Virtually Invisible

The arrival of the Internet ushered in a rapid expansion of library holdings everywhere. Readers suddenly had access not only to what rested on the shelves, but also to countless books and journals from all over the world. It seemed only a matter of time before scholars took this abundance of resources and translated it into broader and more innovative research. But if you’re waiting for that day, don’t hold your breath, advises James A. Evans, a sociologist at the University of Chicago.

Unlikely as it may seem, Evans’s study of more than 30 million articles found that as journals go online, researchers actually see less of their contents. For every additional year of archives a journal makes electronically available at no charge, the number of distinct articles cited in other journals falls by 14 percent on average. Moreover, the articles that are cited tend to be more recent. In other words, if a journal puts more of its older issues online, the effect will be that the newer articles receive more citations—perhaps because scholars are less likely to thumb through the shelved volumes when a journal’s online archive is extensive. For every additional year a journal’s online archive goes back, citations to that journal will reference articles that are, on average, 10 months more recent.

Researchers looking at a journal online may type in a search term or two and find just the article they had in mind. But what they won’t find are the older articles whose content, though perhaps not directly related, complements their research in surprising ways. Scholars have typically unearthed those little gems by manually flipping through the older issues on the bookshelf. Evans writes that his study “ironically intimates that one of the chief values of print library research is poor indexing.” Researching online may be more efficient, but it narrows the window scholars look through.

In a separate study, researchers at Cornell University examined what happens when a journal article is available for free to the public compared to when a subscription is necessary to view it. Philip M. Davis and his team found that providing unfettered access to an article does not increase the quantity of citations it receives. However, they studied not just how often an article was cited but how often it was read, and articles that are available for free are read much more frequently than those requiring a subscription.

As citations converge on newer and fewer articles, scholarly consensus emerges much faster. But, Evans warns, the haste may prove costly. Articles and ideas that don’t become part of the consensus will soon be lost in the never-ending flood of research.

* * *

THE SOURCES: “Electronic Publication and the Narrowing of Science and Scholarship” by James A. Evans, in Science, July 18, 2008.

“Open Access Publishing, Article Downloads, and Citations: Randomised Controlled Trial” by Philip M. Davis, Bruce V. Lewenstein, Daniel H. Simon, James G. Booth, and Mathew J. L. Connolly, in BMJ, Aug. 9, 2008.

Photo courtesy of Flickr/CGP Grey