What This Study Is About

Many people hear an "inner voice" when they talk to themselves in their head. Others say they do not. This is auditory verbal aphantasia: the absence of an inner voice. The study asks whether a listening test can show that difference objectively, without relying only on what people say about themselves.

How They Studied It

This was an online listening experiment with English and French speakers.

Volunteers came through an online aphantasia survey, university mailing lists and online research platforms about mental imagery, with a focus on people with visual aphantasia. Of 240 who started, 80 were excluded, mostly for failing a baseline listening check. That left 160.

Groups came from the auditory verbal part of the Modalities of Thought questionnaire (MOT), a new 7-item self-report measure rated from 1 ("not at all") to 5. People scoring below 2 formed the aphantasia group (28). The other 132 were typical imagers.

Participants heard a sound halfway between "aba" and "ava" while silently repeating ABA, APA, AVA or AFA, then said which sound they heard. APA is made in the mouth much like ABA, and AFA much like AVA.

A "perceptual bias score" counted how often the answer matched the silent syllable or its look-alike. A score of 50% is chance.

What They Found

Typical imagers showed more bias than aphantasic participants (p = 0.016), but the effect size was small (r = 0.171).

Typical imagers were biased well above chance (r = 0.576). The aphantasia group showed no clear difference from chance (p = 0.066).

In typical imagers, silently repeating ABA, AVA or AFA shifted what they heard. Against the authors' expectations, APA did not clearly shift it.

In the aphantasia group, no silent syllable made a clear difference.

How vividly people rated their inner voice did not clearly predict their bias score (ρ = 0.104, not significant).

What This Might Mean

The authors think self-reported auditory verbal aphantasia reflects a real difference in how the mind works, not a mistake in describing one's own experience. They suggest inner speech can happen with no sound quality at all.

The test separated the groups but cannot yet measure one person's imagery. Scores overlapped widely, and some aphantasic participants were biased. Other limits:

The aphantasia group was small: 28 people.

The groups rested on a self-report questionnaire that is not yet validated, and the volunteers were recruited online.

The researchers could not check that people repeated the syllables silently.

The "in-between" sound leaned toward "aba" in both groups, which the authors think may explain the APA result.

The authors call for replication.

One Interesting Detail

People in the auditory verbal aphantasia group did not all lack inner speech. Of the 28, 5 said they never experience any form of inner speech. Another 6 said they experience at least one form very frequently, even though they reported little or no inner voice.

Research L’Année psychologique

Investigating auditory verbal aphantasia with a listening test

Huson and colleaguesSeptember 2026

Read the paper

Abstract
What This Study Is About
How They Studied It
What They Found
What This Might Mean
One Interesting Detail
Authors
Cite this paper

Abstract

This study aimed to assess an objective measure of auditory verbal (AV) aphantasia in a French and English-speaking sample, using a modified online version of the perceptual capture protocol (Scott et al., 2013) and a newly developed self-report measure, the Modalities of Thought questionnaire (MOT). We analysed data from 160 participants, classified into two groups according to presence or absence of AV imagery as assessed with the AV subscale of the MOT. The analysis of the speech perception scores under various inner speech conditions revealed differential effects. Individuals with typical AV imagery (n = 132) were biased to hear ambiguous sounds as similar to the phonological (sensory) content of their IS, unlike participants with AV aphantasia (n = 28). However, results further suggested that this protocol does not yet allow for the assessment of an individual’s level of AV imagery. Implications for these findings are discussed.

Our summary in plain words

What This Study Is About

How They Studied It

This was an online listening experiment with English and French speakers.

A "perceptual bias score" counted how often the answer matched the silent syllable or its look-alike. A score of 50% is chance.

What They Found

Typical imagers showed more bias than aphantasic participants (p = 0.016), but the effect size was small (r = 0.171).

Typical imagers were biased well above chance (r = 0.576). The aphantasia group showed no clear difference from chance (p = 0.066).

In typical imagers, silently repeating ABA, AVA or AFA shifted what they heard. Against the authors' expectations, APA did not clearly shift it.

In the aphantasia group, no silent syllable made a clear difference.

How vividly people rated their inner voice did not clearly predict their bias score (ρ = 0.104, not significant).

What This Might Mean

The test separated the groups but cannot yet measure one person's imagery. Scores overlapped widely, and some aphantasic participants were biased. Other limits:

The aphantasia group was small: 28 people.

The groups rested on a self-report questionnaire that is not yet validated, and the volunteers were recruited online.

The researchers could not check that people repeated the syllables silently.

The "in-between" sound leaned toward "aba" in both groups, which the authors think may explain the APA result.

The authors call for replication.

One Interesting Detail

This summary was written with AI and may contain errors. Check the original paper.

Authors

Nicole Huson
4 papers in the library
Claire Vanbuckhave
School of Psychology, University of Plymouth, Plymouth, U.K.; CNRS, Laboratoire de Psychologie et NeuroCognition, Grenoble, France
6 papers in the library
Téo Pesci
2 papers in the library
A. Chauvin
2 papers in the library
Hélène Lœvenbruck
6 papers in the library

Cite this paper

Huson, N., Vanbuckhave, C., Pesci, T., Chauvin, A., & Lœvenbruck, H. (2026). Investigating auditory verbal aphantasia with a listening test. L’Année psychologique, Vol. 126(2), 159–202. doi:10.3917/anpsy1.262.0159

Topics

Auditory

Most-cited researchers

Recruit participants for your study

We can help you reach people across the imagery spectrum, from aphantasia to hyperphantasia.

Partner with us

More research

All research

2026

Anauralia and aphantasia: Prevalence and distinct associations with personality, well-being and self-regulation in a large, representative sample.

Lambert and colleagues · Neuropsychologia

Researchers found that anauralia and aphantasia share a 0.8% prevalence but have distinct psychological profiles. This suggests that the absence of sensory imagery is modality-specific rather than a single, unified condition.

2026

Multimodal mental comparisons in those with and without aphantasia

Suggate, Milton and Tree · Neuropsychologia

People with aphantasia were more accurate but slower than controls on multimodal mental comparison tasks. This suggests that sensory cognition can be successfully achieved through propositional rather than imagistic strategies.

2026

Disturbing the sound of silence: Bilateral temporal cortex stimulation and auditory mental imagery

Rollo and colleagues · Brain and Cognition

In 49 healthy young adults, 20 minutes of weak, randomly fluctuating electrical current over both temples lowered ratings of how easily they could change one imagined sound into another, compared with a sham session on another day. Ratings of how vivid the imagined sounds were fell by a slightly smaller amount, a trend that did not reach significance. The ratings were self-reported during stimulation, and because no other brain area was stimulated for comparison, the authors cannot rule out a general effect, for example on attention.

Image-Free Insights

A short note from our founder, Tom Ebeyer, every other week, plus occasional news and invitations to take part in studies. Change what you get from any email.