Prediction, control and the challenge to complexity

The dominant discourse in research, management and teaching is one that may loosely be characterised as that of prediction and control. The objective of research is to identify causal correlations within policy, management, teaching strategies and educational outcomes that are sufficiently robust as to be able to predict outcomes and make generalisations across practice. With the construction of a cumulative bank of evidence, given clearly specified educational outcomes, we might see research as supporting the improvement of schools towards 'effective' or 'best' practice. This discourse assumes that education, though complicated, nevertheless takes place within a bounded system of relatively stable, linear and balanced causal interactions. This perspective on schooling may be understood as the 'prediction/control' paradigm. This paper explores the possibility of an alternative perspective, one that is based on principles derived from complexity theory and which may be referred to as the 'complexity' paradigm. Under this paradigm schools are seen as open systems, subject to non-linear and dynamic interactions among the multiple factors of which they are constituted, and often unpredictable. This paradigm, it is argued, is subversive of our ambitions to prediction and control. Researchers need to look to formal enquiry methods adopted in relation to other sites of complexity. In the meantime practitioners and researchers may benefit from various conceptualisations to which complexity theory gives rise, i.e. those of recursive symmetries, attractor states, sensitivity at bifurcation points and the phenomenon of 'lock in'.