Rethinking the choice of filled pause as a sociolinguistic variable
The choice of filled pause has been proposed as a sociolinguistic variable, which reflects a change-in-progress in which “um” is gaining popularity relative to “uh”. This trend is not only observed in English (Tottie, 2011; Acton, 2011; Fruehwald, 2016), but also found across Germanic languages (Wieling et al, 2016). In this study, two potential covariates which could partially explain the variation in the choice of filled pause are further examined: the topic of conversations and the accommodation between interlocutors. Through a large sample of spontaneous telephone conversations compiled from the Fisher corpus, it can be shown that “uh” has greater variability in its distribution compared to “um”, when variations in speaker’s age, conversation topic, and interlocutor’s choice of filled pause are considered. This difference suggests a potential functional difference between the two filled pauses, which further implies possible effects from varying speech planning strategies in different conversation contexts.
