A Working Hypothesis on the Functional Role of REM Sleep

Despite substantial progress in sleep research, the functional role of REM sleep remains an open question. Modern neuroscience associates REM sleep with memory consolidation, neuroplasticity, emotional processing, and several other well-established phenomena. However, these processes are often described individually rather than as components of a single functional mechanism.

This article proposes a working hypothesis that does not seek to replace existing theories. Instead, it attempts to integrate several established observations within a unified functional framework.

Initial Observation

During REM sleep, people continue to experience events almost as if they were awake. Fear, curiosity, joy, anxiety, relief, and other subjective experiences remain vivid. At the same time, critical evaluation of events is markedly reduced, allowing even impossible scenarios to be experienced as plausible.

Proposed Model

This hypothesis proposes that REM sleep places the brain into a mode of internal simulation. The purpose of these simulations is not simply to replay memories, but to obtain subjective feedback that cannot be derived directly from physiological measurements alone.

Within this framework, the brain and consciousness are treated as two different representations of the same underlying processes. The brain operates with measurable physiological variables, whereas consciousness represents those processes as subjective experience. During REM sleep, consciousness functions as an interface that provides subjective feedback while the brain records the corresponding physiological state.

Why Might Such a Mechanism Exist?

One of the brain's fundamental tasks is to establish stable behavioral patterns. Learning, however, is not complete merely because behavior becomes accurate. A behavior may remain physiologically expensive to execute.

According to this hypothesis, REM sleep repeatedly simulates emotionally significant situations, allowing the brain to progressively reduce the physiological cost of executing established behavioral patterns without changing the behavior itself.

In other words, the object of optimization is not the behavior, but the internal cost of performing it.

Possible Implications

If this hypothesis is correct, it offers an alternative perspective on several familiar observations. Increasing emotional stability with age may reflect not only conscious experience but also years of unconscious optimization occurring during REM sleep.

Recurring dreams following psychological trauma may likewise represent repeated attempts to optimize behavioral responses associated with emotionally significant experiences while preserving their adaptive value.

Questions Raised by the Hypothesis

Can the quality of REM sleep be experimentally linked to the optimization of behavioral patterns?

Can changes in the physiological cost of learned behaviors be measured over time?

Could part of the age-related decline in emotional reactivity reflect long-term optimization rather than only age-related physiological decline?

Might recurring traumatic dreams represent an ongoing optimization process rather than solely incomplete emotional processing?

These questions are experimentally testable. For that reason, this work should be viewed as a working hypothesis intended to stimulate discussion and further research rather than as a completed theory.