Hearing Voices as a Function of the Human Brain

Hearing Voices as a Function of the Human Brain

Written on 4 September 2026.

Introduction

The human brain produces thought.

Some thought consists of images. Some consists of memories. Some consists of emotions. Some consists of words. People silently speak sentences in their minds, remember the voices of other people, imagine conversations and dream of people speaking.

There is therefore nothing mysterious about the basic fact that the organic brain can produce words without an external speaker.

Usually verbal thought is experienced simply as thought. But scientific research provides evidence for the possibility that internally generated verbal activity can sometimes acquire the perceptual character of a voice.

This gives us a simple biological starting point:

A human being who experiences a voice may be experiencing verbal activity generated by his own brain.

The existence of the experience does not prove the existence of another speaker.

Thinking in Words

Human beings can think in language.

This is generally called inner speech. A person can internally produce an entire sentence without moving his lips and without any sound entering through his ears.

For example, somebody can silently think:

I need to go outside.

The words exist as a conscious experience even though nobody physically spoke them.

Research has associated inner speech with activity in brain systems also involved in producing and processing spoken language. The left inferior frontal gyrus, including Broca's area, has been implicated in producing inner speech, while temporal regions involved in auditory and speech processing participate in the wider system.

The brain therefore contains machinery capable of both generating verbal information and representing speech auditorily.

How Does the Brain Know "I Thought That"?

An interesting biological question follows.

If the brain itself produces verbal thought, how does it know that the resulting words were self-generated?

Neuroscience has investigated a mechanism known as corollary discharge or efference copy.

When the nervous system produces an action, it can simultaneously generate a prediction concerning the sensory consequences of that action. This mechanism helps an organism distinguish events resulting from its own activity from events originating in its environment.

Corollary-discharge mechanisms are not peculiar to human beings. They have been observed throughout the animal kingdom and participate in distinguishing self-generated sensory consequences from events in the outside world.

Researchers have applied this principle to inner speech.

The proposed mechanism can be simplified as follows:

Brain generates verbal thought → brain predicts its own verbal activity → auditory system receives information that the activity is self-generated → words are experienced as one's own inner speech.

But the subjective result can be different when this coordination operates differently.

The verbal material may acquire greater apparent independence from the person thinking it.

When a Thought Has the Character of a Voice

This is not merely philosophical speculation.

A substantial body of research has investigated whether experiences described as hearing voices can involve internally generated speech being experienced as though it had another source.

Moseley, Fernyhough and Ellison reviewed evidence for inner-speech models and wrote that atypical self-monitoring may result in inner speech being experienced as external.

Research reviewed in that paper includes experiments detecting subvocalization—tiny movements of the vocal musculature associated with inner speech—while people were experiencing voices.

Another neuroscience review describes the proposed mechanism very plainly: frontal areas involved in generating inner speech communicate with temporal speech-reception areas, helping identify the resulting activity as self-generated.

This gives biological substance to an apparently strange possibility:

A person's own verbal thought can potentially be experienced with the apparent character of somebody speaking.

It Is Not Necessarily One Part of the Brain

It would be premature to attribute this entirely to the brainstem or to one anatomical structure.

The evidence instead points toward interactions among several systems.

Research has implicated:

• frontal areas associated with speech generation;
• temporal areas associated with hearing and language;
• auditory cortex;
• connections between frontal and temporal language regions;
• memory systems;
• mechanisms involved in identifying the origin of internally generated information.

The arcuate fasciculus, a white-matter pathway connecting important frontal and temporal language areas, has also been investigated.

Thus brain anatomy and connectivity matter, but current evidence does not justify reducing the phenomenon to one particular anatomical difference.

A better biological model is:

Different activity and interaction within the brain's language, auditory, memory and self-monitoring systems can influence how internally generated verbal information is experienced.

Memories Can Also Become Verbal Experiences

Not every internally generated voice-like experience necessarily has to originate from ordinary present-tense inner speech.

Memory provides another possible source.

The brain stores enormous amounts of verbal information.

A person can remember:

• something his father said thirty years ago;
• an argument;
• an insult;
• a warning;
• somebody shouting;
• a frightening event;
• a familiar person's tone of voice.

