A produção de obstruintes em coda na interfonologia de falantes brasileiros do PB-L1 e inglês-L2

Abstract

The main objective of this study is to analyze the production of obstruents in coda in Brazilian Portuguese (BP) as a result of the development and use of English as a second language (English-L2). To achieve the proposed objective, two phenomena resulting in the production of obstruents in coda were analyzed: deletion in BP (originating from vowel reduction) and vowel epenthesis in English-L2. We assume that varieties of BP in which deletion is more common favor the production of target syllabic patterns in English-L2 (e.g., the deletion of the final vowel in "sete" [sɛt] would facilitate the production of "set" [sɛt] without epenthesis). Thus, speakers of the Potiguar variety of BP, which exhibits higher rates of deletion than others (Nascimento, 2016; Silva, 2019; Silva; Lima Jr., 2021), would have greater ease in producing obstruents in coda in English. Furthermore, we do not support the justification that deletion in BP is a result of influences from learning English-L2, but rather an emerging phenomenon within the L1 system itself (Nascimento, 2016; Silva, 2019; Silva; Lima Jr., 2021). However, we anticipate that the development of English-L2 could influence deletion in BP in speakers immersed in a context where English is the primary language of communication, leading to language attrition (Schmid; Kopke, 2019). To investigate the validity of our hypotheses, data from 61 participants were collected through sentence readings, categorized based on the following criteria: bilingualism (monolinguals and bilinguals), L1 variant (Potiguar or nonPotiguar), and immersion context (Brazil or the United States of America). This research is based on the view of language as a Complex Dynamic System (Larsen-Freeman, 1997; De Bot et al., 2007; Beckner et al., 2009) and adopts Gestural Phonology (Brownman; Goldstein, 1989; Albano, 2020) as a language production model. The data were analyzed through acoustic analysis, and a logistic regression model with random effects was fitted. The results indicate that deletion is an emerging phenomenon from the reorganization of the linguistic system of BP itself, as monolingual speakers produce obstruents in coda in BP. However, we identified that learning English-L2 favors the production of obstruents in coda in BP, indicating language attrition. We also found that the Potiguar variant is marked by more deletion indices than other BP variants, but the presence of obstruents in coda more frequently in L1 does not favor the production of similar patterns in English-L2. Additionally, we found that vowel epenthesis is not a strategy frequently used by Brazilian bilinguals, as they master the syllabic patterns of English without the need for restructuring during production.