"Speakers of German enjoy forming compounds and the German language is infamous for long words like 'Rindfleischetikettierungsüberwachungsaufgabenübertragungsgesetz". Even though compound formation is an easy task for speakers, the linguistic analysis of the semantic relations of the stems of a compound is a complex task. This talk will discuss possibilities of how we can use compound analysis for a deeper understanding of cultural change, discuss data-driven methods, and present empirical evidence from large German newspaper corpora. The talk will present: 1. a quick overview of the different word formation processes in German, 2. different heuristics for the semantic analysis of compounds, 3. analysis of distributional patterns of stems in large corpora, and 4. possibilities of a data-driven identification of the semantic relations between the stems."