Data: Severance Wiki
Dual Voices: Exploring Severance Dialogue Through Data Visualization
A visual analysis of dialogue distribution and character relationships in the Apple TV series
In a world where work and personal life are radically severed, Severance poses not only existential questions but also a narrative rich with nuance and relationships. But what happens when we examine its story through a quantitative lens? This article introduces two data visualizations that uncover hidden patterns in the series’ dialogue—from how character presence shifts across episodes to the web of interactions shaped by shared scenes. A different way to explore the world of Lumon, dividing—just like in the show—the “Innies” from the “Outies.”
Apple TV’s Severance, with its intricate narrative universe and innovative separation between work and personal life, offers fertile ground for the quantitative analysis of its characters’ dialogue. To explore this dimension, two data visualizations have been created to examine how dialogue is distributed and how character dynamics evolve throughout the series. The first visualization is a scatter plot where each point represents a character, with its size reflecting the number of dialogue lines spoken in a given episode. These points are arranged chronologically along the timeline of the episodes, allowing us to observe how each character’s verbal presence shifts over time. Additionally, characters are categorized as either “Innie” or “Outie,” depending on whether their identity belongs to the workplace or the outside world, adding a layer of interpretation regarding how voices are distributed across narrative contexts.
- Point size: number of dialogue lines the character speaks in that episode.
- X-axis position: chronological episode timeline.
- Point color: character category (Innie / Outie).
- Hover over the points to view the dialogues lines of each character.
In the second visualization, we present a network graph in which each node represents a character, scaled according to their total number of dialogue lines across the entire series. The links between nodes indicate shared scenes, and the thickness of those links represents the amount of dialogue exchanged between characters in those scenes. To ensure a cleaner and more readable representation, tertiary characters with minimal screen time or very few lines have been excluded from this visualization. This focus helps highlight the most relevant characters and the most significant dialogue-based relationships within the narrative.
This network graph provides a synthetic view of the relational dynamics among the cast, helping to identify clusters of intensive interaction, dominant relationships, and isolated characters. As in the scatter plot, each node is classified as either an “Innie” or an “Outie,” making it possible to graphically explore the interaction between these two opposing realities at the heart of the series. Together, these visualizations offer a unique and quantifiable perspective on the relational and narrative fabric that shapes the dual universe of Severance.
- Node size: total number of dialogue lines spoken by the character across the series.
- Link thickness: intensity of shared dialogue (lines exchanged in shared scenes).
- Node color: character category (Innie / Outie).
- Hover over the nodes to view a short biography of each character.
- Tertiary characters with minimal dialogue have been excluded for visual clarity.
