This study investigates how Czech instrumentals are used as shell nouns in L2 data, i.e. the learner corpus of CzeSL-SGT and LCSL