Machine learning and the academic journals associated with this rather specialized area of technological study are unlikely candidates for controversy. However, last fall the discipline vaulted into relative notoriety with the mass resignation of 40 members of the editorial board of The Machine Learning Journal (MLJ) and their migration to the competing Journal of Machine Learning Research (JMLR).

It was not a search for a more exciting name that sparked the exodus. According to editorial board member Professor Michael Jordan of the University of California at Berkeley, the move was prompted by a belief the JMLR, which allows authors to maintain copyright in their work and is distributed at low cost over the Internet, better serves the needs of the machine learning community than the privately published MLJ.

The rise of such scholar-controlled electronic journals is a growing phenomenon. In Canada, Acadia University biology professor Phillip Taylor is one of the driving forces behind the electronic journal Conservation Ecology.

He touts the rise of scholar-controlled, peer-reviewed publications as a viable alternative to the high priced scientific journals published by large multinational publishing companies.

And Taylor says the technology now exists for researchers to run their own systems of communications.

"When the publication process was a bricks-and-mortar operation, it might have made sense for researchers to surrender control and sign their copyright away to the big publishing houses," he said, "but with the Internet there is no reason that we shouldn't be doing this ourselves."

Fred Ziegler, a member of CAUT's Librarians Committee and an academic librarian at the University of Alberta, traces the roots of the movement by academics to gain control over scholarly communication not simply to technology, but also to the restructuring that has occurred in the scholarly publishing industry over the last 20 years.

As a result of mergers and consolidations, the industry is now dominated by a handful of large players who have utilized their market control to increase subscription rates at a pace far in excess of inflation.

"The existing system was becoming untenable," Ziegler said. "Journal cancellations in university libraries became an everyday fact of academic life and it was inevitable that alternatives would arise."

One of these alternatives, it seems, is the e-journal. However, they are not yet viable solutions to the crisis in scholarly communication. The archiving and stability of digital content and preserving the integrity of digital back files are two issues of major concern. In addition, in most universities, articles in electronic journals do not carry the weight in tenure and promotion committees that articles in print journals carry.

Nonetheless, a growing awareness of the issues associated with scholarly publishing is percolating out of academic libraries and filtering down to faculty members.

According to Ziegler, initiatives geared to facilitate scholar control over systems of communications, such as SPARC in the United States (www.arl.org/sparc) and "Create Change," its Canadian counterpart (www.carl_abrc.ca), are positive signs.

"The academic community has been slow to awake to the crisis in scholarly communication," Ziegler said, "but now that it is happening, things will improve."