'Comendo o final das palavras': análise variacionista da haplologia, elisão e apócope em Itaúna/MG

Abstract

Do Mineiros "eat" the end of words? This is certainly one of the characteristics of the way people speak in Minas Gerais. The aim of this study is to investigate this linguistic feature through the analyses of the speech produced in Itaúna Minas Gerais. The main objective ofthis study is therefore to present a variationist analysis of the process of vowel deletion and CV unstressed final syllable deletion in paroxytones, seeking to identify, analyze and explain the factors that may favor this processes. The analyses of the results were carried out under thetheoretical background of Variationist Sociolinguistics (Labov, 1972), Autosegmental Phonology (Goldsmith, 1990), Metric Phonology (Liberman and Prince, 1977), Prosodic Phonology(Selkirk, 1984). The data analyzed in this study came from two different sources: spontaneous speech and production tests. Concerning spontaneous speech, the following elements wereanalyzed: vowel and syllable deletion before the consonant (haplology and syllabic elision); the vowel deletion before pause (apocopy) and the vowel deletion before vowel (elision of vowel anddegemination). In order to analyze the data and the selection of statistically significant variables, the model of multilevel regression (logistic and multinomial) was used. The resultsindicate that the processes of extinguishing unstressed final syllable occur in a higher percentage than it was observed in previous studies carried out in different regions in Brazil. The study also shows that processes under analysis can be classified as examples of stable variety, once there is no statistically significant difference among the age groups. The results also pointed to the fact that male gender favors the processes of elision (syllabic and vocalic). This finding is alignedwith Labov (1990), who claims that, in processes of stable variation, men tend to use nonstandard forms more frequently than women. Concerning the analyses of linguistic variables, this study also claims that, in haplology, in syllabic elision and apocopy, the process ofextinguishing vowels is the same process of weakening this vowel (higher vowels are more extinguished). The principle of obligatory contour (Leben, 1973) helps to explain syllable deletion when the involved consonants share place and manner of articulation (haplology). Theresults indicate that, in these processes, a fusion of consonants occurs after the deletion of a vowel, following Sá Nogueira (1958). In haplology, metric prosodic and morphosyntactic aspects did not display further significance. In syllable deletion, in cases in which the involvedconsonants shared only place, only manner or did not share any phonetic property (a process named syllabic elision), the fostering of metric aspects was identified. This study confirms that there are other motivations for syllabic elision besides the phonetic phonological motivation (motivations particularly related to grammaticalization processes). In this regard, the results show that phonetic reductions in items under grammaticalization processes may not be anindication of such processes. Besides being favored by high vowels, the apocopy is associated with syllabic elision, as the effect of the consonant of the CV syllable in the extinguishing of segments in such syllable was proved to be statistically co-related in both processes. The voweldeletion in vocalic elision and in degemination is a consequence of the crash of syllabic nucleus (Bisol, 1996). There was no indication of an association between such extinguishing and the vowel height. The degemination can be explained by the principle of obligatory contour. InItaúna, the elision of [W] is not limited to the inner nature of clitic groups, as proposed by Bisol (2005). In the analysis of test production, a directly proportional relation between speech rate and segments deletion was also found. The existence of compensatory stretching betweenstressed and unstressed vowels was also identified. Also, the study has shown that this phenomenon is more frequent in Belo Horizonte than in Rio de Janeiro, confirming the hypothesis presented by Viegas, Almeida and Dias (2009) that the stressed vowels in Mineiro speech are longer, while the unstressed vowels are shorter, if compared to other speakers of Brazilian Portuguese.