Why Do We Dream? The Science Behind Dreams, Sleep, and the Human Brain

What Is a Dream?

Every night, as contact with the outside world fades, the human brain becomes capable of generating a world of its own.

It can create places, people, conversations, sensations, emotions, and unfolding events, drawing on memories, learned models, bodily signals, and ongoing neural activity. Although the body remains largely still, the dreamer may walk through unfamiliar cities, return to a childhood home, speak with someone who died years ago, or experience fear and happiness with an intensity that feels entirely real.

A dream is a conscious experience that occurs during sleep. It may include images, sounds, thoughts, emotions, bodily sensations, movement, and complete narratives. Some dreams consist of only a few disconnected impressions. Others unfold as detailed stories that seem to last for hours, with recognizable characters, changing locations, and sequences of events.

Dreams are predominantly visual for most sighted people, but they are not limited to vision. They may include voices, music, touch, pain, balance, movement, taste, or smell. External sensory input is greatly reduced during sleep, but it is not eliminated completely. Sounds, temperature, discomfort, breathing difficulties, and signals from within the body can sometimes enter a dream and become part of its story.

For neuroscience, dreaming is a striking example of the brain’s ability to generate conscious experience without continuous guidance from the external world. The environments experienced in dreams do not exist around the sleeper, yet they can feel spatially complete, emotionally convincing, and temporarily indistinguishable from waking reality.

Understanding how the brain produces these experiences, however, is not the same as understanding why they occur.

Researchers have identified sleep stages, brain networks, neurochemical conditions, and memory processes associated with dreaming. Yet the central question remains unresolved: does dreaming perform a specific biological function, several overlapping functions, or no independent function at all?

Dreams have been linked to memory consolidation, emotional processing, learning, social cognition, creativity, predictive simulation, and consciousness. The evidence that these processes occur during sleep is often strong. The evidence that the subjective experience of dreaming is itself necessary for them is far less conclusive.

Modern dream science therefore does not offer one final answer. It presents a collection of partially supported explanations, each describing a different part of a much larger phenomenon.

Why Dreams Are So Difficult to Study

Dream research faces a problem that most areas of neuroscience do not.

Dreams are private experiences.

Scientists can measure electrical brain activity with electroencephalography, observe changes in blood flow with functional magnetic resonance imaging, examine metabolic activity with positron emission tomography, and monitor eye movements, muscle tone, breathing, and heart rate throughout the night.

These methods reveal what the brain and body are doing. They do not directly reveal what the sleeper is experiencing.

Researchers have made limited progress in decoding broad categories of visual imagery from brain activity and predicting whether someone is likely to have been dreaming. No current technology can reconstruct a complete dream as it was subjectively experienced, with its full sequence of images, dialogue, emotion, and meaning.

The most direct method remains surprisingly simple:

Wake the dreamer and ask what was happening.

This means that much of dream science depends on retrospective reports. By the time a person begins speaking, parts of the experience may already have disappeared. Separate scenes may blend together. Events may be rearranged, and the waking mind may unintentionally add connections that were not present during the dream.

Researchers must therefore distinguish between dreaming and remembering a dream. They are related but distinct processes.

Someone who says, “I never dream,” may experience less dream recall rather than less dreaming. A dream can occur without being successfully transferred into a form of memory that survives awakening.

A 2025 study followed 217 healthy adults and found that morning dream recall was associated not only with sleep patterns but also with a person’s attitude toward dreams and tendency to experience mind-wandering during the day. The probability of reporting a dream depended partly on how attention, sleep, and memory interacted around the moment of awakening.

The difficulty goes deeper. Dream reports are usually verbal, while the original experience may be visual, emotional, bodily, or difficult to describe. Two people may use very different language for similar experiences, while similar descriptions may conceal important differences in what was actually felt.

Dream researchers are therefore studying two connected phenomena at once: the conscious experience during sleep and the waking reconstruction of that experience.

Before the Sleep Laboratory: Freud and Jung

Long before scientists could measure the sleeping brain, dreams were interpreted through religion, mythology, philosophy, and medicine. They were treated as messages from gods, predictions of future events, communications from the dead, or signs of physical and spiritual imbalance.

At the beginning of the twentieth century, Sigmund Freud placed dreams at the centre of a psychological theory of the unconscious.

In The Interpretation of Dreams, published with a 1900 date, Freud argued that dreams were not meaningless mental noise. He believed they represented disguised expressions of wishes, conflicts, and impulses that could not be accepted openly during waking life.

Freud distinguished between the manifest content of a dream the people, places, and events the person remembered and its supposed latent content, the unconscious psychological meaning hidden beneath the narrative.

