Abstract
Due to learner corpus research and contemporary lingustic theory it is possible to formulate and verify hypotheses about acquisition of verbs and related syntactic structures in L2 Czech and to draw conclusions about teaching Czech as a foreign language. Previous research of the learner corpus MERLIN (Hudousková, 2023) showed that the knowledge of different meanings and valency frames of individual verbs does not significantly improve with increasing language proficiency and that non-native speakers use a limited number of central verbs even at higher language levels.
Therefore, the goal of language instruction at B2 level and higher is to increase the number of verbs and their meanings that non-native speakers actively use. In this sense, the role of learner corpora (CzeSL and MERLIN for Czech) is crucial, since they give an idea of language competence of non-native speakers at different proficiency levels and it is also possible to use them for didactic purposes so as to identify structures that non-native speakers (don't) use in comparison with native speakers of Czech.