PRODUÇÃO VOCÁLICA DE SUJEITOS COM AUTISMO: UM ESTUDO DE ANÁLISE ACÚSTICA

Abstract

ABSTRACT

People with autism spectrum disorder (ASD), according to Dias et al. (2009), are hardly able to establish a conversation with pretensions of social interaction because, in general, the dialogues deviate from the expected patterns, mainly because they do not usually understand the intentions of the interlocutor. In this sense, it is considered that speech, in addition to the physical elements, carries the speaker's intentions, since it is not only intended to communicate, but also to influence, convince and arouse emotions. They are emitted by means of intonation, tone and volume of the voice, which are prosodic aspects, natural characteristics of the speaker and fundamental for the understanding of speech (LOPES: LIMA, 2009). Thus, we have a hypothesis: considering that study of acoustic parameters is important in the investigation of prosodic aspects, we believe that any changes in the data of autistic subjects, such as non-demarcation of accent, may serve as a clue about the alleged prosodic abnormality in a speech. Based on this research, it is to compare the oral vowels produced by these people with autism in relation to those produced by people without the disorder by means of acoustic analysis to evidence or not, deficiency in the marking of prosodic aspects by artists. To achieve this goal, we compared results of acoustic parameters - formantic patterns, relative duration and fundamental frequency - in the vocal achievements of subjects with and without autism. For this study, we selected six nine-year-old male children, all readers and residents of the city of Vitória da Conquista, BA, three of them with mild autism and three in typical development. The data, resulting from the reading of Paroxyton words with oral vowels in different syllabic positions were collected in the Laboratory's acoustic booth. of Research and Studies in Phonetics and Phonology (LAPEFF-UESB) and submitted to acoustic analysis in the PRAAT software (BOERSMA; WEENINK, 2006). Later, we used some statistical parameters in the analysis of the data that helped us to discuss, with greater precision, the results we obtained in the acoustic analysis, as a Coefficient of Variation (CV) that allowed to verify whether the articulatory particularities of the subject with autism cause greater or lesser variability in the production of oral vowels compared to those of subjects without disorder. In addition, the statistical measurement comparison test was used, the Kruskal-Wallis (AYRES et al., 2014) non-parametric test. By means of e, we checked the measurements of the acoustic parameters analyzed in the vocal performances were significantly different in the different syllabic types, as well as assessing whether these measurements obtained by the subjects with autism were significantly different from those produced by the subject without autism. All statistical tests used in the research were performed using the BioEstat 5.0 program (AYRES et al., 2014). The results showed that the data of the acoustic parameters - standard formantic, duration and fundamental frequency - obtained for subjects with autism are consistent with the literature, that is, the sample of the investigated autistic subjects did not present difference in vocal production and abnormality in the evaluated prosodic level. We conclude that the autistic subjects do not present differences in this aspect when compared to the neurotic subjects. The results meet the difficulties of prosodic demarcation in autistic speech pointed out in the literature. On the other hand, some acoustic behaviors of autistic subjects found in the data analysis, allows us to suggest the need for new studies with a larger sample of participants in order to investigate more precisely the results found in this research, as well - considering the scarcity of works on topic addressed - the need for the language of people with autism to be the target of linguistic research.