Intelligibility of English “voicing” produced by Arabs

Abstract

The implicit claim of traditional approaches to the phonetic and phonological aspects of second language acquisition by adults is that pronunciation difficulty occurs almost exclusively at the abstract level of segmental features. Thus the “contrastive analysis hypothesis” predicts pronunciation difficulty for second language learners when the phonetic inventory of the native language lacks a phoneme of the target language. It would predict, then, that English /p/ and /v/ should be difficult for Arabs since Arabic lacks these two phonemes. An intelligibility test, however, revealed that our native English listeners had no serious difficulty identifying “voiced” from “voiceless” obstruents in the English minimal pairs spoken by ten Arabs. Our experimental approach, which is based on the very theoretical framework of traditional approaches to the analysis of foreign accented speech, nevertheless yielded results which counter the predictions of that framework. It also verifies the results of recent research on foreign accent [F. Mitleb, J. Acoust. Soc. Am. Suppl. 1 73, S29 (1983)] that nonsegmental level differences and/or similarities between languages play an important role in second language learning. [Research supported by Yarmouk University, Irbid, Jordan.]