This study investigated speaker normalization in perception of Mandarin tone 2 ͑midrising͒ and tone 3 ͑low-falling-rising͒ by examining listeners' use of F0 range as a cue to speaker identity. Two speakers were selected such that tone 2 of the low-pitched speaker and tone 3 of the high-pitched speaker occurred at equivalent F0 heights. Production and perception experiments determined that turning point ͑or inflection point of the tone͒, and ⌬F0 ͑the difference in F0 between onset and turning point͒ distinguished the two tones. Three tone continua varying in either turning point, ⌬F0, or both acoustic dimensions, were then appended to a natural precursor phrase from each of the two speakers. Results showed identification shifts such that identical stimuli were identified as low tones for the high precursor condition, but as high tones for the low precursor condition. Stimuli varying in turning point showed no significant shift, suggesting that listeners normalize only when the precursor varies in the same dimension as the stimuli. The magnitude of the shift was greater for stimuli varying only in ⌬F0, as compared to stimuli varying in both turning point and ⌬F0, indicating that normalization effects are reduced for stimuli more closely matching natural speech.
This study reports results of an experiment on Beijing Mandarin tones and stress. There are four lexical tones in Mandarin, plus a pitch pattern referred to as the ‘‘neutral tone,’’ which occurs on atonic or unstressed syllables. The literature has been vague, however, about whether syllables that have undergone tone deletion as a result of destressing and underlyingly atonic syllables show the same surface pitch pattern. This experiment examined these lexical tones in stressed and unstressed contexts, and then compared unstressed tone patterns to that of the neutral tone. The experiment focused on duration and pitch range as effective correlates of stress. Unstressed syllables were found to be significantly shorter in duration than stressed syllables, corroborating previous studies. Pitch range was also significantly reduced in unstressed contexts in all cases except one. The data unexpectedly showed that changes in pitch range are a function of the pitch peak only; pitch valleys did not significantly change. Finally, the results provide evidence that unstressed syllables with underlying tones do not neutralize completely, but remain distinct from the neutral tone.
Segments with identical acoustic characteristics may not be perceived identically, because listeners normalize for acoustic variability due to vocal tract differences among speakers. While normalization studies have been done for vowel perception, the present experiment uses Mandarin Chinese to investigate speaker normalization for lexical tones. In this study, two speakers were identified for whom the low tone (tone 3) of the high- pitched speaker occurred in the same F0 region as the midrising tone (tone 2) of the low-pitched speaker. Synthetic syllables formed three continua used in a series of perception tests. The three continua varied along either F0, timing of the turning point, or both acoustic dimensions. Stimuli were placed after a natural precursor from each of the two speakers. Test items were randomized and presented in a mixed block condition. Results show that category boundaries shift contrastively in the direction predicted; that is, the same stimuli were perceived differently in the two precursor conditions, although only for stimuli varying along the F0 and both F0 and turning point dimension, but not for stimuli varying along the duration dimension (turning point) alone. Further testing is currently being conducted on normalization in the duration dimension. [Work supported by Sigma Xi.]
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
customersupport@researchsolutions.com
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.