Adult bilingualism: what does the big picture tell us?

Research on adult bilingualism shows that selective aspects of grammar become variable in speakers experiencing native language (L1) attrition from long-term exposure to a second language. These are the same aspects that remain variable even in highly proficient non-native (L2) speakers of the same language. Why do we see this convergence between L1 attrition and L2 acquisition when we compare different groups, and only for some language structures? I will consider three possible accounts, considering their strengths and weaknesses in the light of available data from different bilingual contexts and language combinations. At this stage, two general conclusions are possible: first, understanding the big picture requires serious consideration of individual differences; second, it also requires interdisciplinary research on different aspects of bilingualism that combines the insights of both linguistic and cognitive models.