Dynamic Systems Theory Approaches to Second Language Acquisition

Dynamic systems theory (DST) is a theory of change. Starting in the 1960s, the theory has been used extensively in a wide variety of disciplines, from physics to biology and from meteorology to demography. Essentially, DST is an application of mathematics, in which change of complex systems over time is expressed in dynamic equations that describe how these changes take place as a function of time. Each next developmental step (or iteration) is determined by the state of the system at the preceding moment, which leads to unpredictable, nonlinear developmental patterns. DST has provided powerful accounts for the nonlinear, iterative development of natural phenomena. Examples of applications of dynamic systems that show iterative changes over time are found in reaching dynamics (when we reach for a cup on the table, we iteratively adjust the trajectory of our movement) and weather forecasts in unstable atmospheric areas (tomorrow’s weather can be predicted accurately, but long-term predictions are no more than educated guesses due to the large number of dynamic iterations). Recently, applications of DST have been developed within cognitive science (Thelen & Smith, 1994; Port & van Gelder, 1995) developmental psychology (Van Geert, 1998), and language development (Elman, 1995). Since the late 1990s, the theory has also been applied to second language acquisition (SLA) (Larsen-Freeman, 1997; Herdina & Jessner, 2002; De Bot, Lowie, & Verspoor, 2007). These and many other authors have argued that language can be seen as a complex dynamic system and that language development is a nonlinear, chaotic, and highly individual process that cannot adequately be described from a static point of view. Different authors have used different terminologies for complex dynamic systems. This difference is mainly caused by parallel developments in different areas of research. Within the applications to language development, the terms “complex systems,” “complex adaptive systems” (Ellis & Larsen-Freeman, 2009), and “dynamic systems” (De Bot et al., 2007) refer to similar approaches, although “dynamic systems” is more frequently associated with studies that use mathematical modeling (e.g., Caspi, 2010). The arguments for considering (second) languages as complex dynamic systems revolve around three crucial characteristics that can suitably be applied to language and language development: the existence of interconnected subsystems; the tendency to self-organization; and the occurrence of nonlinear, chaotic patterns of development. Each of these characteristics will be discussed in some detail below.