Compare And Contrast Of Filled Pauses And Their Distributions In Lectures Of Liberal Arts And Natural Sciences

Disfluencies are normal in the spontaneous everyday speech. Research on disfluencies is one of the approaches to study speech production. Based on the recorded lectures of 36 university teachers, here was research on filled pauses in Chinese speech. The materiel consists of 116989 Chinese characters and the time of speech in this research is 540 minutes.In study 1, the frequencies of filled pauses in Chinese speech were investigated, with the number of fillers in per 100 Chinese characters as the index. There were 3.93 fillers in per 100 Chinese characters. The common ones were "a", "ne", "zhege", "en", "e" , which compose 92.9% of all the fillers. Among these five, "a" was sensitive to the speech domains. Lecturers in Liberal Arts used more "a" than lecturers in Natural Sciences during their lectures. Male lecturers prelecting in Natural Sciences use more "zhege" than Males prelecting in Liberal Arts.On the basis of Study 1, study 2 focused on the distributions of the most common fillers "a", "ne" and "zhege". This part was accomplished by analyzing the part of speech of the words preceding and following the fillers, and the index was the percentage of a certain word-class. The results follow:(1) In speech production, when lecturers formed the whole thoughts, recalled concepts, or described concepts mentioned formerly, lecturers used "a". When the whole ideas had been in the lecturers' minds, the lecturers tended to use "ne". When lecturers used "zhege", there was the possibility of 63.37% that they were recalling a concept. On the distribution of the fillers, disciples of the lectures and the gender of the lecturers only affected a few word-classes.(2) Disciples of the lectures and the gender of the lecturers did not affect the durations of the pronunciation of "a" and "ne". The duration of "a" was longer than that of "ne".(3) Compared with the results about English speech, taking the distribution and durations together, the filler "a" in Chinese speech was equivalent to "um" in English and "ne" is equal to "uh".(4) Disciples of the lectures and the gender of the lecturers did not affect the repeats and modifications of word or words where fillers happened. There were more repeats with an "a" than where "ne" and "zhgge" were. But the numbers of the modifications are not different where the three fillers happened.Some results of other languages were validated in this study. Our results supported that fillers in spontaneous speech are related to the loads of speech plan. Further studies about disfluencies in Chinese would be necessary to validate the results of this one.