At present, much of the research on bilingual cognition focuses on late second language learners of a small number of languages. In this talk, I show how we can widen the net by integrating advances in the field of bilingualism with the study of endangered languages. Drawing on recent studies from Europe and Latin America, I demonstrate that experimental psycholinguistic methods can be successfully applied outside the lab and, conversely, how data from these understudied populations provide new insights into the adaptive capacities of the bilingual mind. I will discuss, in particular, how bilinguals manage competing conceptualizations of time and space, how their grammars and language mixing patterns adapt to cognitive constraints such as the need for simplification, and how language processing concurrently adapts to their complex bilingual experience.