Scientific researchers have consequently also investigated whether intrusive verbal memories can contribute to voice-like experiences.

This is another completely organic possibility.

The brain does not merely manufacture new thoughts. It reconstructs old information.

A voice-like experience could therefore involve present thought, memory, auditory imagery, spontaneous neural activity or interactions among these processes.

Dreams Demonstrate the Brain's Capacity

Dreaming provides an especially simple demonstration of the principle.

A sleeping person can dream that another person says:

"Come over here."

The dreamer may hear the apparent voice, see the apparent speaker and respond to what was apparently said.

Yet no physical speaker needs to be standing beside the bed.

The organic brain generated the words.

It generated the apparent voice.

It generated the apparent speaker.

It may even generate the dreamer's own reply.

Dreams therefore demonstrate something remarkable about ordinary human biology:

The brain possesses the capacity to generate an apparently independent speaker and apparently heard speech entirely from its own activity.

The fact that related verbal phenomena can occur while awake should therefore not automatically require an external speaker.

The Experience and Its Source Are Different Questions

Suppose somebody says:

"I heard a voice saying that I should leave the house."

There are two separate questions.

The first is:

Did the person experience those words?

The second is:

Where did the words originate?

A truthful answer of "yes" to the first question provides no automatic answer to the second.

This distinction is extremely important.

The person can be completely accurate about what he experienced while being mistaken about what produced the experience.

A Voice Does Not Prove Voice-to-Skull Technology

Suppose somebody experiences a voice and subsequently concludes:

"Somebody transmitted those words electronically into my brain."

The conclusion does not follow from the experience.

The brain itself is already known to generate verbal thought, auditory imagery, dreams, memories and spontaneous mental activity.

Evidence for an external transmitter would therefore have to be independent of the subjective experience.

One would need evidence of the transmitter, signal and causal mechanism.

Otherwise:

I experienced a voice

has simply been transformed without evidence into:

Someone transmitted a voice into me.

Those are different propositions.

A Voice Does Not By Itself Prove a Spiritual Speaker

The same evidential principle applies to spiritual interpretations.

Experiencing words does not, by itself, identify God, an angel, a devil or a demon as their author.

This observation is also consistent with the King James Bible.

The Bible contains accounts in which God speaks to people. It also contains accounts of people possessed with devils.

But Scripture does not establish the modern rule:

A person hears voices in his head = the person is possessed with a devil.

In fact, some people possessed with devils in Scripture have very different manifestations.

Matthew 9:32 describes:

As they went out, behold, they brought to him a dumb man possessed with a devil.

After the devil is cast out:

And when the devil was cast out, the dumb spake: and the multitudes marvelled, saying, It was never so seen in Israel.

Matthew 12:22 similarly describes someone "possessed with a devil, blind, and dumb."

Other passages describe an unclean spirit speaking through a person. Luke 4:33–35, for example, describes a man with "a spirit of an unclean devil" crying out aloud before Jesus commands the spirit to come out.

These passages describe specific events. They do not establish internally hearing a voice as the biblical test for devil possession.

The Bible therefore gives no basis for turning the bare experience "I heard a voice in my head" into proof that a devil was responsible.

The Organic Explanation Requires the Fewest Additional Assumptions

When internally experienced words occur, one fact is already established:

There is a human brain present.

That brain already possesses the biological machinery for:

• thought;
• language;
• inner speech;
• memory;
• auditory imagery;
• dreams;
• spontaneous mental activity;
• and reconstruction of previously heard speech.

Neuroscience additionally provides models explaining how internally generated speech can acquire an apparently non-self or voice-like quality.

Therefore the organic explanation requires no hypothetical transmitter and no unidentified speaker.

It begins with machinery that is already physically present and known to generate verbal experience.

"It Wasn't My Thought"

A person may object:

"But it didn't feel like my thought."

That is important information about the character of the experience.

It is not proof of its external origin.

The entire significance of the inner-speech and self-monitoring research is that the subjective sense of authorship is itself something produced by brain processes.

An internally generated event does not logically have to carry the feeling:

"I deliberately generated this."

People already experience spontaneous thoughts constantly.

A memory can suddenly appear.

