Auditory and Visual Hallucinations: Causes and Symptoms

Auditory and visual hallucinations arise from dozens of different causes, not just the psychiatric conditions most people associate them with. They can appear in schizophrenia, yes, but also in Parkinson’s disease, epilepsy, severe vision or hearing loss, substance use, sleep deprivation, grief, and even childhood development. The common thread across all these triggers is a brain that, for one reason or another, generates sensory experiences without a matching signal from the outside world. What those experiences feel like, how they’re triggered, and what they mean for the person having them varies enormously depending on the underlying cause.

What Auditory and Visual Hallucinations Actually Feel Like

Auditory hallucinations most often take the form of voices, sometimes a single voice making comments, sometimes multiple voices talking to each other or to the person hearing them. Less commonly, they can be non-verbal sounds: music, buzzing, knocking, or environmental noises that nobody else can hear. In schizophrenia-spectrum conditions, auditory hallucinations are by far the most common type, with lifetime rates around 64–80% of patients, while visual hallucinations occur in roughly 23–31%.¹ Tactile and olfactory hallucinations are rarer still.

Visual hallucinations range from simple flashes, shapes, and geometric patterns to fully formed images of people, animals, or complex scenes. The simple versus complex distinction matters clinically: simple visual hallucinations (flashing lights, colored blobs) tend to originate from irritation or damage to early visual processing areas, while complex ones (faces, figures, entire landscapes) involve higher-level brain regions that stitch together meaningful imagery.² Whether the person recognizes that the experience is not real varies a great deal by condition and context, and that insight is not actually the defining feature clinicians use to classify hallucinations.³

When both modalities co-occur, auditory hallucinations usually come first. Among patients who experience visual hallucinations, the vast majority have also experienced auditory ones, but the reverse is much less common. Only about a third of people with lifetime auditory hallucinations also report visual hallucinations.¹

The Brain’s Prediction Machine and Why It Misfires

One of the most influential frameworks for understanding hallucinations across different conditions centers on how the brain constantly predicts what it expects to see and hear. Your brain does not passively receive sensory input and react; it generates ongoing predictions about what should be coming in, then checks those predictions against what actually arrives. When the predictions and the incoming signals match, everything feels normal. When they don’t, the mismatch generates a “prediction error” that forces the brain to update its model of what is happening.

In psychosis, this system may become miscalibrated. Research suggests that the precision assigned to internal predictions gets distorted relative to actual sensory signals, leading the brain to treat its own internally generated expectations as if they were real sensory events.⁴ The brain essentially trusts its own guesses too much and trusts the outside world too little. This framework helps explain why hallucinations can arise in such different conditions: anything that disrupts the balance between internal predictions and incoming sensory data, whether it’s a chemical imbalance, sensory loss, or neural damage, can tip the system toward generating false perceptions.

Why People Hear Voices They Cannot Control

Auditory verbal hallucinations, the experience of hearing voices, have a specific proposed mechanism that goes beyond the general prediction model. When you talk to yourself silently (inner speech), your brain sends a signal called a “corollary discharge” from motor areas to auditory areas, essentially telling the hearing system: “This sound is coming from us, not from outside.” That signal dampens the auditory cortex’s response to self-generated speech, which is why your internal monologue doesn’t feel like someone else talking to you.

Research in neurosurgical patients has pinpointed the source of this corollary discharge to the ventral speech motor cortex, and shown that it fires before speech articulation, predicting the degree of suppression in the auditory cortex during speaking.⁵ In people with schizophrenia who hear voices, this suppression mechanism appears to be disrupted. Rather than dampening the brain’s response to inner speech, the system does the opposite: it amplifies it. This finding provides direct support for the long-standing theory that some voice-hearing experiences are essentially inner speech that the brain misidentifies as coming from an external source.⁶

Visual Hallucinations and the Problem of Too Many Models

Understanding visual hallucinations has been complicated by the sheer number of competing theories. Since 2000, at least eight distinct models have been proposed, each derived from different understandings of how the brain organizes vision. These include deafferentation models (loss of input causes the brain to fill in the gaps), reality monitoring failures (the brain confuses internally generated images with external ones), attentional network problems, and thalamocortical rhythm disruptions, among others. Researchers from each group recently agreed to work toward an integrated framework that reconciles these different approaches.⁷

What the models share is the recognition that visual hallucinations can emerge at multiple levels of the visual processing hierarchy. Damage or dysfunction low in the system (the eyes, optic nerve, or primary visual cortex) tends to produce simple hallucinations. Damage or dysfunction higher up, in the temporal and frontal regions that assign meaning to what we see, tends to produce complex hallucinations with recognizable content like faces and scenes. In Parkinson’s disease, for example, patients who hallucinate visually show a heavier burden of abnormal protein deposits across the frontal, temporal, and parietal cortex compared to Parkinson’s patients who do not hallucinate.⁸

