Mandarin Reading Textual Rhythm Research

Rhythm is one of the important aspects of prosody. This thesis conducts a comprehensive study on the characteristics of the rhythm of read speech in Mandarin. To rhythm, the two important components are the strength of stress and time.1. The strength of stress is about the strength of stress of the rhythmic unit at different levels.2. The time factor is about the organization of rhythmic units related to time. The thesis demonstrates that at the micro level, such as within an utterance, rhythm is organized in a linear fashion while at the macro level, like within a passage, rhythm is organized in a non-linear multi-leveled fashion.The results are as following:I. Degree of stressTwo kinds of factors that influence the perception of the degree of stress contain the inherent strength of a rhythmic unit, like quality of a syllable, lexical tone of a syllable and its placement in the prosodic structure, and the contextual-sensitive variables, like pitch, duration and intensity. The contribution to the perception of stress of the contextual-sensitive factor is greater than that of the inherent factor.1. QualityThe impact of the initials and finals of Mandarin syllables on perception of stress conforms to Sonority Sequencing Principle (SSP). SSP is subject to two factors. The first factor is real intensity or volume or loudness. The greater the volume is, the louder the sound is. And when the two speech sounds are similar in loudness, the one whose sound energy is located within the frequency range that the human ear is sensitive to is perceived louder.2. Lexical tonesThe order of the lexical tones that is most likely to cooccur with stress is high level > high falling > rising > low-dipping > neutral tone. The thesis claims that since lexical tones are the epitome of intonation, the perception of the inherent strength of lexical tones is consistent with the theory of intonation. The strength of lexical tones is embodied as the contrast of pitch height. The contrast within the F0 contour of lexical tones and the contrast of pitch height surrounding the lexical tone influence the perception of the strength of lexical tones.3. Vibration frequency of the glottisOne variable that can measure the strength of a syllable proposed in this thesis is vibration frequency of the glottis. The formula to calculating this variable is the integration of pitch contour with time. The variable involves pitch and time. Thus, this variable has a strong correlation with the strength of the syllable. The order of the strength of correlation with the perception of the strength of the syllable is as follows: vibration frequency of the glottis > duration > the upper limit of pitch > the pitch range > intensity > sound energy.Moreover, the thesis found that the number of syllables in a rhythmic unit has a reciprocal relationship with the average of vibration frequency of the glottis (based on syllables). As with law of conservation of energy, law of conservation of acoustic variables in speaking is formulated. The law states that at the balanced status of speaking, the product of the number of the syllables the speaker produces and the average vibration frequency of the glottis tend to be a constant. The formula is as follows: PULSE=K+T*(1/ N)N represents the number of syllables within a rhythmic unit. PULSE stands for the mean of the vibration frequency of the glottis within this unit. T means slope and K is intercept.It can be induced from this law that the total energy of the particular rhythmic unit tends to be a constant. This assumption will be insightful to understanding the phenomenon of focus stress. And it will help to understand the phenomenon of the negative correlation between the number of the syllables in a rhythmic unit and the average degree of stress of the syllables. II. Prosodic structureThe thesis agrees that the structure of prosody is multi-leveled, with syllable the lowest level, under the node of prosodic word which is in turn under the node of prosodic phrase. And prosodic phrase is under the node of intonational phrase. Finally intonational phrase comprises utterance.1. DurationAt the level of prosodic word and prosodic phrase, the speeding pattern is accelerating at first, and then running at a very high speed and decelerating at the end. However, at the level of intonational phrase and utterance, the speed is on average constant or with little speeding up or down.2. PitchAt different levels, there is a kind of declination. But at the level of utterance, the declination is very gradual due to the fact that the speaker needs to inhale and exhale to make up for energy for speaking. Therefore, the consumption of energy is the physiological reason for the declination.3. BreakThere is a pair of forces at the boundary of the phonological phrase, repellency and attraction. The roles of the two forces is to break and to join, respectively. Specifically, there are two kinds of repellency, the stronger one and the weaker one. The final lengthening is a kind of weaker repellency while pause is a stronger repellency (to break). These two kinds of repellency is complementary. The attraction is embodied as the pitch reset at the initial of the rhythmic unit.Taken together, the relationship among different levels is a continuum from attraction to repellency.III. Interaction between degree of stress and prosodic structure1. The degree of stress at different levels. At the level of prosodic word, four degrees of stress, 0 means weak, 1 means normal, 2 means secondary stress, 3 means primary stress. For prosodic word and prosodic phrase, the primary stress of a syllable can represent the degree of stress for the whole prosodic word and phrase. For the level above intonaitonal phrase, apart from considering the primary stress of the most stressed syllable, the mean strength of stress of the whole unit should also be taken into account. 2. From the level of prosodic word to the level of intonational phrase, the stress pattern shifts from domain-initial stress to domain-final stress.3. Acoustic parameters for stressThe level of prosodic structure and the degree of stress is correlated with the upper limit of pitch range, duration and vibration frequency of the glottis. Specifically,⑴The degree of stress at each level is correlated positively with the upper limit of pitch range, duration and vibration frequency of the glottis. But when the degree of stress is the highest one, ceiling effect will occur.⑵At the level of prosodic word, some data reveals that duration and the upper limit of pitch range is in a complementary relationship in cuing stress. For lower degrees of stress, lengthening plays a major role while for higher degrees of stress, the rising of pitch makes a greater contribution. Vibration frequency of the glottis has a positive linear relationship with the degrees of stress due to the fact this variable involves both duration and pitch.4. Acoustic parameters for stressFor prosodic word and prosodic phrase, the prominence of the syllable within the unit contributes more to the perception of stress whereas for the larger unit above intonational phrase, the contrast between rhythmic units play an important part for the perception of stress.5. The placement of rhythmic units influences the mean strength of stress globally. And the degree of stress embodies the local extreme value (e.g. maximum value). Moreover, the speaking style of the speaker will affect the impact of these factors to some extent.IV. Other conclusions1. The impact of the factors on prosody can be accumulated. One factor will not exclude the existence of the other factors but influence the weighting of the other factors.2. Every language has a multi-leveled prosodic structure. But the language specificity determines the different status for different levels. The rhythmic pattern of the dominant level of the prosodic structure in a particular language represents the rhythmic pattern of that language.