Research in the LeaP project is based on a large corpus of foreign language learners' speech. A multitude of data of various types is being collecetd: the corpus of spoken language consists of 359 recordings of between 2 and 20 minutes length. It comprises four different speech styles:
Scientists