作者: Marina Katerelos
DOI:
关键词: Referent 、 Spatial relation 、 Computer science 、 Generalization (learning) 、 Cognitive psychology 、 Context (language use) 、 Animacy 、 Event (probability theory) 、 Motion (physics) 、 Communication 、 Object (grammar)
摘要: Researchers have postulated word-learning biases to explain infants' effortless acquisition of object words. One such bias, the whole assumption, helps infants acquire new words by narrowing referent a novel label an entire object, rather than actions, spatial relations or parts that (Woodward & Markman, 1998). This bias is surprising given motion particularly salient young (Poulin-Dubois, 1999). The main objective present experiment was determine whether assumption could be overridden when presented in context displaying animate inanimate motion. To address this issue, 18-month-old were with engaged and A paired each these events. On test trials, simultaneously event originally target other label. In generalization original pairings training phase switched, saw performing results study revealed as group, unable learn for current are discussed recent word learning literature