Towards Modelling Pausing Patterns in Adult Narrative Speech
Loading...
Date
2020-12-08
Authors
Kallay, Jeffrey
Journal Title
Journal ISSN
Volume Title
Publisher
University of Oregon
Abstract
The study that is the focus of this dissertation had 2 primary goals: 1) quantify systematic physiological, linguistic and cognitive effects on pausing in narrative speech; 2) formalize a preliminary model of pausing behavior which integrates these effects with those related to individual differences. As a natural consequence of speech-language production, pausing behavior has been of interest to linguistics researchers since at least the mid-20th century. A large number of previous studies have demonstrated systematic effects from factors related to linguistic structure and cognitive processing on speech pausing patterns. Despite that work, a comprehensive understanding of how those factors interact in shaping pausing behavior remains elusive. This is largely due to the highly complex nature of pausing behavior, which also includes more idiosyncratic factors which are difficult to quantify. The current study was aimed at addressing this problem by adopting a more holistic framework.
Three separate within-subject experiments were conducted to alternatively assess effects from respiratory recovery, speech planning, language structure and discourse planning processes on narrative speech pausing behavior. Each of these factors was found to systematically influence both the frequencies and durations of the pauses that were produced. The results also suggested effects on the variability of pause durations, but these effects were primarily limited to the low-level respiratory and speech planning factors. These results together formed the basis of a preliminary model of narrative speech pausing which was tested in the final part of the study. In addition to the experimentally investigated factors, a separate component related to working memory capacity was included in the model to account for the role of individual differences. On the whole, the model was only able to account for a relatively small amount of the variation in the pausing patterns. The language structure component was found to be the most significant contributor. The implications of these results are discussed with reference to the limitations of the preliminary model proposed in the current study. These limitations provide several suggestions on the best directions for future work in attempting to refine the model toward more accurately characterizing narrative speech pausing behavior.
Description
Keywords
Pausing, Planning, Recovery