intellectual effort of a professional indexing staff behind it— it did not provide subject indexing comparable to PubMed or [the physics database] INSPEC."

However Porter, acknowledging that he is at one of the best-equipped private institutions in the US, has sympathy for what the end of PubScience will mean to those far from major research centers. "The casual web surfer, the independent scholar, students at smaller institutions are the ones who will feel the loss most intensely," he said. "Many public libraries, K-12, community colleges, 4-year colleges do not have the wherewithal to provide unlimited access to [engineering database] Compendex, INSPEC, and other major subject databases. PubScience was a free utility to help bridge that gap."