STATIC AND DYNAMIC INFORMATION IN VOWELS PRODUCED BY THE HEARING IMPAIRED (ACOUSTICS, PERCEPTION, VARIABILITY).
Item
-
Title
-
STATIC AND DYNAMIC INFORMATION IN VOWELS PRODUCED BY THE HEARING IMPAIRED (ACOUSTICS, PERCEPTION, VARIABILITY).
-
Identifier
-
AAI8409415
-
identifier
-
8409415
-
Creator
-
RUBIN, JUDITH ANN.
-
Contributor
-
Katherine Harris
-
Date
-
1984
-
Language
-
English
-
Publisher
-
City University of New York.
-
Subject
-
Health Sciences, Speech Pathology
-
Abstract
-
To understand more about the underlying deviant nature of vowel production by the hearing impaired, hearing and hearing-impaired talkers were compared with respect to (1) intrasubject variability in vowel production and (2) the information used by listeners in decoding their vowels. Fifteen repetitions of a set of sven vowels (/i,I, , , , ,u/) were spoken by six hearing-impaired and two normal hearing children. Each vowel target was produced in a fixed consonant context (/bVb/) within a carrier phrase. F1, F2, f0, and duration measurements were obtained as indicators of articulatory precision and the presence or absence of articulatory/acoustic vowel targets. Experienced and inexperienced listeners were asked to identify each vowel token and the hearing status of the talker under conditions providing controlled amounts of dynamic vocal tract information (Sentence, Word and Gated Vowel Conditions). Results of the acoustic analyses suggest that while intrasubject variability was greater than normal, hearing-impaired talkers were not random in their employment of articulatory gestures for vowel production. Instead, they appear to have developed vowel systems based on a reduced number of distinct targets. Three 'styles' of vowel production were identified. In virtually all cases, targets were no better specified in an alternative and/or multidimensional acoustic space (that is, by considering f0 and/or duration measures) than in the standard F1/F2 vowel space; thus providing no support for the notion of a deviant but well-defined phonology. Results of the perceptual experiment suggest that listeners use the same acoustic sources of information in decoding vowels produced by hearing and hearing-impaired talkers. There was no significant difference between experienced and inexperienced listeners in any of the perceptual tasks. Results indicate that while the ability to identify the talker as hearing-impaired decreases as a function of removing dynamic articulatory information, performance in the Gated Vowel condition was significantly better than chance. Thus, there is apparently information in the steady state signal which is relevant to this perceptual task.
-
Type
-
dissertation
-
Source
-
PQT Legacy CUNY.xlsx
-
degree
-
Ph.D.
-
Program
-
Speech and Hearing Sciences
