Lexical, Syntactic, and Stress-Pattern Cues for Speech Segmentation

Abstract

Many sources of segmentation information are available in speech. Previous research has shown that one or another segmentation cue is used by listeners under certain circumstances. However, it has also been shown that none of the cues are absolutely reliable. Therefore, it is likely that people use a combination of segmentation cues when listening to normal speech. This study addresses the issue of how young adults use multiple segmentation cues (lexical, syntactic, and stress-pattern) in combination to break up continuous speech. Evidence that people use more than one cue at a time was found. Furthermore, the results suggest that people can use segmentation cues flexibly such that remaining cues are relied upon more heavily when other information is missing.