hide
Free keywords:
-
Abstract:
In everyday life conversations are riddled with disfluencies: pauses, uhm’s, slow tempo, corrections, repetitions, etc. When assessing the fluency level of a non-native speaker, it has been shown that these acoustic features play a large role. Particularly the pause and speed characteristics of speech contribute much to fluency ratings. But native speakers also portray these symptoms of spontaneous speech and as yet the relationship between native and non-native fluency remains unclear. Native fluency might fundamentally differ from non-native fluency, or it may be a gradient distinction. The current study directly compares the concepts of native and non-native fluency by means of phonetic manipulations. In two experiments, the number and duration of silent pauses (Experiment 1) and the speed of the speech (Experiment 2) were digitally manipulated. Fluency ratings by native listeners on these manipulated speech fragments revealed that increasing the number or the duration of silent pauses both led to a decrease in fluency judgments. Despite the clear gradient difference in fluency level of native versus non-native speakers, no evidence could be found for a difference in the effects of the pause manipulations across native and non-native speech. Results from Experiment 2 will demonstrate whether the same holds for speed manipulations in native and non-native speech. The results from Experiment 1 at least suggest that the notion of fluency is constant across native and non-native speech.
