8 Comments

Summary:

Many links cited by influential science journals and the Supreme Court are broken – the result is a growing memory hole in the places where scholars expect to find an authoritative source of knowledge. The good news is a solution is at hand.

Imagine a research library where most of the books are missing footnotes — where the bottom of the pages are stained or ripped out, making it impossible for scholars to tell the sources of information. That’s the state of many web pages right now, including those for the Supreme Court and The New England Journal of Medicine and Science.

New York Times’ legal reporter Adam Liptak called attention to the problem in September, citing research that showed half of the links on the Supreme Court site don’t work. This pervasive “link rot,” he noted, means the sources and authorities that form the basis of the Court’s decisions are simply missing.

Link rot is a growing issue for both courts and academic journals, but one that is downplayed on the grounds that books and paper are the “real” authorities while internet sources are ephemeral or, at best, unofficial. As the era of print recedes, however, this anti-digital bias looks more and more untenable.

The good news is that libraries have a plan to fix the problem. This weekend, the Times Higher Education website published a feature that looks at Perma CC, a site that is creating etched-in-stone digital references for scholars and lawyers.

It works like this: a scholar (or anyone else) can submit a link to Perma CC, which is managed by a coalition that includes universities, libraries and the Internet Archive. According to Perma CC, the group will create a permanent URL and store the page on its servers and on mirror sites around the world.

Readers who encounter Perma.cc links can click on them like ordinary URLs. This takes them to the Perma.cc site where they are presented with a page that has links both to the original web source (along with some information, including the date of the Perma.cc link’s creation) and to the archived version stored by Perma.cc.

There is also a process for scholars and librarians to “vest” certain URLs so that they become an official, permanent citation for law and science journals.

This process appears to be a long overdue solution. Here are some more stats cited by the Times Higher Education feature:

  • Link rot at influential science journals rises from 4 percent at three months to 10 percent at 15 months to 13 percent after 27 months.
  • 98.3 percent of web pages change in some way within six months, while 99.1 percent do within a year
  • At three Harvard legal journals, over a 12-year period, 70 percent of the links no longer worked

Image by Aleksey Stemmer via Shutterstock

  1. So this is somewhere between DOIs, which many scientific journals already use, and archive.org?

    Share
  2. Timothy W Murray Monday, October 14, 2013

    @Commenter43
    Legal publications (which are the only ones this article give numbers for) do not typically give DOIs. And need a solution. But legal publishing has it’s own complexity. The same case will be “published” by two or three big publishers, yet standard legal citations do not specify the publisher for cases. This makes sense so you can look the case up in either publishers system, but leaves the links unresolved because there are two moving targets not just one.

    When academic journal link to other academic journals DOIs work well. But academic journals sometimes have links to things that are not journal articles. Perma cc could be of help for this.

    Share
  3. Link rot is a big problem. Permamarks launched before Posterous was shuttered by Twitter to enable a personal archiving solution. http://permamarks.org/permamarks-solves-the-problem-of-link-rot-beautifully/

    Share
  4. The NSA probably has a copy.

    Share
  5. It is true that link rot is pervasive on the net and in all documents systems relying on the web infrastructure. There are a variety of other issues related to digital documents that are particularly critical for legal documents, such as provenance and authority control. However specific to broken links, there is no reason why a DOI could not be used in this case. As a stable and fucntional system for persistent linking between documents, it is disappointing to hear of another alternate to a functioning solution. Furthermore, there is no reason why the DOI system would not solve problems in circumstances where the location of either document is mutable as you note.

    Share
  6. What exactly is wrong with DOI’s to solve this problem?

    Share
  7. TinyVox: Tape&Tweet Tuesday, October 15, 2013

    The word “lacuna” springs to mind. Lacunae. Thanks Jeff,! This is key, assuming, y’know, knowledge helps folks and stuff.

    Share
  8. This is a real problem, and I’ve written about it several times, from several perspectives (I’d apologize for the links that follow, but … nah … )

    http://answerguy.com/2013/10/04/influency-links-supreme-court/ speaks directly to the Supreme Court issue

    http://answerguy.com/2013/04/26/photodropper-and-photodropper-influency/ us about one site linking to another … or not

    and http://answerguy.com/2013/04/22/copyright-influency-website-content-trademarks/ is our position on the matter.

    The short of this (and I’m afraid that with so many parties and so many interpretations of both responsibility and law it’s not at all short) is that if you write something and rely on the continued operations and good archival practices of another party, you’re a fool. On the other hand, it’s a lot of work to format everything you write to include self-archival practice and attribution, AND the “I wonder if this is even legal?” aspect is hard to ignore.

    As I said, our position is in that third link, and it covers exactly what I just pointed out: it’s YOUR writing/web site/whatever, so YOU do a complete job; that’s where the value you bring to the table really comes from.

    Share

Comments have been disabled for this post