Diversity in coda /R/ and /S/ under internal migration in Brazilian Portuguese

Work on dialect contact shows that, although mobile speakers generally acquire new variants (e.g., Chambers 1992), their production exhibits substantial inter and intraspeaker variation (e.g., Nycz 2018; Walker 2019; Oushiro 2024). This paper applies a Shannon entropy–based analysis of diversity (Shannon 1948; Bane 2008) to examine migrants’ cohesion and dispersion at the speaker and group levels. Data comprise 60 sociolinguistic interviews, balanced for gender and region/status (20 non-mobile Northeasterners, 20 non-mobile Paulistas, 20 Northeastern migrants in São Paulo). All participants are adults aged 20-50 with no college education, and migrants are further balanced for age of arrival and length of residence. Diversity analyses were performed on 10,700 tokens of coda /R/ (e.g. mulher ‘woman’) and 21,691 tokens of coda /S/ (e.g. desde ‘since’), two dialect-diagnostic variables participating in distinct processes: (i) R-fronting ([h] vs. [ɾ, ɻ]); (ii) R deletion vs. realization [h, ɾ, ɻ] (excluding morphemic-R); (iii) S-depalatalization before /t, d/ ([ʃ] vs. [s]); and (iv) S deletion vs. realization [ʃ, s, h] (excluding morphemic-S). Mixed effects logistic regression models confirm that age of arrival significantly conditions the distribution of variants across migrants, and that dialect indexed processes (R-fronting and S-depalatalization) further correlate with length of residence, with earlier arrival and longer residence favoring Paulista like patterns. Normalized Shannon H (norm_H) quantifies diversity on a 0–1 scale (0 = categorical use of one variant; 1 = equal probable use of variants), allowing direct comparison of the complexity of migrants’ repertoires to those of non-migrants. For each process and group, we summarize intraspeaker diversity as the typical H per speaker (mean_H) and interspeaker differentiation as the standard deviation of H across speakers (sd_H). For deletion vs. realization (Figure 1, right panels), migrants’ mean_H is intermediate between Northeasterners and Paulistas, reflecting intermediate within speaker variability, while their sd_H is high, though generally not significantly different from the control-groups. By contrast, for R-fronting and S-depalatalization (Figure 1, left panels), migrants exhibit both the highest mean_H and the highest sd_H. Migrants show more balanced intraspeaker alternation between Northeastern and Paulista like variants than non migrants, who concentrate on their respective regional variants (lower mean_H). Additionally, migrants display the greatest dispersion in norm_H across speakers, with substantial co occurrence of Northeastern and Paulista like realizations across individual repertoires. This yields an “expanded envelope of variation”: for these socially salient variables, the migrant group does not merely interpolate between origin and destination, but adds diversity to the overall system. Thus, deletion processes, which are more gradient and weakly region indexed in Brazilian Portuguese, respond to migration through relatively monotonic shifts in rates, producing intermediate intraspeaker diversity and maintaining interspeaker dispersion. Contrastively, as dialect specific variants carry strong regional and social meanings, they become loci of recombination of norms, generating greater intraspeaker diversity and interspeaker differentiation among migrants than in non-migrant groups. These results stress the importance of variants’ indexicalities in dialect acquisition and support a view of dialect contact as sites of complexification of the variable space, rather than straightforward levelling. References Bane, Max (2008). Quantifying and measuring morphological complexity. In: Chang, C. B.; Haynie, H. J. (eds.). Proceedings of the 26th West Coast Conference on Formal Linguistics. Somerville, MA: Cascadilla Proceedings Project, 69–76. Chambers, J.K. (1992). Dialect acquisition. Language 68(4), 673–705. Nycz, J. (2018). Stylistic variation among mobile speakers: using old and new regional variables to construct complex place identity. Language Variation and Change 30(2), 175–202. Oushiro, L. (2024). Interaction, confound effect, and collinearity in the analysis of Brazilian internal migrants’ speech. In: Fernández-Mallat, V.; Nycz, J. (eds.) Dialect contact: from speaker to community-based perspectives. Georgetown University Press. Shannon, C.E. (1948). A mathematical theory of communication. The Bell System Technical Journal 27, 379–423, 623–656. Walker, A. (2019). The role of dialect experience in topic-based shifts in speech production. Language Variation and Change 31(2), 135–163.