AI meets ALP in voce

Abstract

Finnegans Wake is the last work written by Irish writer James Joyce. In it, the increasing narrative complexity that characterised his literature reaches extreme levels, verging on the unintelligible. The author not only mixes different languages, he also invents new terms that germinate and are signified, to a large extent, from their sonic dimension. Characters as well as real and fictitious events are mixed together, forming an amalgam in which a perpetual cyclical glocality of mythological archetypes is perpetuated. The female protagonist, Anna Livia Plurabelle, is a mutating character who becomes different natural elements throughout the text. The murmur of the river mingles with her own in a narrative pulse that overrides the shore. Voices are always heard in the background and, although they are not clear, are suggestive.

This literary work is not to be read, it is to be listened to; it is not to be understood, it is to be deduced.

This research is situated within the current context of the boom in the use of Artificial Intelligence (AI) on word processing. It puts the transcription of some spoken passages from Finnegans Wake to the test, evaluating the degree of accuracy and possible biases or deviations in the results. This work is part of the results of a research stay granted by the Moving Minds call from the Campus Mare Nostrum at the University of Murcia.