The Acquisition Of Aspiration Contrast Of Stops In Mandarin

Voice onset time (VOT) has been shown to be the most simple and reliable acoustic cue to differentiate the phonemic stop categories contrasting in voicing. The present study employs the measure of VOT to investigate the voicing contrast acquisition by Mandarin-speaking children. A Mandarin-speaking child growing up in Changsha, Hunan, was observed from nine months old to three years of age. Data drawn for the VOT analysis in this study covered the period from one year and five months to two years and eleven months, involving a total of 21 sessions of hour-long recordings. VOT values of the three homorganic stop pairs at bilabial, alveolar and velar places of articulation are measured from spectrograms and waveforms. Adult VOT values for the stops, which serve as the model of voicing contrast in adult language, were also analyzed from recordings both in careful speech and running speech. The child's VOT statistics in each session was compared to those reported in previous studies on children from other language environments.Our data show that the child went through a period in which no distinction is made in the VOT of the stop consonants to a stage in which a systematic but non-adult-like distinction is made, before reaching a final stage in which the VOT values of stops resemble the adult model value via a process of systematic refinement in VOT. Our study thus provides additional evidence from Mandarin for the three-stage model first proposed in child English and later confirmed by child Cantonese.Special attention is drawn to the VOT data in the final phase of the child's voicing contrast acquisition in that data in this period exhibit systematic fluctuation of VOT values and distributions. Two rounds of lengthening-shortening process in VOT mean values for the voiceless aspirated stop category were observed, each taking a period of about six months from the establishment of the prototype of voicing contrast till the end of our observation. This process is observed unexpectedly more complicated than that reported in the child English. The VOT fluctuation in the child's production is regarded as a reflection of consistent and systematic refinement of the voicing contrast by the subject while approaching to the voicing contrast in adult language.Our data show that the voicing distinction emerged at an earlier age in our Mandarin-speaking subject as compared to those who acquire languages like Spanish, Thai and Hindi, and even earlier than the Cantonese-acquiring child. The early acquisition of voicing contrast by our subject is accounted for by the absence of a voicing lead category and the relative greater saliency of voicing contrast in Mandarin than in Cantonese.As in previous studies, a preference for the short lag stop production in a stage prior to the establishment of voicing contrast and the effect of place of articulation on the child's VOT production were found. Persistent production of voicing lead stop tokens by our subject was also found.The developmental pattern and acoustic characteristics revealed in this study are discussed in the light of language-specific factors and also physiological and aerodynamic constraints.