Internal migrants in Brazil: a few open questions in the study of dialect contact

Resumo

This talk is an overview of studies on dialect contact in Brazil, a setting where most migrants are working-class individuals moving from rural areas to urban centers and dialect contact involves not only different regional varieties, but crucially socioeconomic differences between speakers. I review several works discussing the effect of gender, age of arrival, length of residence, and attitudes on variable patterns in migrants’ speech, from which certain generalizations seem to emerge, along with a number of questions they open for future studies. Gender differentiation in migrant communities has shown quite mixed results. While the first works reported men ahead of women in the acquisition of urban prestigious variants (Bortoni-Ricardo 1985, Rodrigues 1987), more recent studies often show women in the lead (e.g., Cardoso 2009, Silva 2016, Possatti 2020) or lack of correlation with gender (e.g., Guedes 2019, Santana 2021). In two samples of Northeastern migrants in São Paulo, Oushiro (2020a, 2020b) found that women disfavor the original variant in cases of variables associated with ruralness (/t d/ stopping tia ‘aunt’ and nonstandard nominal agreement os menino-0 ‘the boys’) rather than Northeastern-ness (aspirate coda /r/ porta ‘porta’ and double sentential negation não vi não ‘I haven’t seen’), signaling the importance of variables’ indexicalities in understanding these patterns. Length of residence is by far the most analyzed predictor in Brazilian studies, with correlations in the expected direction (the longer the length of residence, the greater use of the host community trait: e.g., Marques 2006, Chacon 2012, Lima and Lucena 2012), but most works have not properly controlled for the confounding effect of age of arrival. When duly controlled, age of arrival stands out as more important a predictor than length of residence for phonetic variables, but not for morphosyntactic ones (Oushiro 2020c). Finally, attitudes are often called upon as a determinant factor in dialect acquisition (e.g. Alves 1979, Cardoso 2009, Lucena 2017), but not infrequently as an ad hoc explanation for unexpected patterns. A challenge for dialect contact studies is how to ascertain migrants’ attitudes, as these are dynamic and speakers are socially expected not to “deny their roots”. Comparing patterns for Northeasterners in the Southeast and Southeasterners in the Northeast for the same variables shows that Southeasterners accommodate at a much lower rate. But this is a proxy for a much larger investigation involving individuals, interlocutors, identities, and social meanings. Thus previous results point to the importance of variables’ indexicalities, but most works so far have focused on patterns of production, while evaluations and perceptions have been mostly analyzed qualitatively. Studies on dialect contact require the control of a possibly larger set of predictor variables, which are frequently non-orthogonal and at times interact with one another. And dialect contact in the global south often require close attention not only to regional dialects, but also to rural-urban and socioeconomic differences embedded in mobile speakers’ rich linguistic repertoire.