He proposed that the sleeping mind transformed latent material through processes such as condensation, displacement, and symbolization. Several people might be merged into one dream character. Emotional importance might be transferred from a threatening subject to a safer one. A conflict might appear through indirect imagery rather than literal representation.

Freud’s work was historically important because it treated dreams as connected to personal experience rather than as supernatural messages or meaningless accidents. His ideas influenced psychology, psychotherapy, literature, cinema, and popular culture throughout the twentieth century.

His detailed theory, however, is not considered scientifically established.

There is no reliable experimental method for showing that a particular image has one fixed unconscious meaning across different people. A snake, a house, or falling may carry very different associations depending on the dreamer’s history, culture, and immediate circumstances. Interpretations can often be adjusted after the fact, making them difficult to disprove.

Modern evidence does support a broader idea that Freud emphasized: dreams frequently reflect emotions, relationships, concerns, and memories from waking life. But this continuity does not validate universal symbol dictionaries or the claim that dreams are primarily disguised wish fulfilments.

Carl Gustav Jung later developed a different psychological account. He agreed that dreams could be meaningful, but he rejected Freud’s strong emphasis on repressed sexual and aggressive wishes.

Jung viewed dreams as part of a compensatory process. He believed they could reveal aspects of the personality that had been neglected by conscious thought, helping the individual move toward greater psychological balance.

He also proposed the collective unconscious, a shared and inherited layer of the psyche containing recurring patterns known as archetypes. Figures such as the Hero, the Mother, the Shadow, and the Wise Old Man were thought to appear across dreams, myths, and religious traditions because they reflected universal structures of the human mind.

Jung’s ideas remain influential in psychotherapy and the humanities, but concepts such as the collective unconscious are difficult to define, measure, and test experimentally. They are not accepted as established mechanisms within mainstream neuroscience.

Freud and Jung changed how modern culture thought about dreams. But the transformation of dream research into an experimental science required a biological discovery.

The Discovery of REM Sleep

In 1953, Eugene Aserinsky and Nathaniel Kleitman reported recurring periods of rapid eye movement during sleep.

These episodes were accompanied by physiological changes and patterns of brain activity that differed from other parts of the night. People awakened during them frequently reported vivid dreams. The state eventually became known as rapid eye movement sleep, or REM sleep.

The discovery changed sleep science.

Sleep could no longer be treated as one uniform condition in which the brain simply became inactive. It consisted of repeatedly changing states with distinct patterns of neural activity, eye movement, muscle tone, breathing, and cardiovascular function.

During REM sleep, the brain can display activity that resembles wakefulness, while most major skeletal muscles are strongly inhibited. This temporary reduction in muscle tone, known as REM atonia, limits the extent to which dreamed movements are physically acted out. The body is not completely motionless: the eyes move, breathing varies, and small twitches can occur.

William Dement and other researchers later helped establish that sleep unfolds through repeating cycles. Across the night, NREM and REM sleep alternate in cycles lasting roughly ninety minutes, although their exact duration varies between individuals and across the night.

Early sleep cycles contain more deep NREM sleep and relatively brief REM periods. Later cycles contain less deep sleep and progressively longer REM episodes. This helps explain why vivid dreams are often reported near morning, when REM sleep is more abundant and awakening may occur directly from a dream.

For several decades, dreaming and REM sleep were treated as nearly equivalent.

That conclusion did not survive.

People awakened from NREM sleep also reported conscious experiences, including imagery, thoughts, emotions, and occasionally complex narratives. At the same time, some awakenings from REM sleep produced no remembered dream report.

Dreaming and REM sleep are therefore strongly associated, but they are not identical. Brain lesions can impair dreaming without eliminating REM sleep, and dream experiences can arise outside REM. Mark Solms’s neurological work was especially influential in demonstrating that the mechanisms controlling REM sleep and those supporting dreaming can be dissociated.

REM sleep is now understood as a state that strongly favours vivid, emotional, perceptually rich, and narrative dreams. It is not the only state in which dreaming can occur.

Sleep Architecture and the Character of Dreams

NREM sleep is divided into stages ranging from light sleep to deep slow-wave sleep.

During deep NREM sleep, slow electrical activity becomes prominent, responsiveness to the environment decreases, and several restorative, regulatory, and memory related processes are supported. REM sleep has a different physiological profile, including wake-like cortical activation, rapid eye movements, irregular autonomic activity, and reduced muscle tone.

Reports obtained from REM and NREM sleep differ on average.

REM dreams tend to be longer, more visual, more emotional, and more narratively complex. They often include sudden location changes, improbable events, social interactions, and intense feelings.

