PATTERNS OF PRONUNCIATION ERRORS IN ENGLISH BY NATIVE JAPANESE AND HEBREW SPEAKERS: INTERFERENCE AND SIMPLIFICATION PROCESSES.
Item
-
Title
-
PATTERNS OF PRONUNCIATION ERRORS IN ENGLISH BY NATIVE JAPANESE AND HEBREW SPEAKERS: INTERFERENCE AND SIMPLIFICATION PROCESSES.
-
Identifier
-
AAI8614653
-
identifier
-
8614653
-
Creator
-
BASSON, SARA H.
-
Contributor
-
Michael Studdert-Kennedy
-
Date
-
1986
-
Language
-
English
-
Publisher
-
City University of New York.
-
Subject
-
Speech Communication
-
Abstract
-
Adult second language (L2) learners often exhibit phonetic deviance in English. Pronunciation errors accounting for reduced intelligibility and perceived accent in native Japanese and Hebrew speakers were investigated. Four native Japanese, four native Hebrew, and four native English male speakers recorded syllables and sentences in English. The Japanese and Hebrew speakers also recorded syllables in their native languages. The segments forming the English syllables were classified in terms of their contrastive status, that is, in terms of whether phonemically identical sounds occur in similar phonetic environments in the speaker's first language (L1). The recorded tokens were randomized and presented to 36 experienced listeners to transcribe and rate in terms of degree of accentedness. Transcriptional errors were interpreted as production errors of the non-native speakers. Two models derived to account for L2 learners' errors were evaluated. The first assumes that errors reflect L1 interference. The second model predicts that certain sounds are inherently more difficult than others, based on cross-linguistic data and findings from L1 acquisition. Of all consonant groups, fricatives, liquids, and clusters were most deviant for Japanese speakers. These findings are consistent with an interference model, though individual consonants within the consonant classes were not all equally deviant. Vowels were inaccurately transcribed more often than consonants for all language groups, and articulatory explanations for vowel errors are discussed. The Japanese speakers performed worse on vowels that do not appear in their native inventory; Hebrew speakers performed poorly on the entire set. Measures of first and second formants reveal that L1 interference largely accounted for vowel deviance in Japanese speakers. The Hebrew speakers centralized the entire back set of vowels irrespective of contrastive status, which may reflect a natural simplification process. The presence of contrastive segments in an utterance did not affect accentedness ratings, though overall intelligibility was impaired. Accentedness ratings for single syllables and sentences correlated highly, suggesting either that segmental information is critical to perceived accent, or that speakers had similarly good or poor control over both parameters.
-
Type
-
dissertation
-
Source
-
PQT Legacy CUNY.xlsx
-
degree
-
Ph.D.
-
Program
-
Speech and Hearing Sciences
