Discourse relations in Czech and their representation in an annotated corpus of texts

The present thesis is a contribution to the widely discussed issue of how the syntactic structure of a sentence and the structure of discourse (text) are related. The syntactic sentence structure along with other language phenomena participates in building a coherent, comprehensible discourse. The author calls the syntactically motivated relations in discourse connective relations. These relations include coordinating relations and some of the subordinating relations within a sentence and, secondly, adjoining of discourse units across the sentence boundary. The explicit means of expressing connective relations are called discourse connectives. It is a group of language expressions that connect or adjoin discourse units while indicating the type of semantic relation between them, i. e. conjunctions, some subjunctions, particles and adverbials, and marginally also some other parts-of-speech. The present thesis describes the semantic category of discourse connectives in Czech on the basis of language data and their syntactic annotation in the Prague Dependency Treebank, and thus aims to contribute to the design of a language corpus annotation scenario capturing the discourse relations in Czech