Attitudes are a fundamental characteristic of human activity. Their main function is the situational assessment of phenomena in practice to maintain action ability and to provide orientation in social interaction. In sociolinguistics, research into attitudes toward varieties and their speakers is a central component of the analysis of linguistic and cultural dynamics. In recent years, computational linguistics has also shown an increased interest in the social conditionality of language. To date, such approaches have lacked a linguistically based theory of attitudes, which, for example, enables an exact terminological differentiation between publicly taken stances and the assumed underlying attitudes. Against this backdrop, the present study contributes to the connection of sociolinguistic and computational linguistic approaches to the analysis of language attitudes. We model a free text corpus of user comments from the RTL.lu news platform using representation learning (Word2Vec). In the aggregated data, we look for contextual similarities between vector representations of words that provide evidence of stances toward multilingualism in Luxembourg. We then contrast this data with the results of a quantitative attitudes study, which was carried out as part of the crowdsourcing project “Schnëssen.” The combination of the different datasets enables the reconstruction of socially pertinent attitudes represented in public discourse. The results demonstrate the central importance of attitudes toward the different languages in Luxembourg for the cultural self-understanding of the population. We also introduce a tool for the automatic orthographic correction of Luxembourgish texts (spellux). In view of the ongoing standardization of Luxembourgish and a lack of rule knowledge in the population, orthographic variation—among other factors like code-switching or regional dialects—poses a great challenge for the automatic processing of text data. The correction tool enables the orthographic normalization of Luxembourgish ...