Science magazine, an international weekly journal, published a letter in early January 2009 coauthored by Christopher von Bartheld, professor in physiology and cell biology at the University. The letter was a challenge to an article published in the July 2008 issue of Science by James A. Evans, called “Electronic Publication and the Narrowing of Science and Scholarship.”

Over the past 15 years, according to von Bartheld, electronic publishing has changed how scientists search and retrieve relevant information, with a shift from browsing for articles in print to searching for them online. Evans’ article claimed that when journals became easily accessible online, citations in scientist’s articles decreased and as a result, the basis of science is being further narrowed rather than expanded.

When von Bartheld first saw Evans’ article the topic piqued his interest, because he is heavily involved in publishing both as a writer and as a reviewer of scientific articles. Von Bartheld authored nearly 100 journal articles and reviews manuscripts for more than 50 different biomedical journals.

“I thought that Evans’ primary data on citation patterns were convincing, but I had serious reservations about some of the conclusions,” von Bartheld said.

Evans examined how a journal’s availability online changed the citation patterns of science articles. After an analysis of his data, Evans concluded that scientists began to include fewer citations for other articles in their own research with the advent of electronic journals. He also found that when scientists did cite, they did not cite as wide a range of books, journals and studies as they previously did.

One of von Bartheld’s major criticisms of the article was that Evans did not consider that a focus on fewer, but more relevant citations may be a positive rather than a detrimental development. Another critique von Bartheld found with the article was Evans’ conclusion that scientists who cite more narrowly follow prevailing opinion.

“He interpreted that as the basis of science becoming more narrow,” von Bartheld said. “He also jumped, I think, to the conclusion that if you cite more narrowly, that you would follow … the prevailing opinion.”

Von Bartheld found this conclusion suspicious and didn’t think that it followed from the data that Evans presented. To write a response, von Bartheld teamed up with fellow scientists Shaun P. Collin from the University of Queensland in Australia, and Onur Güntürkün from Ruhr University in Germany.

“I communicated with several of my colleagues about this article and they agreed that there were some flaws,” von Bartheld said. “We decided that we should make our viewpoint known.”

The letter, headlined “To each citation, a purpose,” pointed out that citation databases do not specify what each citation is for. A citation could be used for to refute a claim, as evidence for a claim, or for technical purposes.

“We felt that it is important to consider what is in the mind of the scientist who actually writes the paper,” von Bartheld said. “We scientists read a lot more papers than we actually cite, so how do we make the decision whether a citation should be included or should not be included.”

But why would this behavior change with electronic access to scientific articles? Citations can be classified on the basis of their purposes, but such data are difficult to obtain and not available in current citation databases.

“This approach [classification of citations] would help to resolve whether narrowly citing authors are indeed more likely to follow prevailing opinion,” von Bartheld said. “Such an approach – lacking in Evans’ article – would eventually either support or refute Evans’ claim of a narrowing of science.”

By publishing his letter in the Science magazine, von Bartheld is taking part in the long and fruitful tradition of scientific debates.