Detour, Digress, Destroy (on random walks): Demoing a Platform for Text Manipulation
Proposal Type
Individual Talk
Location
Hypertexts & Fictions
Start Date
July 2026
End Date
July 2026
Abstract
Detour, Digress, Destroy (on random walks)^1: Demoing a Platform for Text Manipulation
Literary works authored with some recourse to algorithm are especially ripe for computational recreation and expansion. Authors affiliated with, or working under the influence of, the mostly French Oulipo group or the L=A=N=G=U=A=G=E poets of North America, use algorithmic routines to create or alter existing literary works, inventing new forms or concepts within the texts that embody them. This talk will present ongoing work to build an interactive literary platform for applying selected statistical transformations to selected texts. These transformations will range from relatively simple (or deceptively simple) “mechanical” routines, such as the selective transfer of words used by Jackson Mac Low to fill his own poem “Call Me Ishmael” with words from Moby-Dick^2, to more complex textual re-configurations using the increasingly sophisticated natural language models (word vector relationships, but not AI) that may transform us in turn. The Oulipian “N+7 Machine” which replaces nouns with the entry seven spots farther on in the dictionary can be recreated alongside more subtle détournements using algorithmically generated word embeddings to subtly tune texts in particular ways. (What if Jane Austen’s texts were rendered in words, say, 30% angrier?) The ability to tune a text in subtle or drastic ways can estrange language and defamiliarize texts. What’s more, beyond offering the chance to create works of startling newness, a platform to perform repeatable, experimental assays on literary texts within the web browser, is also an instructive opportunity to illustrate the methodological operations being performed, highlighting the inner workings of the algorithms in play.
1. In the spirit of Tristan Tzara’s Dada Manifesto 1918: “Always destroy what you have in you. On random walks. Then you will be able to understand many things.”
2. See Brian McHale’s discussion in “Poetry as Prosthesis,” Poetics Today Spring 2000 (21:1).
Detour, Digress, Destroy (on random walks): Demoing a Platform for Text Manipulation
Hypertexts & Fictions
Detour, Digress, Destroy (on random walks)^1: Demoing a Platform for Text Manipulation
Literary works authored with some recourse to algorithm are especially ripe for computational recreation and expansion. Authors affiliated with, or working under the influence of, the mostly French Oulipo group or the L=A=N=G=U=A=G=E poets of North America, use algorithmic routines to create or alter existing literary works, inventing new forms or concepts within the texts that embody them. This talk will present ongoing work to build an interactive literary platform for applying selected statistical transformations to selected texts. These transformations will range from relatively simple (or deceptively simple) “mechanical” routines, such as the selective transfer of words used by Jackson Mac Low to fill his own poem “Call Me Ishmael” with words from Moby-Dick^2, to more complex textual re-configurations using the increasingly sophisticated natural language models (word vector relationships, but not AI) that may transform us in turn. The Oulipian “N+7 Machine” which replaces nouns with the entry seven spots farther on in the dictionary can be recreated alongside more subtle détournements using algorithmically generated word embeddings to subtly tune texts in particular ways. (What if Jane Austen’s texts were rendered in words, say, 30% angrier?) The ability to tune a text in subtle or drastic ways can estrange language and defamiliarize texts. What’s more, beyond offering the chance to create works of startling newness, a platform to perform repeatable, experimental assays on literary texts within the web browser, is also an instructive opportunity to illustrate the methodological operations being performed, highlighting the inner workings of the algorithms in play.
1. In the spirit of Tristan Tzara’s Dada Manifesto 1918: “Always destroy what you have in you. On random walks. Then you will be able to understand many things.”
2. See Brian McHale’s discussion in “Poetry as Prosthesis,” Poetics Today Spring 2000 (21:1).
Bio
Devin Higgins builds digital collections and supports digital scholarship at Michigan State University Libraries. His previous publications have focused on literary texts as data, visualizing collaborative networks, text mining in libraries, and linked open data.