Publications / The Shapes of Stories
The Shapes of Stories: Sentiment Analysis for Narrative
Cambridge Elements in Digital Literary Studies, Cambridge University Press, 2022.
Bibliographic record
- Author
- Katherine Elkins
- Title
- The Shapes of Stories: Sentiment Analysis for Narrative
- Series
- Cambridge Elements in Digital Literary Studies
- Publisher
- Cambridge University Press
- Year
- 2022
- ISBN
- 9781009270397
- DOI
- 10.1017/9781009270403
- Access
- Cambridge Elements, subscription and purchase; front matter freely available
- Last updated
- 14 July 2026
Summary
The Shapes of Stories develops an ensemble methodology for diachronic sentiment analysis of full-length narrative and connects computational modeling with close and middle reading. It shows how model selection, smoothing, scale, and segmentation affect the story curve a researcher sees and the interpretation built from it.
The book introduces a vocabulary for narrative shape, including storyteller curves, curves-on-a-hill, curves-in-a-hole, tragic curves, person-on-the-plain, and the distributed heroine, and proposes middle reading as a methodological bridge between distant reading and close reading. Rather than treating a sentiment model as a measuring instrument that simply reports a result, it treats the choice of model, the degree of smoothing, and the scale of analysis as interpretive decisions that a reader must own.
It also treats translation and transmission, asking how narrative patterns persist, drift, or change as stories move across versions and cultural contexts. That question becomes the explicit subject of the later work on Proust in translation and on Cinderella across languages.
What this work contributed
- Established an ensemble approach to diachronic sentiment analysis of narrative, benchmarking many models rather than trusting one.
- Named and defined middle reading: reading at more than one analytical distance, so that a computational trajectory is always returned to the passage that produced it.
- Showed that smoothing, model selection, and scale are interpretive choices, not neutral preprocessing.
- Supplied a shared vocabulary of narrative shapes that later work could compare against.
- Opened translation and transmission as a computational question about what persists when a story travels.
Relationship to earlier and later work
It builds on the 2019 study of a nonlinear novel, Can Sentiment Analysis Reveal Structure in a “Plotless” Novel?, where middle reading was first tested, and on the ensemble and trajectory-comparison instruments developed with Jon Chun.
It is the foundation for In Search of a Translator (2024), which applies whole-narrative comparison to translation, for The Shapes of Cinderella (2025), which extends it across languages and cultures, and for Beyond Plot (2025), which states the methodological argument directly. See Scholarly Reception for the citation record.
How to cite
Chicago
Elkins, Katherine. The Shapes of Stories: Sentiment Analysis for Narrative. Cambridge Elements in Digital Literary Studies. Cambridge: Cambridge University Press, 2022. https://doi.org/10.1017/9781009270403.
MLA
Elkins, Katherine. The Shapes of Stories: Sentiment Analysis for Narrative. Cambridge UP, 2022. Cambridge Elements in Digital Literary Studies.
APA
Elkins, K. (2022). The shapes of stories: Sentiment analysis for narrative. Cambridge University Press. https://doi.org/10.1017/9781009270403
BibTeX
@book{elkins2022shapes,
author = {Elkins, Katherine},
title = {The Shapes of Stories: Sentiment Analysis for Narrative},
series = {Cambridge Elements in Digital Literary Studies},
publisher = {Cambridge University Press},
year = {2022},
isbn = {9781009270397},
doi = {10.1017/9781009270403}
}