Search for a command to run...
Isolated reports have long suggested a similarity in content and thought processes across mind wandering (MW) during waking, and dream mentation during sleep. This overlap has encouraged speculation that both "daydreaming" and dreaming may engage similar brain mechanisms. To explore this possibility, we systematically examined published first-person experiential reports of MW and dreaming and found many similarities: in both states, content is largely audiovisual and emotional, follows loose narratives tinged with fantasy, is strongly related to current concerns, draws on long-term memory, and simulates social interactions. Both states are also characterized by a relative lack of meta-awareness. To relate first-person reports to neural evidence, we compared meta-analytic data from numerous functional neuroimaging (PET, fMRI) studies of the default mode network (DMN, with high chances of MW) and rapid eye movement (REM) sleep (with high chances of dreaming). Our findings show large overlaps in activation patterns of cortical regions: similar to MW/DMN activity, dreaming and REM sleep activate regions implicated in self-referential thought and memory, including medial prefrontal cortex (PFC), medial temporal lobe structures, and posterior cingulate. Conversely, in REM sleep numerous PFC executive regions are deactivated, even beyond levels seen during waking MW. We argue that dreaming can be understood as an "intensified" version of waking MW: though the two share many similarities, dreams tend to be longer, more visual and immersive, and to more strongly recruit numerous key hubs of the DMN. Further, whereas MW recruits fewer PFC regions than goal-directed thought, dreaming appears to be characterized by an even deeper quiescence of PFC regions involved in cognitive control and metacognition, with a corresponding lack of insight and meta-awareness. We suggest, then, that dreaming amplifies the same features that distinguish MW from goal-directed waking thought.