Box 1. Open house

Online pioneers they are not, but traditional publishers are not entirely stuck in the past. Publishing online often means bundling supplementary information with a mirror copy of the print article, but the web is now being used to open up some journals to more interactive discussions — previously only possible at conferences.

The BMJ website led the way in allowing readers to post 'rapid responses' to published articles. But in June this year, the BMJ changed its criteria for accepting online contributions — adding heavier moderation. Journals thinking of adding companion blogs (see main text) will also want to moderate comments.

Atmospheric Chemistry and Physics (ACP), published by Copernicus, uses online discussion to open up the peer-review process. Papers are published online quickly and referees post comments online, anonymously if they wish. Authors, and other researchers, can chip in as long as they identify themselves. After the discussion is closed, editors use it to shape the final version of a paper.

Advocates say the online debate improves the final product. "It lets others see what the leading people in the area are thinking and forces editors, referees and authors to work at a higher standard," says Scot Martin, an environmental chemist at Harvard University and an editor at ACP.

Arne Richter, managing director at Copernicus, has high hopes for the journal, which has gained a healthy impact factor of 2.7 since its 2001 launch. But Richter admits that of six Copernicus journals with online discussion, not all have been welcomed by users. Hydrology and Earth System Sciences added open peer review seven years after its launch. "A tribe of very conservative scientists keeps asking why there has to be a discussion feature," says Richter. "They just don't want it."

The editors of a new online journal to be published by BioMed Central think biologists are ready for open peer review. Biology Direct authors have to solicit their own reviews from an editorial board, and the comments appear online for all to see.

"In many areas of biology there's roughly a one-in-three chance one of your reviewers just won't like your point of view," says editor-in-chief David Lipman. If that were to happen to a Biology Direct paper, it would still be published. But anyone could read the naysayer's comment.