Situation models and the default mode network

Cognitive neuroscientists continue to puzzle over how the default mode network (DMN) contributes to cognition. A renewed interest in studying narratives has led to a compelling possibility: the DMN tracks our evolving understanding of the current ‘state of affairs’ — a situation model. In this piece, we draw upon work in discourse processing to characterize what a situation model is and then use this framework to evaluate recent neuroimaging evidence. Recent functional magnetic resonance imaging studies demonstrate multivariate DMN activity is sensitive to an individual’s construal of the current situation. We also propose that (i) DMN situation representations are not specific to social or narrative content and instead summarize the current task state; and (ii) DMN representations are relatively low resolution and work alongside the hippocampus in service of event comprehension. Overall, we contend that situation models provide a productive framework for understanding the role of the DMN in human cognition. • Multivariate patterns of activity in the DMN may track our evolving situation models. • Situation models reflect the current 'state of affairs' from incoming experience. • DMN activity patterns shift alongside changes in event-level semantics. • DMN regions may track a generic 'state of affairs' beyond narratives (e.g. task set). • Interactions between DMN and hippocampus are key for event comprehension and memory.