Welcome back. The previous lesson distinguished fuller from condensed inner speech, and one-sided from conversational forms. Those distinctions describe what an episode is like. This lesson asks a different question: how often does such an episode arise without prompting, and what can someone do when a task calls for language?
The distinction matters when interpreting claims about having “little inner voice.” A person might seldom notice spontaneous words while still being able to form a sentence inwardly on purpose. Someone might also use speech-related information to solve a problem without experiencing an audible-seeming voice. Allow about 35–40 minutes for the lesson, video, and reading.
Frequency is not capacity
Consider three observations about the same person. During an ordinary afternoon, they report almost no unprompted inner speech. When invited to do so, they can silently form the sentence “I should leave early.” Later, they can judge whether two pictured objects have names that rhyme. These observations need not conflict: they concern different circumstances and do not all measure the same thing.
Spontaneous frequency concerns how often inner speech is reported to occur without an instruction to produce it. Deliberate production concerns whether someone can form an inward utterance when they intend to. Neither is an all-or-nothing trait. A person may be able to produce a phrase briefly but not sustain a lengthy inner monologue; they may also use deliberate self-talk frequently in one setting and rarely in another.
The figure below places deliberate and spontaneous inner speech along one axis of a proposed framework. Its other axis concerns whether speech is oriented toward oneself or another person—a different distinction from the frequency-versus-capacity contrast. Treat the depicted brain patterns as part of a research framework, not as a scan that can classify an individual’s everyday thinking.

A report such as “I don’t have a running inner monologue” is therefore narrower than “I cannot think in words.” It may mean that verbal episodes seldom arise on their own, that they are brief or condensed, or that the person does not experience them as a voice. The previous lessons showed why those descriptions should not be treated as interchangeable. Even a sincere report of very low frequency does not, by itself, establish what happens during every activity or what the person can do on request.
Prompted inner speech is a different research situation
Researchers sometimes investigate inner speech by instructing participants to repeat a sentence silently. That creates a fairly clear task: participants know when to begin, and researchers know when to examine their responses. But it is task-elicited inner speech, not a sample of what happened to arise during an otherwise ordinary moment.
Continue the Royal Institution lecture from the previous lesson. Charles Fernyhough describes a study that contrasts instructed inner speech with episodes reported as spontaneous during scanning.
The Science of the Voices in your Head – with Charles Fernyhough
In The Science of the Voices in your Head, Charles Fernyhough explains why inner speech produced on command need not be an adequate stand-in for inner speech that arises without prompting.
Watch the comparison. At the start, note how participants were instructed to produce inner speech and how spontaneous episodes were identified. In the latter part, focus on the reported contrast in activity in speech-production-related and auditory regions. The important point here is the difference between the two research conditions, not memorizing brain-region names.
The study reports different patterns of activity for its instructed and spontaneously reported episodes. That makes it risky to take a result from “silently say this now” and assume it describes every naturally occurring verbal thought. It does not follow that every spontaneous episode has one fixed brain pattern, or that a scan alone can reveal exactly what a participant experienced. The instruction, the surrounding task, and the way episodes are identified all matter.
There is also a useful distinction between spontaneous and uncontrollable. An utterance can come to mind without a researcher’s prompt yet still be something the person can redirect. Conversely, someone who often experiences unprompted self-talk may find a particular instructed phrase awkward to sustain. Frequency and control are related questions, not synonyms.
A task can recruit language without revealing a “voice”
Return to the pictured-object rhyme task. To decide whether the objects’ names rhyme, a participant might silently name them and compare their sounds. They might instead say the names aloud, or use another workable strategy. Success on the task does not tell us which strategy they used, and difficulty does not prove they lack the ability to form inner speech.
This is the third distinction: speech-related processing during a task concerns the use of information such as word names or sound patterns. It does not automatically establish a conscious, voice-like experience. A person can know that two words rhyme without reporting that they inwardly heard either word.
Read the following short selections from Nedergaard and Lupyan’s study. Its participants differed in self-reported inner speech; the study did not continuously observe every participant’s spontaneous experience. Pay particular attention to what happened when participants reported saying words aloud.
Not Everyone Has an Inner Voice: Behavioral ...
Nedergaard and Lupyan compare self-reports with performance on language-related tasks. Read these passages to see both why a task might draw on verbal processing and why the availability of an overt verbal strategy matters.
In “The Present Study,” read the task rationale, from the description of recruitment through the predictions about word memory. Then find “Strategies for Rhyme Judgments” (around printed p. 619). Read the rhyme results, from the comparison of who reported speaking aloud through the authors’ interpretation. Finally, in “Strategies for Verbal Working Memory” (around printed pp. 619–620), read the memory results, from the comparison of reported strategies through the disappearance of the group difference among those who reported speaking aloud. Focus on the difference between an observed performance pattern and an inference about an unobserved private strategy.
In the study, performance differences associated with self-reported inner speech diminished or disappeared in the relevant comparisons when participants reported talking aloud. This shows why “uses little inner speech” should not be equated with “cannot use words to help with a task.” It is consistent with the possibility that some participants who reported more inner speech were using a verbal strategy covertly, as the authors suggest. But talking aloud is not the same act as silently producing an utterance. The result does not demonstrate that every lower-reporting participant could produce the same inner speech on command, nor does it directly reveal what anyone experienced during silent trials.
The three claims can now be kept separate:
| Claim | What would bear on it? | What it would not establish alone |
|---|---|---|
| “I rarely experience unprompted inner speech.” | Reports about ordinary moments | Inability to produce it deliberately |
| “I can form an inward sentence on request.” | A prompted report or task | Frequent spontaneous inner speech |
| “I use speech-related information for this task.” | A task strategy or pattern of performance | An audible-seeming inner voice |
The wording matters. “I rarely experience” is a report about experience across occasions; “I can” concerns capacity under an instruction; “I use” concerns a strategy in a particular setting. None can be safely substituted for either of the others.
Takeaways
How often someone reports spontaneous inner speech, whether they can produce it deliberately, and whether they use speech-related processing in a task are three distinct questions. Evidence about one can inform the others, but it does not settle them. In particular, low reported frequency is not proof of permanent absence or incapacity, while language-related task performance is not a direct recording of a conscious inner voice.
This completes the module’s account of what inner speech can be like and when it might occur. The next module examines how researchers gather reports of private experience—and why a general questionnaire, a sampled moment, and a detailed interview can support different conclusions.