As of the seventh edition of the MLA Handbook, “the MLA no longer recommends the inclusion of URLs in the works-cited-list entries for Web publications.”
The logic behind this is that URLs are often long and unwieldy (ironically, the worst offenders in this department are often major bibliographic indexes and article repositories, not to mention the mile-longers offered up by Amazon.com), and that retyping URLs is not nearly as efficient as simply Googling for the author and title of a work. (On this second point, I agree; but as I suggest below, omitting URLs is not the solution.)
There are two significant flaws with this logic.
The first is that the MLA seems to be imagining a web governed by restrictive copyright, one that is free from Creative Commons and other forms of permissive content licensing. Only in a locked-down copyright scenario does the assumption hold that a given document on the web would have one single, solitary location–not to mention one single, solitary version. In reality, neither is true: not only because of Creative Commons licensing (which leading web-available scholarly journals such as Kairos and First Monday openly encourage their authors to use), but also because of RSS feeds, JSON, and other modes of content syndication that make one piece of content easily published in multiple locations. Add to this sites like the World Wide Consortium that explicitly and meticulously version all of their pages, and the “Just Google it!” idea behind omitting URLs quickly reveals its limits.
The second flaw in the MLA’s logic is that it misses the entire point of URLs (or more generally URI: Uniform Resource Identifier) on the web. URLs are the only means for reliably referring to a particular item on the web and other parts of the internet. URLs are also the only means for accessing pages stored in the Internet Archive’s Wayback Machine, which is an essential scholarly tool. I cannot say how many times I have read articles, particularly in the fields of computers and writing and digital humanities, that cited now-defunct websites and web pages, only to find them via Wayback. Without a URL/URI in those articles’ lists of references, this type of crucial digital archive reading would be impossible. And while one could argue that Wayback needs a text search, it would reveal the same problems of uniqueness in referring to live pages.
What I find peculiar and frustrating with citation style guides generally, not just MLA’s, is the extraordinary care that is expected of writers who cite books. The intricacies of citing books within series, or that are publisher imprints, have always left me puzzled and uncertain. Needlessly so (at least for books less than 40 years old), I believe, when there is an internationally recognized standard numbering system for books, ISBN (and its counterpart ISSN for periodicals).
The URL/URI is to the web what the ISBN and ISSN are to books and periodicals. They are absolutely, unambiguously unique both in form and to the resource they identify; only one web page or other web-available artifact can have the address http://www.mla.org/style (for example). That is what makes URLs useful: they are inherently unique and unambiguous (that is why the also find their place as in XML Namespaces). If URLs were not unique, the web would cease to function reliably.
Which brings me back to ISBNs: these have been around since the mid-1960s and are a true international standard. And I have often wondered why they are not used in citation, but instead primarily for commerce. ISBNs, and to a lesser extent ISSNs, could eliminate much of the ambiguity and needless information in book (and periodical) citations. For older or non-mainstream-published books, the additional publication information could be used instead.
Ultimately, though, the MLA’s decision to make URLs optional is a blow to digital information literacy. Not just for researchers writing works cited pages (and their eventual readers), but for those who make and maintain websites of scholarly value. There is no good reason for URLs to be anything less than memorable and shareable in email, on Twitter, and yes, in lists of references.
Sites like tinyurl demonstrate that simple, short, unique URLs are possible (thought: we should have a scholar’s tinyurl with guaranteed link permanence). And projects like WordPress and even the favorite target of some scholars’ ire Wikipedia show that so are simple, unique and human-readable URLs.
Modifications to citation styles should, as the MLA has attempted, make it easier to list sources of information and, more importantly, to help readers to find them. But ease cannot come at the expense of accuracy and removing ambiguity from a citation, which is exactly what happens when the URL is omitted from citations.
10 Responses to “Respect the URL in Reference Lists”
Amen brother!
Good points, all Karl. It seems to me that the MLA is particularly unable to take into account experiences outside of a very traditional view of textual production and circulation — one that doesn’t take into account even the basic practices of composing on word processors (remember the angle brackets around URLs?). I don’t teach MLA style in any of my writing classes any more because it has limited use outside of literature courses and I find it to be increasingly less relevant as a useful scholarly resource.
Doug: The problem is that MLA style, and here I’m thinking about the Handbook in particular, is taught in just about every middle- and high-school English class. So even if you’re not teaching it, chances are students will have encountered it long before they walk through the doors of your classroom.
