A speech audiogram of Setswana consonants

Abstract

Setswana is a Bantu language used in several regions of Southern Africa, and its phonemic inventory includes a diverse range of consonants. To date, very little information exists about the acoustic correlates of Setswana speech sounds. The present study was aimed at analyzing these speech sounds in order to develop a clinically useful speech audiogram for Setswana. Three native speakers of Setswana produced 45 target phonemes in word initial and medial positions, in three vowel contexts (/a/, /i/, and /u/). The stimuli were then analyzed for peak frequency and intensity. A preliminary analysis revealed an effect of word position and vowel context on peak frequency, and hence, only phonemes in the word medial position and in the /a/ vowel context were used to plot the speech audiogram. The speech audiogram hence developed was then compared to speech audiograms of American English phonemes commonly employed in clinical settings. Consonants common to both Setswana and American English inventories were used to establish the validity of the study’s procedures. The speech audiogram developed can potentially be used to provide partial estimates of the audibility of Setswana phonemes to native speakers with various types of hearing loss. a)Currently at the Dept. of Communication Pathology, Univ. of Pretoria, South Africa