Children expect tool functions to be culturally determined: evidence from the phenomenon of scale errors

• In Event: 3-184 - Integrating Perception, Cognition, and Action: The Case of Scale Errors

Integrative Statement

It has been suggested that children’s proneness to commit scale errors may originate from their propensity to reason about artifacts as existing to fulfill a specific function (Casler, Eshleman, Greene & Terzyjan, 2011). In this view, children overlook size information because they focus on the function the given tool category is associated with. If this is true, scale errors should be sensitive to the circumstances of the learning episode in which children become familiar with the object function. Since tool functions are cultural constructs, we hypothesized that if children witnessed a function demonstration from a member of a different social group, they would not encode the performed action as the normative function of the object category. Consequently, faced with oversized exemplar of the same category they would be less likely to commit scale errors.

In our experiment, 3-year-old children (n=37) were presented with object sets consisting of a target and a potential tool to bring about a certain goal. First, a model speaking either in children’s native language or a foreign language demonstrated the object functions. After this, the model left and children were handed the object sets for exploration, however, the sets contained an important alteration. The original tool was always replaced by one that was too big to efficiently bring about the goal but that was otherwise identical to the original one. Children were also provided with a novel, affordant tool. Results show that children committed scale errors less frequently when the object functions were demonstrated by a non-native model (30%) than when the model had spoken their native language (48%).See Fig. 3.1

In a follow-up study, we tested whether 4-year-old children (n=39) understand the cultural aspect of tool functions and would selectively assign a specific function to a novel tool only when the function was demonstrated by an in-group model. In the first phase, children watched as a person speaking either in their native or a foreign language introduced novel tool functions. Then, another experimenter, who always spoke the child’s native language, entered the scene and expressed a wish to achieve a different goal on the same apparatus. For this, she could either use a color variant of the original tool or choose a novel tool. Children were asked to help the second experimenter make the choice. We hypothesized that children in the In-group condition would opt for the novel object since the original tool would be reserved for the function presented by the first experimenter (“teleofunctional stance” – Casler & Kelemen, 2005). However, if children understand that functions are culturally determined, they would not have any specific preferences in the Out-group condition. Our results confirmed this hypothesis showing that children chose the novel tool more often in the In-group than the Out-group condition (63% and 43%, respectively– see Fig. 3.2.)

Together, these results suggest that children from an early age view tool functions as normative but only within the boundaries of specific groups. The results also support the teleo-functional account of scale errors.

Authors

• Katalin Oláh, MTA-ELTE Momentum Social Minds Research Group
  Presenting Author

• Fruzsina Elekes, MTA-ELTE Momentum Social Minds Research Group / Central European University
  Non-Presenting Author

• Réka Pető, Eötvös Loránd University
  Non-Presenting Author

• Ildikó Király, MTA-ELTE Momentum Social Minds Research Group / Central European University
  Non-Presenting Author