Related Experiment Videos
Perceptual normalization for inter- and intratalker variation in Cantonese level tones.
Patrick C M Wong1, Randy L Diehl
1University of Texas at Austin, USA. pwong@uchicago.edu
Journal of Speech, Language, and Hearing Research : JSLHR
|January 1, 2004
Summary
Listener perception of Cantonese tones is affected by talker variation. Listeners better identified tones when talker information was grouped, and context influenced tone perception in sentences.
Area of Science:
- Phonetics
- Psycholinguistics
- Acoustic phonetics
Background:
- Lexical tone perception can be challenging due to variations in how individuals produce sounds.
- Acoustic overlap between tones may lead to listener confusion.
Purpose of the Study:
- To investigate if acoustic overlap in Cantonese lexical tones causes perceptual ambiguity.
- To determine if listeners can use contextual cues to reduce this ambiguity.
Main Methods:
- Experiment 1: Native Cantonese speakers identified isolated tones from 7 talkers, with blocked vs. mixed presentation.
- Experiment 2: Listeners identified target tones in sentences with altered preceding context (f0 patterns).
- Experiment 3: Listeners identified target tones with independent alterations to the first and second halves of the context.
Main Results:
- Tone identification accuracy improved when stimuli were blocked by talker compared to mixed presentation.
- The perceived identity of a target tone varied based on the f0 pattern of the sentence context.
- Listeners placed greater importance on the latter half of the contextual f0 pattern for tone identification.
Conclusions:
- Inter- and intratalker variations in lexical tone production can create perceptual ambiguity.
- Listeners utilize contextual information, particularly the immediately preceding acoustic cues, to resolve tonal ambiguity.
- Findings highlight the dynamic and context-dependent nature of lexical tone perception.