Researchers at UNSW Sydney report evidence that auditory verbal hallucinations in schizophrenia-spectrum disorders may involve a breakdown in the brain’s normal ability to dampen responses to self-generated inner speech, causing internally generated thoughts to be processed more like external sounds.
A study led by psychologists at UNSW Sydney reports evidence that hearing voices in schizophrenia-spectrum disorders may be tied to a disruption in how the brain distinguishes internally generated “inner speech” from sounds coming from the outside world.
Published in Schizophrenia Bulletin, the research tested a long-debated idea in psychiatry: that some auditory verbal hallucinations (AVH) may occur when a person’s inner speech is misperceived as external speech.
Professor Thomas Whitford of the UNSW School of Psychology described inner speech as “the voice in your head that silently narrates your thoughts – what you're doing, planning, or noticing.” The researchers say that, in typical brain function, the auditory system shows a reduced response to predicted, self-generated speech-like signals. In people currently experiencing hallucinations, the study found an opposite pattern.
How the experiment worked
The team used electroencephalography (EEG) to measure participants’ brain responses while they listened to brief syllables through headphones and, at specific moments, imagined producing syllables silently.
Participants were divided into three groups:
- 55 people with schizophrenia-spectrum disorders who had experienced AVH within the past week,
- 44 people with schizophrenia-spectrum disorders who either had no history of AVH or had not experienced them recently,
- 43 healthy control participants with no history of schizophrenia.
During the task, participants were asked to imagine saying a syllable such as “bah” or “bih” while hearing a syllable played aloud. Sometimes the imagined and audible syllables matched; other times they did not.
What researchers observed
In healthy controls, EEG responses showed a reduction in early auditory processing when the imagined syllable matched the sound played aloud—an effect consistent with the brain correctly predicting and suppressing responses to expected speech-like input.
Among participants who had recently experienced AVH, the researchers observed the opposite: brain responses were stronger when imagined and heard syllables matched.
“Their brains reacted more strongly to inner speech that matched the external sound, which was the exact opposite of what we found in the healthy participants,” Whitford said.
The group without recent hallucinations showed a different pattern from both the hallucinating group and healthy controls, with responses that the researchers reported as falling between the other two groups overall.
Implications and next steps
Whitford said the findings support a theory that has been discussed for decades but has been difficult to test because inner speech is private and cannot be directly observed. The study’s results suggest an EEG-based measure of this “inner speech” prediction mechanism could be explored as a potential biological marker related to psychosis risk.
“This sort of measure has great potential to be a biomarker for the development of psychosis,” Whitford said.
The researchers said they plan further work to assess whether these brain-response patterns could help predict who may later develop psychosis—an approach that, if validated, could support earlier identification and intervention.
