The Emotion Dynamics of Literary Novels

Krishnapriya Vishnubhotla,Adam Hammond,Graeme Hirst,Saif M. Mohammad

2024-03-05

Abstract:Stories are rich in the emotions they exhibit in their narratives and evoke in the readers. The emotional journeys of the various characters within a story are central to their appeal. Computational analysis of the emotions of novels, however, has rarely examined the variation in the emotional trajectories of the different characters within them, instead considering the entire novel to represent a single story arc. In this work, we use character dialogue to distinguish between the emotion arcs of the narration and the various characters. We analyze the emotion arcs of the various characters in a dataset of English literary novels using the framework of Utterance Emotion Dynamics. Our findings show that the narration and the dialogue largely express disparate emotions through the course of a novel, and that the commonalities or differences in the emotional arcs of stories are more accurately captured by those associated with individual characters.

Computation and Language

What problem does this paper attempt to address?

The paper primarily explores the issue of emotional dynamics in literary novels. Specifically, it attempts to address the following core questions:

1. **Emotional Dynamics of the Novel**: The study examines the emotional changes throughout the entire novel, the narrative parts only, and the emotions of various characters from beginning to end.

2. **Emotional Differences Between Narration and Dialogue**: It investigates whether there are distinct emotional trajectories between the narrative parts and the dialogues of different characters. Previous literature often views the entire novel as a single emotional trajectory or focuses on the narrative, neglecting the emotional dynamics in character dialogues.

3. **Diversity of Character Emotional Trajectories**: It analyzes the diversity of emotional trajectories of characters within a single story. For example, in a story about a hero and a villain, the two main characters may exhibit opposing emotional trajectories; whereas in a story about a group of friends on an adventure, the characters' emotional trajectories may be more correlated.

4. **Impact of Gender on Emotional Trajectories**: Is there a difference in the consistency of emotional trajectories among characters of different gender groups? How is this difference reflected in the works of different authors?

To answer these questions, the authors conducted experiments using a range of resources, including the PolyU Discourse Novel Corpus (PDNC), emotion lexicons, and tools such as the Utterance Emotion Dynamics (UED) framework to quantify the patterns of emotional changes in the text.

Key findings of the paper include:

- Novels tend to express higher positive emotions (Valence), lower arousal (Arousal), and moderate dominance (Dominance) overall.
- Character dialogues are more likely to express higher positive emotions and dominance compared to the narrative parts.
- The emotional trajectories of characters are significantly different from the narrative parts, and there are notable differences between the emotional trajectories of different characters.
- Characters written by female authors tend to express higher positive emotions and dominance, while male characters written by male authors exhibit higher arousal.

Through these studies, the paper reveals some underexplored narrative aspects in computational literary research and provides quantitative evidence supporting the importance of placing characters at the center to gain a more nuanced understanding of novels.