German research literature shows a great lack of empirical studies of sentence adverbs. Indications about their significance and linguistic application in dictionaries are sometimes inappropriate or even misleading. Apart from that, there is little corpus linguistic research about synsemantica in the German language. Initially, however, we have to question whether words that are poor in content such as sentence adverbs can actually be empirically studied with the support of a corpus, for instance the Deutsches Referenzkorpus. In our opinion, they undoubtedly can. The following essay will demonstrate where corpus based linguistic research can lead. Step by step, the methodical approach, being at the basis of the research of sentence adverbs in modern German, will be illustrated in the discussion (chapter 2). The two examples glücklicherweise and paradoxerweise will furthermore exemplify the applied methods (chapter 3). Finally, the comparison of newly acquired results with entries in dictionaries clearly proves the great value of such research in the fields of lexicography and German as a foreign language (chapter 4).