A Comparative Case Study Of Disfluency In Spontaneous Speech

Oral English has long been a subject for English teaching and research. It has been investigated from many different perspectives, including disfluency. This thesis made a comparative case study of disfluency in spontaneous speech between native speakers and advanced non-native speakers. Hopefully, the results we got will be valuable to future oral English teaching and research.From a cognitive point of view, disfluency is the discrepancy caused by various reasons from ideal speech production. Though, it has been recognized that a certain amount of disfluencies are inevitable, too many disfluencies can be harmful to oral fluency. Besides, disfluency can also be regarded as an overt manifestation of 'an interlanguage' and its types, quantity and frequency, to some extent, can reflect the speaker's development of oral fluency in the interlanguage continuum. This thesis put the native speakers in the position of 'reference', and made a comparative study to find out the differences between native speakers and advanced non-native speakers in terms of disfluency.Our study selected 8 subjects, 4 native speakers and four advanced non-native speakers and recorded what they said towards three topics we provided. Thereafter, we marked all the disfluencies including pauses, repeats, self-repairs, errors not repaired (Though they are not part of disfluency, they are prerequisite in that without them, part of the analysis would be misleading.), and classified them into corresponding categories and subcategories based on which a quantitative and qualitative analysis was carried out.The results were, in general, in accordance with our previous assumption, i.e. the average percentage of each category of native speakers was lower than that of advanced non-native speakers which may serve as evidence for the gap between advanced non-native speakers and native speakers. The gap was considerably large in terms of unlexicalized fillers, self repairs and errors not repaired. However, when we looked into the subcategories, it was interesting to notice a few phenomena which were in contrary to our previous assumptions and to other research findings. The most significant one was the extensively use of lexicalized fillers by advanced non-native speakers. We attempted to explain it in the sense that the four advanced non-native speakers we selected had overcome the first fossilization stage and were still approaching native-like fluency.