Understanding the pattern of interactions between the languages of multilingual speakers, and the mechanisms supporting efficient learning of a foreign-language are imperative in today's global society. For one, the majority of individuals in the world use more than one language in their daily lives. Second, many individuals are required to learn a foreign-language as adults, to be able to acquire higher education, to advance in the work place, and the like. Developing efficient instruction methods is thus an essential endeavour. The proposed work will address both of these issues, focusing on learning and representation of words.The mapping between lexical-form and word meaning(s) is often ambiguous both within a language and across languages. Such indirect mappings pose particular difficulty for vocabulary learning, because learners need to achieve stable links between the to-be-learned lexical form and an unequivocal specification of its meaning. The proposed project focuses on this challenge, and examines it's consequences for first-language representation. The first objective is therefore to explore the consequences of foreign-language vocabulary learning on meaning representation of first-language words. Such backward influences are theoretically important because they speak to the dynamic and interconnected nature of the bilingual lexicon. The second objective is to improve foreign-language vocabulary learning by (1) identifying ambiguity types that create challenges for learning, and (2) test different instruction methods tailored at alleviating the difficulty associated with these ambiguity types. The third objective is to examine how learner's characteristics interact with this particular complexity. We focus on (1) individual differences in working memory span and the ability to ignore irrelevant information and on (2) language experience, comparing how monolingual and multilingual speakers deal with ambiguity in learning.