Analysing the Role of Pauses in Literary Text Using Deep Learning and Emotion Analysis

This study analyses the role of pauses in Samuel Beckett's Waiting for Godot using deep learning-based emotion analysis. Previous emotion analysis studies have mainly focused on explicit emotional expressions, but Beckett’s work is distinctive in that emotions are not directly revealed, and pause plays a crucial role in the text. To address this, the study proposes a new approach to analyse the role of pauses on emotions in literary texts. In this study, the dialogues of literary texts were converted into a form of structured data, and multiple emotion classification was performed using the ELECTRA deep learning model. The results indicate that pauses have a tendency to keep emotions in their existing state and thus play a role in regulating the emotional balance of the tragicomedy genre. Furthermore, it was found that pauses regulate the emotional rhythm of the play, maintaining neutrality and moderating extreme emotional changes. These results provide a new perspective through the deep learning emotion analysis on the interpretation of pauses as emotional and structural mechanisms in literary studies, rather than as simple spaces between lines of dialogue. Therefore, this study is significant for quantitatively assessing the emotional impact of non-linguistic elements in literary analysis using deep learning techniques and highlights the need for future emotion analysis models to consider factors such as pauses. Lastly, this study contributes to the methodological expansion of emotion analysis research and suggests possibilities for the convergence of literary studies and digital humanities.