NREM experiences are, on average, more likely to resemble thoughts, fragments, or direct extensions of recent concerns. However, this is a statistical tendency, not a strict boundary. Detailed dreams can occur in NREM sleep, and short, uneventful reports can follow REM awakenings.

Dream quality depends on more than the label of the sleep stage. It can be influenced by the precise moment of awakening, local brain activity, the depth and stability of sleep, how much time has passed since an experience occurred, and whether the dream was successfully remembered.

This shifted the scientific question. Instead of asking only which stage produces dreams, researchers began asking which patterns of brain activity are associated with the presence and content of conscious experience during sleep.

The Neural Signature of Dreaming

In 2017, Francesca Siclari and her colleagues used high density EEG to investigate neural activity immediately before awakenings from both REM and NREM sleep.

Participants were repeatedly awakened and asked whether they had been experiencing anything. When a dream was reported, they were also asked to describe its content.

Dreaming was associated with reduced low frequency activity in a collection of posterior cortical regions. These areas are involved in visual, spatial, and sensory processing and became known in this context as the posterior hot zone.

High-frequency activity in more specific regions was associated with particular types of dream content. Reports involving faces, movement, speech, or spatial environments corresponded with activity in areas involved in processing similar information during wakefulness.

By monitoring the posterior regions in real time, the researchers could predict the presence or absence of dream experience above chance, including during NREM sleep.

This does not mean that scientists discovered one isolated “dream centre.” Dreaming involves distributed networks, and the importance of different brain regions remains debated.

The study did show that global sleep stage alone is insufficient. Parts of the brain can be in different functional conditions at the same time. Conscious experience may depend on whether specific cortical regions reach a state capable of supporting perception-like activity, even while the person remains behaviourally asleep.

What Happens Inside the Dreaming Brain?

Functional imaging studies show that REM sleep involves an unusual pattern of activation and deactivation.

Emotionally and motivationally important regions, including parts of the amygdala and limbic system, often display increased activity. This may contribute to the emotional intensity of dreams, in which relatively ordinary situations can feel terrifying, exhilarating, humiliating, or deeply significant.

The hippocampus and connected memory systems may contribute fragments of recent and older experience. These fragments are rarely replayed as exact recordings. A familiar person may appear in the wrong house, an adult relationship may unfold inside a childhood school, or several periods of life may be combined within one scene.

Visual association regions can remain active while the primary systems that receive information from the outside world are relatively disconnected. This helps explain how internally generated images can appear vivid despite the absence of an external scene.

Regions involved in autobiographical thought, social interpretation, and self-related processing may also contribute to the personal character of dreams.

At the same time, parts of the frontoparietal control system involved in working memory, deliberate planning, and critical evaluation are often less active than during wakefulness.

This imbalance offers one explanation for the weak reality testing of ordinary dreams.

You may travel instantly from one country to another, accept that two incompatible locations are the same place, or speak calmly with someone you know is dead. The dream does not necessarily lack structure, but it is being produced under different cognitive constraints.

Emotion, association, and imagery exert greater influence. Sustained logical control and reflective awareness exert less.

The strange quality of dreams therefore does not require the brain to become chaotic. It may result from familiar neural systems operating in an unfamiliar balance.

The Neurochemistry of Dreams

The chemical environment of REM sleep differs sharply from wakefulness.

Acetylcholine-related activity is relatively high, supporting cortical activation and internally generated perceptual experience. By contrast, neurons that release serotonin and noradrenaline become much less active.

Noradrenaline is involved in vigilance, attention, and responses to stress. Its low level during REM sleep has inspired theories suggesting that emotionally significant memories may be reactivated under conditions in which the brain is less dominated by immediate defensive responses.

One influential model proposes that sleep may help preserve information while gradually reducing some of its emotional intensity. This is sometimes summarized as “sleep to remember, sleep to forget.”

It remains a hypothesis rather than an established universal rule. Studies examining REM sleep and emotional reactivity have produced mixed findings, and some emotional memories may be strengthened rather than weakened.

Dopamine related motivational systems may also be relevant. Solms observed that damage to particular forebrain regions could eliminate reported dreaming even when REM sleep remained present. This suggested that dreaming depends not only on the brainstem systems that regulate REM sleep but also on forebrain networks involved in motivation, salience, and goal directed experience.

Dreaming cannot be explained by one neurotransmitter. It emerges from interactions among acetylcholine, serotonin, noradrenaline, dopamine, GABA, glutamate, and the wider systems controlling sleep, arousal, memory, perception, and emotion.

Activation-Synthesis: Dreams as the Brain’s Interpretation of Internal Activity

In 1977, J. Allan Hobson and Robert McCarley proposed the Activation-Synthesis Hypothesis.

