User Manual Privacy Policy Disclaimer Contact us
  Advanced SearchBrowse





When is speech fluent? The relationship between acoustic speech properties and subjective fluency ratings.

There are no MPG-Authors available
There are no locators available
Fulltext (public)
There are no public fulltexts available
Supplementary Material (public)
There is no public supplementary material available

Bosker, H. R., Pinget, A.-F., Quené, H., Sanders, T., & De Jong, N. H. (2011). When is speech fluent? The relationship between acoustic speech properties and subjective fluency ratings. Poster presented at the Workshop Production and Comprehension of Conversational Speech, Nijmegen, The Netherlands.

Cite as: http://hdl.handle.net/11858/00-001M-0000-0023-D7CA-3
The oral fluency level of an L2 speaker is often used as an important measure in assessing language proficiency. In order to improve the objectivity of such language tests, previous studies have attempted to determine the acoustic correlates of fluency (e.g., Cucchiarini et al. 2002). Many of these studies have used multifaceted global measures making the results often difficult to interpret. An example of such a measure is overall speech rate which is confounded because it relates both to speed of articulation and to the use of pauses. Also there is within the literature much diversity in the type of instructions raters were given. Arguing that fluency ratings are dependent on the perception of the acoustic characteristics of speech, Experiment 1 investigated fluency perception by establishing what speech properties raters are capable of perceiving. Three groups of listeners rated the same set of L2 Dutch speech stimuli on either the use of pauses, speed of delivery or the use of repairs (corrections and repetitions). Stimuli were 20sec excerpts from turns in a simulated discussion. Using linear mixed models the subjective ratings were modelled by non-confounded acoustic measures which only measured one aspect of fluency (pause, speed or repairs). Explicit and very specific test instructions resulted in high interrater reliability. Most of the variability of the ratings from the pause group and the speed group was accounted for by pause or speed measures, respectively. Concluding that raters are capable of perceiving and rating pause and speed phenomena (but repair phenomena to a lesser extent), a fourth group of listeners rated the same stimuli on overall fluency. The variability of these ratings was best modelled by pause and speed measures. It is concluded that pause and speed measures are better acoustic correlates of fluency than repair measures. Considering the strong effect of pause measures on fluency perception, Experiment 2 investigates the independent effects of the number of silent pauses and the duration of silent pauses, both in L1 and in L2 speech. Instead of looking at correlations, this experiment attempts to establish a clear causal relationship between these two acoustic speech properties and fluency ratings. By comparing the ratings on identical stimuli differing only in the number or the duration of silent pauses, this experiment reveals whether the number of silent pauses and/or their duration have any effect on fluency perception, both in L1 and in L2 speech.