This paper describes how an event-based multilinear description of German phonotactics which was originally designed for a speech recogniser (cf. CarsonBerndsen 1993, Hiibener & Carson-Berndsen 1994) can be used in a multimedia application for pronunciation teaching. The event-based phonotactic description (cf. Carson-Berndsen 19926) has been developed within the paradigm of finite-state phonology, is declarative, and can be employed in procedurally different tasks such as recognition and synthesis in audio and visual media. A number of applications have been developed thus far. The application described in this paper has been designed with two didactic goals in mind:

• for teaching articulatory phonetics in an introductory course in linguistics, • for use in the teaching of pronunciation in second language learning.