Subsídios para análise prosódica do dialeto de Curitiba : uma contribuição ao projeto AMPER-POR
Resumo
Resumo: Esta pesquisa pretende investigar as características prosódicas do dialeto de Curitiba (Paraná) a partir de enunciados assertivos neutros e interrogativas totais (que comportam resposta sim/não). Trata-se de um estudo vinculado ao Projeto do Atlas Multimídia Prosódico do Espaço Românico do Português (AMPER-POR). Analisamos os enunciados divididos em padrões acentuais (oxítonos, paroxítonos e proparoxítonos) e também por gênero do informante (femi-nino e masculino). Nossas perguntas de pesquisa pretendem responder como se caracteriza o dialeto de Curitiba, como os falantes desse dialeto distinguem as modalidades interrogativas totais e assertivas neutras e, ainda, se curitibanos identificam seu modo de falar ouvindo outros curitibanos. Por isso, queremos saber: (i) O pré-núcleo dos enunciados do dialeto de Curitiba tem a mesma configuração melódica para as declarativas e as interrogativas totais ou apresenta contornos melódicos diferentes? (ii) Quais são as notações autossegmentais métricas que des-creveremos para caracterizar o núcleo dos enunciados do falar de Curitiba? A taxa de inclinação das sílabas nucleares pode contribuir para apontar a diferença entre modalidades, juntamente com a notação Autossegmental Métrica (AM)? (iii) A gama tonal e a intensidade das sílabas do falar curitibano, na região nuclear, são parâmetros acústicos que distinguem modalidades? (iv) A taxa de elocução fornece pistas prosódicas de discriminação dialetal comparando a veloci-dade do falar de Curitiba com a de outros? (v) Com relação à melodia da fala, na análise per-ceptual, curitibanos são capazes de discriminar seu dialeto, tanto em comparação a outros como quando não há um dialeto diferente para contrastar? E, ainda, são capazes de identificar seu falar ao ouvir apenas um enunciado isolado? (vi) Quantos semitons são necessários para que os ouvintes curitibanos percebam que uma palavra é mais aguda que outra? Para responder a nos-sas perguntas de pesquisa gravamos dois curitibanos, uma mulher e um homem, que nasceram em Curitiba, sempre residiram nesta localidade e se reconhecem como falantes do dialeto da capital paranaense. Cada um elaborou 66 frases (33 assertivas e 33 interrogativas) a partir de imagens e repetiu quatro vezes esta tarefa (das quatro repetições, analisamos as três que melhor foram ditas, sem pausas na fala, focalizações, entre outros). Observamos a curva entoacional dos enunciados, propondo notações autossegmentais métricas para a região nuclear. As decla-rativas receberam a notação H+L*L% e as interrogativas a notação L+H*L%, embora as inter-rogativas oxítonas do informante masculino tenham tido uma curva melódica distinta no núcleo (L+L*H%). Calculamos a taxa de inclinação das sílabas nucleares (em semitons por segundo), o que nos facilitou observar numericamente a distinção entre modalidades por meio da inclina-ção da curva. Os valores de intensidade analisados nas sílabas nucleares também indicaram diferenças entre declarativas e interrogativas totais, o que não ocorreu, porém, com a gama tonal que se manteve semelhante entre as modalidades. No pré-núcleo, observamos os valores de F0 e percebemos que eles são significativamente diferentes e que a curva melódica de cada moda-lidade também se comporta de distintas maneiras. Além disso, um teste perceptual sobre o pré-núcleo demonstrou que curitibanos distinguem auditivamente as duas modalidades ouvindo apenas o início do enunciado. A taxa de elocução dos informantes curitibanos apontou que eles falam mais rápido que os informantes paulistas, porém, mais devagar que os informantes flori-anopolitanos. Os testes de percepção dialetal revelaram que curitibanos são capazes de identi-ficar seu dialeto e também de discriminá-lo em comparação a outros (embora não com a mesma eficácia da identificação). Um outro teste de percepção apontou, ainda, que ouvintes curitibanos conseguem perceber auditivamente a mudança de três ou mais semitons em uma palavra ou apenas na sílaba tônica. Palavras-chave: prosódia curitibana; assertivas neutras; interrogativas totais; AMPER-POR. Abstract: The aim of this research is to identify the prosodic features of the dialect