This study evaluates issues of validity in measuring fricative voicing using instrumental acoustic correlates. Ten different gradient instrumental measures as well as four impressionistic evaluations were made of the voicing of 170 tokens of the Buenos Aires Spanish palato-alveolar fricative, which is in the process of devoicing. Content-related and criterion-related evidence for the validity of each measurement is presented. The available evidence calls into question the validity of consonant-to-vowel intensity ratios as correlates of gradient fricative voicing due to their dependence on the non-constant intensity of the adjacent vowel, which varies due to the height and stress of the vowel. The evidence presented shows that other measurements such as Praat's voice report, a low-frequency-to-total intensity ratio, and center of gravity, on the other hand, are relatively good measures of gradient fricative voicing. Impressionistic coding of fricative voicing is shown to have a relatively low rate of inter-rater reliability. The results of this study point to the importance of evaluating the validity of instrumental measurements in representing activity in the vocal tract.
Selected Proceedings of the 5th Conference on Laboratory Approaches to Romance Phonology
edited by Scott M. Alvord
Table of contents