Their model challenged the idea that dreams begin as disguised psychological messages.

They argued that REM sleep produces internally generated activity, particularly through brainstem mechanisms, which activates the cerebral cortex. The cortex then attempts to interpret and organize that activity using memories, emotions, perceptual systems, and its general tendency to construct coherent experience.

Dreams, in this model, are a form of synthesis.

The brain receives activity that is not being directed by a stable external environment and turns it into scenes, characters, movement, and narrative. Sudden transitions and impossible combinations arise partly because the information being organized does not follow the structured sequence of normal sensory input.

The theory is often simplified into the claim that dreams are the brain’s interpretation of “random signals.” That description can be misleading. Internally generated activity is not necessarily meaningless or completely random. It is constrained by brain anatomy, emotional priorities, memory networks, and the sleeper’s personal history.

Hobson and McCarley did not need to claim that dream narratives contained no meaning. Their central point was that meaning could be constructed during the brain’s attempt to synthesize internally generated activation rather than being encoded in advance as a disguised message.

Activation-Synthesis helped move dream research away from unrestricted symbolic interpretation and toward testable brain mechanisms.

It was not a complete theory. Dreams often show long term personal consistency, emotional relevance, and structured themes that cannot be explained adequately by undirected activation alone.

The AIM Model

Hobson and his colleagues later developed a broader framework known as the AIM Model.

AIM describes conscious states through three dimensions:

Activation refers to the overall level and pattern of brain activity.

Input–output gating describes how strongly the brain is connected to external sensory information and voluntary motor action.

Modulation refers to the neurochemical conditions under which neural processing occurs.

Wakefulness, NREM sleep, REM sleep, anaesthesia, and other altered states occupy different positions within this multidimensional space.

During ordinary wakefulness, brain activation is high, sensory input is extensive, voluntary movement is available, and monoaminergic systems are active. During REM sleep, activation can also be high, but external input and motor output are strongly restricted, while the neurochemical environment changes.

The model explains dreaming not as an isolated switch but as a conscious state produced by a particular configuration of brain activation, sensory disconnection, motor inhibition, and chemical modulation.

AIM is a descriptive framework rather than a final explanation for the evolutionary purpose of dreams. Its importance lies in showing that dreaming belongs within a wider biology of consciousness.

The Continuity Hypothesis: Why Dreams Feel Personal

If brain physiology explains how dreams become possible, it does not fully explain why their content is so personal.

The Continuity Hypothesis proposes that dreams reflect ongoing concerns, emotions, relationships, interests, and experiences from waking life.

This does not mean that dreams replay the day literally.

A person worried about an examination may dream about being unprepared for a performance, missing a train, or being unable to find the correct room. The imagery changes, but the underlying concern may remain recognizable.

Dreams frequently contain family members, partners, friends, workplaces, schools, unfinished tasks, social conflicts, fears, and personal goals. People who spend much of their time in particular environments or activities tend to incorporate related themes into their dreams.

Continuity can be emotional, conceptual, social, or perceptual. A dream may reproduce the feeling of an event without reproducing the event itself.

G. William Domhoff’s neurocognitive theory builds on large collections of dream reports. His work argues that recurring dream characters and themes often correspond to stable waking concerns, relationships, and conceptions.

From this perspective, dreams do not require fixed symbolic codes. Their meaning comes from continuity with the dreamer’s cognitive and emotional life.

The same image may mean different things to different people because it belongs to a different personal network of memories and associations.

Domhoff also connects dreaming with neural systems involved in imagination, autobiographical memory, and internally directed thought, including parts of the default network. Dreams may therefore be understood as a form of spontaneous cognition generated under the conditions of sleep.

Dreams, Memory, and Neural Replay

The evidence that sleep contributes to memory consolidation is strong.

The evidence that conscious dreaming is itself necessary for consolidation is much less certain.

During waking life, new information is initially encoded in networks that include the hippocampus. These memories may remain unstable and vulnerable to interference. During sleep, recently active neural patterns can be reactivated, reorganized, and gradually integrated with existing knowledge.

In a landmark 1994 study, Matthew Wilson and Bruce McNaughton recorded hippocampal neurons in rats exploring an environment. Patterns of activity associated with waking exploration reappeared during later sleep.

The sleeping brain was re-expressing information related to recent experience.

Later research found forms of replay across animals and humans. Replay does not function like watching a complete recording. Neural sequences may be compressed, reversed, selectively strengthened, or integrated with other information.

This could help explain why dreams combine familiar elements in unfamiliar ways.

A dream may contain a real person, an invented building, an emotion from a recent conflict, and a setting remembered from childhood. Each component has a source, but the resulting scene never occurred.