from Curitiba (state of Paraná, southern Brazil) based on the analysis of neutral statements and neutral yes/no ques-tions (which have yes/no answers). This is a study linked to the project Multimedia Prosodic Atlas for Romance Languages - Portuguese (AMPER-POR). We have analyzed the statements based on stress patterns (oxytones, paroxytones and proparoxytones) and also on sex of inform-ants (male and female). Our research questions are aimed at characterizing the dialect from Curitiba, finding out how speakers of that dialect distinguish neutral yes/no questions from neutral statements, and also verifying whether or not people from Curitiba recognize their own dialect when they listen to other speakers from the same city. Therefore, these are our research questions: (i) Does prenuclear pitch accent of utterances in the dialect from Curitiba have the same melodic configuration for neutral statements and neutral yes/no questions, or does it have different melodic contours? (ii) What are the autosegmental-metrical notations that we will de-scribe to characterize the nuclear pitch accent of the utterances of the dialect from Curitiba? Can the slope rate of nuclear syllables provide input into the difference between modes, with the use of autosegmental-metrical notation? (iii) Can pitch range and syllable intensity of the dialect from Curitiba be used as acoustic parameters for distinction of modes, in the nuclear pitch accent? (iv) Does speech rate provide prosodic cues for dialectal discrimination when comparing the speech rate of the dialect from Curitiba with that of other dialects? (v) In per-ceptual analysis, with respect to the melody of speech, can speakers from Curitiba distinguish their own dialect, both in the presence and in the absence of different dialects? Moreover, are they able to identify their own dialect when they only listen to an isolated utterance? (vi) How many semitones are necessary for listeners from Curitiba to realize that a word is higher-pitched than another? To answer our research questions, we recorded two people (a female and a male) born and bred in Curitiba, who acknowledge themselves as speakers of the dialect from Curitiba. Each speaker produced 66 sentences (33 neutral statements and 33 neutral yes/no questions) using images as prompts, and they repeated this task four times. We analyzed three out of the four repetitions: the ones that were best pronounced (no pauses in speech, no focalizations, etc). We observed the intonation curve of the utterances, and proposed autosegmental-metrical no-tations for the nuclear pitch accent. Neutral statements received the notation H+L*L% and, yes/no questions the notation L+H*L%, although the melodic curve of the questions with oxy-tone pattern produced by the male informant was different in the nuclear region (L+L*H%). We calculated the slope rate of nuclear syllables (in semitones per second), which made it easier to observe, in numerical terms, the distinction between modes through the curve slope. The intensity values analyzed in nuclear syllables also indicated differences between neutral state-ments and neutral yes/no questions; however, this did not occur with pitch range, which re-mained similar across the modes. In prenuclear pitch accent, we observed the F0 values and realized that they are significantly different, and the melodic curve of each mode also behaves in different ways. In addition, a perceptual test on the prenuclear region showed that people from Curitiba distinguished the two modes aurally by listening only to the beginning of the utterance. The speech rate of the informants from Curitiba indicated that they speak faster than the informants from São Paulo, but more slowly than the informants from Florianópolis. The perception tests showed that listeners from Curitiba can identify their own dialect and also rec-ognize it when compared with other dialects (although not as effectively as in the identification task). Another perception test also showed that listeners from Curitiba can aurally perceive the change of three or more semitones in a word or just on the tonic syllable. Keywords: prosody of people from Curitiba; neutral statements; neutral yes/no questions; AM-PER-POR.
Collections
- Teses & Dissertações [10146]