In this paper we analyze the variable regressive palatalization of /t, d/ in a variety of Brazilian Portuguese (BP) in contact with Italian dialects. We follow Coetzee’s (2016) assumptions regarding the distinctive effects of grammatical and non-grammatical factors on the grammar. The aim of the study is to test such claims with Noisy-HG (BOERSMA; PATER, 2008; COETZEE, 2012; 2016; COETZEE; KAWAHARA, 2013). Palatalization in BP affects /t, d/ and it is triggered by a following underived /i/ (t/i/jolo ‘brick’, d/i/nheiro ‘money’) or derived [i] from /e/ in unstressed word positions (t/e/atro ® t[i]atro ‘theater’, nád/e/ga ® nád[i]ga ‘buttock’). We focus on palatalization triggered by derived [i]. Besides tokens of non-palatalized consonant plus non-raised vowel ([te]atro, ná[de]ga) and of palatalized consonant plus raised vowel ([ʧi]atro, ná[ʤi]ga), the data set comprises tokens of non-palatalized consonant plus raised vowel ([ti]atro, ná[di]ga), an innovation of the analysis regarding prior studies (BATTISTI; DORNELLES FILHO, 2010; GUTIERRES; BATTISTI; DORNELLES FILHO, 2018). The data were extracted from sociolinguistic interviews by Battisti et al. (2007). The linguistic constraints interacting in the grammar of palatalization come from Battisti and Dornelles Filho (2010). The analysis demonstrates that a non-grammatical factor as ‘place of residence’ works as a scaling factor on faithfulness constraints, moving their weights up or down and so affecting the variable raising of /e/, a process that feeds palatalization.
Read full abstract