Are names ever mapped onto preexisting categories?
Support for Mervis's (1987) proposal that youngsters map new common nouns onto attribute cluster categories was obtained. In 3 studies, preschoolers interpreted a novel noun as if it designated a preexposed category. Experiment 3 included a group of adults who behaved similarly. Mapping was independent of the tendency to generalize according to shape. Neither age group mapped onto preexisting categories when making object similarity judgments, but preexposure did bias their naming of the objects' colors. Those who had examined sets containing 2 colors, both of which covaried with other bivalent dimensions, used more distinctive names for these hues than those preexposed to sets containing 6 colors, none of which covaried with bivalent dimensions.