LEARNING AND FORGETTING A SECOND LANGUAGE: THE ACQUISITION, LOSS AND RE-ACQUISITION OF HINDI-URDU NEGATIVE STRUCTURES BY ENGLISH-SPEAKING CHILDREN

This study addresses three questions: (1) how do first language structures affect the acquisition of second language syntax by children? (2) what is the effect, if any, on reacquisition of having previously known and forgotten a language? and (3) how does the sequence of structures traversed in forgetting a second language compare with the acquisitional sequence?;Persuasive evidence is found for first language influence in the acquisition of word order for negated structures in the target language. A prominent early stage of the English-speaking learners is the placement of the negative particle before the object, as in English, rather than after it, as required in Hindi-Urdu. The result is an ordering of constituents in the learners' speech which is unacceptable in Hindi-Urdu, S-NEG-O-V. First language learners of this language do not show this prepredicate placement of the negative particle, indicating that the English-speaking children are operating under the influence of the prepredicate negator placement rule of their first language. A comparison of the acquisitional sequences for negation of the first and second language learners, however, reveals a number of developmental commonalities. With regard to negator choice, both tend to prefer 'nahiiN,' the most frequent of the three Hindi-Urdu negative particles. They make similar errors of negator placement with two-part verbal forms, apparently perceiving these verbs as single units.;Evidence on the effect of prior experience with a language in relearning comes from the older of the two subjects in this study. Data from three periods of learning and relearning after apparent total loss of the second language were examined. This examination reveals a clear facilitation resulting from previous experience, even when the learner has virtually no conscious memory left of the language.;With regard to language forgetting, the data from each of the two subjects indicate a recapitulation in reverse of the acquisitional sequence. The findings suggest that careful "forgetting" studies may provide insights not only into the sequence of developmental stages in learner speech, but into the prominence and duration of each stage in the acquisitional record as well.;Data of three types were used in examining these questions: spontaneously produced speech samples collected from two learners longitudinally, elicited speech samples from these same learners, and elicited speech samples collected only once from a larger group of learners. The longitudinal data on which the study is mainly based were collected from two American children during several sojourns on the Indian subcontinent: three periods for a child who was eight years of age at the completion of the study, and two periods for another who was six. Between these periods of exposure, there was apparent total loss of the second language so that each subsequent exposure required a relearning of the language. The spontaneous speech records are supplemented by speech samples that were elicited from them during their last stay in India and for a year-and-a-half following that. Elicited speech samples were collected from twenty-five additional English-speaking children who were also learning Hindi-Urdu in South Asia; these cross-sectional data supplement the longitudinal data that are the primary evidence used in this study.