Adaptation to Accents by learning from Mistakes Event Related Potentials during error-based Phonological Learning
No Thumbnail Available
If people are exposed to an unfamiliar accent, they can usually learn to understand it and adapt to its specific patterns of pronunciation very quickly. However, the mechanisms involved in this learning process are still not fully understood. In this study we suggest that such phonological learning is based on external feedback processing and internal monitoring. When we make errors in our phonological perception and receive feedback, we process it and feed it into our internal monitoring, which should become better at detecting errors itself over time, and lead to better performance. We tested a group of Dutch native speakers in a novel accent learning task. Participants were presented with accented forms of Dutch words, pronounced according to an artificially created, novel accent. From two visually presented words, which differed from one another and the accented word only in their vowel, participants had to choose the correct word that corresponded to the accented word they had heard. They received corrective feedback on their choice to learn the mapping of the phonemes of the novel accent. We measured the electrophysiological activity (EEG) and analyzed response-locked (ERN & Pe) and feedback-locked (FRN & P300) data to investigate the mechanisms of this learning process. Participants adapted to the accent and they showed internal monitoring and feedback processing throughout the entire task. Internal monitoring was already present in the beginning of the experiment and no change over time could be found, indicating that learning had already taken place in the very beginning of the experiment and thus that phonological learning is very quick. The results suggest that phonological learning can indeed be described in terms of development of specific internal monitoring processes based on external feedback on erroneous responses.
Faculteit der Sociale Wetenschappen