An Acoustical and Perceptual Study of Vowels Produced by Alaryngeal Speakers of Cantonese

Abstract

Objective:The present study determines the acoustical and perceptual differences between alaryngeal and laryngeal speakers of Cantonese in the production of Cantonese vowels. Subjects and Methods:Formant frequencies (F1 and F2) associated with eight Cantonese vowels (/i, y, 3, a, q, $, œ, u/) produced by 40 alaryngeal (10 esophageal, 10 tracheoesophageal, 10 with pneumatic artificial larynx, 10 electrolaryngeal) and 10 laryngeal speakers were obtained. To assess how formants affect vowel perception, the vowels were identified by 20 naïve listeners, based on which confusion matrices were established. Results: Spectral analysis revealed that F1 and F2 values of Cantonese vowels produced by the alaryngeal speakers were significantly higher than those produced by laryngeal speakers, with the exception of F2 of /3/, /i/ and /y/ produced by alaryngeal speakers using pneumatic artificial larynx, and /3/ produced by electrolaryngeal speakers. Conclusion: The acoustic findings suggest a general shortening of the effective vocal tract length for resonance in alaryngeal speakers after laryngectomy. Percent correct identification of vowels from the perceptual experiment indicated similar error patterns in vowel identification between alaryngeal and laryngeal speaker groups. Among the different alaryngeal speaker groups, naïve listeners tended to perceive the vowels produced by electrolaryngeal speakers more accurately.