Search

This study looks at the syllable onset interval (SOI) patterning in Taiwan Mandarin spontaneous speech and its relationship to discourse and syntactic units. Monologs were elicited by asking readers to tell stories depicted in comic strips and were transcribed and segmented into Discourse Segment Units (Grosz & Sidner, 1986), clauses, and phrases. Results showed that the degree of final lengthening was modulated by boundary types. Lengthening before discourse boundaries was longer than that before clausal boundaries, which was in turn longer than that before phrasal boundaries. Final SOI lengthening also seemed to reflect cognitive load. At the discourse and clausal levels, the degree of lengthening is modulated by narration order. First narrations tended to have longer final SOIs than second narrations. In addition, there was also a mild lengthening effect in nonfinal SOIs, as was evidenced by the length differences in initial and medial SOIs and the differential lengthening effect regarding the positioning of phrases in a clause. (Contains 7 tables, 6 figures and 12 footnotes.)

This study focused on durational cues (i.e., syllable duration, pause duration, and syllable onset intervals (SOIs)) at discourse boundaries in two dialects of Mandarin, Taiwan and Mainland varieties. Speech was elicited by having 18 participants describe events in The Pear Story film. Recorded data were transcribed, labeled, and segmented into clauses. Discourse boundary indices were used to label discourse disjuncture levels. Results showed that the scope of lengthening included both the final and the penultimate syllables. Pause was a robust but optional indicator for discourse boundaries, and Taiwan Mandarin preferred unfilled over filled ones, while the Mainland variety did not show such strong preferences. Discourse hierarchy corresponded consistently with occurrences of pauses and duration of SOIs, within which pause was the main contributor. Higher discourse levels were more likely to be accompanied by pauses, and were indicated by longer pauses and SOIs. Syllable duration only played a secondary role in indicating discourse disjuncture size when pause was absent. When there was an accompanying pause, Mainland Mandarin relied solely on it to indicate discourse hierarchy, while Taiwan Mandarin used both syllable and pause duration, a dialectal difference that seemed to be rhythm-related. By shortening the degree of lengthening in the boundary syllable and lengthening the following pause at the same time, Taiwan Mandarin increased the absolute and relative duration of pause and maximized its role in indicating discourse hierarchy. The results of this study implied that the use of pause in discourse disjuncture demarcation was more of a language-specific choice while its role in discourse hierarchy encoding is more language-universal. [ABSTRACT FROM PUBLISHER]