AI-Based Theoretical Validation of Emotion Architecture: A Systematic Analysis of Emotion Generation Based on the Three-Layer Semantic Network Model

This study conducts a systematic empirical validation of Emotion Architecture (Nakahama, 2026a), an applied theory built upon the foundational Hierarchical Model of Inner Meaning (Nakahama, 2026b). Using a novel methodology termed AI-assisted theoretical validation, the study demonstrates the explanatory power of Emotion Architecture. The theory reconceptualizes emotion as a two-story structure consisting of an outer layer (primary emotions and sensory input) and an inner layer (a semantic network composed of surface, value, and worldview layers). It further proposes a unified model in which higher-order emotions (secondary and tertiary emotions) emerge through two pathways: (Pathway A) primary emotion × semantic network, and (Pathway B) semantic network alone. In this study, an AI system was instructed to internalize Emotion Architecture and subsequently generate 50 emotional phenomena that are difficult to explain using conventional theories but can be structurally accounted for by Emotion Architecture. Ten of these phenomena were selected for detailed analysis. The results show that five categories of higher-order emotions—memory-reactivation emotions (surface-level secondary), normative emotions (value-layer secondary), value-based emotions (value-layer secondary), narrative-transformative emotions (worldview-layer tertiary), and worldview-transformative emotions (worldview-layer tertiary)—can all be coherently explained through the three-layer semantic network and the five types of semantic updating (addition, reinterpretation, reorganization, value shift, worldview transformation). Furthermore, the AI-generated phenomena and theory-based explanations jointly demonstrate both the breadth (ability to account for diverse emotional phenomena) and depth (ability to provide precise, layer-specific explanations) of Emotion Architecture. This study shows that Emotion Architecture offers an integrative account of higher-order emotional processes that have remained insufficiently explained by foundational emotion theories, appraisal theories, constructivist approaches, moral emotion research, and aesthetic emotion research. It also highlights the methodological value of AI-assisted theoretical validation.