Why Talking to Yourself Out Loud Is a Sign of a Working Brain

The boundary condition is that speaking only helps when the word and the thing are well matched. Push them apart and saying it aloud starts making performance worse.

Vygotsky's account

Lev Vygotsky observed in the 1920s that children solving problems narrate what they are doing out loud, and argued this was not immaturity but a functional stage. On his account, thinking begins as dialogue with others, becomes self-directed private speech spoken aloud, then goes quiet, and finally becomes silent inner speech.

This ran directly against the prevailing view that children's self-talk was noise from an incompetent mind. Vygotsky's position was that private speech does work — regulating attention and organising action — and that adult reasoning is this same process run silently.

The framework has held up well, which is not something you can say about most 1920s psychology. Private speech in children increases with task difficulty rather than decreasing with age alone, which is what a functional account predicts and a maturity account does not.

The visual search experiment

The cleanest adult demonstration comes from Gary Lupyan and Daniel Swingley. Participants searched for objects in cluttered displays, sometimes saying the target's name aloud and sometimes not.

Speaking facilitated search. The proposed mechanism is label feedback: hearing yourself say a word activates visual properties associated with it, temporarily tuning your visual system toward that category. Saying "keys" primes shape, size and metallic sheen in a way that thinking it does not do as efficiently.

The part that gets dropped

The facilitation depended on a strong association between the name and the visual target. As the discrepancy between label and object grew, speaking began to impair performance rather than help it.

That is the difference between a technique and a slogan. Saying "keys" while looking for your keys helps. Saying "keys" while looking for an unusual object you would not naturally call keys biases your visual system toward the wrong thing. Almost every popular write-up of this study reports the benefit and omits the cost.

Where else it helps

Several converging lines of evidence, with varying strength.

• Reading instructions aloud rather than silently improved task performance in a controlled experiment, though with a modest sample.
• Interfering with covert verbalisation impairs the ability to switch flexibly between tasks, suggesting inner speech supports task-switching.
• Suppressing inner speech impairs performance on inhibition tasks, implicating self-directed speech in self-control.
• In sport psychology, instructional and motivational self-talk interventions have been meta-analysed and show moderate effects on performance.

The underlying theme is that speech externalises something that was distributed and vague, and an externalised representation is easier to hold, manipulate and act on. This is exactly why saying a goal out loud before crossing a threshold protects it — the spoken word survives the boundary because it no longer lives only in working memory.

How much inner speech is there really?

One finding worth knowing, because it complicates the tidy picture of adults as silent narrators of their own lives.

Descriptive experience sampling — beeping people at random moments and asking what was in their awareness immediately before — consistently finds inner speech present in a minority of samples rather than running continuously. People substantially overestimate how much of their mental life is verbal, which matters for any account treating inner speech as the default medium of thought.

More strikingly, some people report having no inner speech at all. Recent work on this, which has been given the name anendophasia, suggests it is associated with measurable differences on verbally mediated tasks while leaving general reasoning intact. That is early-stage research and should be read as such, but it argues against any strong claim that thinking requires an inner voice.

Which puts the audible version in perspective. Speaking aloud is not a degraded form of a universal inner monologue; it is one point on a range that varies considerably between people.

The pronoun question

A related line of work concerns how you address yourself. Research on distanced self-talk suggests that using your own name or "you" rather than "I" — "you can handle this" rather than "I can handle this" — improves emotion regulation and performance under stress, apparently by creating psychological distance.

This is the least secure claim in the article and should be treated accordingly. It has generated considerable enthusiasm and its effect sizes and robustness across contexts are less settled than the coverage implies. It sits in the region of social psychology that has had the most difficult decade, and it deserves the same scepticism this site applies elsewhere rather than an exemption for being appealing.

The intelligence myth

A claim circulating widely holds that talking to yourself is a marker of high intelligence. There is no good evidence for this, and the articles making it generally cite studies about the cognitive benefits of self-talk, which is a different proposition entirely.

Showing that a behaviour improves performance on a task does not show that people who do it more are more intelligent. The genuine finding is that self-talk is functional and normative — not that it is diagnostic of anything about the person.

The older folk belief in the other direction is equally unfounded. Talking to yourself is not a sign of instability; it is what a cognitive tool looks like when it surfaces, and it surfaces most when the task is hard. It belongs alongside the other things people do with no audience present — including gesturing at a listener two hundred miles away, which turns out to be part of producing the sentence rather than delivering it.