Changes to research methodology motivated by the adoption of a complexity theory perspective on language development are considered. The dynamic, nonlinear, and open nature of complex systems, together with their tendency toward self-organization and interaction across levels and timescales, requires changes in traditional views of the functions and roles of theory, hypothesis, data, and analysis. Traditional views of causality are shifted to focus on co-adaptation and emergence. Context is not seen as a backdrop, but rather as a complex system itself, connected to other complex systems, and variability in system behavior takes on increased importance. A set of general methodological principles is offered, and an overview of specific methods is given, with particular attention to validity in simulation modeling.