有限統合主義による分析窓の異事例適用可能性と統合コスト ――Silicon Valley Bank破綻を用いた競合理論との比較

This paper examines whether the analytical windows generated by Finite Integrationism (FI) remain applicable across substantively different cases and whether FI offers any residual integrative advantage once comparable analytical functions are reconstructed from existing theories. Using the 2023 failure of Silicon Valley Bank (SVB) as a case, the study applies six pre-specified analytical windows (W1–W6) concerning judgment, social input, shared expectations, technological mediation, responsibility, and institutional feedback. Five windows were judged to be established in the case, while W3 was only partially established. However, the corresponding analytical functions could also be reconstructed through combinations of existing research on bank runs, depositor networks, digital banking, responsibility and accountability, organizational learning, and institutional change. The comparison was then strengthened through broader stress tests using frameworks including Actor-Network Theory, STAMP/CAST, and Resilience Engineering. These tests substantially reduced the differences initially observed in conceptual translation and analytical-sequence continuity. The main residual candidate difference concerned the additional bridging assumptions required to connect transformations across analytical levels within a single sequence. This result does not establish the general theoretical or integrative superiority of Finite Integrationism. The remaining difference is based on author coding, has not yet been independently reproduced, and is sensitive to the construction of the competing model and to the scope of theory search. The paper therefore advances a narrower claim: Finite Integrationism may function as an integrative analytical lens that pre-arranges transformations across levels and may thereby reduce some of the additional bridging work required when applying multiple theoretical resources to a new case. Whether this constitutes a stable methodological advantage remains an open empirical question.