Tweet, Loc. Cit.

The Modern Language Association has now issued its official, authoritative, and precisely calibrated guidelines for citing tweets – a matter left unaddressed in the seventh edition of the MLA Handbook (2009). The blogs have been -- you probably see this one coming -- all a-Twitter. The announcement was unexpected, eliciting comments that range from “this is really exciting to me and i don’t know why” to "holy moly i hate the world read a damn book." (Expressions of an old-school humanistic sensibility are all the more poignant sans punctuation.) Somewhere in between, there’s this: "when academia and the internet collide, i am almost always amused."

Yet the real surprise here is that anyone is surprised. The MLA is getting into the game fairly late. The American Psychological Association has had a format for citing both Twitter and Facebook since 2009. Last summer, the American Medical Association announced its citation style after carefully considering "whether Twitter constituted a standard citable source or was more in the realm of ‘personal communications’ (such as e-mail),” finally deciding that tweets are public discourse rather than private expression.

The AMA Style Insider noted that a standard format for Twitter references should “help avoid citations sounding like a cacophony of Angry Birds.”

How long was the possibility of an MLA citation format been under consideration? Was it a response to MLA members needing and demanding a way to bibliograph tweets, or rather an effort to anticipate future needs? Rosemary Feal, the organization’s executive director, was the obvious person to ask.

"The release of the tweet citation style,” she said by e-mail, “came in response to repeated requests from teachers, students, and scholars (most of them received, perhaps unsurprisingly, over Twitter). We debated the particulars on staff for some weeks. We're certain that the format we've announced is just a first step; user needs will change over time, as will technologies.”

Having exact, authoritatively formulated rules is clearly an urgent, even an anxiety-inducing matter for the MLA’s constituency. “Every time people asked me on Twitter about citing tweets,” Feal said, “I told them MLA style was flexible. Just adapt the format.” And as a matter of fact, the current MLA Handbook does have a format for citing blog entries – which would seem to apply, given that Twitter is a microblog.

“But because people wanted something very specific,” Feal said, “I asked staff to think about it…. Our hope is to remain nimble enough to respond to circumstances as they develop.” In that case, it might be time to start brainstorming how to cite Facebook exchanges, which can certainly be recondite enough, if the right people are involved. At least the Twitter citation format will be part of the eighth edition of the MLA Handbook -- though Feal indicated it would take at least another year to finish it.

Directing scholarly attention to the incessant flow of 140-character Twitter texts can yield far more substantial results than you might imagine, as explained in this column almost two years ago. Often this involves gathering tweets by the thousands and squeezing them hard, via software, to extract raw data, like so much juice from a vat of grapes. Add the yeast of statistical methodology, and it then ferments into the fine wine of an analogy that’s already gone on far too long.

So let’s try that again. Social scientists have ways of charting trends and finding correlations in tweets en masse. Fair enough. But recent work by Kaitlin L. Costello and Jason Priem points in a different direction: towards Twitter’s role in the more narrowly channeled and discussions taking pace within scholarly networks.

Costello and Priem, who are graduate students in the information and library science at the University of North Carolina at Chapel Hill, have been gathering and analyzing information about academics who tweet. Their findings suggest that Twitter has become a distinct and useful -- if exceedingly concentrated -- mode of serious intellectual exchange.

In one study, they examined the departmental web pages at five universities in the United States and Britain, compiling “a list of all the scholars (defined as full-time faculty, postdocs, and doctoral students) at each one, yielding a sample of 8,826.” Through a process of elimination, they were able to generate a pool of 230 scholars with active Twitter accounts. Out of the initial pool, then, they found one scholar in 40 using Twitter – not a lot, although it’s definitely an underestimation. Some in the pool were removed because Costello and Priem could not establish a link between faculty listing and Twitter profile beyond any doubt. (In the case of people with extremely common names, they didn’t even try.)

The most striking finding is that the scholars who used Twitter were almost indistinguishable from those who didn’t. Status as faculty or nonfaculty made no difference. Natural scientists, social scientists, and humanists were represented among the Twitterati at rates nearly identical to their share of the non-tweeting academic population. Scholars in the formal sciences (math, logic, comp sci, etc.) proved less likely to use Twitter than their colleagues – though only slightly.

A large majority of tweets by academics, about 60 percent, were of a non-scholarly nature. A given tweet by a faculty member was about twice as likely to have some scholarly relevance than one by a nonfaculty person. While the share of traffic devoted to strictly scholarly matters is not enormous, its importance shouldn’t be underestimated – especially since a significant portion of it involves the exchange of links to new publications.

In an earlier study (archived here) Costello and Priem conducted interviews with 28 scholars – seven scientists, seven humanists, and 14 social scientists – as well as harvesting more than 46,000 of their tweets. For each subject, they created a set of the 100 most recent tweets containing links that were still active. (A few didn’t reach the 100 mark, but their data was still useful.)

Six percent of the tweets containing hyperlinks fell into the category of what Priem and Costello call “Twitter citations” of peer-reviewed scholarly articles available online. One of their subjects compared linking to a scholarly article via Twitters to citing it in a classroom or seminar setting: “It’s about pointing people in the direction of things they would find interesting, rather than using it as evidence for something.”

At the same time, tweeting plays a role in disseminating new work in particular: 39 percent of the links were to articles less than a week old -- with 15 percent being to things published the same day.

The researchers divided citation tweets evenly into two categories of roughly equal sizes: direct links to an article, and links to blog entries or other intermediary pages that discussing an article (usually with a link to it). Not surprisingly, 56 percent of direct links lead to open-access sources. About three-quarters of the indirect links went to material behind a paywall. “As long as intermediary webpages provide even an abstract-level description,” write C&P, "our participants often viewed them as equivalent.”

One scholar told them: “I don’t have time to look at everything. But I trust [the people I follow] and they trust me to contribute to the conversation of what to pay attention to. So yes, Twitter definitely helps filter the literature.” Another said, “It’s like I have a stream of lit review going.”

At this level, Twitter, or rather its users, create a quasi-public arena for the distribution of scholarship – and, to some degree, even for its evaluation. Costello and Priem suggest that harvesting and analyzing these citations could yield “faster, broader, and more nuanced metrics of scholarly communication to supplement traditional citation analysis,” as well as strengthening “real-time article recommendation engines.”

At the MLA convention in January 2011, Amanda French gave a talk that summed up, in its title, a major implication of Priem and Costello’s work: “Your Twitter Followers and Facebook Friends Won’t Read Your Peer-Reviewed Article If They Have to Pay For It, and Neither Will Strangers.” This is true. And its obvious corollary – that open-access and scholarly tweeting can magnify an article’s impact considerably – is demonstrated by Melissa Terras, the co-director of the Center for Digital Humanities at the University College London.

On October 16, she made one of her papers available through the UCL online repository. Two people downloaded it. She tweeted and blogged about it on a Friday, whereupon it was downloaded 140 times in short order, then re-tweeted it on Monday, with the same effect. “I have no idea what happened on the 24th October,” she writes. “Someone must have linked to it? Posted it on a blog? Then there were a further 80 downloads. Then the traditional long tail, then it all goes quiet.”

In all, more than 800 people added the article to their to-read collections in a couple of months – which, for a two-year old paper called "Digital Curiosities: Resource Creation Via Amateur Digitisation," from the journal Literary and Linguistic Computing, is not bad at all.

That may be another reason why citation formats for Twitter are necessary. One day, and it might be soon, an intellectual historian narrating the development of a theory or argument may have to discuss someone’s extremely influential tweet. Stranger things have happened.