Master of Arts / Department of Modern Languages / Earl K. Brown / This paper reports on a study that analyzes how a sequential bilingual speaker (L1 Mexican Spanish, L2 American English) uses stylistic phonetic variation in different speech types during an interview (short answer, spontaneous speech, dramatic anecdote, reading) to construct a dynamic gay persona. There are many stylistic variables that can interact when an individual is creating a persona in an interaction, and this becomes even more complex when analyzing L1 speech as well as L2 speech as there are two collections of stylistic phonetic variables (indexical fields) interacting from two different cultural ideologies available to the interlocutors. It is problematic to assign one distinct variable to an identity, such as gay, as it homogenizes a diverse social group of individuals and underestimates members' ability to manage perceptual salience of their identity as a gay individual based on context and social pressure(s). While the field of Lavender Linguistics (language use associated with the LGBTQ community) has shown that there are many resources that can be used to "sound gay," this case study focuses on how a speaker stylistically creates a gay persona throughout the interview through stylistic variation of two principle variables: 1) word-final /s/ duration, and 2) center of gravity of word-final /s/.

Submitted by Fernando Souza (fernandoafsou@gmail.com) on 2017-08-03T14:14:18Z No. of bitstreams: 1 arquivototal.pdf: 2709686 bytes, checksum: 972411d4415db11a6a470c2b58219f77 (MD5) / Made available in DSpace on 2017-08-03T14:14:18Z (GMT). No. of bitstreams: 1 arquivototal.pdf: 2709686 bytes, checksum: 972411d4415db11a6a470c2b58219f77 (MD5) Previous issue date: 2016-05-05 / The main objective of this study was to evaluate the existence of phonetic characteristics of speech read by male speakers that can be used to identify them as having a "gay" or a heterosexual orientation. Thus, two groups were constituted, bringing together 14 informants from the city of Recife-PE. In the first group, 7 individuals, who declared themselves gays, were investigated; In the second, the other 7, self-declared heterosexuals. In order to better understand the phonetic characteristics, the audio recording of the voice of these men was subjected to tests of perception and later judged by hearers lay judges. As a specific objective, it was sought to investigate which acoustic clues, such as phrasal aspects and vowel and consonant sounds, could contribute to the fact that the listener's lay judges, non-specialists belonging to the area of Linguistics or corresponding, could attribute values to the sexual orientation of the 14 informants. The lay judges were allocated into three groups: a group of 25 gay men; A group of 25 heterosexual men; And a group consisting exclusively of 25 women whose sexual orientation was unknown. Another objective was to verify the efficiency of each of these three different groups of evaluating judges to identify the sexual orientation of the 14 informants, subdivided into the two groups. As theoretical basis for the thesis, the theoretical assumptions of Variationist Sociolinguistics (LABOV, 1972, 1974, 1982, 1994) and Eckert (1989, 1997, 2000, 2003, 2005, 2010, 2012) were used and, for the analysis of the perception were used Gaudio (1994); Smyth, Jacobs and Rogers (2003); Levon (2007); Campbell-Kibler (2011); Lopes (2012) and Tracy, Bainter and Santariano (2015). In the research, significant differences in the duration of the seven tonic oral vowels /a/, /Ɛ/, /e/, /i/, /Ɔ/, /o/, /u/ produced by gay informants were found, especially in the mid near front /e/ and / Ɛ/ and in the central open low /a/. Differences, were also found in the fricative /s/ when in the final coda position, the productions of the group of gay men were longer than that of heterosexuals group. About the formants, the mean of F1 and F2 production of gay men were considerably higher than those of heterosexuals in all vowels, especially about the values of /i/, /Ɛ/, /e/ and /a/. Therefore, among the gay speakers, there was wider opening of the mandible and anterioration of the tongue in the production of the vowels. Significantly greater averages of pitch variability among gay informants were also verified. In most aspects, the averages of homosexual informants, whose were part of the research, were closer to the averages of gay informants participating in English-language surveys than the averages presented by the heterosexual informants. Thus, due to the similarities found in the results of some surveys, obtained among English and Portuguese speaking gay men, it seems plausible to affirm that there are some universal typical aspects that characterize a so-called "gay speech". / O presente trabalho teve por objetivo principal avaliar a existência de características fonéticas de falas lidas por falantes do sexo masculino que podem ser usadas para identifica-los como de orientação sexual “gay” ou heterossexual. Para isso, foram constituídos dois grupos, reunindo 14 informantes da cidade de Recife-PE. No primeiro, grupo, foram investigados 7 indivíduos, que se autodeclararam gays; no segundo, os outros 7, autodeclarados heterossexuais. Para melhor compreender as características fonéticas, o registro em áudio da voz desses homens foi submetido a testes de percepção e, posteriormente, julgado por ouvintes juízes leigos. Como objetivo específico, procurou-se investigar quais pistas acústicas, tais como aspectos frasais e sons de vogais e consoantes, poderiam contribuir para que os ouvintes juízes leigos, participantes não especialistas pertencentes à área de Linguística ou correlata, pudessem atribuir valores à orientação sexual dos 14 informantes. Os juízes leigos foram alocados em três grupos: um grupo formado por 25 homens gays; um grupo formado por 25 homens heterossexuais; e um grupo formado exclusivamente por 25 mulheres, cujas orientações sexuais nos eram desconhecidas. Outro objetivo foi verificar o grau de eficiência de cada um desses três diferentes grupos de juízes avaliadores, para identificar a orientação sexual dos 14 informantes, subdivididos nos dois grupos. Como base teórica para a tese, foram utilizados os pressupostos teóricos da Sociolinguística Variacionista (LABOV, 1972; 1974; 1982; 1994) e Eckert (1989, 1997, 2000, 2003, 2005, 2010, 2012) e, para análise dos testes de percepção foram utilizados Gaudio (1994); Smyth, Jacobs e Rogers (2003); Levon (2007); Campbell-Kibler (2011); Lopes (2012) e Tracy, Bainter e Santariano (2015). Na pesquisa foram encontradas diferenças significativas na duração das sete vogais orais tônicas, /a/, /Ɛ/, /e/, /i/, /Ɔ/, /o/, /u/ produzidas pelos informantes gays, principalmente nas vogais médias anteriores /e/ e / Ɛ/ e na baixa central /a/. Diferenças, também, foram encontradas na fricativa /s/ quando em posição de coda final, sendo as produções do grupo de homens gays mais duradouras que a dos heterossexuais. Com relação aos formantes, as médias de produção de F1 e F2 dos homens gays foram consideravelmente mais altas que as dos heterossexuais em todas as vogais, especialmente com relação aos valores de /i/, /Ɛ/, /e/ e /a/. Portanto, entre os falantes gays, houve maior abertura da mandíbula e anteriorização da língua na produção das vogais. Médias consideravelmente maiores de variabilidade do pitch dos informantes gays também foram verificadas. Na maioria dos aspectos, as médias dos informantes homossexuais brasileiros utilizados na pesquisa aproximaram-se mais das médias dos informantes gays participantes de pesquisas em língua inglesa do que das médias apresentadas pelos informantes heterossexuais. Assim, devido às semelhanças encontradas nos resultados de algumas pesquisas entre homens gays falantes de língua inglesa e os aqui obtidos, pode-se supor, sem tentar generalizar, que existem aspectos da fala que caracterizam uma chamada “fala gay’.