The purpose of this study was to examine the efficacy of three days of listen-and-repeat training on the perception and production of vowel duration contrasts. Generalization to an untrained vowel and a non-linguistic sound was also examined. Twelve adults underwent four sessions of listen-and-repeat training over two days with the pseudoword contrast /tite/-/ti:te/. Generalization effects were examined with another vowel contrast, /tote/-/to:te/ and a sinusoidal tone pair as a non-linguistic stimulus. Learning effects were measured with psychophysiological (EEG) event-related potentials (mismatch negativity and N1), behavioral discrimination tasks and production tasks. The results showed clear improvement in all perception measurements for the trained stimuli. The effects also affected the untrained vowel by eliciting an N1 response, and affected the behavioral perception of the nonlinguistic stimuli. The MMN response for the untrained linguistic stimuli, however, did not increase. These findings suggest that the training was able to increase the sensitivity of preattentive auditory duration discrimination, but that phoneme-specific spectral information may also be needed to shape the neural representation of phoneme categories.

Phonological duration differences in quantity languages can be problematic for second language learners whose native language does not use duration contrastively. Recent studies have found improvement in the processing of non-native vowel duration contrasts with the use of listen-and-repeat training, and the current study explores the efficacy of similar methodology on consonant duration contrasts. 18 adult participants underwent two days of listen-and-repeat training with pseudoword stimuli containing either a sibilant or a stop consonant contrast. The results were examined with psychophysiological event-related potentials (mismatch negativity and P3), behavioral discrimination tests and a production task. The results revealed no training-related effects in the event-related potentials or the production task, but behavioral discrimination performance improved. Furthermore, differences emerged between the processing of the two consonant types. The findings suggest that stop consonants are processed more slowly than the sibilants, and the findings are discussed with regard to possible segmentation difficulties.

In quantity languages, the durations of segments affect the meanings of words. This can present problems for second language (L2) learners who do not already have this feature in their native language. This study examines the effects of an intensive, four-week language course with a communicative focus on the perception and production of non-native vowel duration contrasts. A total of 68 students of Finnish, divided into speakers of quantity or non-quantity languages, took part in identification and production tests before and after taking part in the course. The course produced a significant improvement on identification, but not production. Furthermore, a slight advantage was found for speakers of quantity languages in the identification task. Comparison to native control groups revealed significant differences between groups in both tasks. The results are discussed in relation to the interaction of perception and production, L2 learning models and relevance to L2 teaching.