A Hierarchical-Dynamic integration model for the formation of interpersonal neural synchrony during spoken communication.

Interpersonal neural synchrony (INS) is widely observed during spoken communication, yet the cognitive and neural processes through which it emerges remain incompletely understood. This review synthesizes evidence on the functional significance, representational organization, and neural implementation of INS and proposes a Hierarchical-Dynamic Integration Model (HDIM) of its formation. HDIM conceptualizes INS as the cross-brain manifestation of ongoing coordination between two interdependent processes: multilevel linguistic alignment and contextual-intentional integration. The hierarchical organization of linguistic alignment encompasses functionally differentiated yet interacting phonological/prosodic, lexical-syntactic, semantic, and higher-order representational levels, without assuming a fixed bottom-up sequence. Its dynamicity arises from recurrent processes of prediction, imitation, and adaptation that continuously regulate interpersonal coordination as communication unfolds. At the neural level, HDIM proposes that linguistic alignment engages distributed bilateral language systems, whereas contextual-intentional integration recruits coordinated activity across language-related and social-cognitive/mentalizing systems. The model yields four empirically testable predictions concerning partial dissociation among representational levels, functional dependencies across levels, temporally differentiated yet interacting regulatory processes, and partner-sensitive integration of linguistic and non-linguistic information. We further identify uneven empirical support across components of the model, together with causal and methodological challenges and limitations in its current scope. We conclude by outlining priorities for causal and interaction-specific testing, computational formalization, and evaluation of the model's boundary conditions and generalizability across communicative contexts.