Controlling the Triple Network Model: Salience Network Modulatory Roles, Default Mode Dynamic Functions, and Central Executive Network Heterogeneity

Abstract

The triple network model has become a widely adopted framework for studying large-scale brain organization in basic and clinical neuroscience; however, accumulating evidence complicates this taxonomy. The default mode network, once labelled "task-negative", is now recognized as dynamically engaged across both internally directed and task-related processes. The salience network, often described as a "switching" hub, exerts context-dependent modulatory influences on both default mode and central executive network activity, supported by oscillatory coupling and directed connectivity. The central executive and frontoparietal networks, traditionally viewed as a unitary task-positive network, display heterogeneity across parcellations and frequency bands, suggesting finer-grained functional subdivisions than previously assumed. Multimodal studies integrating electroencephalography, functional magnetic resonance imaging, and other approaches converge on the view that networks interact through flexible, frequency-specific dynamics, but diverge in their anatomical assignments depending on atlas, modality, and task. In this scoping review synthesizes evidence on network definitions, methodological inconsistencies, and putative modulatory mechanisms, highlighting the heterogeneity of control networks and underscoring the need to incorporate temporal dynamics into future taxonomies. We propose that geometric hypotheses of neural control describing the global coordination of local activities, such as those used in neural manifolds, may improve our understanding of disordered cognition by integrating the many existing methodologically constrained maps into a coherent stable atlas that is more robust to approximating biases and errors associated with sampled population variability.