When Losing a Sense Creates a New One

Some of the most striking hallucinations occur in people who are otherwise psychiatrically healthy but have lost significant sensory input. Charles Bonnet syndrome affects people with severe vision loss: as the visual cortex receives less and less information from the eyes, it begins generating its own imagery, sometimes vividly detailed scenes of people, animals, or patterns the person has never encountered. The leading explanation involves homeostatic plasticity, the brain’s attempt to maintain a stable level of activity. When external input drops, internal activity ramps up to compensate, but in Charles Bonnet syndrome, this compensation overshoots.⁹

The auditory equivalent is Musical Ear Syndrome, where people with significant hearing loss begin hearing music, voices, or other sounds that are not there. The proposed mechanism mirrors Charles Bonnet syndrome: reduced inhibitory signaling in the auditory cortex leads to neural hyperexcitability, and that excess activity gets experienced as phantom sound.¹⁰ People with these conditions are typically aware that what they are experiencing is not real, which distinguishes their hallucinations from psychotic ones phenomenologically, if not mechanistically. The experience can be alarming at first, and many people delay reporting it out of fear they’ll be diagnosed with a psychiatric condition.

Epilepsy and the Geography of Seizures

Seizure activity can trigger hallucinations, and the type of hallucination reliably maps onto where in the brain the seizure starts. Occipital lobe seizures produce elementary visual hallucinations: flashing lights, colored spots, and geometric shapes. Complex visual hallucinations, the kind involving formed images of people or objects, do not arise from purely occipital seizures. They require involvement of temporal lobe structures that handle higher-level visual processing.² In clinical comparisons, visual symptoms were exclusive to patients with occipital lobe epilepsy and did not appear in patients whose seizures originated in the temporal lobe alone.¹¹

Auditory hallucinations from seizures are less common but do occur, usually in temporal lobe epilepsy. These tend to be brief and stereotyped, unlike the extended conversational voices that characterize schizophrenia. This geographic mapping is clinically useful: the character of a hallucination can help neurologists pinpoint where seizure activity is originating, which matters for treatment planning.

Drugs, Serotonin, and Altered Visual Processing

Classic psychedelics like psilocybin produce visual hallucinations through a well-characterized receptor pathway. Psilocybin activates serotonin 2A receptors, and blocking those receptors with a specific antagonist completely prevents the visual effects, confirming that this receptor is the key gateway.¹² The drug appears to shift the visual cortex into a mode where spontaneous internal activity overwhelms the brain’s normal responses to actual visual input. Researchers have observed that serotonin 2A activation increases spontaneous rhythmic oscillations in visual cortex, essentially turning up the volume on the brain’s own internally generated activity.¹³

Neuroimaging under psilocybin has shown a pattern where early visual regions increase their self-inhibition (dampening incoming signals) while higher visual-association regions reduce their normal inhibitory control over those early areas. The net effect is that top-down imagery signals flow more freely while bottom-up sensory signals are suppressed, creating a brain state where internally generated images dominate.¹⁴

On the other end of the substance spectrum, alcohol withdrawal can trigger delirium tremens, a dangerous state that includes agitation, confusion, and sometimes vivid hallucinations. This affects a small percentage of people with alcohol dependence, but when it occurs, it constitutes a medical emergency.¹⁵ The underlying chemistry is different from psychedelics: chronic alcohol use suppresses excitatory brain activity, and when alcohol is suddenly removed, the brain rebounds into a hyperexcitable state that can produce hallucinations across multiple senses.

Sleep, Isolation, and Sensory Deprivation

Hallucinations at the boundary between sleep and wakefulness are remarkably common and almost always benign. Hypnagogic hallucinations (while falling asleep) and hypnopompic hallucinations (while waking up) can involve seeing figures in the room, hearing someone call your name, or feeling a presence. They are especially likely when sleep is disrupted or insufficient. In rare cases, they co-occur with sleep paralysis, the terrifying inability to move upon waking, which can intensify the hallucinatory experience.¹⁶

Extended sensory deprivation, whether from solitary confinement, prolonged isolation, or extreme environments, can also induce hallucinations in otherwise healthy people. The neurobiological changes involve shifts in dopamine signaling, stress hormone regulation, immune activation, and disrupted circadian rhythms. Duration matters enormously: brief voluntary sensory restriction (like a float tank session) rarely causes problems, while prolonged involuntary isolation can produce lasting vulnerability to hallucinations, depression, and cognitive impairment.¹⁷ Individual factors like baseline anxiety, prior psychiatric history, and personal resilience strongly influence who develops hallucinations under these conditions and who does not.

