Preschool children's use of novel predicates to make inductive inferences about people

Preschool children's use of novel predicates to make inferences about people was examined in three studies. In a procedure adapted from Gelman and Markman Cognition 23 (1986) 183.], participants (ages 3 years 5 months–4 years 11 months) saw line drawings of three different faces. In Study 1 (N=16), the drawings were described as depicting children, and participants were asked to predict whether one of the children would share properties with a child who has the same novel predicate (e.g., “is zav,” which is never defined for participants) but is dissimilar in appearance, or with a child who has a different novel predicate but is similar in appearance. Participants tended to use the novel predicates rather than superficial resemblance to guide their inferences about people. In Study 2 (N=16), in which the line drawings were described as depicting dolls rather than children, participants showed no such emphasis on the novel predicate information. Study 3 (N=38) replicated the results of the first two studies. The results suggest that children have a general assumption that unfamiliar words hold rich inductive potential when applied to people but not when applied to dolls.