An exploratory study of voice quality in esophageal speech

This study was designed to investigate the relationship between judged pleasantness of esophageal vocal productions of an isolated vowel and the amount of jitter present in the acoustic signal. Secondary variables considered included- the relationship between fundamental frequency and judged vocal pleasantness of phonations, and the amount of "wetness" that listeners perceived in the phonations. Eight esophageal speakers phonated the vowel /æ/ several times. Two phonations from each subject were chosen for exploration. Master tapes were compiled in two designs- one, in pairs for comparison, and one, for rating on a seven-point scale. The two phonations were subjected to Visicorder analysis. Hand measures of the tracings were used to compute fundamental frequency and jitter ratios. The jitter ratios were computed arithmetically by dividing adjacent frequencies, always using the smaller figure as denominator. Thirty-four judges performed three listening tasks. They were asked to judge which of each pair of stimuli was most pleasant; rate 80 stimuli according to degree of pleasantness; and rate the amount of wetness present in the phonation. Parametric procedures were followed in analyzing the rating scale data- nonparametric computations were used for the paired-comparisons data. Results of this study indicated that the amount of wetness perceived during the phonation of an isolated vowel was a strong determinant of judgments of unpleasantness in esophageal speaker. The amount of jitter was less important to the judgments than was wetness. Fundamental frequency did not appear to be related to the pleasantness judgnents. This study was limited in scope and additional research on voice quality in esophageal speech should be carried out.