constrained writing

We report on Renderings, which focuses on translating highly computational literature into English. This has involved (1) locating literature of this sort that is written in other languages, (2) applying techniques that are typical of literary translation, (3) using programming and other Web development work to port and reimplement older works that are not easily accessed today, and (4) bringing literary and computational thinking together when the interaction of language and computing demand it. All four of these reveal cultural aspects of computational literature, including the one related to typical translation practices. The need to think in literary and computational terms as seen in (4) is particularly interesting, as is the search described in (1). Translators do not usually frame their search for work to translate as part of the translation task, but this is an explicit part of Renderings, which involves culturally specific investigations and considerations of different communities of practice.

The Renderings project began in summer 2014. During its first phase the collaborators were all based in the Trope Tank, the laboratory that Montfort founded and directs, for some of the time. The collaborators met weekly and were joined at four meetings by literary translators Robert Pinsky, Marc Lowenthal, John Cayley, and David Ferry. Seven core Renderings collaborators (Patsy Baudoin, Andrew Campana, Qianxun Chen, Aleksandra Małecka, Piotr Marecki, Nick Montfort, and Erik Stayton) worked on the first phase of the project, which concluded in December 2014 when 13 translations and bilingual works, from six languages (Chinese, French, German, Japanese, Polish, and Spanish), were published in Fordham University’s literary journal Cura.

The project has so far not only dealt with e-lit across languages, but also in different national contexts (Argentine and Spanish language generators, for instance), from different communities of practice and literary movements and groups (including, in French work, the Oulipo and Mutantism), and of different historical eras. The earliest pieces translated were originally BASIC programs that generated Spanish and Polish texts, and were published in magazines for readers to type in and run. Adaptation to the Web was important, because the project aims to give access to today’s English-language readers using the typical Web context.

Members of the group are now actively seeking interesting computational literature in languages other than English, and additional Renderings collaborators are also being sought. The next phase of the project is being done in a more distributed fashion using simple systems for collaboration, including a mailing list and a wiki, which will hopefully allow a broader range of participation while still providing collaborators with common ground for discussion. Initially, Renderings focused on translating small-scale but complex projects; the project is now expanding to new languages and genres, including longer-form work. We anticipate including games, interactive fiction, and bots as the project continues.

Our discussion will address the question of how Renderings offers a new, broader perspective on electronic literature, how our search for computational literature suggests new directions for scholars, editors, and readers, and how our practice of the literary translation of computational works extends current concepts of translation.

(Source: ELO 2015 Conference Catalog)

As chronicled on the “Beard of Bees” website, authors involved with Gnoetry, “an on-going experiment in human/computer collaborative poetry composition”, have collectively engaged with digital textual processing for more than a decade (see http://beardofbees.com/gnoetry.html and also http://gnoetrydaily.wordpress.com/). In 2011, the group published their first anthology, Gnoetry Daily, Vol. 1, a 52 page collection of verse spun with programs named Gnoetry, charNG, Infinite Monkeys, ePoGeeS, welatanschauung, and JanusNode, with accompanying commentary by Eric Elshtain, eRoGK7, Matthew, edde addad, nathanielksmith, and DaveTolkacz.

The software program engineered by the group, Gnoetry, synthesizes language based on its analysis of existing texts, thus mimicking the “statistical properties” of its input texts; users filter language by applying constraints in each of the programs they favor. Concluding the “Methodological Notes” included in Gnoetry Daily, Vol. 1, addad writes, after highlighting capabilities of their preferred programs, “Generally, we just want to write good poetry”. Emphasizing the
contents of Gnoetry Daily, Vol. 1, we analyze successes and failures in this pursuit and discuss how the group’s practice interestingly falls in line with what could be called a “post-TRAVESTY” continuum. TRAVESTY is a text processor constructed in the 1980s by Hugh Kenner and Joseph O’Rourke that had great influence on digital poetry’s development in the United States, by inspiring subsequent important works by Jackson Mac Low, Charles O. Hartman, and others, and initiating dialog between practitioners.

Efforts of those involved with Gnoetry group not only recall the communal rapport initiated by artists who appreciated and worked with TRAVESTY, but also some of the program’s aesthetic and computational agenda as well. In addition to pursuing such topics, we will, using the conference paper as an input, create spontaneous poems that offers us, and the audience, an opportunity to evaluate the program’s qualities in real time—as a way of suggesting the key to interpreting the significance of these alliances and aesthetic directions is to imagine the authorial process as a mode of interactivity.

(Source: Author's abstract, 2012 ELO Conference site)