Paper - Action Anticipation Through Attribution of False Belief by 2-Year-Olds (Southgate)

  • Metadata:
    • author: Victoria Southgate, Atsushi Senju, Gergely Csibra
    • title: Action Anticipation Through Attribution of False Belief by 2-Year-Olds
    • year: 20072007

  • Theory of mindTheory of mind
    .
    : The ability to attribute mental states to others.
  • The paper starts of with a seeming contradiction in the research field dedicated to "Theory of MindTheory of Mind
    .
    " : Children under the age of 4 have broadly been shown to fail classic false-belief tests, however, two year olds already exhibit behavior that requires the attribution of mental state to others (lying, playing pretend)
  • Onishi and Baillargeon showed that 15 old infants may already attribute false beliefs.
    • Switch position of items around, infants looked longer when the agents actions were incongruent with a false belief (unexpected while having false belief).
    • Problem: When attributing ignorance to an agent, young children expect the person to get the answer wrong. (-> this study did not attribute false beliefs but rather ignorance)
  • Classic false-belief tasks are flawed, as they require other abilities than understanding mental states:
    • inhibitory control (not act upon urges)
    • reality bias (knowledge about situation interferes with ability to select correct response)
  • This leads us to believe that the failure are based on performance limitations rather than competence limitations.
  • Why did this not also apply to 15 month old then? -> The verbal nature of question and answer elicits the reality biased response. (questions shown to be hard to understand)
  • This study builds upon the onishi study by also using nonverbal tasks (dodging the verbal problem) and also adresses some pitfalls of the previous study
    • Object is always removed, so that the location of the object would not attract gazes (reality bias removed)
    • experimental design was chosen in order to rule out a) the last position of the object b) the last position of the actors attention
  • Experimental Design:
  • Results: 9/10 infants in one condition 8/10 gazed to the correct window (according to false belief of actor) and on average spent twice as long looking at the correct window.
  • This shows that 25 month old infants can correctly attribute false beliefs to others.
  • Cannot be explained by attributing ignorance (would've looked randomly)
  • Rule based a possibility (agents tend to search in places where they last saw things regardless of belief), but not likely
  • The notion that children undergo a conceptual revolution between the ages 3 and 4 seems to be incorrect.