DIT’s Dynamic Speech Corpus

Given the current focus of language learning on learner autonomy, the Technological University Dublin is using its FLUENT project (commencement date 7/1/’08) to produce a Dynamic Speech Corpus, which will be a valuable resource for learners, teachers, authors and researchers. The resultant database will afford access to a unique audio resource based on unscripted dialogues between friends, characterized by industry-standard audio quality (for analysis purposes) and a high degree of naturalness. By means of search string and/or tags such as speaker intention, users will be able to find samples of native-to-native speech (various L1 varieties) on a chosen topic, particularly speed-induced elisions, reductions and vowel centralizations. They will also be able to study sequences of interest in their full, natural, dialogic environment. They will be able to listen to and contrast what Cauldwell (2002) calls the ‘blur of natural speech’ with its orthographic, ‘static’ counterpart, thus allowing them to focus on the manner in which native speakers produce those difficult reduced forms of English. At all stages the user will be able to slow down sequences for heightened intelligibility. The presentation will outline how a scaffolding approach will allow students following any learning method to derive benefit from the Dynamic Speech Corpus