Some researchers propose that dreams are the conscious expression of memory reactivation and integration. Others argue that memory consolidation takes place independently and that dream experience merely reflects some of the activity accompanying it.

The distinction matters.

Improved memory after sleep does not prove that dreaming caused the improvement. A person may benefit from neural replay without remembering any dream at all.

Day Residue and the Dream-Lag Effect

Waking experiences can enter dreams on different timescales.

The day-residue effect describes the incorporation of events from the previous day. A conversation, location, person, or emotional concern may reappear that night, often in altered form.

The dream-lag effect refers to reports in which elements of an experience appear several days later, sometimes around a week after the original event.

Research suggests that personally significant and emotionally meaningful experiences may be more likely to show delayed incorporation than routine daily activities. However, dream-lag findings have not been identical across studies, participants, or sleep stages. It should be treated as a recurring research pattern rather than a universal timetable for memory processing.

Both effects support a central point: dreams do not simply reproduce waking life. The sleeping brain selects, transforms, delays, and recombines its material.

Targeted Memory Reactivation

Researchers can sometimes influence which memories are reactivated during sleep through Targeted Memory Reactivation, or TMR.

In a typical experiment, participants learn material while it is paired with a sound or smell. During later sleep, researchers quietly present the same cue without fully waking them.

Under specific conditions, the associated memory may be strengthened relative to uncued material.

In a well-known 2009 study, spatial memories linked to sounds were selectively improved when the sounds were replayed during sleep.

TMR does not allow complex new knowledge to be inserted into an unconscious mind. Its effects depend on prior learning, cue timing, sleep stage, task type, and the stability of sleep.

The relationship between TMR and dreams is also uncertain. Cued material sometimes appears in dream reports, but memory improvement can occur without conscious dream incorporation.

TMR nevertheless demonstrates that the sleeping brain remains responsive and selective. Sleep is not a passive period in which memory systems simply stop receiving information.

Dreams and Emotional Processing

Memories are rarely stored as neutral facts. Important experiences carry emotional meaning.

A breakup, bereavement, achievement, humiliation, accident, or conflict is remembered partly through the feelings attached to it. Because emotional memories frequently appear in dreams, researchers have asked whether dreaming helps regulate them.

REM sleep is particularly interesting because emotional and memory related regions can remain active while noradrenaline levels are low.

One hypothesis proposes that this combination allows difficult material to be revisited without the full neurochemical conditions of waking stress. The brain may reactivate the event while temporarily reducing the need for immediate action.

This could help separate information about what happened from the intensity of the original response.

The idea is plausible, but the evidence remains mixed.

Some studies find that sleep reduces later emotional reactivity. Others find that it preserves, reorganizes, or strengthens emotional memory. It is also unclear whether the dream experience performs the emotional work or merely accompanies underlying sleep processes.

Dreams may sometimes help emotional adaptation. They can also maintain distress.

People experiencing grief may repeatedly dream of someone who died. A person under chronic stress may dream of failure, pursuit, conflict, or loss. After trauma, dreams may reproduce fragments of danger rather than reducing their emotional force.

Dreaming is therefore not automatically therapeutic.

Nightmares and Recurring Dreams

A nightmare is more than a dream containing something unpleasant. Clinically, nightmares are vivid dysphoric dreams that produce strong negative emotion and often lead to awakening with clear recall.

Fear is common, but nightmares can also involve shame, disgust, grief, helplessness, or despair.

Occasional nightmares are widespread. They become a clinical concern when they recur, disrupt sleep, create fear of going to bed, or impair daytime functioning.

Nightmares are particularly associated with stress, anxiety, trauma, and post traumatic stress disorder, although they can also occur without an identifiable traumatic event. Nightmare disorder is recognized as a treatable sleep condition rather than a sign that a dream contains one hidden symbolic message.

Recurring dreams similarly resist simple interpretation.

A repeated dream may reflect a persistent concern, emotion, memory, or expectation. But repetition does not prove that the brain is sending a coded warning or attempting to reveal one specific repressed event.

Recurring content may persist because the underlying emotional and cognitive networks remain highly accessible. Each time sleep enters similar conditions, the brain may return to related themes.

The existence of effective treatments for recurrent nightmares also shows that dream narratives are not fixed. Techniques such as imagery rehearsal therapy can help some people rewrite and practise a less distressing version of a recurring nightmare, reducing its frequency or intensity.

The Threat Simulation Theory

In 2000, Finnish cognitive neuroscientist Antti Revonsuo proposed the Threat Simulation Theory.

The theory asks whether dreaming may have offered an evolutionary advantage by allowing the brain to rehearse dangerous situations.

