Segmentation of Reduced Speech by Native and Non-native Listeners of Dutch

This study investigates whether there is a difference between native and non-native speakers of Dutch in the segmentation of reduced and unreduced Dutch forms. For this purpose, two segmentation tasks were constructed, measuring both speed and accuracy of segmentation. Participants were presented with full and reduced auditory input and had to indicate how many and which words they heard. Native speakers were faster and more accurate than non-native speakers in segmenting unreduced and reduced speech. This may be ascribed to non-native speakers' inability to effectively use Dutch segmentation cues. They may also be less able to use reconstructive processes. More notably, non-native speakers were better at counting words than identifying them. This may represent a phase in the - what seems to be a - gradual process of the acquisition of segmentation skills. This study constitutes a stepping stone toward bridging the gap between studies of segmentation and studies of reduction processes.