Does seeing another's viewpoint need working memory?
When people had to hold letters in mind, they got worse at judging what an on-screen avatar could see, suggesting even the 'automatic' part of perspective taking leans on working memory.
Source
Executive function underlies both perspective selection and calculation in Level-1 visual perspective taking
Study at a glance
- Design
- Human experiment — Within-subject dual-task experiment: Samson dot perspective task done alone or while holding a letter span in mind, crossing consistency (avatar sees same/different dots) and perspective (self/other); process-dissociation reanalysis also applied to earlier inhibitory-control data
- N
- N=54 · 54 undergraduate students (47 female); the reanalysed earlier inhibitory-control dataset is separate
- Population
- Right-handed university students aged 18-22 participating for course credit
- Outcome
- Error rates and response times on self- and other-perspective dot judgments; process-dissociation estimates of controlled and automatic processing
Structured fields used in claim comparison tables when every cited study has a complete layer.
What they did
Fifty-four students did a computerized dot task: they saw a room with dots on the walls and an avatar facing one wall, and verified how many dots either they or the avatar could see. On half the trials the avatar saw the same dots as the participant (consistent) and on half it did not (inconsistent). Each block was done either alone or while holding a letter string one item below the person's maximum span, which was recalled after each response. The team also used process-dissociation equations to separate a controlled component (answering from one's own perspective) and an automatic component (computing the avatar's view), and applied the same analysis to an earlier dataset that had loaded inhibitory control.
What they found
Holding letters in mind raised error rates overall, and inconsistent trials produced more errors than consistent ones, except for other-perspective trials under load, where the consistency effect vanished. Under load, errors on consistent other-perspective trials rose above those for self-perspective trials, suggesting working memory was especially needed to compute the avatar's view. Process-dissociation estimates showed that working-memory load reduced both the controlled and the automatic component, whereas in the reanalysed inhibitory-control data only the controlled component dropped. The authors conclude that working memory contributes to both perspective calculation and perspective selection, while inhibitory control mainly supports selection.
The limits
What it doesn't show
The dot task requires counting dots and comparing them with a number, so the working-memory load may have disrupted counting rather than perspective taking itself, a possibility the authors explicitly cannot rule out. Only a verbal letter-span load was used, and the sample was small, mostly female undergraduates. The power calculation was based on effect sizes from a different (inhibitory-control) study, and only Level-1 perspective taking was tested, so it is unclear whether results extend to more complex theory-of-mind tasks. Comparisons with the inhibitory-control data come from a separate sample and study, not a direct within-study contrast.
Key terms
- Level-1 visual perspective taking
- Judging whether or what another person can see from where they are, as opposed to how it looks to them.
- Perspective calculation vs. selection
- Calculation is working out what another person sees; selection is choosing which perspective (yours or theirs) to use for a response.
- Dual-task paradigm
- An experiment in which people do a main task while also doing a secondary task, to see whether both draw on the same mental resource.
- Altercentric interference
- Slowing or errors when judging your own view because the other person's differing view is processed anyway.
- Process-dissociation procedure
- A method that uses accuracy on consistent and inconsistent trials to estimate separate controlled and automatic contributions to performance.
Flashcards
0 of 10 answers reviewed
Research intelligence for this paper
See its role on concept claims, tensions it is part of, placement history, and related discoveries.
Quiz yourself
What was the main manipulation in this experiment?
Common questions
Why would working memory matter for something infants can do?
The authors suggest basic perspective tracking may be a core ability present early on, which becomes more efficient as working memory develops and suffers when working memory is stretched.
Why was the letter load set one below each person's maximum span?
To make the working-memory demand similar across people while avoiding floor effects from impossible loads.
Does this show perspective calculation is not automatic at all?
Not necessarily. The findings suggest working memory helps it run efficiently, but the counting demands of the task and the single type of load leave room for other explanations.
More on Social cognition