Why the brain enters rapid eye movement (REM) sleep, and why it accompanies vivid, predominantly visual dreaming, remains one of sleep neuroscience's unresolved questions. The defensive activation theory proposes that REM sleep circuitry periodically amplifies visual-cortex activity to defend the visual system's cortical territory against takeover by the other senses during the extended visual deprivation of night; its published support is purely correlational (cross-species primate correlations between plasticity proxies and REM proportion, and parallel declines of plasticity and REM with aging), and the direct causal tests its authors specify — blindfold/visual-deprivation neuroimaging across species differing in plasticity, REM-suppression neuroimaging, and comparative analyses using direct neurophysiological plasticity measures — remain unperformed as of 2026. The general question of what function REM sleep and dreaming serve is the motivating background, and the concrete, expert-reviewable open core is the causal adjudication of the defensive-activation account, which is scoped to 'the biological reason underpinning visual dreaming' rather than to 'the numerous other features of REM sleep'.

Background

Rapid eye movement (REM) sleep is a sleep stage characterized by an activated, wake-like electroencephalogram, rapid eye movements, muscle atonia (paralysis of the skeletal muscles), and — in humans — the vivid hallucinoid experiences we call dreams, which are strikingly visual. This poses two linked puzzles. First, what is the biological function of REM sleep and dreaming? Many proposals exist — energy conservation, memory consolidation, emotional regulation, brain maturation — but none is causally established, and it is actively debated whether REM sleep has any single function at all; Kirov's 2013 commentary in Behavioral and Brain Sciences, responding to Llewellyn's proposal that dreams serve elaborative memory encoding, is one explicit statement of this non-reductionist position. Second, why are dreams so specifically visual, strongly activating the occipital (visual) cortex? A key neurophysiological fact is that during REM sleep, waves of activity travel from the pons (a brainstem region) through the lateral geniculate nucleus of the thalamus to the occipital cortex — the ponto-geniculo-occipital (PGO) waves — so REM circuitry specifically targets visual cortex, for reasons that remain unexplained. Eagleman and Vaughn's defensive activation theory (DAT) addresses both puzzles at once. Its premise is competitive neuroplasticity: cortical territories maintain their claim on the brain through continuous activity, and when activity slows or stops the territory is progressively taken over by neighboring senses — in blind individuals the occipital cortex is colonized by touch and hearing, and remarkably, in sighted people as little as 40–60 minutes of blindfolding suffices for touch-related activity to emerge in primary visual cortex, as summarized by Eagleman and Vaughn (2021). Because planetary rotation casts the visual system into roughly 12 hours of darkness per cycle while the other senses stay active, the visual cortex faces a recurring nightly threat of encroachment. DAT proposes that REM sleep — the visual hallucinations of dreaming and the PGO-wave circuitry driving them — evolved as the visual system's periodic defensive amplification, keeping its territory from being taken over during sleep. The theory's published support is correlational: across 25 primate species, three behavioral proxies of plasticity (time to locomotion, time to weaning, and time to adolescence) correlate positively with the proportion of sleep spent in REM; plasticity and REM proportion rise together with evolutionary recency to humans; and in humans both decline in parallel with aging. The causal tests the theory proposes remain unperformed as of 2026: suppression of REM sleep — pharmacological, with monoamine oxidase inhibitors or tricyclic antidepressants, or non-pharmacological, by contingent awakening when a REM episode begins in a dark room — with functional imaging of occipital responses to touch and hearing before and after sleep; blindfolding studies comparing the speed and extent of occipital takeover across species that differ in plasticity, such as humans versus the less-plastic Senegal galago; and re-examinations of the cross-species plasticity–REM relationship using direct neurophysiological indices of plasticity such as paired associative stimulation, long-term potentiation or dendritic spine density instead of behavioral proxies. Notably, classical experimental work borders this question without settling it: in kittens during the developmental critical period (the early window of heightened plasticity), suppressing REM sleep or selectively eliminating PGO waves during monocular deprivation (closing one eye) amplifies the shrinkage of deprived-eye cells in the lateral geniculate nucleus (Shaffery et al., 1999) — REM-related phasic activity buffers visual-pathway plasticity during development — but no study has tested the adult, cortical, cross-modal takeover that DAT posits, and a later study found REM deprivation does not rescue developmental forms of long-term potentiation (a laboratory model of synaptic strengthening) in mature rat visual cortex (Shaffery and Roffwarg, 2003). A recent mechanistic review (Knopper and Hansen, 2023) argues via the locus coeruleus (a brainstem nucleus producing noradrenaline, involved in both plasticity and REM regulation) that DAT remains unproven and difficult to test, with recent large studies of aging partially conflicting with its original support. Meanwhile new unitary theories of REM function continue to appear (an alertness-based 'sentinel' theory; a 'dummy-model of the world' simulation framework), none causally adjudicated — confirming that the function question is genuinely open.

Problem Statement

Determine whether the defensive activation theory is correct: does the circuitry underlying REM sleep — the PGO-wave-driven visual hallucinations of dreaming — amplify visual-cortex activity periodically throughout the night, defending the visual cortex's cortical territory against takeover by the other senses during the recurring visual deprivation that planetary rotation imposes (roughly 12 hours per cycle) while touch, hearing and smell remain active? The broader question of what benefit — if any single benefit — REM sleep and its vivid visual dreaming provide is the motivating background: the field has not causally settled it, and new candidate theories continue to address it without decisive experiments. The open target is the causal testing of this specific hypothesis, whose published support is purely correlational and whose authors limit its reach: the theory 'intends to explain the biological reason underpinning visual dreaming; it does not seek to directly explain the numerous other features of REM sleep'. The target is accordingly the causal adjudication of this account of visual dreaming, not a field-wide verdict over every rival account of REM function. Deciding whether the theory is correct turns on what the existing evidence supports and on which outcomes of the three direct tests below, and of the OPGO prediction, would confirm or refute it.

