Universal contrastive analysis as a learning principle in CAPT

Cross-linguistic comparison is a good starting point for computer-assisted pronunciation training (CAPT). A comparison between the segment inventories of a learner’s mother tongue (L1) and the target language (L2) can be made on the basis of the IPA categories. Since these are claimed to reflect universal phonemic distinctions, mastering the contrasts in the target language in perception and production ensures communicative effectiveness for learners at the segmental level. The ComputerAssisted Listening and Speaking Tutor (CALST) implements contrastive analysis in two types of exercises for phonetic and abstract listening. In these exercises users can practice with word pairs/sets demonstrating unfamiliar sound contrasts of the target language to improve their perceptual discrimination. Since substitutions for unfamiliar sounds depend on L1, the selection of sound contrasts which are trained in the exercises should also depend on L1. We shall argue for a pragmatic approach to the selection of exercises.