Libraries combine to preserve vanishing sources online

Source: Getty

Out of sight: websites cited as primary sources can very quickly disappear

Add this to the heretofore staid lexicon of the academic library: “Link rot.”

It is a term that reflects the downside of technology that puts footnotes a click away from any reader: links to primary source material that vanish into cyberspace.

“It’s such a pervasive problem,” said Kim Dulin, co-director of the Harvard Library Innovation Lab - particularly in the law, for which citations are essential to future cases.

“The key to law is to look at the citations, especially in court opinions,” Ms Dulin said. “If it is an internet source and later the lawyer or another court sees no support for the proposition, it diminishes the authenticity of the court’s decision.”

Now an international coalition of law libraries has a plan to make online citations permanent: Perma CC, a site that proposes to host primary source material for ever.

“One of the things we like to say is that libraries have always been in the for ever business,” Ms Dulin said. “In the past we’ve acquired and made available print sources authors and publishers cited, so it seems logical that we also do that for internet sources.”

The consortium so far includes the law libraries at the universities of Harvard, Columbia, Georgetown, Yale, California-Los Angeles, Oxford, Melbourne and about 20 others.

It envisions ultimately expanding into other disciplines to protect footnotes from link rot. Authors and publishers can enter a URL for archiving and it will be protected permanently on Perma CC servers and mirror sites worldwide, the organisers promise.

The problem is as fast-growing as it has been little noticed, researchers say: in 1994, according to a study at John Marshall Law School, there were just four internet citations in law reviews; 10 years later, there were about 100,000. Even the US Supreme Court has cited internet sources 555 times.

But websites are often altered, moved, renamed or deleted. The John Marshall study found that 98.3 per cent of web pages are changed in some way within six months, 99.1 per cent within a year.

Other research has revealed that link rot in The Journal of the American Medical Association, The New England Journal of Medicine and Science - among the most influential journals in their fields - rose from 4 per cent at three months to 10 per cent at 15 months to 13 per cent after months.

Perma CC remains focused on legal citations, which are particularly prone to link rot, Ms Dulin said. A review of three Harvard law journals over a period of 12 years found that 70 per cent of the links did not work, she said. Forty-nine per cent of the internet citations in US Supreme Court decisions have also fallen prey to attrition.

“In the old days we could just look up everything in print,” Ms Dulin said. “In the internet world, if you can’t find it, it casts doubt on the validity of the decision.”