UNIVERSIDADE ESTADUAL PAULISTA
"JÚLIO DE MESQUITA FILHO"

Faculdade de Ciências e Letras - Campus de Araraquara
  Agenda Pós-Graduação - Lingüística e Língua Portuguesa

Aluno(a) Marcus Garcia de Sene
Titulo A percepção sociolinguística de gênero e sexualidade: efeitos da duração de /s/ e do pitch médio
Orientador(a) Gladis Massini-Cagliari
Data 26/05/2022
Resumo Esta tese é resultado de uma pesquisa sociolinguística que trata do efeito de
duas variáveis linguísticas – a duração de /s/ em coda final e o pitch médio – na
percepção de quão masculino soa um determinando homem. Em outras
palavras, o objetivo é verificar se perceber um homem como alguém que soa
menos masculino está correlacionado a percebê-lo como um homem que soa
mais gay. Trata-se, portanto, de uma tese que se insere no campo das relações
entre linguagem, gênero e sexualidade. Para o atendimento do objetivo
proposto, desenvolveram-se experimentos, com base na técnica de matchedguise (LAMBERT et. al., 1960), a fim de compreender de que modo a alteração
do pitch médio (de original – F 0 mais baixo, para voz mais aguda - F
0 mais alto) e da duração de /s/ (de original para mais longo) afeta a percepção de vozes
masculinas, em termos de gênero e sexualidade.
Primeiramente, as variáveis aludidas foram testadas isoladamente, ou seja,
cada uma em um experimento diferente; após isso, foram combinadas no
mesmo experimento, com o objetivo de avaliar se elas se potencializam
mutuamente na indiciação do soar (mais ou menos) gay/masculino, ou se elas
têm efeitos que são independentes. No tocante ao experimento da duração de
/s/ em posição de coda final, os estímulos foram criados no Praat (BOERSMA,
WEENINK, 2015), e duas versões do mesmo trecho foram geradas: uma em que
todos os tokens de /s/ permaneceram originais (0.07 a 0.13 milissegundos) e outra
em que a duração de /s/ de todos os tokens foi digitalmente alongada para 0.27
a 0.32ms. Os resultados gerais mostram que a duração de /s/ em posição de
coda final tem efeito na percepção de quão masculino e quão gay soam os
falantes que foram recrutados para participar da pesquisa. Em outras palavras,
quando os falantes eram ouvidos em seu disfarce cuja duração de /s/ foi
alongada digitalmente, todos eles foram percebidos como homens que soam
mais gays e menos masculinos.
De igual modo, em relação à variável pitch médio, criaram-se dois disfarces:
um com o valor do pitch original e outro com o valor alterado em +30Hz. Sete
dos oito falantes recrutados para o experimento tiveram sua percepção alterada
quando ouvidos em seu disfarce em +30Hz, o que significa dizer que eles foram
percebidos como homens que soam mais gays e menos masculinos.
No experimento em que ambas as variáveis foram combinadas, os
resultados gerais atestaram que, quando combinado pitch médio +30Hz e
duração de /s/ alongada digitalmente, a percepção de gênero e sexualidade é
potencializada. Do ponto de vista estatístico, o tamanho do efeito (ȓ) é maior
quando essas variáveis são ouvidas em conjunto. Sendo assim, todos os oitos
falantes utilizados nesse experimento soaram mais gays e menos masculinos
diante dos disfarces em que as variáveis foram combinadas. Esses resultados
apontam para uma coesão social dos participantes da pesquisa no modo como
eles avaliam as variáveis em questão. Nesse sentido, é possível concluir que
essas variáveis estão ligadas a um certo esquema ideológico, sendo-lhes
conferido significados sociais, significados esses que as tornam socialmente
identificáveis.
Palavras-chave: Percepção sociolinguística; gênero; sexualidade; linguagem

Abstract: This thesis is the result of a sociolinguistic research that deals with the effect
of two linguistic variables - the duration of /s/ in final coda position and the
average pitch - on the perception of how masculine a man can sound. In other
words, the objective is to examine whether perceiving a man as someone who
sounds less masculine is correlated to perceiving him as a man who sounds
gayer. It is, therefore, a thesis that is inserted in the field of relations between
language, gender, and sexuality. To meet the proposed objective, experiments
were developed, based on the matched-guise technique (LAMBERT et. al., 1960),
in order to understand how changes in pitch (from original – lower F
0 , for higher voice - F 0 higher) and duration of /s/ (from original to longer) affect the
perception of male voices, in terms of gender and sexuality.
First, the mentioned variables were tested separately, that is, each one in a
different experiment; after that, they were combined in the same experiment,
with the objective of evaluating if they are mutually potentiated in the
indexicality of sounding (more or less) gay/masculine, or if they have effects
that are independent. Regarding the experiment of duration of /s/ in final coda
position, stimuli were created in Praat (BOERSMA, WEENINK, 2015), and two
versions of the same passage were generated: one in which all tokens of /s/
remained original (0.07 to 0.13 milliseconds) and another where the duration of
/s/ of all tokens was digitally lengthened to 0.27 to 0.32ms. Overall results show
that the duration of /s/ in final coda position influences the perception of how
masculine and how gay the speakers who were recruited to participate in the
research sounded. In other words, when speakers were heard in their disguise
whose /s/ duration was digitally lengthened, they were all perceived as more
gay-sounding, less masculine-sounding men.
Likewise, in relation to the average pitch variable, two guises were created:
one with the original pitch value and another with the value changed by +30Hz.
Seven of the eight speakers recruited for the experiment had their perception
altered when heard in their guise at +30hz, which is equivalent to say that they
were perceived as more gay-sounding and less masculine men.
In the experiment in which both variables were combined, the general
results attested that, when combined +30Hz average pitch and digitally
lengthened /s/ duration, the perception of gender and sexuality is enhanced.
From a statistical point of view, the effect size (ȓ) is larger when these variables
are heard together. Thus, all eight speakers used in this experiment sounded
more gay and less masculine in the face of the disguises in which the variables
were combined.
These results point to a social cohesion of the research participants in the
way they evaluate the variables in question. In this sense, it is possible to
conclude that these variables are linked to a certain ideological scheme in a way
that gives them a high symbolic value, a value that makes the variables socially
identifiable.

Keywords: Sociolinguistic perception; gender; sexuality; language
Tipo Defesa-Doutorado
Texto Completo

APG 2.0
Copyright 2014 (c) UNESP - Faculdade de Ciências e Letras do Campus de Araraquara