Infant babbling and ambient language.
Item
-
Title
-
Infant babbling and ambient language.
-
Identifier
-
AAI9946158
-
identifier
-
9946158
-
Creator
-
Edmondson, Daisy.
-
Contributor
-
Adviser: Gerald Turkewitz
-
Date
-
1999
-
Language
-
English
-
Publisher
-
City University of New York.
-
Subject
-
Psychology, Developmental
-
Abstract
-
This dissertation was intended to explore the question of whether infants begin to incorporate aspects of the adult target language to which they are exposed into their own vocalizations prior to the emergence of first words and, if so, whether adult listeners are sensitive to such changes. A further question was which specific aspects of the ambient language emerge earliest in babbling, the phonetic or the prosodic.;In each of eight experiments, English-speaking adult judges were asked to listen to tape-recorded samples of vocalizations of infants and to pick out which infants they believed were from English-speaking homes. Samples were used from 8-month-old Arabic, Chinese, French, and American English-environment infants.;Samples were selected on the basis of duration of utterance alone; there were no other restrictions on the characteristics of the samples. As a result, samples differed on a number of attributes, including consonant number. In the first study, which contrasted English and Arabic, easy discrimination was found. However, the English and Arabic samples differed widely with respect to number of consonants. Subsequent studies were designed to examine the possibility that consonant number served as the basis for discrimination. Limited support was found for this explanation: Some samples that did not differ significantly with regard to consonant number were discriminated, while others that did differ significantly were not. The possible role of segmental and nonsegmental information in discriminating between the vocalizations of infants from different language backgrounds was assessed by low-pass filtering or scrambling some samples. The filtered samples were discriminated as easily as the full-spectrum versions, while scrambled samples paired with their unscrambled counterparts were not discriminated. The findings that the discriminability of some samples was affected by the context (i.e., that a sample might be more easily discriminated when contrasted with French than with Arabic) and that the masking of specific segmental information through low-pass filtering affected judges' ability to correctly label the samples seem to require an explanation that takes into account the influence of the ambient language.
-
Type
-
dissertation
-
Source
-
PQT Legacy CUNY.xlsx
-
degree
-
Ph.D.
