Language destabilization and (re-) learning from a Complexity Theory perspective: timescales and patterns across four studies

The growing interest in Complexity Theory (CT) within SLA has refocused our attention on development and change, as exemplified by the non-linear processes of language destabilization and (re-) learning in multilingual systems. It is clear that this theoretical perspective has important implications for empirical research from design to analysis and interpretation. In this chapter, I take a first step towards applying the theoretical apparatus of CT to findings obtained in four different studies of multilingual development. I show how the findings can be interpreted using CT as a post-hoc theoretical prism and, in turn, how the findings support and explicate CT claims. The four studies involved multilingual adult L2 learners and users in Ireland, and were designed to address different timescales ranging from weeks to years, and time windows from one year to decades. Thus, the study goes some way towards de Bot’s (2014) suggestion to study complex dynamic systems by focusing on adjacent timescales. Although the studies differ in their purpose and design, they partially overlap in methodology and analysis, and exemplify how, through convergence of methodological and analytical procedures, one may generalise from the particular.