Why Are Dreams So Weird? The Neuroscience of Dream Bizarreness

By Ron van Cann · May 2026 · 7 min read

You are in your childhood home, which is also somehow a hotel, and you have an exam in twenty minutes for a subject you never studied — but that is fine because you can fly, and the exam proctor is your old boss who is also, somehow, a horse. None of this strikes you as unusual until you wake up.

Dreams are bizarre in a way that is qualitatively different from other unusual experiences. It is not just that improbable things happen — it is that you do not notice they are improbable. The flying is accepted as matter-of-fact. The person who is simultaneously your mother and a stranger raises no internal alarm. The scene that shifts without transition is simply the new scene.

This is not a failure of memory or a quirk of the way dreams are stored. It is a feature of the specific neural state that produces dreaming. The weirdness of dreams has a precise explanation.

---

The Prefrontal Cortex: The Monitor That Goes Offline

The most important fact about why dreams are strange is this: the prefrontal cortex — the brain region most responsible for logical monitoring and reality-testing — is significantly suppressed during REM sleep.

The prefrontal cortex performs several functions that are critical for the coherence of waking experience:

• It continuously evaluates the plausibility of thoughts and perceptions ("this doesn't add up")
• It inhibits implausible associations before they reach conscious awareness
• It maintains the distinction between internally generated content and external reality
• It tracks narrative coherence and flags when a story jumps without explanation
• It provides the metacognitive capacity to think about thinking — to recognise that you might be wrong, that something might be a dream, that a conclusion is not following from the evidence

During REM sleep, prefrontal activity is dramatically reduced. EEG studies show decreased delta and beta activity in prefrontal regions during REM. Neuroimaging confirms the relative deactivation.

The result: the monitor is offline. The dreaming brain generates imagery, associations, and narratives — but the system that would flag impossibilities as impossible, transitions as incoherent, or identities as inconsistent is not running. The dream is accepted as reality precisely because the machinery for evaluating it as reality is suppressed.

---

The Hyperassociative Brain

Beyond the suppression of monitoring, REM sleep places the brain in a distinctive hyperassociative mode — a state in which connections between remotely related concepts and memories are more easily activated than during waking.

In waking cognition, activation spreads along well-established associative pathways — from "dog" to "pet," to "fur," to "walking." Remote associations ("dog" → "hot dog" → "summer" → "the summer I turned twelve") exist but are less readily activated under normal cognitive conditions.

During REM sleep, this constraint relaxes. Activation spreads more broadly and less predictably — concepts that share emotional resonance, visual features, or distant semantic connections activate each other more readily than those that are causally or temporally related.

This is the neural basis of the surreal, free-associative quality of dream content. The dream does not follow the logical pathways of waking narrative. It follows associative and emotional pathways — which is why a dream about being late for work can transition, with its own internal logic, to a scene on a beach from ten years ago, to a conversation with someone now deceased, to a completely impossible physical environment.

The connections are not random — they have internal logic. But the logic is associative rather than causal.

---

Activation-Synthesis: The Best Story the Brain Can Tell

The activation-synthesis model of dreaming, proposed by neurophysiologists Allan Hobson and Robert McCarley in 1977, offers a structural account of where dream content comes from that helps explain its bizarreness.

In this model, the brainstem generates periodic bursts of neural activation during REM sleep (the "activation") — somewhat randomly, in the sense that they are not driven by external sensory input. The cortex, receiving these activation signals, attempts to synthesise them into coherent experience by drawing on memory, expectation, and stored schemas (the "synthesis").

The result is "the best story the brain can construct from the available signals" — a narrative built from memory fragments, emotional states, and associative connections, shaped by whatever activations arrive from below. Since the activations are not systematically coherent, the story the cortex builds will have gaps, juxtapositions, and discontinuities that a waking narrative would not.

The dream's bizarreness, in this model, is the residue of the synthesis process — the seams where the brain has done its best to connect signals that don't have a logical connection, using the only tools available to it: memory and association.

---

Why We Don't Notice the Impossibilities

One of the most striking features of dream experience is not just that impossible things happen, but that we do not notice they are impossible while they are happening.

