The project deals with theoretical linguistic research in derivational morphology of Czech, taking advantage of interconnecting derivations with inflectional morphology. The research team, including linguists and experts in Natural Language Processing, concentrates on three closely connected topics. First, vowel and consonant alternations occurring during derivation are modelled on the basis of alternations present in inflectional paradigms. Second, a data-based approach to morphemic analysis is developed, profiting from insights from historical grammar of Czech and from comparative Slavic linguistics. Third, a semantic classification of relations between base words and derivatives is proposed and applied to the core part of the lexicon of Czech. The results are applied on other two highly topical issues: the concept of intercomprehension, well-established in research in Romance languages, is addressed with regard to Slavic languages, and the benefit of a formalized description of derivational morphology in Machine Translation between related languages is tested.