Advanced search
Start date

Monophthongization, diphthongization and hiatus resolution: a study with real and nonsense words in the Portuguese spoken in São Paulo

Full text
Lucas Pereira Eberle
Total Authors: 1
Document type: Master's Dissertation
Press: Campinas, SP.
Institution: Universidade Estadual de Campinas (UNICAMP). Instituto de Estudos da Linguagem
Defense date:
Examining board members:
Maria Filomena Spatti Sândalo; Andrew Nevins; Seung Hwa Lee
Advisor: Maria Filomena Spatti Sândalo

This dissertation has as its study object the vowel sequences, diphthongs, and hiatuses of Brazilian Portuguese. It consists of the analysis of the following phenomena: (i) the reduction of oral diphthongs to a simple vowel from the deletion of the glide; (ii) the epenthesis of [j] in syllables ending in /s/; (iii) the epenthesis of a homorganic glide between hiatus; (iv) the elevation of V1 in hiatus. The first phenomenon is called Monophthongization, the second is Diphthongization, and the third and fourth is Hiatus Resolution. Such analysis sought to investigate whether these phenomena were sensitive to the positional prominence and acoustic dispersion of vocoids. According to Beckman (1998, 1999), Smith (2005), and Becker et al. (2018), stressed syllables, initial syllables, and monosyllables are positions of high psycholinguistic and phonetic prominence, hence, they demand prominent segmental material, i.e., in Monotongation, those positions would have more resistance to losing the glide, in Diphthongization, they would facilitate epenthesis, and in Hiatus Resolution, they would prefer resolution through epenthesis rather than raising the V1, as the diphthong is more prominent than a single vowel. Regarding the acoustic dispersion, according to Flemming (2004), Nevins (2012), Becker et al. (2018), and Casali (2011), low contrast vowel sequences (both in sonority and in the acoustic dimensions F1 and F2) have a lower preference in languages because they are perceptually less distinctive and, consequently, they are often avoided. In order to test the hypotheses that positional prominence and acoustic dispersion indeed influence those phenomena, experiments were conducted to judge the acceptability of Monophthongization, Diphthongization, and Hiatus Resolution in real words and nonsense words (invented words that obey the phonotactics of the language). The experiments consisted of a Yes/No test in which participants were presented with different pronunciations of the same word and instructed to rate them as natural or unnatural. The obtained results with the statistical analysis, although asymmetric, corroborated the hypotheses. Monophthongization was often avoided in strong positions and the diphthongs in which the reduction was considered acceptable were those with small contrasts in F2. Diphthongization was more accepted in strong positions, and the Hiatus Resolution through epenthesis was often welcomed in strong positions while resolution through V1 elevation was more accepted in weak positions. The Maximum Entropy Grammar was the reference to model the grammar that explains these phenomena, unlike Optimality Theory, it allowed the modeling of grammar with variation, i.e., with the possibility of more than one output. By assigning weights to the constraints and using the obtained results, it is possible to predict which output will have the highest occurrence using the MaxEnt Grammar Tool. Although the effects of positional prominence and acoustic dispersion are not uniform and symmetrical in those three phenomena, they demonstrate how acoustics and suprasegmental factors influence the perception of vowel sequences in Brazilian Portuguese (AU)

FAPESP's process: 19/27204-1 - Diphthongization and monophthongization: the effects of positional privilege and acustic dispersion between vowels in Brazilian Portuguese grammar
Grantee:Lucas Pereira Eberle
Support Opportunities: Scholarships in Brazil - Master