Title

Sensitive period effects on the acquisition of language: A study of language development

Date of Completion

January 2003

Keywords

Language, Linguistics|Education, Educational Psychology

Degree

Ph.D.

Abstract

A sensitive period for first language acquisition has been proposed and previously supported primarily by case studies of social isolates and studies with Deaf adults who were exposed to American Sign Language (ASL) during mid- to late-childhood. Although informative, case studies with hearing, social isolates are confounded by physical abuse experienced by the children. Studies with Deaf adults do not show the development of language acquisition under the condition of delayed input. There is now new evidence for sensitive period effects on first language acquisition from two unrelated children, MEI and CAL. MEI and CAL were not exposed to a first language until approximately 6 years of age. There is no history of physical abuse—just a misdiagnosis of mental retardation instead of deafness. MEI and CAL, once exposed to language, were immersed in ASL. ^ The results of filming MEI and CAL for 3 1/2 years, from the beginning of their language acquisition process, suggest that sensitive period effects are seen with at least one specific aspect of language—the formal syntactic features (Chomsky 1995). Formal syntactic features are found in different domains of language, including verb agreement, word-order changing mechanisms, and null referents. Analyses of MEI's and CAL's naturalistic language production data, along with preliminary experimental results, reveal difficulties with precisely these domains. MEI and CAL have a higher overall percentage of errors per sample than the two native-signing Deaf comparison children. MEI and CAL made most of their errors with agreeing verbs. This class of verbs is the only one that marks syntactic features in ASL. MEI and CAL attempted fewer utterances with word order variations, suggesting a difficulty with the formal features that trigger some of the word order change mechanisms. Finally, MEI and CAL produced utterances with incorrectly null referents more often than the native signers, again implicating a difficulty with the formal features needed to trigger the syntactic licensing of null elements. ^ The results from this present study, combined with the results from the studies with Deaf, adult late-learners, suggest that sensitive period effects exist, are specific, and are long-lasting. ^