| At our best, we reason flexibly about how two items are alike. For example, we can notice that a map and a city share the same configuration, or that both a shark and a tiger are predators. Across cognition, there is substantial empirical evidence and computational modeling support for the basic importance of relational structure in similarity, categorization, memory retrieval, and inductive reasoning, among other psychological processes. Yet, a robust developmental finding is that children have difficulty understanding relations (e.g., Gentner & Rattermann, 1991; Halford, 1993). For instance, young children are likely to interpret the metaphor “a cloud is like a sponge” as meaning they are both soft and fluffy, rather than comment on the relational commonality that both can hold and release water (Gentner, 1988). The central claim tested in this dissertation was that two experiential processes lead to learning relations: making comparisons and hearing relational language.;Seven experiments and computer modeling results examined how preschool children learn to notice and use similarities based on spatial relations. Children played mapping games in which they were asked to find a toy in a test array after seeing one placed in an initial array. The correct choice was always the one in the same relative position as the one in the initial array. Both comparison and relational language helped children remember the initial toy's location and find the toy in the second array. Thus, relational language and comparison helped children to encode and map on the basis of spatial relations. These gains were attributed to learning relations, because the comparison and relational language advantages persisted when children were tested with an analogous task, and on the initial task after several days' delay. These findings are consistent with the claim that making comparisons and hearing relational language are means by which children learn relations from experience. |
