Investigating Native English Speakers’ Perception of Novel Arabic Phonemes after First Exposure

Abstract

This study reports on an experiment carried out to investigate native English speakers' perception of selected Arabic phonemes after first exposure to a controlled naturalistic input of a weather report. It closely follows Brown's model of L2 speech perception and L1 feature geometry (1998), which seeks to relate theories of segmental phonology to L2 speech perception and the first exposure treatment of Gullberg et al. (2010). Eight Arabic sounds were carefully selected for the experiment: /b/-/d/ which are found in both English and Arabic; /x/-/ɣ/ which are not found in English but are distinguished by features which are distinctive in English [dorsal, voice, continuant], and lastly the contrastive pairs /ʔ/-/ʕ/ and /h/-/ħ/, where the latter phoneme in each pair is alien to the phonemic inventory of English. These pairs are distinguished by the feature [RTR (retracted tongue root)] which is lacking in the feature geometry of English. Participants were divided into an Arabic control group, English+ group with prior exposure to Arabic, and an English group with no prior exposure to Arabic. The results from an AX discrimination task confirmed Brown's hypothesis that L2 perception of non-native contrasts is constrained by the L1 feature geometry.