Flying feels natural. The person who is your colleague and your childhood friend simultaneously raises no inconsistency-flag. The house that is also a spaceship is simply where you are.

This is metacognitive suppression — the failure of the metacognitive capacity that would, in waking life, recognise the experience as implausible.

Metacognition — thinking about thinking, monitoring your own mental states — depends heavily on the prefrontal and frontal systems that are suppressed during REM. In waking, metacognition is what lets you think: "Wait, this doesn't make sense" or "I might be wrong about this" or "Is this real?"

Without functional metacognition, there is no comparison available. The dream is not measured against waking reality because waking reality is not accessible from inside the dream. The dream is the only reality at hand, and without a monitoring system that would evaluate its consistency with known physical and logical law, it is accepted as such.

This is why the strangeness of a dream is typically only apparent in retrospect — once you have woken up and waking metacognition has been restored. During the dream, there is nothing to notice.

---

Specific Types of Bizarreness and Their Neural Roots

The different varieties of dream bizarreness have specific neural correlates:

Identity Fluidity

People who are simultaneously two different people, or who shift identity mid-scene. This reflects reduced activity in the temporal-parietal junction (TPJ) — a region involved in distinguishing self from other, maintaining stable person representations, and representing the perspective of specific individuals. Reduced TPJ engagement during REM allows identity categories to blur.

Scene Discontinuity

Abrupt transitions between unrelated settings without narrative continuity. The hippocampus, normally responsible for constructing temporally ordered episodic narratives, is operating in a replay-and-recombination mode during sleep — drawing on memories that share associative or emotional features regardless of their temporal relationship. The discontinuities reflect the seams between recombined memory fragments.

Physical Impossibility

Flight, walking through walls, impossible scale, bodies that don't obey physics. These reflect the disengagement of the parietal spatial processing systems from normal physical constraint when there is no real sensory input grounding the simulation. The brain's spatial model runs unconstrained.

Emotional Non-Sequitur

Intense fear in response to something trivially unthreatening; profound joy without apparent cause. The amygdala — active during REM at levels comparable to waking — drives emotional tone without the prefrontal modulation that would normally ensure emotional responses are proportionate to their triggers. The emotion is real and intense; its attachment to specific content is governed by association rather than rational evaluation.

---

Why Some Dreams Are More Coherent

Not all dreams are maximally bizarre. Some are relatively ordinary, nearly boring. And lucid dreams — in which the dreamer is aware they are dreaming — often have a notably more coherent character than ordinary REM dreams.

The reason is directly predicted by the prefrontal suppression model: lucid dreaming is associated with partial reactivation of prefrontal regions during REM sleep.

Neuroimaging of lucid dreamers shows increased activation in prefrontal and frontal areas compared to non-lucid REM. The metacognitive monitoring that is suppressed in ordinary dreaming has, in lucid dreaming, partially come back online — which is precisely what allows the lucid dreamer to recognise "I am dreaming."

This reactivation also produces a more coherent dream experience: the lucid dreamer can evaluate what is happening, notice inconsistencies, and (in skilled practitioners) exercise some control over the content. The weirdness is still present — the brain is still in REM — but the monitor is running again.

---

The Dream as Meaning-Making

The hyperassociative, prefrontally suppressed dreaming brain is not simply malfunctioning. It is in a genuine cognitive state — one that is different from waking, with different strengths and weaknesses.

The hyperassociativity of REM sleep may be useful precisely because it connects material that waking cognition would not connect: finding relationships between distant experiences, combining emotional and semantic material across memory boundaries. This is likely why REM sleep is associated with creative insight and the integration of emotionally significant material.

The bizarreness of dreams is, in this sense, the cost of admission: the hyperassociativity and reduced monitoring that allow the dreaming brain to make unexpected connections also produce the flying horses and identity-fluid strangers and scenes that shift without reason.

The weird dream is the brain doing something genuinely useful — just not in a way that maps onto waking logic.

---

The Hypnos app supports recording dreams in the moment of waking — before the prefrontal monitoring that makes the dream seem strange has had time to overwrite the dream experience itself.