Fast Mapping and Slow Mapping in Children's Word Learning

When young children encounter a word they do not know, their guesses about what the word might mean are often surprisingly accurate. This is true not only with respect to the particular instance that the speaker refers to at that moment but also with respect to the entire category of things, states, situations, or events to which the word may refer in the language. For more than 30 years, understanding how this is possible has been the central empirical and theoretical concern of most of the developmental psychologists and linguists who study the process of word learning experimentally. This attention has not been misplaced. The domain of word learning has provided a fertile ground for testing competing accounts of children’s understanding of reference in language, their use of ontological divisions and other world knowledge in categorization, and their grasp of syntactic regularities. On the whole, children have performed surprisingly well in these experiments—by the age of two or three, they make efficient and appropriate use of a wide range of sources of information in determining what speakers are referring to in the moment, and in evaluating how novel words may be used in future situations. Most such tests have taken place in contrived but well-controlled experimental situations in which a brief exposure to a novel word, in a particular social or linguistic context, is revealed to lead children to choose one object or scene rather than another as a referent of the word. What gives such studies their force is the careful manipulation of the precise content of the introducing event, and the selection of the alternatives offered to the child, which pit one possibly tempting interpretation against another. The point is not that the experimental situation closely mimics children’s daily lives but rather that children’s interpretations reveal antecedent knowledge either innately specified or gained in development. The experiments acknowledged as the primary intellectual ancestors to this research tradition are those reported in Carey and Bartlett (1978; see also Brown, 1957; Katz, Baker, & MacNamara, 1974). Carey and Bartlett (1978) introduced the term “fast mapping,” which has become central to developmental psychology’s narrative about how words are learned. In this narrative, it is children’s accuracy in fast mapping that cries out for explanation. How can children arrive at the correct meaning of a word given only indirect and incomplete evidence? Yet in Carey and Bartlett’s famous “chromium” study, fast mapping was not so successful. Fewer than one in ten of the 3-year-olds appeared to have linked the word to its intended meaning (olive green). The children who had been exposed to the word in the study’s naturalistic teaching context (“bring