Estatísticas de frequência fundamental da população masculina da cidade de São Paulo para uso em fonética forense

Abstract

This study aimed to produce reference statistics for the fundamental frequency (f0) of speech in the male population of the city of São Paulo (SP), to be used as a resource in forensic phonetics practice. The samples analyzed in this research were taken from the SP2010 Project database, developed by the Sociolinguistics Research Group at the University of São Paulo (USP). The samples were stratified according to sociodemographic factors such as level of education, geographic zone, and age group. During the course of the study, audio excerpts of approximately 45 seconds were selected from each speech sample, from which f0 contours were extracted. Based on these contours, statistical descriptors of f0 were calculated and aggregated according to the speakers’ sociodemographic factors of education level, geographic zone, and age group. The resulting statistical summaries constitute the main outcome of the study and may be used in speaker comparison examinations that adopt a likelihood-ratio-based methodology for the expression of conclusions. In addition, the influence of the sociodemographic stratification factors defined by the SP2010 Project was examined for some of the statistical descriptors. The results indicated statistically significant effects for the variables age group and geographic zone, suggesting sociolinguistic variation among speakers from different geographic zones, as well as lower f0 values in the group of speakers aged over 60.