Social brain
How does the brain produce internal conversations?
Open access · cc by · source: Europe PMC
Silent internal conversations recruit brain areas involved in understanding others' minds alongside traditional language regions, unlike simple internal monologues.
How this study connects
Role on claims
Each row is a claim on a concept or method page where this paper supports, challenges, or qualifies the statement. Roles are hand-checked — not a model guess.
Complex internally generated tasks, such as silent dialogic inner speech, engage core DMN regions like the precuneus and posterior cingulate in addition to traditional language networks.
Evidence for the claim as stated.
Engaging in silent internal conversations (inner dialogue) recruits cortical regions associated with social perspective-taking, specifically sharing a precise activation zone in the posterior superior temporal gyrus with traditional Theory of Mind tasks.
Evidence for the claim as stated.
Engaging in silent, dialogic inner speech (internal conversations with oneself) recruits the right posterior superior temporal gyrus, sharing a functional neural substrate with tasks designed to measure Theory of Mind.
Evidence for the claim as stated.
Study at a glance
- Design
- Human experiment — fMRI inner dialogue vs inner monologue plus mock-scanner articulatory-suppression ratings
- N
- N=19 · 21 recruited; 19 in the fMRI analysis; 20 in the mock-scanner behavioral study
- Population
- Right-handed native English-speaking university adults with no neurological history
- Outcome
- fMRI activation for dialogic vs monologic inner speech, overlapping right pSTG with theory-of-mind
Structured fields used in claim comparison tables when every cited study has a complete layer.
Inspect source passages
How does the brain produce internal conversations?
These passages occur verbatim in the stored paper text. They support the points listed below; they do not verify every statement in the explanation.
19 participants in the final analysis
Thus, the results later display data from a sample of 19 participants (5 male, age m (s.d.) = 24.63 (7.01) years).
20 participants in the mock scanner behavioral study
A separate set of 20 participants [2 male; age m (s.d.) = 19.65 (1.31) years] attempted the original inner-speech task and then rated a variety of phenomenological characteristics for each dialogic and monologic scenario (see Supplementary Materials for an example response sheet).
Key findings
Generating internal dialogues led to greater activation in the bilateral superior temporal gyrus, precuneus, posterior cingulate, and left inferior frontal gyrus compared to internal monologues. A conjunction analysis showed that both internal dialogues and theory-of-mind reasoning shared a specific activation zone in the right posterior superior temporal gyrus. Behavioral testing in the mock scanner confirmed that articulatory suppression significantly reduced the self-reported vividness of both inner speech and internal voices.
Methodology
Researchers scanned 19 healthy right-handed participants using functional magnetic resonance imaging while they generated internal monologues and internal dialogues based on written scenarios. Participants also completed a theory-of-mind task and a visual perspective-switching task in the scanner to compare brain activation patterns. To verify task compliance, a separate group of 20 participants completed the tasks in a mock scanner under normal conditions and during articulatory suppression.
Limitations
The study cannot objectively verify if participants were actually generating inner speech in the scanner since it relies primarily on post-scan self-reports. The visual perspective-switching task failed to produce significant brain activation differences, meaning the researchers could not fully test whether internal dialogue simply demands more general mental imagery control. Additionally, the small sample size of 19 participants limits the generalizability of the individual differences analysis.
Related papers in this topic
Same topic cluster — not a recommendation engine.