Abstract
The prosodic stress unit (prosodic word) represents a relevant factor for segmentation of the sound continuum in Czech. Research has shown that for the percipient, the cohesion of this unit is above all based on internal cohesion of acoustic characteristics in the unit's entirety especially on the course of F0.
Now it seems that the recognition of a stress unit may also be to some degree dependent on the context extending beyond the unit itself. The paper describes an experiment in which the listeners decided whether an ambiguous sequence of syllables represents one prosodic word or a sequence of two units.
In one test the syllable sequence was placed within a short context, in the other not. The results prove that the presence of a context results in greater ratio of correctly recognized items.
Owing to the close relationship between a word and a stress unit in Czech this result may be of interest as a view on the process of perception in spoken communication as well