Britannica objects to allegations of inaccuracy by science journal
Encyclopedia Britannica, stung by a December finding in the science journal Nature that the venerable reference work was only slightly more accurate than the online encyclopedia Wikipedia, has released a paper denouncing the Nature study as shoddy, misleading, and wrong in details and conclusions. Britannica called on Nature to retract its article.
''Dozens of inaccuracies attributed to the Britannica were not inaccuracies at all," said the paper, which was posted on Britannica.com and e-mailed to 5,000 librarians and educators this week, ''and a number of the articles Nature examined were not even in the Encyclopedia. The study was so poorly carried out and its findings so error-laden that it was completely without merit."
In a short statement on Nature.com, the London-based journal stood by its findings and refused to retract the article. ''We reject those accusations," the statement said, ''and are confident our comparison was fair."
Encyclopedia Britannica, founded in 1768 in Scotland, is now an independent US publisher based in Chicago. While its multivolume printed book version is still available, the company has made its 65,000 articles available online to subscribers, in response to the growing importance of the Internet.
Wikipedia, founded by Jimmy Wales in 2001, is the fast-growing free online reference with more than 1 million articles in English, written by volunteers and open to editing by anyone. A wave of controversy broke over the project in November when it was discovered that someone had rewritten Wikipedia's article on Tennessee publisher John Seigenthaler Sr., alleging falsely that he had been implicated in the assassinations of President John F. Kennedy and Senator Robert F. Kennedy.
In its article, which has been widely quoted and reported, Nature said it had provided 50 pairs of scientific articles from Wikipedia and Encyclopedia Britannica to anonymous qualified scientific reviewers, without revealing the origins of the articles, and asked for assessment of accuracy. The reviewers found a total of 162 ''factual errors, omissions, and misleading statements" in Wikipedia and 123 in Britannica. The implication was that Wikipedia, written by volunteers, was nearly as accurate as the reference written by experts.
While Wikipedia's article on Encyclopedia Britannica cites the Nature article, Wales himself has never endorsed its findings or claimed his project compares in accuracy.
In its paper, Britannica complains that, along with articles from its main reference, Nature sent reviewers articles from the Britannica Student Encyclopedia and from a 1998 edition of Britannica's annual Book of the Year. In its response, Nature acknowledges using those other publications but defends such use by saying all the articles were on Britannica's website.
''They're using doubletalk to justify comparing the children's Britannica and an eight-year-old yearbook with an adult encyclopedia," Britannica executive editor Theodore Pappas said by phone yesterday. ''I think this study has been a nuisance because it has been so widely publicized and is so fundamentally wrong."
The Britannica paper also alleged that Nature stitched together passages from different Britannica articles, with transitional text written by Nature editors. It alleged that in one instance where reviewers criticized a serious omission in a Britannica article on lipids, Nature had only provided a 350-word introduction to the 6,000-word article. Finally, Britannica complained that one section of text criticized by a reviewer was ''not from any Britannica publication."
A call to the London office of Nature was not returned yesterday. In its written statement, the journal acknowledged and defended using excerpts from articles, and did not dispute stitching together sections from different articles, but denied reviewing any material that was not from a Britannica publication.
Meanwhile, the Wikipedia article on Encyclopedia Britannica has already been rewritten to include this week's controversy.
David Mehegan can be reached at mehegan@globe.com. 
