Three-year-old Children’s Use of Category Labels and Motion in Drawing Inferences about Animal Kinds

Abstract

Three-year-old Children’s Use of Category Labels and Motion in Drawing Inferences about Animal Kinds Benise S.K. Mak (benise@hku.hk) Department of Psychology, The University of Hong Kong, Pokfulam Road, Hong Kong, PR China Alonso H. Vera (avera@arc.nasa.gov) Cognition Lab, NASA Ames Research Center, Mailstop 262-4, Moffett Field, CA 94035-1000, USA Lap Yan Lo (h9605361@hkusua.hk) Department of Psychology, The University of Hong Kong, Pokfulam Road, Hong Kong, PR China Abstract Linguistic labels, information about category memberships, have been found to be more important than perceptual information in guiding young children’s inferences about animal kinds. However, perceptual information of static shape cues has often been stressed. A recent study has shown that young children tended to use dynamic perceptual cues, such as motion, more often than static shape cues to make categorical judgments. The overriding effects of linguistic labels over perceptual information in young children’s inferences need to be re-examined. This paper was an attempt to examine how 3-year-old children use category labels and motion cues to draw inferences about animal kinds. Data showed that preschool children tended to use motion more often than labels when confronted with a choice between labels and motion. This provides support for our view that the role of category labels in young children’s categorical judgments is not as important as what has been suggested in previous studies. Introduction The Importance of Category Labels The importance of linguistic information in preschool children’s inductive inferences about animal kinds has well been demonstrated in a series of studies by Gelman and her colleagues (e.g., Gelman & Markman, 1986, 1987; Gelman & Coley, 1990). In their studies, the extent to which children use category labels and perceptual appearance was examined. For instance, 2½ years old children were found to be more likely to assign a property, e.g., “lives in a nest”, from a bluebird to an atypical bird, dodo, which looked different from the target but shared the same label “bird” than to a dinosaur which looked similar to the bluebird but carried a different label “dinosaur” (Gelman and Coley, 1990). However, when the stimuli were not labeled, they tended to draw inferences based more on perceptual similarity than on the similarity of verbal labels. The young children were able to go beyond perceptual appearance and use linguistic labels to draw inferences, suggesting that linguistic information is more powerful than perceptual information for preschoolers to draw accurate categorical judgments. As Gelman and Coley (1990) stated that for young children “language conveys important information beyond that which meets the eye” (p.804). Despite of this, the role of category labels remains to be determined as perceptual characteristics have generally been taken as static perceptual cues in Gelman et al.’s studies. Perceptual cues should also include dynamic properties, such as motion. The Importance of Motion Information In a recent study by Mak and Vera (1999), 4- and 7- year-old children have been found to categorize animals based more on motion similarity than on static shape similarity. For instance, they were more likely to categorize, for example, a donkey with an antelope than with a horse when the antelope and the donkey were shown to jump in the same manner even thought the donkey looked more similar to the horse than to the antelope. The children tended to infer that the donkey shared the same property of “having poor vision” that was ascribed to the antelope rather than the property of “having good vision” ascribed to the horse. The role of motion information has been stressed. As such, this paper tried to look further into the role of category labels and include motion information to study young children’s inductive inferences about animal kinds. We will argue that the effectiveness of verbal labels is not as important as Gelman et al. have suggested.