Allusions, Algorithms, and the Mind: Observations from a Computational Approach to Reading Poetry
From Lisa Otty
Allusions in epic poetry are a game played between the poet and the reader. The epic poet modifies a famous passage, and invites the reader to make a comparison between their changed version and the original. A terrifying shipwreck in Virgil’s Aeneid becomes proof of a hero’s courage when the scene is borrowed, centuries later, by Voltaire. A reader who recognizes Voltaire’s allusion is impressed with the poet’s knowledge of the classics–they are also pleased with themselves for perceiving it, since the same education is required.
This game has become much harder to play since classical languages and epic poetry have dropped off the standard curriculum. However, digital tools can assist the modern reader interested in hunting for allusions. This talk will present one such tool, Tesserae-OBVIL, which attempts to locate points of similarity between French epic poetry and Latin.
This cross-language context constitutes a particularly interesting data-set. In order to design software that can reproduce the expert reader’s experience of allusion, we need to determine how epic poets secure a reader’s attention, inspire recollection of a source passage, and overcome the barrier of second-language suppression during first-language reading. Tesserae-OBVIL only represents a first step in designing this ‘mechanical reader,’ yet it has already been used to discover allusions previously un-discussed in the scholarly literature. Stylistic analysis of search results and known allusions not only suggests next steps in the development of Tesserae-OBVIL, but also shows us how poets leverage the mechanics of human cognition to control a reader’s attention and recall.
James Gawley has published a number of articles on the computational analysis of epic poetry in French in Latin, as well as best practices in digital humanities and experiment design. He is currently a postdoctoral researcher at Sorbonne University, where he is developing a research tool called Tesserae-OBVIL designed to locate allusions in French poetry, especially when those allusions cross the language barrier from French to Latin.