Trauma, Dissociation, and Hearing Voices

Childhood trauma, sexual abuse, combat exposure, and other severe stressors are linked to hallucinations through a pathway that does not require a psychotic illness. Dissociation, a psychological process in which aspects of experience become disconnected from conscious awareness, appears to mediate some of this relationship. Researchers distinguish between two forms: “detachment” (feelings of being unreal or disconnected from yourself) and “compartmentalization” (where aspects of mental function become inaccessible). The detachment type, which is more closely tied to trauma, appears to be the form more strongly associated with voice-hearing.¹⁸

That said, dissociation does not explain all hallucinations in trauma-exposed individuals.¹⁹ Multiple pathways appear to connect traumatic experience to voice-hearing, including changes in stress-response systems, altered threat processing, and the brain’s tendency to interpret ambiguous internal signals as external threats. This is an area where the boundary between “psychiatric” and “normal” hallucination blurs: some trauma survivors who hear voices never meet criteria for schizophrenia or any psychotic disorder.

Children Hear and See Things More Often Than Adults

Hallucinations in children are surprisingly common and usually transient. Auditory hallucinations occur in roughly 12–13% of children and adolescents, compared to about 6% of adults and under 5% of elderly people.²⁰ Most childhood hallucinations resolve on their own and are not associated with psychotic illness. They are more common during periods of stress, sleep disruption, or high fever, and in children with anxiety or mood disorders.

The challenge for parents and clinicians is distinguishing the benign developmental kind from those that signal something more serious. Persistence (lasting more than a few months), distress, commanding content (voices telling the child to do something harmful), and accompanying disorganization in thinking or behavior are all red flags that warrant further evaluation. A child who occasionally hears their name called when falling asleep is in very different territory from a child who hears sustained critical voices throughout the day.

Shared Genetics Across Hallucination Types

There is growing evidence that susceptibility to hallucinations has a genetic component that cuts across diagnostic categories. Subclinical auditory hallucinations (the kind experienced by people in the general population who do not have a psychiatric diagnosis) share significant genetic overlap with schizophrenia and major depression.²¹ Subclinical visual hallucinations show a similar pattern of genetic correlation with schizophrenia and depression, though notably neither auditory nor visual hallucination proneness showed a significant genetic link with bipolar disorder after correcting for multiple comparisons.²¹

Perhaps most striking, auditory and visual subclinical hallucinations were almost perfectly genetically correlated with each other, suggesting that the biological predisposition to hallucinate in one modality largely overlaps with the predisposition to hallucinate in the other. Some specific genetic variants appear to confer risk for psychotic symptoms regardless of diagnosis. For example, a variant in the CACNA1C gene, which encodes a calcium channel involved in neural signaling, was associated with psychotic symptom risk across different psychiatric conditions.²² Both the dopamine and glutamate neurotransmitter systems have been implicated in schizophrenia specifically, playing distinct but interacting roles in how psychotic symptoms emerge.²³

How Culture Shapes What People Hallucinate

Culture does not just affect whether hallucinations are interpreted as illness; it affects their actual content and character. Cross-cultural comparisons have found that the rates and types of hallucinations differ across populations. Auditory and visual hallucinations occur more frequently in some African populations, while specific types of delusions show different patterns in Pakistani, Latino, and White-British groups.²⁴ In community samples (people without psychiatric diagnoses), a study comparing the Middle East and Europe found significantly higher scores for both auditory and visual hallucination phenomenology in Qatar compared to the Netherlands.²⁵

The implications go beyond academic interest. In settings where voice-hearing is culturally framed as spiritual communication or ancestral contact, the experience tends to be less distressing and may even be valued. In settings where it is immediately pathologized, the same experience becomes frightening and stigmatizing. Research suggests these cultural framings can actually influence clinical outcomes for people who develop psychotic disorders, potentially affecting how disabling the hallucinations become.²⁶ This has pushed some researchers and clinicians toward more culturally sensitive assessment approaches that do not automatically equate hallucination with disorder.

Brain Stimulation as a Treatment for Persistent Voices

For people whose auditory hallucinations resist standard medication, repetitive transcranial magnetic stimulation (rTMS) has emerged as a promising option. The technique delivers magnetic pulses to specific brain regions, typically targeting the left temporoparietal cortex, an area consistently linked to voice-hearing. A recent randomized trial found that patients receiving active rTMS showed significantly greater reductions in auditory hallucination scores compared to sham treatment, with the benefits still evident a month after treatment ended. Patients who received a stronger magnetic field within the brain network linked to voice-hearing showed greater improvement.²⁷

Even in patients whose hallucinations had resisted clozapine, a medication typically reserved as a last-line treatment, twice-daily low-frequency rTMS produced a meaningful reduction in hallucination severity, with about a third of patients achieving at least a 50% reduction and most of the rest showing partial improvement.²⁸ A meta-analysis pooling results across many trials found that rTMS outperformed sham treatment for auditory hallucinations, though the effect was modest and the authors cautioned that the result was not fully stable across different analytical approaches.²⁹ The field is still working out which patients benefit most and which stimulation parameters are optimal, but for people living with treatment-resistant voices, brain stimulation represents a genuinely new avenue that did not exist a generation ago.