Portuguese-Spanish Language Mixing in the Venezuelan Diaspora in Brazil: A Variationist Account of Cognate Languages in Contact

Abstract

This dissertation examines the presence of Spanish in the Portuguese variety used by Venezuelan migrants residing in Brazil and analyzes the relationship between these elements and linguistic (cognate status and semantic classification of the NP head) and extralinguistic factors (gender, length of residence, social integration and Portuguese use index and age of arrival to Brazil). This study enhances our understanding of language mixing patterns in L2 varieties, specifically of lexical borrowings (Poplack, 2018), code-switching (Poplack, 1980), loanblends (Haugen, 1953), and congruent lexicalizations (Muysken, 2000). Additionally, it investigates the frequency and diffusion of these borrowings.Results are based on 32 sociolinguistic interviews conducted with Venezuelans residing in the Campinas region in Brazil. Tokens were extracted from all noun phrases (NPs) in subject positions. The variants considered in the analysis were: a) a NP entirely in Portuguese, b) a NP in Portuguese with the head borrowed from Spanish (lexical borrowings), c) a NP in Portuguese with a loanblend in head position, d) a NP with stretches of Spanish, or code-switchings, and finally, e) NPs with elements and structure from both Spanish and Portuguese (congruent lexicalizations). 100 NPs per speaker were extracted from the transcribed corpus, which encompassed a total of 350,000 words, resulting in 3,132 tokens in total. The data was submitted to nine multivariate logistic regression analyses and several raw frequencies analyses with the objective of shedding light on to what extent were the different types of language mixing used, and what linguistic and extralinguistic influenced the production of borrowings, code-switching, loanblends, and congruent lexicalizations. Contrary to expectations of a high prevalence of hybrid language mixing phenomena due to high cross-linguistic permeability between Spanish and Portuguese (Lipski, 2009), the findings indicate that NPs entirely in Portuguese are the most common variant (approximately 65%). Code-switching follows at 15.1%, while lexical borrowings, congruent lexicalizations, and loanblends account for 10.3%, 6.4%, and 3.2%, respectively. The main predictors for borrowings, code-switching, loanblends, and congruent lexicalizations are length of residence and social integration and Portuguese use index, two extralinguistic factors operationalized to account for different degrees of exposure to Portuguese. As expected, those who have lived in Brazil for shorter durations (4 months or less and 5 to 14 months) and have limited interactions with native Portuguese speakers tend to use all language mixing phenomena more frequently. In addition, women are more inclined towards borrowings, aligning with their role as language innovators and agents of linguistic change (Labov, 2001). In terms of linguistic factors, cognate status significantly influences the use of borrowings (favored by cognates and semicognates) and loanblends (favored by semicognates). The results also reveal the prevalence of recurrent and widespread borrowings, aligning with Poplack’s (2018) findings, despite the language circumstances and the social contexts being different. This observation further emphasizes the influence of community norms within this particular speech community. The presence of code-switching is favored by a slightly different set of factors. Surprisingly, cognates also trigger code-switching, suggesting that they not only prompt code-switching immediately before the code-switching stretch (as explained by Broersma & De Bot, 2006) but also when functioning as NP heads in code-switches. Interestingly, men exhibit a higher usage of code-switching, while women employ more lexical borrowings. This suggests that while women lead in adopting linguistic innovations through borrowings, they still adhere to more standard language patterns compared to men, as code-switching represents a more radical form of language mixing by incorporating not only lexical items from the LD (Spanish) but also following its structure, whereas borrowings adhere to the structure of the LR (Portuguese). Finally, congruent lexicalizations show similar social patterns to code-switching. In addition to shedding light on the formation of community bilingualism, this study adds to the field of L2 acquisition of cognate languages by submitting L2 data to multiple variationist analyses that reveal systematicity underlying lexical variation. Overall, this dissertation contributes to the growing body of knowledge on language mixing, providing empirical evidence that deepen our understanding of the complex dynamics of language contact between cognate languages in an L2 acquisition context of immersion.