Native English speakers' perception of Arabic emphatic contrasts

Abstract

In Arabic, emphasis (secondary velar/pharyngeal constriction) distinguishes some consonants. Native Jordanian Arabic speakers have been shown to rely more heavily on the rime than the onset of CVC syllables when identifying plain versus emphatic onsets (Jongman et al. 2011). We investigated whether native English speakers similarly rely on the rime when discriminating Arabic plain-emphatic pairs. We also investigated the influence of vowel quality on discrimination performance. Native English speakers (no Arabic experience) performed an AXB task involving cross-spliced CVCs with plain/emphatic onsets/rimes. Our subjects also relied more heavily on the rime than on the onset; this effect was most robust when the V was /æ/, followed by /u/ and /i/. A cross-language vowel identification task revealed that subjects identified Arabic /æ/ in emphatic contexts as systematically different English vowels than in plain contexts, with only 10% overlap in vowels identified. The overlap for /i/ and /u/ was much higher, at 84% and 91%, respectively. We thus found that native English listeners, like native Arabic listeners, rely on the rime to make onset emphasis judgments, this effect is moderated by vowel, and the influence of the preceding vowel may be related to the mapping between vowel allophones and English vowel categories.