An investigation of hearing infants' preferences for American Sign Language and nonlinguistic biological motion.

摘要

The existence of a language bias, a tendency for infants to pay attention to language more than to other signals in the environment, has powerful implications for the development of linguistic capacities during the first year of life. In this dissertation, I investigate the robustness of the language bias by measuring hearing infants' preference for a completely unfamiliar language perceived in a visual modality (American Sign Language) over a nonlinguistic visual control (pantomime). In this way, I differentiate humans' inherent predisposition to pay attention to the patterning and structure human language in general from their natural attraction to speech---a highly familiar, and acoustically complex, stimulus. Six-month-old (Study 1) and 10-month-old (Study 2), hearing infants were shown ASL short stories and pantomime sequences in a preferential-looking task. The 6-month-olds indicated a preference for the ASL, indicating a language bias that is not specific to speech. The 10-month-olds demonstrated a more equal preference for both types of motion, reflecting either a decline in interest of the linguistic properties of a nonnative language or an increased interest in the intentional properties of the pantomime. In Study 3, the stimuli were reduced to point-light displays in order to investigate whether the movement properties of the ASL in isolation may be sufficient to drive the infants' attentional preferences. A group of 6-month-olds run on the point-light version of the experiment preferred the pantomime. In Study 4, the stimulus features and patterns of features of the full signal ASL and pantomimed stimuli were carefully described and compared in the context of linguistic constraints and infant preferences. The signing made use of more local features of handshape and facial movements than the pantomime, with a higher frequency of transitions between features. It is likely that this relatively detailed information carried in the hands and face that is unique to sign was particularly salient to the infants in Study 1. In conclusion, I propose that infants have an innate interest in attending to these unique properties of language and that this language bias plays a fundamental role in driving the language learning process.