Transforming the Northwest frontier: development discourse in Republican China through computational analysis of the historical press

This study examines the discourse surrounding Northwest China’s development during the Republican era (1911–1949). Drawing on 5,461 newspaper and periodical articles from the Quan Guo Bao Kan Suo Yin (CNBKSY) and the Shenbao database, we developed a multi-stage text extraction workflow leveraging Google Gemini to convert complex historical scans into machine-readable text. After standardising historical character forms, we applied structural topic modelling (STM) to recover 26 coherent themes covering infrastructure, resource extraction, state governance, ethnic relations, and cultural mobilisation. Topic-correlation analysis reveals three higher-order clusters—infrastructure and resources, governance and industry, and cultural-educational development—integrated by overarching concerns for strategic geography and state-led economic planning. Temporal analysis shows dramatic inflections: discourse initially gained momentum through regional development initiatives led by warlords in the 1920s, before the 1931 Manchurian Incident amplified a security-centred rhetoric that fully shifted the focus from exploratory surveys to urgent state-led nation-building after the 1937 Japanese invasion—a focus that gradually ebbed post-1945 as wartime imperatives faded. These patterns illustrate how national crises transformed the Northwest from a peripheral frontier into a strategic heartland in Republican-era Chinese media and policy discourse. By leveraging a large-scale corpus and cutting-edge computational methods, this study moves beyond traditional historiography to provide the most comprehensive analysis yet of how Republican China envisioned, prioritised, and rhetorically shaped its interior frontier during a pivotal period of modern Chinese state-building.