Non-segmental structure in foreign accent

Abstract

Typological studies have shown that “voicing” contrast for obstruents is more marked in syllable-final position and least marked in initial position. Also, implicational universal relations suggest that fricatives are more marked than stops and voiced segments are more marked than voiceless ones. The markedness differential hypothesis (MDH) whose underlying claims stem from such studies predicts that syllable-final voicing contrast should be difficult for Germans speaking English since German neutralizes voicing in a marked position. MDH, also, predicts more difficulty for Arabs with English /v/ than with /p/, which are absent from Arabic phonemic inventory, on the basis of markedness. A spectrographic test of English minimal pairs spoken by ten Germans, however, revealed that Germans have acquired control of English voicing in syllable-final position. Another spectrographic test of English minimal pairs spoken by eight Arabs revealed that they had considerable difficulty learning the correlates of English voicing. These results prove that MDH can not correctly predict the relative degree of difficulty in second language aquisition. Instead we propose that nonsegmental similarities and/or differences between languages play an important role in second language acquisition. Thus, due to the similarity between English and German and the differences between English and Arabic with respect to vowel duration difference as a function of voicing, Germans contrasted English obstruents for voicing but Arabs did not. [Research supported by Yarmouk University.]