Distinct but cooperating brain networks supporting semantic cognition

Abstract

Semantic cognition is a complex brain function involving multiple processes from sensory systems, semantic systems, to domain-general cognitive systems, reflecting its multifaceted nature. However, it remain unclear how these systems cooperate with each other to achieve effective semantic cognition. Here, we investigated the neural networks involved in semantic cognition using independent component analysis (ICA). We used a semantic judgement task and a pattern matching task as a control task with two levels of difficulty to disentangle task-specific networks from domain-general networks and to delineate task-specific involvement of these networks. ICA revealed that semantic processing recruited two task-specific networks (semantic network [SN] and extended semantic network [ESN]) as well as domain general networks including the frontoparietal network (FPN) and default mode network (DMN). Specifically, two distinct semantic networks were differently modulated by task difficulty. The SN was coupled with the extended semantic network and FPN but decoupled with the DMN, whereas the ESN was synchronised with the FPN and DMN. Furthermore, the degree of decoupling between the SN and DMN was associated with semantic performance. Our findings suggest that human higher cognition is achieved by the neural dynamics of brain networks, serving distinct and shared cognitive functions depending on task demands.