Old Comedy Through New Lenses, A Computational Study of Personal Satire in Aristophanes

Abstract

For centuries, the prevailing method for analyzing classical text has been close reading–extracting qualitative insights through detailed scrutiny of selective text excerpts. However, this approach, despite its depth, may overlook broader patterns that span across the texts. The emergence of "distant reading" and the recent proliferation of Natural Language Processing (NLP) tools present an opportunity to complement traditional analysis with data-driven objectivity, offering empirical evidence to bolster literary discourse. This thesis seeks to integrate NLP and distant reading techniques into the study of ancient Greek comedy, an endeavor largely unattempted in Classics. This study concentrates on the extant corpus of 4th and 5th century BC ancient Greek comedy, which comprises eleven plays written by the playwright Aristophanes. Within this genre, I focus on a form of personal satire, known as ὀνομαστὶ κωμῳδεῖν (“to ridicule by name”), that is characteristic of Aristophanes’ works. While this aspect of his comedies has been recognized by scholars, a systematic, quantified examination has been absent. Consequently, there is a scarcity of quantitative insight regarding the prevalence of this humor, Aristophanes' preferences for satirizing certain types of individuals (such as politicians, poets, mythological figures, etc.), and how these patterns evolved over time. Through the development of Named Entity Recognition (NER), this study compiles a comprehensive dataset of characters mentioned across Aristophanes' works, addressing the gap between digital humanities scholars and traditional Classicists in the application of Natural Language Processing (NLP) techniques to ancient texts. Subsequently, by correlating the types of characters with the competitive success of his plays, this thesis offers fresh insight into Aristophanes' satirical tactics and their impact on his audience. The findings provide a nuanced understanding of his use of personal satire and challenge existing perceptions of the comedic intent underlying his works. Through this novel approach, the thesis aims not only to advance the field of Classics with NLP techniques but also to shed light on the interplay between comedic artistry and audience reception in ancient Greek theater.