Research Statement

My work circles a single question: why does identity form robustly even when information is abundant, oppression and censorship are known? States narrate, platforms filter, individuals participate. Yet the narrative frameworks that organize collective self-understanding prove remarkably durable, even among people who can articulate their constructed nature.

I approach this question across three scales, each of which emerged because the previous one reached its limits.

Three Scales

Intellectual history. My undergraduate thesis examined how Japan’s Hua-Yi distinction and Western international law jointly constructed the modern imagination of “Asia.” Asianism positioned itself as a counter-thesis to Eurocentric universalism, but it reproduced the spatial-order logic it claimed to transcend: a failed antithesis. The structural entrapment I found there (frameworks shaping actors even as those actors try to oppose them) set the question I have pursued since.

Macro-behavioral observation. A synthetic control study of China’s Double Reduction policy tried to use satellite nighttime light as an objective proxy for student activity. Three identification strategies failed; a follow-up Shanghai GIS exercise showed that VIIRS data at 463-meter resolution cannot separate student behavior from urban infrastructure lighting. The project confirmed a methodological intuition: objective remote-sensing proxies cannot penetrate individual-level states. What looks like a clean behavioral signal dissolves under scrutiny into seasonality, urbanization, and measurement noise.

Micro-level digital expression. A study of the 2025 “Mourning Ming” wave on Xiaohongshu asked which narrative directions within a single discourse event face differential search-visibility friction. The finding was layered and counterintuitive: feminist-lineage framing and Dream of the Red Chamber (Honglou Meng) material were the most robustly associated with reduced visibility, more so than the anti-Manchu revisionism one might expect to be most sensitive. Platform governance operates through friction rather than deletion, and the friction is not where conventional accounts of Chinese censorship would predict.

Current and Future Work

The current projects bring these scales together. A course proposal asked whether patriotic education achieves broad coverage of official nationhood premises without comparable integration of their logical dependencies. My M.A. thesis proposal extends this: using the Mourning Ming wave as a diagnostic case, it tests whether ordinary citizens actually rely on the Qing-continuity premise the state treats as load-bearing when justifying territorial integrity, or whether they reach for entirely different reasons. Both projects draw directly on the intellectual-historical groundwork of my Asianism thesis and the measurement lessons of the nightlight and Xiaohongshu studies.

Methodological Growth

My trajectory has moved from pure intellectual history (close reading, Schmitt, Maruyama, Foucault) through large-scale empirical work (satellite data, synthetic control, web scraping, LLM-assisted content coding) toward the recognition that neither approach alone is sufficient. The questions I care about require combining archival and conceptual analysis with computational methods. Not because mixed methods are intrinsically better, but because identity formation sits at the intersection of historically inherited frameworks and present-day individual cognition, and no single method reaches both.

Gender is a second thread that emerged empirically. In the Xiaohongshu study, feminist-lineage direction (the claim that maternal descent is the unbroken line) was the single most stable predictor of reduced search visibility. I did not set out to study gendered dimensions of nationalist discourse, but the data pointed there insistently. This finding now shapes how I think about identity formation more broadly: patrilineal assumptions about national continuity may be more politically load-bearing, and more actively policed, than the ethnic-classificatory or temporal-causal dimensions that dominate the existing literature.