Zwischen theoretischer Modellierung und praxisnaher Anwendung. Zur korpusgesteuerten Beschreibung usueller Wortverbindungen

This paper shows how a corpus-driven approach leads to a new perspective on central issues of phraseology and on lexicographical applications. It argues that a data-driven pattem search (applying Statistical methods), an a posteriori interpretation of the data and a user oriented documentation of the usage of multi-word units (e. g. in lexicographical articles) constitute a step-by-step process where each step has its own informational value and useflilness. The description of multi-word units (Usuelle Wortverbindungen) presented in this paper focuses on the second Step, the high quality analysis and interpretation of collocation data, exemplified by the fields of multi-word units centered around the word formslIdee/Ideenl(idea/ideas)