Abstrakt
This paper deals with the perceptual impact of using word-initial glottalization as a boundary signal. Research into glottalization is discussed before presenting a reaction time experiment based on a word-monitoring paradigm.
Natural and acoustically manipulated sentences selected from a spontaneous political debate were used as stimuli. The low level of control of the material makes the design relatively innovative in comparison with similar previous studies.
Fifty listeners were instructed to press a button as soon as they heard a specified target word embedded in a sentence, where a preceding carrier word included glottalization or linking. The results support the hypothesis that the presence of glottalization facilitates the processing of subsequent words, but the temporal scope of the effect varied.
The experiment also raises new methodological issues and offers suggestions for further research.