Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Analysis of Mismatch Negativity as an index of dialect discrimination
Poster Session E, Friday, October 2, 11:00 am - 1:00 pm, Wangari Maathai
Pedro Santos1, Aniela França1; 1Federal University of Rio de Janeiro
Language perception has been examined from multiple disciplinary perspectives. Within Variationist Sociolinguistics, however, the focus is not simply how sounds are perceived, but how they are socially interpreted. To investigate the temporal and physiological dynamics of dialect discrimination, this study used Event-Related Brain Potentials (EEG-ERP) methodology to test whether the presentation of sociolinguistically variable stimuli in the oddball paradigm would elicit the Mismatch Negativity (MMN) component in native speakers from Rio de Janeiro, Brazil. The linguistic phenomenon under investigation was the palatalization of the alveolar stop /t/ before /i/, such as in the word 'tio' (“uncle”). The salient non-palatalized variant is categorically absent from the speech community of Rio de Janeiro. As a control condition, we examined the pre-stressed nasalization /ã/ in the word 'canal' (“channel”), a variable whose oral and nasal realizations may coexist even within the speech of the same Rio speakers. After the EEG experiment, participants also completed explicit behavioral measures (questionnaires probing attitudes and regional knowledge) and an implicit measure (Implicit Association Test – IAT) allowing us to explore potential correspondences between neural responses and behavioral patterns. The results revealed an initial peak between 200 and 300 ms corresponding to the classic MMN, that was not modulated by either variant of either phenomenon in deviant position when identity MMN (iMMN) was calculated by subtracting each variant in standard position from the same variant in deviant position. However, two later effects were identified: A first between 316–388 ms, for both palatalization and nasalization; then, most strikingly, a second broader effect, spanning 444 to 654 ms, was found but only for the palatalization condition. We interpreted this later response as reflecting delayed recognition of the non-self, that is, a realization that the stimulus was not produced by someone from the Rio de Janeiro speech community. This late effect was not modulated by either the IAT or the explicit questionnaire. However, the earlier peak, shared by both variables, was modulated by whether participants had friends from outside Rio who use the relevant variant. This suggests that positive social affect toward speakers of other varieties may enhance sensitivity to deviant sociolinguistic stimuli. Taken together, these findings reveal a neural time course for dialect discrimination and open a promising line of research right at the interface between Sociolinguistics and Neuroscience of Language.
Topic Areas: Speech Perception, Phonology