Human ancestors lived in environments containing predators, physical injury, conflict, unstable social groups, and limited resources. Rapid detection of danger could determine survival.

Dreams frequently contain pursuit, attack, falling, entrapment, being lost, arriving too late, or failing to protect someone. These experiences activate genuine fear and urgency despite occurring in a simulated environment.

Threat Simulation Theory proposes that such dreams provide practice in recognizing and responding to danger without real world physical risk.

Evidence from people exposed to war, violence, and severe adversity shows that threatening environments can increase threatening dream content. This supports a relationship between real danger and simulated danger.

It does not prove that threat rehearsal is the original or primary purpose of all dreams.

Many dreams contain no clear danger. Some are peaceful, social, humorous, sexual, mundane, or difficult to categorize. Threatening dreams may represent one prominent class rather than the complete explanation for dreaming.

The theory also faces a deeper evolutionary question: even if dreams simulate danger, does that simulation improve waking behaviour enough to affect survival and reproduction? Direct evidence remains limited.

The Social Simulation Theory

Humans evolved not only by escaping threats but also by navigating relationships.

Cooperation, attachment, competition, trust, status, communication, and recognition of other people’s intentions have always been central to human survival.

The Social Simulation Theory proposes that dreams provide an environment in which social situations can be modelled and rehearsed.

Dream reports contain a large number of social characters. People converse, argue, cooperate, comfort one another, experience rejection, seek approval, protect relatives, and interact with strangers.

These scenes often involve relationships that matter in waking life.

Someone beginning a romantic relationship may dream frequently about the new partner. A parent may dream of protecting a child. A person experiencing workplace conflict may dream about authority, exclusion, or failure in an unrelated setting.

The theory does not require dreams to provide literal social training. Internally generated interaction may repeatedly activate systems used to interpret emotion, intention, and social consequence.

As with Threat Simulation Theory, evidence that dreams contain social material does not prove that social rehearsal is their evolved purpose. It may instead reflect the fact that the human brain is deeply social during both waking and sleep.

Sleep, Dreams, and Creativity

Dreams have long been associated with artistic and scientific inspiration.

Famous stories describe ideas appearing in dreams, but anecdotes cannot establish a general biological function. People tend to remember successful dream inspired insights and forget the far more numerous dreams that produce nothing useful.

Controlled research does suggest that sleep can support creativity and insight.

In a 2009 experiment, REM sleep improved participants’ ability to connect previously unrelated information and solve problems requiring remote associations compared with quiet rest and NREM sleep.

This fits the cognitive character of dreaming.

During dreams, the brain combines material more loosely. Strong executive control is reduced, remote memories can become associated, and ideas that would remain separate during deliberate waking thought may appear together.

But the distinction between sleep and dreaming remains essential.

Evidence that REM sleep improves creative performance does not prove that a remembered dream generated the solution. Creativity may benefit from neural processes occurring during sleep whether or not they become conscious dream content.

Dreams may occasionally reveal the result of those processes, but they are not guaranteed sources of accurate or useful insight.

Dreaming as Predictive Simulation

Many contemporary theories describe the brain as a prediction generating system.

During wakefulness, the brain does not passively record sensory information. It continuously generates expectations about what it is likely to see, hear, and feel, then updates those expectations using incoming signals.

Dreaming changes the balance.

External input becomes weak and unstable, while internal models remain active. Without continuous correction from the environment, the brain’s predictions can unfold more freely.

Memories, emotions, expectations, and imagined possibilities become the main material from which experience is generated.

From this perspective, a dream is not simply a replay of the past. It may be an internally generated model in which the brain explores possible situations, social outcomes, threats, and emotional responses.

The default network is relevant because it supports autobiographical memory, imagination, internally directed thought, and construction of possible events. Researchers have proposed that parts of this network contribute to both waking mind wandering and dreaming.

Predictive processing does not yet provide a complete theory of why dreams evolved. It offers a framework for understanding how the brain can generate convincing worlds when sensory evidence is reduced.

Dreams may be what perception like modelling looks like when prediction continues but external correction becomes weak.

Protoconsciousness and the Developing Brain

Hobson later proposed the Protoconsciousness Theory.

According to this view, REM sleep provides a primitive virtual model of the world that helps establish and maintain the neural systems required for waking consciousness.

The theory is especially concerned with development.

REM like active sleep occupies a much larger proportion of early life than adulthood. The developing brain must learn to coordinate perception, bodily movement, emotion, and internal models before it has extensive experience of the external world.

Internally generated activity during active sleep may help organize these systems.

In adults, dreaming may continue to exercise and maintain networks involved in constructing a coherent world and a self located within it.

The theory is difficult to test directly. Infants cannot provide dream reports, and active sleep in early development should not automatically be assumed to contain adult-like dreams.

