Browsing by Author "Nurmi, Pietari"

The aim of this study is to examine the learning mechanisms and acquisition of non-native phoneme contrasts in young adults using neurophysiological and behavioral methods. According to the traditional view, acquiring novel phonemes after the sensitive periods in the early childhood is very difficult. However, later findings have shown that foreign phoneme contrasts can be learned at a later age, too. Acquiring new phonemic categories requires neuroplastic changes in the brain. Neurophysiological studies have examined the brain’s ability to differentiate between closely related phonemic categories at the early stage of spoken language processing by measuring, for example, event-related mismatch negativity responses (MMN). MMN, or its magnetic equivalent MMNm, is elicited when the brain registers a difference in a repetitive sensory stimulus. Studies have shown that even a moderate amount of auditory training with closely related foreign phonemes improves the brain’s ability to discriminate between them resulting in enhanced MMN or MMNm responses. In this experiment the neural mechanisms of foreign language phoneme acquisition and the learning-related neuroplastic changes were studied using magnetoencephalography (MEG) and neuromagnetic evoked responses (MMNm in particular). 20 Finnish subjects were measured in the experiment. Their task was to learn to differentiate between acoustically closely related Russian fricatives Ш /ʂ/ and Щ /ɕ(ː)/. The subjects’ differentiation skills were first tested in a behavioral task where Russian pseudoword minimal pairs were presented to them auditorily. The first phoneme in the word pairs was varied and the subjects had to report whether they heard a difference between the words or not. The same stimuli were then presented in a passive MEG task where the brain’s change detection responses were tested in an unattended situation as the subjects were watching a silent film. After the measurement the subjects practiced the phonemes at home for approximately one week by playing a learning game by computer. After training they were measured again. Structural magnetic resonance images of the subjects’ brain were also measured for MEG source localization purposes. Behavioral discrimination ability of the experimental phonemes was considerably worse than with familiar control phonemes. The discrimination skills seemed to improve by training, but the difference was not statistically significant. Contrary to the hypotheses, statistically significant MMNm responses were not found before or after training. No significant differences were found in other auditory MEG responses or their neural source current distributions between the measurements either. However, individual differences in learning were sizeable. For the subjects who improved their performance in the behavioral task a modest training-related boost in the auditory responses supporting the hypotheses could be observed. Although very small and statistically insignificant, the effect was opposite for control stimuli and did not exist in the non-learner group suggesting some sort of change in neural processing in the learner group. This study was not able to replicate the findings from various previous studies on phoneme acquisition in adulthood. Although it is likely that certain methodological limitations (e.g. small number of stimulus repetitions, challenging stimuli) affected the significance of the results, based on this study the generalizability of some of the previous findings can be called into question.