A word can suddenly appear.

A solution to a problem can suddenly appear.

A song can suddenly begin repeating internally.

A forgotten event can suddenly return to consciousness.

The absence of a conscious decision to generate something does not demonstrate that somebody else generated it.

Hearing Voices Can Therefore Be Viewed as Something a Brain Can Do

The simplest biological perspective does not begin by assigning some larger meaning to voice-hearing.

It asks what an organic brain is capable of doing.

The answer is already extensive.

A brain can think in words.

A brain can reproduce another person's voice from memory.

A brain can imagine conversations.

A brain can create speakers during dreams.

A brain can generate unwanted thoughts.

A brain can spontaneously retrieve verbal memories.

A brain can generate auditory imagery.

And scientific evidence supports models in which internally generated verbal activity can acquire the subjective character of a voice.

Consequently:

Hearing a voice can simply be something that an organic human brain is capable of producing.

The experience does not contain within itself proof of anything beyond the experience.

Conclusion

Human thought is biological activity of extraordinary complexity.

The brain does not merely receive speech. It produces language internally.

It does not merely hear voices. It remembers voices.

It does not merely observe conversations. It can construct conversations in imagination and dreams.

Scientific research into inner speech, auditory processing and corollary discharge provides a plausible biological explanation for how internally generated verbal activity can sometimes be experienced with the character of a voice.

This does not mean that science has reduced every possible instance of voice-hearing to one mechanism. Several mechanisms have been proposed, including inner speech, memory processes and spontaneous activity within auditory systems.

But one conclusion is secure:

The experience of a voice does not require an external speaker in order to exist. The organic human brain itself is capable of generating verbal experience.

Therefore hearing a voice cannot, by itself, prove who or what supposedly spoke.

It cannot by itself prove technological transmission.

It cannot by itself prove demon possession.

It cannot by itself prove divine communication.

Those are additional claims requiring additional evidence.

The biological starting point is much simpler:

A brain produces thoughts, memories, language and internally generated sensory representations. Some of those products can be experienced as voices.

Hearing Voices Does Not Determine Ability to Work

The experience of hearing voices should not be confused with an inability to function or work.

Population research indicates that voice-hearing is considerably more widespread than might commonly be assumed. Estimates vary according to definitions and methods. A systematic review and meta-analysis reported a median lifetime prevalence of hallucinations of approximately 5.8% among adults, while other studies using broader definitions of voice-hearing have produced substantially higher estimates.

This means that an experience such as hearing a voice cannot by itself tell us what somebody is capable of doing in everyday life.

Research has specifically described people who experience voices while nevertheless functioning in ordinary life. Studies comparing different groups of voice-hearers have found substantial variation in such things as frequency, emotional response, perceived control and effect upon everyday activities.

The important practical distinction is therefore:

Hearing a voice describes an experience. Ability to work describes functional capacity. They are not the same measurement.

One cannot logically infer from:

"This person experiences voices."

that:

"This person cannot work."

The person's actual ability to perform activities must be considered separately.

This distinction can also be seen in the Swedish system of sickness compensation (sjukersättning). Försäkringskassan bases eligibility upon long-term reduction of work capacity. Compensation can be granted at different levels—one-quarter, one-half, three-quarters or full compensation—depending upon the extent to which work capacity is considered reduced.

Thus even an administrative system concerned with sickness compensation ultimately has to address the practical question of what the individual can actually do. The existence of one particular subjective experience cannot answer that question by itself.

The population evidence makes the same basic point from another direction. If voice-hearing occurs among people who continue functioning in ordinary life, then voice-hearing itself cannot logically constitute evidence of inability to work.

A person may experience voices and still work, manage ordinary activities and function independently. The existence of the voice does not determine the person's functional capacity.

This fits the broader biological argument of this article. Human brains can produce different forms of internally generated verbal experience. Knowing that somebody experiences one of those forms tells us something about that person's subjective experience; it does not, by itself, tell us what that person is capable of doing.

References

AI Disclosure: Parts of this page may have been created, edited, or assisted by artificial intelligence tools (such as ChatGPT or other language models). All AI-assisted content is reviewed by a human before publication. For questions, contact the site administrator.