Its importance lies in reframing dreaming as part of the development of consciousness rather than solely as a reaction to daily memories.

Lucid Dreaming: Awareness Within a Dream

A lucid dream occurs when the dreamer becomes aware that they are dreaming without immediately waking.

Lucidity and control are not the same thing.

A person may recognize that a dream is unreal while remaining unable to change it. Others can deliberately alter the setting, choose an action, speak to dream characters, or wake themselves.

Lucid dreaming was once difficult to verify because all evidence appeared to depend on reports made after waking.

In 1981, Stephen LaBerge and his colleagues showed that experienced lucid dreamers could make predetermined left right eye movements while remaining in verified REM sleep.

Most skeletal muscles are strongly inhibited during REM, but eye movements can still be performed voluntarily. The signals allowed dreamers to mark the moment at which they became lucid from inside the dream.

This provided objective evidence that reflective awareness can emerge during an ongoing dream.

Lucid dreaming is sometimes described as a hybrid of sleep and wakefulness. More recent research suggests greater caution. It may be better understood as an activated form of REM sleep in which aspects of metacognition and self awareness become unusually available, rather than a simple mixture of being asleep and awake.

Experiencing at least one lucid dream is not rare. A meta-analysis estimated that approximately 55 percent of people have experienced one at some point, while about 23 percent reported lucid dreams at least monthly. Frequent, reliable, and highly controlled lucidity remains much less common.

Researchers are investigating whether lucid dreaming could help with recurrent nightmares, motor rehearsal, and experimental communication during sleep. The field remains limited by small samples, variable definitions, and the difficulty of inducing lucidity consistently in a laboratory.

Do Animals Dream?

Animals cannot provide verbal reports, so their subjective experience during sleep cannot be confirmed directly.

Researchers instead look for physiological and behavioural evidence.

Many mammals and birds display sleep states resembling human REM and NREM sleep. During REM-like sleep, they may show rapid eye movements, reduced muscle tone, irregular breathing, and small twitches.

Dogs may move their paws or vocalize. Cats and rodents may display brief movements consistent with ongoing motor activity.

These behaviours are compatible with internally generated sleep experiences, but they do not prove that the animal is consciously dreaming.

Neural replay offers stronger but still indirect evidence.

When rats navigate an environment, hippocampal neurons represent locations and movement through space. Related activity patterns can reappear during later sleep, showing that the animal’s brain is reprocessing waking experience.

Replay is not identical to a dream. It could occur without subjective experience.

Nevertheless, the evolutionary continuity of sleep architecture, memory systems, and internally generated neural activity makes it plausible that at least some non human animals experience dream-like states.

What those experiences feel like remains unknowable with current methods.

How Blindness Changes Dream Experience

Blind people dream, but the sensory composition of their dreams depends partly on developmental history.

People who lose their sight after years of visual experience often continue to report visual dreams. Their brains retain representations of faces, colours, objects, and spatial scenes, sometimes for decades after vision is lost.

People born blind generally report a much greater contribution from sound, touch, movement, smell, taste, and spatial awareness. Their dreams can be vivid and emotionally complex without resembling the primarily visual experience common among sighted dreamers.

A 2014 study found that blindness significantly changes the sensory composition of dreams and that the age at which blindness begins matters. Blind participants reported fewer visual elements and more auditory, tactile, olfactory, and gustatory sensations.

Whether some experiences in congenitally blind dreamers should be classified as visual imagery, spatial imagery, or another form of representation remains debated. Recent work continues to examine how people describe forms, layouts, movement, and space without previous visual experience.

The broader lesson is clear.

Dreaming is not fundamentally a visual phenomenon. It is the brain’s capacity to generate conscious experience using the sensory and conceptual resources available to it.

What Science Can and Cannot Conclude

More than seventy years after the discovery of REM sleep, science knows far more about dreaming than it once did.

Researchers can identify brain activity associated with dream experience, observe memory replay, influence selected memories during sleep, communicate with some lucid dreamers, and measure how sensory history changes dream content.

They can explain why dreams are often emotional, why their narratives tolerate contradiction, why recent experiences appear in transformed forms, and why vivid dreams are especially common during REM sleep.

They still cannot identify one universally accepted function of dreaming.

The strongest conclusions concern sleep rather than dreams themselves.

Sleep supports memory consolidation, learning, emotional regulation, physiological maintenance, and cognitive performance. Dreams occur while many of these processes are active and frequently reflect their content.

What remains uncertain is whether subjective dreaming performs essential work of its own.

Several possibilities remain open.

Dreams may contribute directly to memory integration, emotional adaptation, social rehearsal, threat simulation, creativity, or the maintenance of conscious models.

