Human languages across the world have a wide range of variations in their grammatical properties. But, is there any limit to such variations? The search for properties holding across languages, or language universals, is a fundamental goal in linguistics, as it offers a window into our shared linguistic ability as a human kind. Meanings of *single words* is one area in which the research on language universals has been highly successful. However, the current literature is crucially limited in its empirical scope; it is largely restricted to traditionally logical words (e.g., “every”). This collaboration extends the research on word-meaning universals into an uncultivated domain: modal words, such as "may"/”must” and "believe"/"know". Recent semantic theories of *question sentences* suggest that the semantic behavior of modal words in questions reveals non-trivial cross-linguistic universals about modal meanings. To explore this line of research, it is crucial to systematically investigate a wide range of cross-linguistic data. In this project, we will entertain and evaluate hypotheses about modal lexical-semantic universals in view of language samples including Dutch, German, French, Hungarian, Romanian, Turkish, Ngamo (Nigeria), Japanese and Zapotec (Mexico). Given the empirical work undertaken in this project, the network will develop a proposal for the ERC Synergy Grant to further investigate the comprehensive theoretical explanations of word-meaning universals. This future project explores linguistic and cognitive-psychological explanations for why the universals exist in the first place, integrating methodologies from humanities, cognitive science and computer science.