Longitudinal voice changes: Facts and interpretation

To study intraindividual vocal changes over time, a longitudinal research d esign is preferred. However, this procedure reveals many restrictions conce rning methodology and interpretation, especially in cases of (semi)retrospe ctive studies. This is illustrated by an investigation of the voices of 20 male Dutch-speaking reporters of the Belgian Radio Broadcasting service. Th ese men read the same text they did 30 years previously for a digital regis tration. After having copied the high-quality archival recordings digitally , the results of the acoustical analysis of both recordings (recent and arc hival) were compared. This was done to search for the contribution of aging . In this article we report on the results of the speaking fundamental freq uency (SFF) and the standard deviation of the SFF (SFFsd) as a measure of i ntraindividual variability, and the voice onset time (VOT). Results indicat e that, in connected speech, SFF decreased, SFFsd became more restricted, a nd the VOT increased strongly. The interpretation of the results compels us to conclude that it is extremely difficult to define the specific effect o f aging alone, since it is only one element among many others, including me thodological concerns.