Children clearly learn language from the ambient input, but not from input alone. One of the main questions in theoretical linguistics and language acquisition is how much is provided by an innate endowment and how much must be learned from the primary linguistic data. The present project addresses this central issue within a new approach to language acquisition and attrition, the micro-cue model (MCM) that the PI has developed in a number of publications in recent years (e.g. Westergaard 2009a, 2014) based on data from monolingual acquisition. The MiMS project will extend this theoretical approach to multilingual situations, and new data will be collected from several populations of bi- and multilingual children and adults, e.g. German-Russian bilingual children, Norwegian-American heritage speakers, and bilingual children learning English as an L3. The focus is on (morpho-)syntactic micro-variation in Norwegian, Russian, German and English, related to word order and certain aspects of nominal structure (grammatical gender, determiner use). These languages and these syntactic constructions are chosen as they represent an interesting combination of challenges, both for children acquiring and adults maintaining these systems in multilingual contexts. According to the MCM, both acquisition and diachronic change take place in very small steps (e.g. Westergaard 2008, 2009d). By identifying the small steps in these processes (referred to as micro-cues), the project will increase our understanding of the human language faculty and the building blocks of language. It will also investigate the effect of more general factors such as complexity, frequency and economy in the acquisition and attrition processes. The MiMS project will thus make important contributions to current research in the fields of language acquisition, multilingualism and theoretical linguistics, providing new insights with potential to change the current dominance of the two opposing schools in the field, generativism and constructionism.