Computer-based auditory training improves second-language vowel production in spontaneous speech.

The current study examined the effectiveness of computer-based auditory training on Greek speakers' production of English vowels in read sentences and in spontaneous speech. Another group of Greek speakers served as controls. Improvement was evaluated pre- and post-training via an identification task performed by English listeners and by an acoustic analysis of vowel quality using a combined F1/F2 measure. Auditory training improved English vowel production in read sentences and in spontaneous speech for the trained group, with improvement being larger in read sentences. The results indicate that auditory training can have ecological validity since it enhances learners' production beyond the (read) sentence level.

The effect of native vowel processing ability and frequency discrimination acuity on the phonetic training of English vowels for native speakers of Greek.

The perception and production of nonnative phones in second language (L2) learners can be improved via auditory training, but L2 learning is often characterized by large differences in performance across individuals. This study examined whether success in learning L2 vowels, via five sessions of high-variability phonetic training, related to the learners' native (L1) vowel processing ability or their frequency discrimination acuity. A group of native speakers of Greek received training, while another completed the pre-/post-tests but without training. Pre-/post-tests assessed different aspects of their L2 and L1 vowel processing and frequency acuity. L2 and L1 vowel processing were assessed via: (a) Natural English (L2) vowel identification in quiet and in multi-talker babble, and natural Greek (L1) vowel identification in babble; (b) the categorization of synthetic English and Greek vowel continua; and (c) discrimination of the same continua. Frequency discrimination acuity was assessed for a nonspeech continuum. Frequency discrimination acuity was related to measures of both L1 and L2 vowel processing, a finding that favors an auditory processing over a speech-specific explanation for individual variability in L2 vowel learning. The most efficient frequency discriminators at pre-test were also the most accurate both in English vowel perception and production after training.

Perceptual assimilation and L2 learning: evidence from the perception of Southern British English vowels by native speakers of Greek and Japanese.

This study examined the extent to which previous experience with duration in first language (L1) vowel distinctions affects the use of duration when perceiving vowels in a second language (L2). Native speakers of Greek (where duration is not used to differentiate vowels) and Japanese (where vowels are distinguished by duration) first identified and rated the eleven English monophthongs, embedded in /bVb/ and /bVp/ contexts, in terms of their L1 categories and then carried out discrimination tests on those English vowels. The results demonstrated that both L2 groups were sensitive to durational cues when perceiving the English vowels. However, listeners were found to temporally assimilate L2 vowels to L1 category/categories. Temporal information was available in discrimination only when the listeners' L1 duration category/categories did not interfere with the target duration categories and hence the use of duration in such cases cannot be attributed to its perceptual salience as has been proposed.