The angle-brackets issue is, believe it or not, originally part of an RFC on URL syntax: http://www.ietf.org/rfc/rfc2396.txt Remarkably, key web inventor Tim Berners-Lee is the first author of it. While I think it’s a terrible practice, the MLA adopted it on good (if short-sighted) authority.
The other kind of horrible thing about this is that clearly the MLA is not imagining a world in which books and articles and papers are published online. Books need to be networked objects! A citation basically *is* a link, when you think about it — it’s a pointer to another work, and it should be the job of a bibliography to make citation checking as easy as possible. Books need to become networked objects.
I was speaking recently with someone at a University Press who was interested in how they could get all their books’ bibliographies to link together, and it seemed to me that the only way was to require that authors put a URL for *all* cited works, even if they had consulted the print. For instance, for books, you could put the link to the WorldCat record.
I was wishing, in fact, that there was an online ISBN registry somewhere, so that authors could put in an ISBN and have that link to a stable record of that particular edition. Amazon is about the closest thing we have to that, but come on — people shouldn’t have to link to a commercial site.
So, on the angle brackets, rather than accepting the feedback from specialists in their own discipline, who have expertise in actual writing practices (and, in particular, expertise in digital writing practices), MLA sought some other authority without regard to the consequences. It seems like a similar approach is being taken here. So I’m not inclined to give them a pass on either issue.
However, I do take your point about MLA style as taught in 6-12 (or even earlier). I don’t mind that so much, because then students understand the mechanics of citation styles — and when they get to my class, we can look at the rhetorics of citation/reference and critically evaluate both the uses and contexts that drive our decisions about which styles to use (and how to modify them if necessary).
I’m of two minds on this: sometimes the URL is a unique identifier; sometimes it is only valuable to people at my own institution–outsiders will never be able to use it to find the same item. If the world was open source and Creative Commons, then URLs would more frequently be useful. Alas, the world is [STILL!!] proprietary.
ISBNs and ISSNs are supposed to be unique, and in most cases are, but there is no mechanism in place to ensure that they are.
As a cataloger, I can assure you that ISBNs are used again, for varying reasons. To presume that an ISBN is unique is to make a mistake. To make the following argument is to fully misunderstand the sociotechnical environments in which these technologies function and the differences between them:
“The URL/URI is to the web what the ISBN and ISSN are to books and periodicals. They are absolutely, unambiguously unique both in form and to the resource they identify; only one web page or other web-available artifact can have the address http://www.mla.org/style (for example).”
As to the overall point of your post, I fully agree.
“(thought: we should have a scholar’s tinyurl with guaranteed link permanence)”
Maybe DOIs or PURLs?
http://www.doi.org/
http://purl.org/
As you seemed to acknowledge, a given URL, while unique, does not map to a fixed target. Even if you specify the time a URL was retrieved, you usually can’t reliably determine what was viewed at any point in the future. Most of the web doesn’t give you a timestamped version history like Wikipedia, things disappear and change without notice, and most of the useful web involves ongoing comments and dialogue at a given page. Forget broken URLs and proxies and that bag of problems, the static monograph is going extinct. If I was an instructor, I’d require inclusion of an archive of all digital resources and data analyzed with any paper, if I actually cared what people are doing. For now, I don’t see an alternative (and if you think you or your student will get prosecuted for copyright infringement for that, please just leave the profession now). If someone ever gets to the point where they will publish something that requires some citation format then they can learn/apply it at that point, and if that means re-doing much of anything they are not using a reasonable research system, but otherwise give me something that actually enables me to view what is cited, not ducks in a row. And as far as scholarly publishing goes, I’ll start respecting it when it is exclusively digital, breaks from journals that are destroying the ability of the library to evolve, includes attached data sets for alternative analysis, provides enough info for replication, etc. At that point, the MLA style will be irrelevant, whether or not there are unique pointers to snapshots of whatever is being analyzed, however peer review is implemented, etc.
[...] Karl Stolley, apparently as disturbed as I am, quickly responded with some insight: The logic behind this is that URLs are often long and unwieldy (ironically, the worst offenders in this department are often major bibliographic indexes and article repositories, not to mention the mile-longers offered up by Amazon.com), and that retyping URLs is not nearly as efficient as simply Googling for the author and title of a work. [...]
Add a Response to This Post