Tracing the Signature Dynamics of EFL Learners’ L2 Motivation in an AI‐Enhanced Learning Context: A Retrodictive Study on College Students in China

ABSTRACT Adopting a retrodictive qualitative modeling approach, this study explored the signature dynamics of Chinese college EFL students’ L2 motivation in AI‐enhanced learning contexts. From an initial pool of 118 non‐English major students in one Chinese university, three distinct motivational archetypes were identified based on the result of hierarchical clustering analysis of the questionnaire regarding their motivational conglomerate. Two prototypical students from each archetype were then selected to take part in retrodictive interviews exploring the developmental trajectories of their motivational intensity across the past three semesters and the effects of AI on their motivational states in different phases. Aligned with Complex Dynamic Systems Theory, students’ motivational dynamics exhibited the inherent characteristics of complexity and nonlinearity. Contrary to prevailing techno‐optimism in the literature, the findings demonstrated the dual‐function of AI in shaping various motivational states of students. The results yield important pedagogical implications for effectively integrating AI tools in college English curriculum.