Tone production by alaryngeal speakers of Cantonese : a replication study

Abstract

The present study investigated the acoustic and perceptual characteristics of tones produced by alaryngeal speakers of Cantonese in comparison to laryngeal speakers. Forty-nine male native laryngeal and alaryngeal speakers of Cantonese including 15 laryngeal (NL), 12 esophageal (ES), 15 tracheoesophageal (TE) and 7 electrolaryngeal (EL) speakers were were instructed to produce two sets of Cantonese monosyllabic syllables (/si/ & /fu/) at six lexical tones using comfortable loudness. Speech samples produced were randomly presented to 15 native Cantonese listeners for tone identification, and at the same time average and range of fundamental frequency (F0) values were obtained. Results indicated significant difference in percent correct tone identification (PCTI) between alaryngeal and laryngeal speakers, with NL tones perceived most accurately, followed by ES and TE, and EL with the lowest accuracy. Acoustically, significantly lower mean and range values of F0 were found for EL against all other speaker groups. Significantly greater F0 range was found for ES compared to other speaker groups. Specific tone confusion patterns to ES and TE were found: T1/T4 and T3/T4. All EL tones were perceived as T1, T3 or T6. The present study provided clinical implications to the competence of tone production associated with various alaryngeal speech.