Phonetic Training Strategies for Non-Native Speech Perception

Native language experience affects the perception of non-native sounds and can inter- fere with learning sounds in a new language. Targeted training with acoustic dimensions that underlie phonemic contrasts can help improve non-native speech perception. The assumption of most phonetic training studies to date is that training strategies should mirror the perceptual strategies of native speakers. The ways in which a learner’s first language may influence their ability to utilize specific acoustic cues is rarely taken into account when deciding on a phonetic training strategy. This dissertation research directly addresses this oversight. Two perceptual training experiments were run in order to test the efficacy of acoustic training dimensions that differed in regards to prior perceptual experience of the learners. Specific mechanisms of category learning are demonstrated through a variety of statistical approaches, including a novel application of multi-dimensional scaling. The results of these experiments show that it is important to consider both prior perceptual experience and the specific features of target contrasts when developing a phonetic training strategy. These results have implications for models of speech perception based on attention to acoustic cues, and can inform theories of how native language experience affects non-native speech perception