Recognition of fricatives in normal hearing and simulated hearing-impaired conditions

Abstract

Previous studies on consonant recognition and simulated hearing impairment have shown that consonants most often get confused in noise conditions (Miller and Nicely, 1955, Maniwa et al., 2008). This study follows these previous studies by investigating how our linguistic perception of fricatives can be affected by various auditory conditions. In particular, the four voiceless fricatives /f/, /θ/, /s/ and /ʃ/ have been embedded in the words fin, thin, sin and shin and recorded in the carrier sentence “I say ______ now” by a female and a male speaker of Scottish English. The 32 sentences were presented to students at the University of Edinburgh twice in random order in a normal hearing and a simulated hearing-impaired condition with and without noise. The sentences were played over a headset with the 4 target words simultaneously presented on the computer screen. The listeners‟ task was to select the word they heard. Analyses were conducted on talker‟s voice, normal and impairment conditions regarding the listeners‟ results. Findings showed that the talker‟s voice, noise and reaction time all had an effect on the listener‟s recognition of fricatives. As expected the noise condition produced higher incorrect results. By trying to replicate earlier studies in a different language environment, some of the results confirmed the findings by Miller and Nicely (1955) and Maniwa et al. (2008) with regard to consonant recognition and confusion. The study also revealed that noise affected the recognition more than the impairment condition and that fricatives in the male voice condition were overall recognised better.