Digital Digest: Meta-Analysis Goes Mainstream

Adrian Versteegh
From the May/June 2012 issue of
Poets & Writers Magazine

No book is a closed system. Pull up Small Demons (, a website launched last November that catalogues the people, places, and things referenced in literature, and choose a work from the virtual shelves—say, Julian Barnes’s Flaubert’s Parrot (Knopf, 1985), which announces its intertextual investment right in the title. Scroll down past the gallery of individuals mentioned (each of whom gets a dedicated page chronicling walk-ons and shout-outs in other books), past the map of places where the story unfolds (each location can be selected to disclose other literary action set there), and note finally that a Peugeot makes a passing appearance in Barnes’s novel. A click through to the entry for the automobile brand reveals that it also crops up in Joan Didion’s A Book of Common Prayer (Simon & Schuster, 1977), which in turn cites, among other things, Moby-Dick (Harper & Brothers, 1851). From Flaubert to Melville in three clicks, by way of a French car manufacturer.

While books have never been the hermetic, linearly consumed media we sometimes imagine them to be, the proliferation of digital literary tools is increasingly underscoring that fact.

Intended as “a concordance of the world’s literature,” as founder Valla Vakili wrote in Publishing Perspectives last fall, Small Demons is only the latest, spiffiest example of the kind of meta-analytical approach that is becoming easier for readers and critics to adopt as digital technology evolves. The user-propagated site (at present, visitors are encouraged to contribute references and feedback via an online contact form, but more advanced tools are in development), whose content is managed by publishing innovator Richard Nash, has echoes in the academy. Since the 1960s, when work in what’s often referred to as the “digital humanities” began, scholars have been taking advantage of computers to amass greater and greater bodies of text (called corpora) and using software to locate linguistic patterns that can help shine a light on the contextual world of a literary work. While books have never been the hermetic, linearly consumed media we sometimes imagine them to be, the proliferation of digital literary tools is increasingly underscoring that fact. And as these sorts of applications make their way into consumer reading platforms, the connections among books are playing a more prominent role at the site of reading itself.

Take the X-Ray feature of Amazon’s e-readers, introduced for the Kindle Touch last fall. By clicking on selected phrases, concepts, characters, and places, readers have access to a continually expanding, cross-referenced index that tracks the occurrence of key terms across the entirety of the book. Content for X-Ray entries is pulled from Wikipedia and Shelfari, a “community-powered encyclopedia for book lovers” acquired by Amazon in 2008. As the number of Kindle titles that come packaged with X-Ray metadata grows, this intermeshing is changing the way users read, so that the question becomes whether—and how—books are changing in tandem.

The relative timidity of new publishing approaches accounts for how underwhelming most mainstream experiments in multimedia book production have been, with publishers struggling to make their “enhanced” offerings feel like more than a bundle of bells and whistles. But what’s distracting in one genre has been a boon to another, and it’s no surprise that digital literature has found its greatest success so far in reference works or, less common, in enhanced versions of familiar print classics. A sign of how complete the shift here has been came in March, when Encyclopaedia Britannica announced that it would cease production of the print editions of the encyclopedia altogether. (Britannica, incidentally, gets a mention in Flaubert’s Parrot.) For most readers, it seems, the style of attention fostered by the web—flitting from one stimulus to the next, careening through unrelated topics, and everything implied in the term “browsing”—is better suited to a stroll through the reference section than an evening curled up with a novel. And in its still-fledgling state (retail and social-network integration are expected soon), Small Demons can feel like a prettified index.

Some readers may find this reference framework uninspiring, reductive, or even stifling. At an event introducing the Kindle Touch last September, Amazon CEO Jeff Bezos announced that X-Ray “lets you look at the bones of the book.” That’s fine for readers who accept that the “bones” of a work are the bits that happen to correspond to Wikipedia entries, or the elements that, according to Amazon’s algorithms, have been determined beforehand to warrant attention. “When you download [a Kindle edition],” Bezos said, “we’ve calculated—we’ve pre-calculated—all the interesting phrases, including historical figures and so on.”

In the case of Small Demons, those bones are a bit more stylishly articulated. Vakili says the website gleans an average of two hundred fifty significant references from each title that is added to its database, and each of those references is highlighted in the text and provided a brief identifying blurb. But with this more controlled curating process (and streamlined user-contribution features still in development), only a selection of each book’s myriad reference points is currently displayed on the website.

But the database can be revealing in places where it falls short. To return to Flaubert’s Parrot, for instance, as of this writing the listing in Small Demons that records the appearance of “Sade” in the novel points not, as expected, to the French marquis, but rather to the modern-day recording artist. An error in the strictest sense, yes (and one that users of the site will undoubtedly set straight), but also, by another token, a moment of serendipity. By leveraging technology to model the process of discovery, tools like Small Demons are introducing a weightier media issue that carries particular import for writers. The question becomes not simply how we can curate an endless barrage of information, but also how reliable this connection building can be when it comes to managing overload, and what type of meaning these seemingly infinite webs of knowledge can offer the literary experience itself.

Adrian Versteegh is a journalist and PhD candidate at New York University, where he runs Listen to Academia, a project that explores the relation between writing and sound. He lives in Paris and New York City.