Justus Liebig University Giessen

Fluency in Native and Non-native English speech: Theory, Description, Application(s)

Most of the fluency-related research so far has focused on one of the following aspects: (1) temporal variables in speech production, (2) the use of prefabricated units, (3) certain performance phenomena which contribute to a perception of naturalness in speech. This project combines these three approaches and aims at defining functional categories abstracted from features of spoken language that are relevant for a native speaker's fluency. Once these features are established, a contrastive analysis of the learner data and comparable native data is possible. As a database for the analysis, the 90.000-word German learner corpus LINDSEI-Ger (Louvain International Database of Spoken English Interlanguage) and the comparable 100.000-word native speaker corpus LOCNEC (Louvain Corpus of Native English Conversation) will be used. To test the relevance of the findings to native speakers' assessments of the fluency of learners of English, the corpus analysis will be combined with an experimental set-up in which native speakers' ratings of the fluency of learners will be compared with the prediction of the degree of fluency of the learners according to the data-analysis.