Literary systems as complex adaptive systems: a computational study of motif self-organization in Agatha Christie’s detective fiction

Abstract This study analyzes the fifty-six-year evolutionary process of core motifs in Agatha Christie’s sixty-six full-length detective novels from the perspective of complex systems theory, providing empirical confirmation of the hypothesis that literary works undergo self-organization processes as complex adaptive systems. Through an approach combining Term Frequency-Inverse Document Frequency with genre-specific domain knowledge, we identified twenty-five core motifs and analyzed them using a multi-layered analytical framework consisting of time-series analysis, clustering, network analysis, and critical point detection. The analysis confirmed that Christie’s works exhibit systematic evolution with strong inter-motif co-evolutionary phenomena. The concentrated occurrence of critical points around 1935 (simultaneous changes in twelve motifs) demonstrates phase transition phenomena characteristic of complex systems, and the continuous decrease in Shannon entropy (4.382 → 4.177) quantitatively demonstrates the system’s self-organization. Network analysis showed that the motifs are organized into five semantic clusters, distinct from the six categories initially proposed by the experts. This indicates the existence of emergent reorganization tendencies. We also observed sociocultural changes and genre evolution in the mid-twentieth century with increasing psychological and atmospheric motifs while traditional detective motifs diminished. This study presents a new theoretical framework for understanding literary works not as static texts but as dynamically evolving complex adaptive systems and demonstrates the potential of a new literary research methodology involving motif system analysis of a single author’s complete works.