Sponsor(s): Chinese Psychological Society; Institute of Psychology, Chinese Academy of Sciences
12 issues per year
Current Issue: Issue 09, 2014
Journal official website:http://journal.psych.ac.cn/xlxb/EN/0439-755X/home.shtml
Acta Psychologica Sinica is a scholarly journal sponsored by Chinese Psychological Society and Institute of Psychology, Chinese Academy of Sciences, co-sponsored by Department of Psychology, Chinese University of HongKong, published monthly by the Science Press. It is to publish original empirical studies and theoretical papers in the broad field of psychology including cognitive and experimental psychology, developmental and educational psychology, physiological and medical psychology, management social psychology, psychological measure, psychological history and method et al.
ZHANG Kan, Institute of Psychology, Chinese Academy of Sciences, China
FUNG Helene Hoi Lam, Department of Psychology, Chinese University of Hong Kong, China
HAU Kit-Tai, Department of Educational Psychology, Chinese Univ
Acta Psychologica Sinica,2014,Vol 46,No. 09
Sounds are physically continuous but linguistic phonemes are discrete and limited. Categorical Perception(CP) offers a useful approach to address this mismatching issue between the physical and perceptional features of speech. It is well established that consonants are perceived categorically, while vowels are perceived continuously. Earlier studies showed that the speech CP was restricted to segmental features. However, as Abramson(1961) firstly pointed out that CP mode could be applicable to the suprasegments. His limited investigation revealed that tone perception in Thai is categorical. Likewise, a similar result has been confirmed by Wang(1976) that the perception of Mandarin tones is also categorical, in the sense that native Chinese speakers have a linguistic boundary when perceiving the tones. The studies was based on a synthesized acoustic continuum, which was superimposed on vowel /i/, consisting of 11 tonal variants from Mandarin Tone 2(yi2 ‘aunt') to Tone 1(yi1 ‘clothing'). There is still no conclusion that what effect vowels have on the boundary of tonal perception. Regarding this, current study is to address this issue. In this research, tests of tonal CP are extended from /i/(Wang 1976; Peng et al. 2010) to six Mandarin monophthongs(/a//o//r//i//u//y/) to investigate the effects of vowels' intrinsic pitch on tonal boundary. Firstly, a continuum with a rising tendency was synthesized from Mandarin Tone 2 to Tone 1 using the acoustic software Praat. This continuum was superimposed on six Mandarin monophthongs, which were then used as stimulus in the identification test. Secondly, the DMDX software was used to randomly and with nine repetitions present the stimuli to the subjects in the identification test. Finally, the participants, twenty native Mandarin speakers(nine male and eleven female), were asked to identify the stimulus as either Tone 1 or Tone 2 by pressing different button on the computer. The mean category boundary position and width in terms of stimulus number were obtained by Probit Analysis of individual identification curves. To investigate the effects of different monophthongs on categorical boundary position and width, our results were analyzed using repeated measures with SPSS software. The results show that the categorical boundary position differs significantly between Mandarin monophthongs(p < 0.001): the boundary position of the tone continuum of vowel /a/(stimulus step number 5.13) is smaller than that of tone continuum of other vowels(stimulus step numbers of /o//r//i//u//y/ are 6.02; 6.31; 6.59; 6.27 and 6.54 respectively). The finding that boundary position of low vowel /a/, whose intrinsic pitch is low, is consistently lower than the five other monophthongs(p < 0.001) supports the validity of the “tongue-pull” hypothesis(Ohala & Eukel 1976: 44). Categorical boundary width across the six vowels was not significantly different(p = 0.303). To sum up, in this study, the influence of different monophthongs on the perception of pitch contours in the framework of CP was examined. The results of this experiment indicate that 1) identification of Tone 1 and Tone 2 is clearly categorical; 2) stimuli of lower vowel tends to require steeper F0 contour slope for subjects to make a "Tone 2" response than the stimuli of higher vowels did; 3) the intrinsic pitch of vowels plays an important role in identification of Tone 1 and Tone 2.