Media Theory for the 21st Century

March 4, 2008

Database and Poetry, Pt. 2

Filed under: Discussion — jeremy @ 9:16 pm
Tags: ,

Part 1 is here.

What modern and contemporary poetries crave in the absence of narrative and traditional metrical forms is structure. That structure can be minimal, procedural, or constraint-based, but it’s always-sorta-kinda there even if primarily as an absence. One viable way of reading the history of poetics from the late-19th century through today is as a, well, narrative of progressive dissolution or atomization. Poets have shattered text and language into smaller and smaller pieces. This is a trend Adalaide Morris touches on when in her introduction to New Media Poetics she discusses “literal art” or those poems that “feature not the stanza, the line, the phrase, or the resonant word but… letters” (20). Think of Kittler’s reading of Christian Morgenstern in Discourse Networks. Of Lewis Carroll and Edward Lear. Of Futurists fonts. Of Mallarmé’s words-in-freedom. Of the stubby lines of HD, WCW, and George Oppen. Of Louis Zukofsky. Of E. E. Cummings’s typographical games.

The central problem confronted by contemporary poets then becomes a very basic one: How do we build or rebuild poems of any size and substance from the tiniest bits of language? To repeat, borrowing from Espen Aarseth and Brian Kim Stefans as well as Gertrude Stein and Morris: How do we build “non-trivial” poems that can help us know (intellectually) what we already know (in our bodies and habits)? More specifically, how do we do so without relying on the crutch of (submerged) narrative as so many of the medium-length Imagist poems did and so much contemporary thin “free verse” does?

One solution is digital animation or manipulation in the vein of The Dreamlife of Letters or Univers Revolved. Another is database. Database provides a way to structure a long poem without resorting to the internal organization scheme of narrative. At one end of the spectrum, Susan Wheeler can explore, in “The Debtor in the Convex Mirror” (a relatively traditional poem that does incorporate fragments of narrative), the obsessive keeping-track-of mentality of our collector culture. She does this through an interplay between her own words, quotations, a reading of Ashbery, a reading of a Dutch painting, and a covertly organized list of source texts at the back of the aptly named collection Ledger (2006). Or she can drop a bunch of worthless HTML code at the back of Source Codes (2001)—a strangely poignant admission that she’s thinking about stuff she does not understand at the technical level.

Meanwhile, Lyn Hejinian can simply use dates (specifically those from October 6, 1986 to January 21, 1989) to structure a series of one- and two-page Language poems about the body and writing in The Cell (1992). In 2008, her sequence looks like a series of bizarre blog entries. It’s unity is achieved less through tonal consistency or unified subjectivity than through the simple meta-data act of dating: the word “data” coming, like the word “date,” from a form of the Latin dare, “to give,” as in data (epistola) for “(letter) given or delivered.”

At the other end of the spectrum, conceptual poets like Kenneth Goldsmith and Christian Bök can select massive databases and devise onerous constraints/algorithms. The data source can be Goldsmith’s daily activities as in Fidget (2000); all the words from a single copy of the New York Times as in Day (2003); a transcription of a year’s worth of radio weather reports as in Weather (2005); a list of phrases ending in schwa as in his No. 111 2.7.93-10.20.96 (1997); or all the mono-vocalic words in English as in Bök’s Eunoia (2001). In these Oulipo-on-steroids writers, the data approach is invested with an X-Games mentality. The goal is, at least in part, to have the audience gape in awe. Oh, shit! He alphabetized what?!??

These various styles of datapoetics, then, raise a number of questions about authorship, gender, formalism, extremity, data vs. database and the like. I’ve gone on far too long already, but one central concern worth emphasizing is the transformation of the trivial or “banal” (Sianne Ngai’s word) into the “non-trivial” by way of database. Whether it’s chatroom jabber, ads from the NYT, bizarre word lists, or daily experience, the algorithmic transformation of data seems to link the poetics of Listening Post to at least a few camps of book-bound poets. Together, these practitioners are doing for “data” and “database” what Galloway and Thacker attempt to do for “network.”

–Jeremy Schmidt

Database and Poetry, Pt. 1

Filed under: Discussion — jeremy @ 8:53 pm

Galloway and Thacker’s critique of network discourses (to reverse Kittler) in The Exploit is one piece of a larger movement to bring the knowledge of networks found in our bodies and habits up to our heads. What would a poetics look like that participates in this same movement?

Here I’m thinking in part of Rita’s discussion of Listening Post as “literary” and specifically as expressing a “poetics.” Spoken word and Whitman were mentioned in relation to the tone and style of the installation’s output. To that list we might add confessional poetry. But what’s interesting about Mark Hansen’s piece is largely, I’d argue, the ways it subverts the apparent immediacy of these earlier styles and foregrounds performativity via rigorously mechanical data manipulation. In the following, I hint at a literary-historical context for what might be termed “datapoetics.” I also mention some strains of recent datapoetics that are bound up with what we find in Listening Post.

If, as Kate argues, database and narrative have a “tempestuous” but ultimately symbiotic relationship (Lindsay’s Bert and Ernie), then poetry and narrative might be thought of as doddering divorcées who occasionally share a cocktail. They have a long history, one they relive via the occasional fling, but they don’t depend on or need each other the way database and narrative seem to.

The tired tale of the breakup between poetry on the one hand and narrative, the lyric “I,” and referentiality on the other has been recapitulated many times. (See, for example, Ron Silliman’s polemical essay collection The New Sentence.) But what about poetry and database?

In one sense, their relationship is the opposite of that between database and narrative. If narrative seems, at least on the surface, to fear and evade databasing, then poetry craves it. Of course, Kate picks up on Lev Manovich’s use of the word “natural” (from “natural enemies”) to deepen the relationship between narrative and database, punning to bring in the biological metaphor of symbiosis, and argues that the tension between the two is productive rather than hostile. This pun helps her get at Manovich’s strangely ambiguous use of the word “structured.” He defines database as “a structured collection of data” (40) but later undermines this computer science definition by arguing that “database represents the world as a list of items, and it refuses to order that list” (44). This is the blurred distinction between data and database that we discussed in class.

A key issue at stake here seems to be that of internal ordering/structuring versus external ordering/structuring. Narrative relies on relationships between elements that are internal and temporal: a character does one thing, which leads to another. Databases, I think, rely on relationships that tend to be external, spatial, or abstract: alphabetization, numerical ordering, tree-like structures, etc. This is true (again, I think) of the various types of computer databases (hierarchical, relational, network, object-oriented), but I’d love to here from someone with more technical knowledge.

Of course, the internal/external binary is problematic and might be more usefully thought of as a continuum a la Crary. Even the basic question I posed in class—“Does alphabetizing a group of items transform it into a database?”—was met with some confusion. Kate said “no,” Rita “yes.” [CORRECTION: Apparently I goofed and reversed these responses, or else they both said “no.” This uncertainty serves to emphasize my point nicely.] In any case, the relationships expressed in the organization of databases tend to be syntactical or logical rather than semantic or referential. This is key.

Please see Part 2 for a discussion of specific poets and poems.

–Jeremy Schmidt

Blog at