Reconciled with complexity in research on cognitive systems

The causes of human behavior cannot be simple. Every move we make has a nested hierarchy of causes that affect its direction, timing and form. The billiard-ball type of causality that is usually assumed to explain human action cannot give sufficient justice to this complexity. In this paper, I point to those perspectives that respect the complexity of cognitive systems and recognize that cognition involves changes on many nested time scales and in many nested systems. A brief overview of methods that are suitable for dealing with such interaction-dominant complex systems is presented and used as a background for describing a specific research program with the aim of clarifying the role of language as one of the nested factors shaping cognition. I illustrate this endeavor with two studies: one concerning the development of language as interaction control and another detailing how language may shape cognitive processes on several timescales. Reconciliation with complexity leads us to ask slightly different questions and expect different answers than when using simplified componential models of cognition and helps demarcate the limits of predictability.