The development of reality and constitution in concepts of animals, artifacts, and social institutions

• In Event: Categories and Concepts

Abstract

According to psychological essentialism (Gelman, 2003; Medin & Ortony, 1989), participants infer kinds possess internal essences. This belief leads participants to infer a category is both naturally constituted and real. The category is naturally constituted as the internal essence causes its properties rather than people and their intentions. And, the category is real because it maps onto a real-world structure: the essence. Consequently, psychological essentialism suggests people inflate natural with real and social with nominal (not real), corresponding to the presence or absence of an essence. However, we hypothesized that children and adults would distinguish between constitution and reality. If children and adults distinguish real from natural and nominal from social, then we would expect their judgments about the reality of diverse categories (e.g., animals, artifacts, and social institutions) to differ from their judgments about constitution.
We measured realness using an existing paradigm (Kalish, 1998; Rhodes & Gelman, 2009). On each trial, children saw pictures from two categories labelled using conventional American labels (e.g., “turtle” and “frog”); children then learned that the novel community considered them members of the same category. Children were asked “are they maybe right?” If children believe the category boundary is real, they should reject alternative classifications (as alternative classifications would fail to map onto the corresponding real-world structure). To measure constitution, we adapted a task based on the transformation paradigm (Rips, 1989; Keil, 1989). A member of category A is transported (accidentally) to a new community who intends for the individual to be a member of category B. If children believe category membership is constituted by social processes, then children should report that category membership has changed once the individual becomes embedded in a new pattern of social interactions. Otherwise, they should reject that category membership has changed.
The categories presented to children were: (animals) turtle / frog, goat / pig, and wolf / lion; (artifacts) car / train, hammer / screwdriver, and table / bookcase; (institutions) license / badge, police officer / judge, and artist / teacher. We interviewed 270 children from a broad age range (4-10). We also collected adult data (N = 360 total).
Confirming our hypothesis, the profile of children’s reasoning differed by measure, particularly for artifacts and institutions, such that there was a domain by measure interaction, b=1.14, SE=.24, p<.001, OR=3.12. Children judged that social-institution categories were significantly more socially constituted than artifact categories, b=0.79, SE=0.17, p<.001, OR=2.19, such that they said membership to social-institutional categories could change with a community’s intentions. In contrast, they judged that artifact categories were more nominal than institution categories, b=0.35, SE=0.17, p=.037, such that a community’s alternative classifications could be right too. There was no developmental change in reasoning about the reality of categories (ps>.10). But there was change in reasoning about social constitution. At first, children said all categories were slightly naturally constituted: With age, this pattern strengthened for animals, b=-.61, SE=.25, p=.013, OR=1.85, held constant for artifacts, b=.10, SE=.11, p=.395, OR=1.10, and decreased for social-institution categories, b =.20, SE=.09, p=.018, OR=1.23.

Authors

• Alexander Noyes, Yale University
  Presenting Author

• Yarrow C Dunham, Yale University
  Non-Presenting Author

• Frank C Keil, Yale University
  Non-Presenting Author