Abstract
In the present article, we focus on valency and synonymy of verbs in a bilingual, Czech-English setting. Our research of semantic equivalence of verbs is based on the FGD theory on the syntactic side (including valency), and gets main inspiration from FrameNet and VerbNet on the semantic side.
As the main source of evidence, we use the Prague Czech-English Dependency Treebank 2.0. We consider this "bottom-up" approach a novel and appropriate approach to study verbal synonymy.
Synonymous Czech and English verbs are being grouped into cross-lingual synonym classes and captured in the new CzEngClass lexicon. This lexicon contains not only mappings of valency arguments to semantic roles for each member of the synonym group, but also links them to individual verb entries in FrameNet, VerbNet, Vallex(es) and Czech and English WordNets, making CzEngClass also a richly interconnected lexicon.