The Evolutionary Purpose of Dreaming: Rehearsal for the Waking Social World

Dreams are strange enough that every culture that has left records has tried to explain them. The ancient explanations — divine communication, prophetic vision, access to another world — have been replaced by scientific ones, but the scientific ones have not achieved the clarity that replaced earlier models of, say, disease or celestial motion. Dreams remain genuinely puzzling. We spend roughly two hours in dreaming sleep every night, burning metabolic resources, generating vivid and often disturbing experience, with no agreed explanation of why. The most persuasive current accounts share a common structure: they propose that dreaming is functional, that it does something for the waking organism that would otherwise not get done. Among these accounts, the one with the strongest explanatory reach is what might be called the social simulation hypothesis.

The Threat Simulation Hypothesis and Its Extension

Antti Revonsuo, a Finnish neuroscientist and philosopher of mind, proposed the threat simulation theory of dreaming in 2000. His observation was that dream content is systematically skewed toward threatening scenarios relative to waking life. Across multiple cultural contexts and dream diaries, we are chased, attacked, humiliated, lost, unprepared for exams we have not studied for, late to events that matter. Positive dreams exist, but the distribution is not neutral. Revonsuo's argument was evolutionary: dreaming is the brain's threat rehearsal system. By running simulations of dangerous scenarios during sleep — when the motor output is blocked and the simulations are therefore safe — the organism practices detecting and responding to threats without real-world consequence. This is the same logic behind play behavior in juvenile mammals, who practice fighting and fleeing in contexts where the stakes are low. The theory has been extended significantly. Researchers at the University of Turku have argued that the simulation function extends well beyond physical threats to the full domain of social challenge. Dreams overwhelmingly feature people — typically people the dreamer knows or has emotional investment in — navigating situations that carry social stakes: conflict, abandonment, reunion, betrayal, status negotiation. The brain appears to be running social simulations as much as threat simulations.

The Social Brain During Sleep

This makes sense from the perspective of what we know about why a large brain is metabolically worth maintaining. The social brain hypothesis, developed by Robin Dunbar and others, proposes that primate brain expansion was driven primarily by the demands of social living — tracking relationships, modeling other minds, managing coalitions, predicting behavior. If this is right, then the most computationally demanding thing human brains do is not spatial navigation or tool use but social cognition. A system that invests so heavily in social processing would have strong reason to maintain that processing during sleep. The dream as social simulation would be the brain maintaining and extending its core competency during the hours when no external social demands are placing constraints on what gets practiced. Research from the Max Planck Institute for Human Cognitive and Brain Sciences using neuroimaging during REM sleep found activation of the medial prefrontal cortex — a region centrally involved in mentalizing, the modeling of other minds — during dreaming. The brain, during the most vivid dream states, is running its social cognition systems at high intensity.

The Tangent: What Recurring Dreams Tell Us

Recurring dreams have fascinated psychologists and lay observers alike, and there is now reasonable evidence about what they represent. The most common recurring dream themes across cultures — being chased, falling, being unprepared for a test, teeth falling out, appearing naked in public — all share the structure of social inadequacy or threat of exposure and judgment. This distribution is not random. If dreams are social simulations, then recurring dreams are the brain's evidence of unresolved social challenges — scenarios in which the simulation has not yet reached a resolution that allows the brain to file the case as closed. The anxiety exam dream persists because the underlying concern about being found inadequate has not been metabolized. This is consistent with the finding that recurring nightmares tend to decrease when the waking situation driving them is resolved.

What This Means for Waking Social Capacity

If the social simulation hypothesis is correct, then sleep quality and dream quality are not just matters of rest but of social competency maintenance. A person who is chronically sleep-deprived is not just tired — they are running their social simulation system at reduced capacity, with less practice at the scenarios that waking life will present. Studies consistently find that sleep deprivation impairs social cognitive performance — the ability to read emotional expressions accurately, to model others' intentions, to regulate one's own emotional responses in social contexts. These findings are usually reported as showing that sleep deprivation impairs cognition generally. The social simulation hypothesis suggests the impairment is not general but specific: what sleep does most critically for the brain is run the social simulations that keep the social cognition system calibrated.

The Meaning of the Strange Content

The bizarre quality of dream experience — the ways that people morph into other people, that locations blend, that impossible events unfold without surprise — may also make sense within the simulation framework. A rehearsal system benefits from exploring edge cases and combining elements in ways that waking experience does not present. The strangeness is the brain's version of stress-testing: what if this person I trust suddenly became hostile, what if this familiar place suddenly became unfamiliar, what if the social rules I rely on suddenly stopped applying? Dreams may be the brain asking itself the hard questions it does not get to ask in the constrained space of ordinary daily life. The answers it works out there — and the fact that it is asking the questions at all — may be doing more for our waking social lives than we have recognized.