Next Generation Literary Machines: The “Dynamic Network Aesthetic” of Contemporary Poetry Generators

Abstract (in English): 

This dissertation investigates the current state of digital poetry generators. The study argues that today’s poetry programs embody a new stage distinct from second- generation digital literature. The new stage projects a “dynamic network aesthetic,” reflective of four trends in Web 2.0 cultural production: the potential for mass aggregation; the adoption of participatory platforms; the instability of textual material; and the unpredictability of individual actors. The chapters interrogate the ways in which the four characteristics emerge in three categories of poetry generators: programs that manipulate user input; programs that manipulate Internet text; and programs that generate autonomous content. Corresponding with the three categories, three theorists of the machine inform the discussion. Donna Haraway, Jean Baudrillard, and Gilles Deleuze each provide apt approaches to twenty-first-century technology. However, each theorist also demonstrates the ways in which recent models of the machine do not yet fully account for the “dynamic network aesthetic” of the Internet era. In addition to the theoretical discussion, each chapter addresses the design and output of eight poetry generators. The readings suggest that poetry generators, like contemporary technologies in general, advance towards increasingly autonomous cognitive abilities. Emphasizing the cultural value and literary merit of poetry generators, the discussion aims to validate poetry generators’ place in literary studies, in addition to demonstrating a strong connection between contemporary cultural trends and current poetic practice. The dissertation concludes by imagining a potential field of study dedicated to poetry generators. The conclusion calls for further investigations of a “dynamic network aesthetic,” and foregrounds the need to consider future literary machines.

Critical writing referenced:

The permanent URL of this page: 
Record posted by: 
Scott Rettberg