| Intonation,the variations in pitch,stress,and timing of speech,plays various communicative roles such as to attract attention,to convey affective state,or the distinguish kinds of speech acts(e.g.,questions versus statements).In tonal language,intonation and tones can be quantified in terms of fundamental frequency(F0),duration and intensity,and they are mainly determined by the F0.To some extent,the acquisition of tone and intonation is strongly associated with children’s speech experience,speech motor control and cognitive development.The Parallel Encoding and Target Approximation model(PENTA),an articulatory-functional view of intonation,claimed that surface F0 contours transmit multiple communicational functions in parallel by implementing the encoding schemes of each of communicative function,through a dynamic articulatory process of Target Approximation(TA).This raises the questions of(1)in which age,could children realize multiple functions in intonation?and(2)whether the number of tone in languages will affect the order in which the use of F0 and duration is acquired for intonation?.In order to answer the questions above,an experiment was performed by Mandarin and Miao 4-11-year-old children.Using a picture-naming task,spontaneous production of sentences in varies sentence types(yes-no questions and statements)are elicited from five age groups in each language(including one adult group).Four tones in mandarin and eight tones in Miao are posed on the final position of sentences.Results from Mandarin and Miao children and adults show that:(1)The interrogative intonation in Miao is marked by rising terminal F0 with duration prolonged significantly.F0 contours of the eight tones remained consistent in the final position of question and statement,with the exception of contours of tones in the final position of question.Compared to Miao,in addition to a final rise in question intonation,the F0 of non-final tones in the sentence also are raised,and the amount of raising is greater when the closer the tone is the end of the sentence.Besides,tone contours are also largely maintained with raising question intonation in Mandarin.(2)For tones produced by children,4-11-year-old Miao and Mandarin-speaking children produced appropriate tone contours for tones in their mother tongue in sentence context.However,all groups are not fully adult-like.In Mandarin,to be more precise,4-year-old children produced falling or rising tones with flatter slope than adults while 7-and 9-year-old children produced tones with oversize contours.Similarly,falling tone produced by 4-11 year-old Miao-speaking children are faster than adults,and there is a shallower dipping contour produced by children.Additionally,Mandarin and Miao children produced tone with longer duration.(3)For the question intonation,Miao-speaking children are earlier than Mandarin-speaking children in how to use pitch-related cues to distinguish question from statement.We found that question intonation was difficult for 4-year-old Mandarin children but mastered by 7,9 and 11-year-old groups.Differently,there was no significant difference in using pitch to mark question interrogative intonation between Miao-speaking children groups and adults.This paper argues that the development of pitch control are not affected by the number of tones in languages but the encoding strategy of tone and intonation.Generally speaking,pitch realization for tone and intonation functions varies across languages,which leads to children producing tones and intonation with different strategies.To be more specific,when the pitch contours of tone and intonation go opposite,Mandarin children tend to maintain contours of tones while Miao children tend to moderate tone contour to realize interrogative intonation.Overall,the development of speed motor control inMandarin and Miao children is that:duration of syllable is decreasing and the speed of F0 change is increasing with age,indicating that children’s ability to control the vocal cords and the coordination of various organs are increasing.In general,the results of this paper support PENTA model from the perspective of language development. |