Finding the right words: an investigation with electrophysiological measures of the role of statistical cues on language acquisition

• Grant number: 21/13731-0
• Support Opportunities: Scholarships in Brazil - Master
• Start date: May 01, 2022
• End date: April 07, 2024
• Field of knowledge: Humanities - Psychology - Human Development Psychology
• Principal Investigator: Débora de Hollanda Souza
• Grantee: Isabella Toselli Prequero Sbrocco
• Host Institution: Centro de Educação e Ciências Humanas (CECH). Universidade Federal de São Carlos (UFSCAR). São Carlos , SP, Brazil
• Associated research grant: 14/50909-8 - INCT 2014: Behavior, Cognition and Teaching (INCT-ECCE): relational learning and symbolic functioning, AP.TEM
• Associated scholarship(s): 23/02238-6 - Statistical language learning and electrophysiological measures: an international collaboration with the Brainvitge Group at Universitat de Barcelona, BE.EP.MS

Abstract

Recent evidence shows that individuals detect regularities present in speech and the environment from a very early age, and they use them to overcome some challenges of the language acquisition process, such as speech segmentation and the mapping of words to their referents in the world. In a series of unprecedented experiments with adult Brazilian-Portuguese speakers, researchers investigated the effects of the interaction between transitional probability and phonotactic probability in a simultaneous word extraction and mapping task with visual stimuli across ambiguous trials. When exposed to a language with misaligned transitional and phonotactic probabilities, participants were not able to segment speech, but they correctly mapped the words to their referents. These results suggest that participants learned the words without being aware of what was learned, given that they were not successful in the segmentation test alone, and such process is described in the literature as implicit learning. Although there is still much debate about what implicit language learning is and how it occurs, the literature suggests that neuroimaging and neurophysiological recording measures can contribute to a better understanding of the phenomenon. Therefore, the present project aims to integrate the recording of electrophysiological measurements into a speech segmentation task. Two languages will be used: the first will have aligned transitional and phonotactic probabilities and the second one will have misaligned probabilities. The procedure will be divided into three phases: familiarization, implicit test (violations of the language words), and explicit test. It is expected that the electrophysiological data (N200 and FN400 ERPs) will reveal whether participants identify the structure of the languages and learn the words, regardless of their performance in the explicit test. (AU)