How to Cite e-Resources without Stable URLs

It used to be said, especially by the Internet’s nay-sayers, that the insuperable barrier to publishing and citing online is that links are never stable. The number of pages that appear and disappear every day means that even a year-old list of sites is likely to contain significant link rot. There is a significant movement to promote both stable and cool URLs (see for example [van Kesteren 2004] and [Berners-Lee 1998]), and most of those of us who publish online take great pains to have URLs that are both predictable and will not need to change.

For example, we recently published The Inscritions of Roman Tripolitania, a digital edition based very closely on the 1952 printed volume by Reynolds and Ward-Perkins, at:

Because this is largely a reprint, there is obviously more work to be done, and we hope to add a new edition fairly soon (incorporating, for example, Arabic translation and new digital photographs). When we do so, the new site will be labeled “irt2011” or similar, all internal links will include this date, and the old site will not need to be removed or renamed. No links will be broken in this process.

Similarly, good electronic journals have URLs that reflect date and/or issue number in the directory structure:

Again we can see that they don’t need to change when new issues come along or the site is restructured. Additionally, you can guess from these URLs what the address of DM issue 5, or DHQ issue 4.2, would be. Additionally, I can remember (or guess) the URLs of individual papers within DM by their authors’ surnames. I strongly suspect that in a year and in ten years, these URLs will still work.

All of which makes is especially surprising that an institution like the Center for Hellenic Studies, which is in so many ways a field-leader and standard-setter in Digital Humanities matters, has a website whose URLs seem to be generated by a content management system. These URLs (including that of their flagship online journal Classics@, and of the magisterial Homer Multitext) are ungainly, arbitrary, and almost certainly not stable. Even worse, many individual pages within the site have URLs that contain a session-specific hash, and so cannot be cited at all:

One might argue that these pages should be cited as if they were paper publications, and readers are then left to their own devices to track them down, but surely that isn’t good enough? Are there any solutions to citing electronically, and linking to, a page whose URL is likely to be itinerant? A persistent redirect? A Zotero biblio URL that you can update if you notice it’s broken?

This entry was posted in General, Publications, Standards. Bookmark the permalink.

5 Responses to How to Cite e-Resources without Stable URLs

  1. Michael McAllister says:

    Iterasi allows you to take a copy of a page, and share it. That may be useful for links intended to reference a single page rather than an entire site. Of course, you’re dependant on Iterasi not going away.

    Side note, because I’m a pedant, “Resourse” in the title should be “Resource.”

  2. Tom Elliott says:

    As to the Zotero URL idea … I fear that’s just pushing the rot down the pike. It still depends on an arbitrary user periodically checking on the location of a resource and updating it, instead of the issuer taking responsibility for stability and curation.

  3. Michael–thanks for the correction. I’ve fixed in it the title.

    Tom–yes, but if you’re having to cite the URL *in print*, at least it gives you the opportunity to update, which is something you don’t have on paper normally. I agree it doesn’t solve the bigger problem, though.

  4. I don’t think there is an obvious answer to your question. I would put the bar even higher: I would like to be able to cite with a stable URL every part of the text, down to the word level (if it’s a digital edition) or at least the paragpaph for reference works, including revisions that may have taken place. How far away are we from something like this?

  5. WebCite® it’s an attempt to fix this problem developed in the medical field

    It’s an “on-demand archiving system for webreferences (cited webpages and websites, or other kinds of Internet-accessible digital objects), which can be used by authors, editors, and publishers of scholarly papers and books, to ensure that cited webmaterial will remain available to readers in the future”, see

    What’s interesting is that WebCite® – as member of the nternational Internet Preservation Consortium – has made a commitment for long-term preservation of archived webreferences.

    Perhaps it’s not the definitive solution but at least it’s a step towards.

Leave a Reply

Your email address will not be published. Required fields are marked *