New High German Texts for Evidence of Phrasemes

Most dictionaries containing phraseological information are restricted to a synchronic perspective. Diachronic information on structural, semantic, and pragmatic change over time has to be reconstructed by a time-consuming consultation of various dictionaries providing only punctual insights. In the OLdPhras, project we construct an online dictionary for diachronic phraseology in German from ca. 1650 to the present by combining dictionary exploration with corpus-based methods. This paper highlights some challenges we have met: How to select the “interesting” phrasemes, i.e., those that underwent some change? How to deal with historical corpora? How to include different kinds of phraseme variation? We present a semi-automatic corpus-based approach for the investigation of phraseme development. We argue for a combination of dictionary exploration and corpus-based methods to provide reliable and extensive information about the diachronic development of German phrasemes.