Análise do mismatch negativity como índice de discriminação dialetal

Abstract

Linguistic perception is a topic studied across various disciplines. In the case of Variationist Sociolinguistics, the focus is not purely on the perception of sounds, but on what is associated with them: social meaning. To understand how sociolinguistic perception of dialect discrimination works, this research used Electroencephalography, with the Event-Related Potentials (ERP) methodology, in order to analyze whether the presentation of sociolinguistically variable stimuli in the paradigm called ‘oddball’, with native speakers from Rio de Janeiro, would elicit the bioelectrical component Mismatch Negativity (MMN). The variable linguistic phenomenon analyzed was the palatalization of alveolar stops /t, d/ before /i/, in the word ‘tio’, which has a salient variant not used in the speech community of Rio de Janeiro. As a control stimulus, the phenomenon of variable pre-stressed nasalization in the word ‘canal’ was used, whose variants coexist in the city of Rio de Janeiro. After the EEG experiment, explicit (questionnaires on attitudes and knowledge of the region) and implicit (Implicit Association Test – IAT) behavioral tests were also conducted, in order to understand how the neural data correlates with participants' behavior. As a result, we identified an initial peak between 200 and 300 ms corresponding to the classic MMN, which was not modulated by the change in the variant during the identity MMN (iMMN) computation. However, two other peaks were identified: one at 296–358 ms, present in both phenomena analyzed (palatalization and nasalization), and another later and distributed one at 316–388 ms, which only occurred in the case of palatalization and was interpreted as the later perception that the variant does not belong to the Rio speech community. This later peak was not modulated by the information from the IAT nor by the explicit questionnaire; however, the peak shared between the two phenomena was modulated by having friends who use the variant from outside Rio, which suggests that having positive affect toward people who speak that way would increase the capacity for perception of the stimulus when it is in the deviant position. Therefore, these data reveal a temporality of access to sociolinguistic information that provides a basis for new research at the interface between Sociolinguistics and Neuroscience, in addition to offering evidence that sociolinguistic information is accessed after the linguistic information from auditory input. Thus, this research expands the field of interface between the two areas and the understanding of how the brain deals with human language.