The direct tests the theory specifies, and which remain unperformed, are: (i) a causal test of whether reducing or eliminating REM sleep accelerates cross-modal takeover of the visual cortex, with functional imaging of occipital responses to touch and hearing before and after sleep; (ii) a test of whether the speed and extent of occipital takeover after blindfolding scales with a species' plasticity, comparing species differing in plasticity with functional imaging — comparative data of this kind currently exist only in humans; and (iii) a re-examination of the cross-species plasticity–REM relationship using direct neurophysiological indices of plasticity rather than behavioral proxies, carefully tracking developmental stage. Belonging to the same hypothesis is the theory's further untested anatomical prediction of a threshold-monitoring feedback loop — axons projecting from visual cortex back to the pons, directly or indirectly, which would make these occipito-pontine-geniculo-occipital (OPGO) waves rather than a one-way pontine directive. The theory's remaining speculations carry explicitly non-decisive weight and are context rather than components of the target: of elephant REM sleep on poorly lit nights the authors write that 'a positive result would buttress our hypothesis; the absence of a correlation would simply suggest that the amount of REM sleep may not be responsive to immediate environmental conditions', and beyond primates they remark only that 'further research will be required to understand the strength of the present hypothesis across animal species, including species that are nocturnal or crepuscular'.

These components jointly address the single objective — explaining why REM sleep exists and why dreams are so visual — and the evidence that would settle it is causal evidence about the visual cortex, which the existing correlational support does not supply.

References

  1. Eagleman DM, Vaughn DA (2021). The Defensive Activation Theory: REM Sleep as a Mechanism to Prevent Takeover of the Visual Cortex. Frontiers in Neuroscience 15:632853. DOI: 10.3389/fnins.2021.632853. https://doi.org/10.3389/fnins.2021.632853
  2. Eagleman DM, Vaughn DA (2020). The Defensive Activation theory: dreaming as a mechanism to prevent takeover of the visual cortex. bioRxiv preprint. DOI: 10.1101/2020.07.24.219089. https://doi.org/10.1101/2020.07.24.219089
  3. Kirov R (2013). REM sleep and dreaming functions beyond reductionism. Behavioral and Brain Sciences 36(6):621-622. DOI: 10.1017/S0140525X13001362. https://doi.org/10.1017/S0140525X13001362
  4. Llewellyn S (2013). Such stuff as dreams are made on? Elaborative encoding, the ancient art of memory, and the hippocampus. Behavioral and Brain Sciences 36(6):589-607. DOI: 10.1017/S0140525X12003135. https://doi.org/10.1017/S0140525X12003135
  5. Shaffery JP, Roffwarg HP, Speciale SG, Marks GA (1999). Ponto-geniculo-occipital-wave suppression amplifies lateral geniculate nucleus cell-size changes in monocularly deprived kittens. Developmental Brain Research 114(1):109-119. DOI: 10.1016/S0165-3806(99)00027-9. https://doi.org/10.1016/S0165-3806(99)00027-9
  6. Shaffery JP, Roffwarg HP (2003). Rapid eye-movement sleep deprivation does not 'rescue' developmentally regulated long-term potentiation in visual cortex of mature rats. Neuroscience Letters 342(3):196-200. DOI: 10.1016/S0304-3940(03)00279-9. https://doi.org/10.1016/S0304-3940(03)00279-9
  7. Knopper RW, Hansen B (2023). Locus coeruleus and the defensive activation theory of rapid eye movement sleep: A mechanistic perspective. Frontiers in Neuroscience 17:1094812. DOI: 10.3389/fnins.2023.1094812. https://doi.org/10.3389/fnins.2023.1094812
  8. Horton CL (2023). The neurocognition of dreaming: key questions and foci. Emerging Topics in Life Sciences 7(5):477-486. DOI: 10.1042/ETLS20230099. https://doi.org/10.1042/ETLS20230099
  9. Parrino L, Rosenzweig I (2024). The futuristic manifolds of REM sleep. Journal of Sleep Research 34(2):e14271. DOI: 10.1111/jsr.14271. https://doi.org/10.1111/jsr.14271
  10. Spinassi RB (2026). The Sentinel Sleep Theory: Unweaving the biological function of REM sleep. Sleep Medicine: X 11:100186. DOI: 10.1016/j.sleepx.2026.100186. https://doi.org/10.1016/j.sleepx.2026.100186
  11. Censi ST, Granzotto A, Sensi SL (2026). REM sleep as a dummy-model of the world: A theoretical framework. Progress in Neurobiology 263:102939. DOI: 10.1016/j.pneurobio.2026.102939. https://doi.org/10.1016/j.pneurobio.2026.102939
  12. Naik Olson RK, Wichlinski L (2026). Brain and Environment Interaction: Neuroplasticity of the Visual Cortex and the Roles of Sensory Deprivation, Sleep, and Dreams. Psychiatry 89(3):293-306. DOI: 10.1080/00332747.2026.2680659. https://doi.org/10.1080/00332747.2026.2680659