ISCA Archive SpeechProsody 2012
ISCA Archive SpeechProsody 2012

Prosodic encoding and perception of focus in Tibetan (anduo dialect)

Ling Wang, Bei Wang, Yi Xu

The prosodic realization of focus and its perception in Tibetan (Anduo dialect) were experimentally investigated. Using the question-and-answer paradigm, the speakers were asked to read aloud two target sentences in different focus conditions. Systematic acoustic analysis and statistical tests showed that, (1) On-focus F0 was raised sharply in medial and final focus conditions, but not much in initial focus. In addition, post-focus compression (PFC) occurred in initial and medial focus conditions. (2) Duration lengthening was found (about 11%) in focused words, but not in pre-focus or post-focus words. (3) Intensity was increased significantly (about 1.2 dB) in on-focus words, and decreased in post-focus words (about 0.5 dB). (4) In perception, correct focus identification was near 80% for medial focus, 63.3% for final focus, but only about 40% for initial focus. Overall, except for initial focus, the production and perception of focus in Tibetan were similar to those in Mandarin and English. Key words: Tibetan, focus, intonation, perception