Your Inner Voice Is Harder to Measure Than It Sounds

“I talk to myself all day” and “my thoughts do not sound like a voice” are descriptions of experience. Turning them into measurements takes more than asking which sentence feels familiar.

That is what makes inner speech interesting. Researchers study silent verbal thought through reports, tasks, and observations, while trying to understand how those forms of evidence relate. An account of someone's experience and a result on a memory test can inform each other without becoming the same measurement.

The recent debate is a useful correction to the idea that everyone has one standard internal narrator, or that a quiet narrator reveals a simple deficit in thinking.

Inner speech is a varied experience

A major 2015 review by Ben Alderson-Day and Charles Fernyhough brought together research on inner speech's development, cognitive functions, subjective qualities, and neurobiology. The breadth matters: silent verbal thought is studied across several tasks and ways of describing experience, rather than through one universal narration meter. The research review

For a simple example, silently rehearsing a short phrase and experiencing a spontaneous internal conversation need not feel identical. Asking how often someone has an “inner voice” may bring one of those experiences to mind more readily than the other.

That creates a measurement problem before we reach a brain scan. Researchers need to explain what they are asking about and how participants interpret the question.

It also makes casual comparisons between people difficult. A person using “voice” to mean a vivid sound may answer differently from someone using it to mean any thought expressed in words. Their answers may concern different qualities of experience.

A 2024 study found differences on particular tasks

Johanne Nedergaard and Gary Lupyan compared adults reporting high and low levels of inner speech. Their 2024 paper found that the low-reporting group performed less well on a verbal working-memory task and had more difficulty judging rhymes. Differences were not found in the task-switching and perceptual measures reported in the abstract. The original study

That is more informative than a headline saying that people without an inner voice have worse memory. It identifies how the groups were formed and which outcomes differed. It also keeps the outcomes that did not differ inside the account.

The result does not justify ranking people's overall intelligence or assuming how they manage every part of daily life. A task measures performance under its conditions. A broad claim about someone's mind needs more evidence.

The authors introduced the term “anendophasia” for a lack of experienced inner speech. The relationship between that label, low questionnaire scores, and complete absence of inner speech then became part of an explicit scientific disagreement.

Low frequency and complete absence are different claims

In a 2025 commentary, Andreas Lind argued that the study made an important contribution while failing to establish that some participants entirely lacked inner speech. His central distinction was between reporting relatively little of an experience and demonstrating none at all. Lind's commentary

That distinction is useful beyond this one debate. If a question groups people by frequency, a difference between the groups does not automatically establish a separate category defined by complete absence.

It does not follow that a person's report should be dismissed either. The challenge is to decide what the report can support and what additional observation would be needed for a stronger claim.

I would keep the descriptive language close to the method: people who reported low levels, on this questionnaire, performed differently on these tasks. That sentence leaves room for interpretation without pretending the experiment has already resolved every question about subjective experience.

The 2026 discussion asks what kind of evidence is needed

Russell Hurlburt's January 2026 commentary distinguishes measuring a psychological construct from investigating how faithfully a method captures an experience. He argues for closer attention to descriptions of experience and considers people with little or no inner speech plausible, while acknowledging that separating “mostly” from “completely” may be difficult in practice. The 2026 commentary

This is not a new experiment proving one universal answer. It is a methodological argument about what would count as appropriate evidence.

That makes the disagreement productive to read. The researchers are not simply choosing whether to believe a colorful anecdote. They are asking how to connect a reported experience, an experimental measure, and the conclusion drawn from them.

The debate should therefore remain visible in an explainer. Replacing it with a confident prevalence figure would give readers a tidier story than the evidence currently supports.

Let the description stay larger than the label

My original note treated the inner voice too much like one speech system running silently and extended that account into clinical explanations. That compressed several different questions into a neat mechanism. The research on variation deserves a more careful treatment.

For ordinary conversation, the useful starting point is curiosity about how someone describes thinking. Words, images, and other reported qualities should not have to fit a single familiar metaphor before they are taken seriously.

For a scientific claim, ask the next question: how was the experience investigated, and which conclusion follows from that method? A questionnaire, a task result, and a close account of a particular moment offer different kinds of information.

Inner speech can be useful and important without being a universal soundtrack that plays the same way for everyone. Keep the experience, the measurement, and the interpretation distinct. The differences become easier to understand when the explanation stops trying to make every mind sound like the narrator we know best.