Abstract

Sound reduction and deletion have been studied across many languages for some time. Usage-based approaches suggest that the more often a word is used, the more likely it is that some of the sounds are reduced. Phonetic environment, stress, and speech rate have all been studied as reasons for sound reduction or deletion. Most recently, frequency in reducing context (FRC) has been included when studying sound reduction and deletion. FRC in this thesis measures the portion of word tokens of a given word type that are followed by a reducing context. This thesis focuses on word-final /t/ reduction and deletion in German. Audio and transcriptions of six native German speakers were time aligned with the Montreal Forced Aligner for the analysis. Word frequency, phonetic environment, stress, and FRC were analyzed as factors that condition reduction and deletion. A linear mixed-effects regression model with the dependent variable of word-final /t/ duration found significance between a shorter /t/ duration and (1) a shorter duration of the preceding sound and (2) a consonant preceding the reduced /t/. A logistic mixed-effects regression model with the dependent variable of word-final /t/ deletion found significance between deletion and (1) a consonant preceding the deleted /t/ and (2) word frequency. Though FRC was not found to be significant in this study, perhaps measuring FRC with a different reducing context would be significant in a future study.

Degree

College and Department

Humanities; Linguistics and English Language

Rights

https://lib.byu.edu/about/copyright/