Different types of dreams may serve different functions.

Dreaming may also emerge naturally from processes that are useful for other reasons. A brain that continues to reactivate memories, generate predictions, regulate emotion, and maintain perceptual systems during sleep may produce conscious experience without dreaming itself having evolved as a separate adaptation.

These explanations are not mutually exclusive.

Activation Synthesis may describe how internal activity becomes a narrative. The Continuity Hypothesis may explain why the narrative reflects personal concerns. Memory models may explain why old and new experiences are combined. Emotional theories may explain why certain themes remain intense. Simulation theories may describe how dreams explore possible situations.

Each may capture a different layer of the same phenomenon.

Dreams are therefore neither proven messages from a hidden symbolic code nor meaningless collections of neural noise.

They are structured conscious experiences generated by a sleeping brain shaped by physiology, memory, emotion, motivation, perception, and personal history.

Every night, as external sensory information becomes less dominant, the brain continues to build models of experience. It revisits fragments of the past, combines them with present concerns, and generates situations that have never occurred.

Dreams may be the conscious expression of some of this work. But current evidence does not show that they possess one single, indispensable function.

That uncertainty is not a failure of dream science.

It may be its most important conclusion.

We understand increasingly well when dreams occur and how the brain makes them possible.

The deeper question remains open:

Did dreaming evolve to perform something essential or is it what consciousness feels like when the brain continues its work without a stable external world?

Last updated: July 2026

References

Aserinsky, E., & Kleitman, N. (1953). Regularly occurring periods of eye motility, and concomitant phenomena, during sleep. Science, 118(3062), 273–274.

Carskadon, M. A., & Dement, W. C. (2017). Normal human sleep: An overview. In M. H. Kryger, T. Roth, & W. C. Dement (Eds.), Principles and Practice of Sleep Medicine (6th ed.). Elsevier.

Domhoff, G. W. (2003). The Scientific Study of Dreams: Neural Networks, Cognitive Development, and Content Analysis. American Psychological Association.

Domhoff, G. W. (2018). The emergence of dreaming: Mind-wandering, embodied simulation, and the default network. International Journal of Dream Research, 11(2), 127–134.

Freud, S. (1900/2010). The Interpretation of Dreams. Basic Books.

Hobson, J. A. (2002). Dreaming: An Introduction to the Science of Sleep. Oxford University Press.

Hobson, J. A. (2009). REM sleep and dreaming: Towards a theory of protoconsciousness. Nature Reviews Neuroscience, 10(11), 803–813.

Hobson, J. A., & McCarley, R. W. (1977). The brain as a dream state generator: An activation-synthesis hypothesis of the dream process. American Journal of Psychiatry, 134(12), 1335–1348.

Jung, C. G. (1964). Man and His Symbols. Doubleday.

LaBerge, S. (1985). Lucid Dreaming. Ballantine Books.

LaBerge, S., Nagel, L., Dement, W. C., & Zarcone, V. (1981). Lucid dreaming verified by volitional communication during REM sleep. Perceptual and Motor Skills, 52(3), 727–732.

Nir, Y., & Tononi, G. (2010). Dreaming and the brain: From phenomenology to neurophysiology. Trends in Cognitive Sciences, 14(2), 88–100.

Revonsuo, A. (2000). The reinterpretation of dreams: An evolutionary hypothesis of the function of dreaming. Behavioral and Brain Sciences, 23(6), 877–901.

Rasch, B., Büchel, C., Gais, S., & Born, J. (2007). Odor cues during slow-wave sleep prompt declarative memory consolidation. Science, 315(5817), 1426–1429.

Rasch, B., & Born, J. (2013). About sleep's role in memory. Physiological Reviews, 93(2), 681–766.

Scarpelli, S., Bartolacci, C., D’Atri, A., Gorgoni, M., & De Gennaro, L. (2025). Individual differences in dream recall: Evidence from a longitudinal study. Communications Psychology.

Siclari, F., Baird, B., Perogamvros, L., et al. (2017). The neural correlates of dreaming. Nature Neuroscience, 20(6), 872–878.

Solms, M. (1997). The Neuropsychology of Dreams. Lawrence Erlbaum Associates.

Stickgold, R., & Walker, M. P. (2013). Sleep-dependent memory triage: Evolving generalization through selective processing. Nature Neuroscience, 16(2), 139–145.

Walker, M. P. (2017). Why We Sleep: Unlocking the Power of Sleep and Dreams. Scribner.

Wilson, M. A., & McNaughton, B. L. (1994). Reactivation of hippocampal ensemble memories during sleep. Science, 265(5172), 676–679.