Humanities rely on both field research data and databases but rarely have the means necessary for employing them. Crowdsourcing on the Web using social media specifically designed for the purpose offers a promising alternative. This article reports about two endeavors of this kind: enriching an art history database with semantic interpretations and collecting and assessing the regional and social origins of parts of speech for a linguistic investigation. The article motivates and describes the approach and further introduces into the semantic analysis method based on higherorder singular value decomposition specially designed for the project.