This project sits at the intersection of digital humanities and social science, using text analysis and survey methods to study how national identity is constructed and transmitted through education.
Korean national history textbooks are how the state, through public education, narrates the nation and represents dominnate national identity frames that inform nation-building. For generations of students, these textbooks shape early understandings of what it means to be Korean, whether defined primarily by ethnic heritage, democratic values, or some combination of both. This project asks: do those formative narratives persist into adulthood?
The project proposes to work with a corpus of South Korean natinoal history textbooks spanning the republic's entire existence, from 1948 through 2016. Collected from the National Institute of Korean History and the Georg Eckert Institute, these materials cover the post-liberation authoritarian regimes and the democratic era, each presenting different grand narratives of the (South) Korean nation.
The project proceeds in two stages. First, we use AI-assisted text analysis to identify how the nation is narrated across different periods and what type of national identity is articulated in each era's textbooks. Second, we design a survey experiment that draws directly from these textbook-derived narratives, asking South Koreans to evaluate competing versions of the nation rather than respond to abstract hypotheticals. This design allows us to test whether individuals favor the version of the nation that reflects what they learned during their school years, as theories of long-run political socializaiton would predict.
Keywords: Korean national identity, history textbooks, text analysis, political socialization, survey experiments
- How is the Korean nation narrated in state-written or authorized textbooks, and how has this changed over time?
- What types of national identity is articulated across different political regimes and generations?
- Do the identity frames encountered during childhood and adolescence persist into adulthood?
Text Analysis: The project develops an AI-assisted pipeline using Korean language-trained large language models to extract and classify national identity narratives from the textbook corpus. This approach enables an analysis of a large volume of historical texts while validating results against established text-as-data methods.
Survey Experiment: The text analysis directly informs a nationally representative survey experiment in South Korea (n ≈ 2,500). Rather than asking abstract questions about national identity, respondents will evaluate vignette-style passages drawn from actual textbooks—each presenting a distinct vision of the Korean nation. By linking respondents' ages to the textbooks used during their school years, we can test whether people prefer the national identity frames they encountered during their formative education.
Duration: 12 months (February 2026 – January 2027)
| Phase | Period | Activities |
|---|---|---|
| Corpus preparation | Spring 2026 | Finalize corpus; build and validate text analysis pipeline |
| Text analysis | Summer 2026 | Analyze textbooks; design survey based on findings |
| Data collection | Fall 2026 | Field survey experiment in South Korea |
| Synthesis | Winter 2026/27 | Integrate results; draft manuscripts; release data and code |
Expected Deliverables:
- Academic article on how national identity is narrated in Korean history textbooks
- Academic article on whether textbook-derived identity frames persist into adulthood
- Methodological contribution on AI-assisted text analysis for Korean-language corpora
- Publicly available dataset and analysis code
Primary Funding: Academy of Korean Studies (한국학중앙연구원)
Grant Program: Overseas Korean Studies Grant
Grant Number: KS-2026-R-094
Award Amount: $19,262 USD
Principal Investigator: Steven Denney (Assistant Professor, Korean Studies & International Relations)
Institution: Leiden University, Institute for Area Studies
Current Phase: Awarded; project launch February 2026
Next Milestones: Corpus finalization and pipeline development (Spring 2026)
This repository will contain project materials as they are developed:
- Project documentation and methodology notes
- Analysis scripts and pipeline code
- Survey instruments and experimental materials
- Cleaned metadata and summary data
- Figures, tables, and publication drafts
This project builds on ongoing collaborative research with Aron van de Pol (PhD Candidate, Korean Studies and Computer Vision, Leiden University) on the digitization and computational analysis of Korean history textbooks.
Steven Denney
Assistant Professor, Korean Studies & International Relations
Institute for Area Studies, Leiden University
Email: s.c.denney@hum.leidenuniv.nl