On the nature of complexity in cognitive and behavioural science

The term 'complexity' is often found in the behavioural and cognitive sciences. This paper compares the common-sense usage of the world with attempts by information theorists to provide a formal definition. The concepts of randomness, predictability, organization and complexity are reviewed and we advocate that a distinction be made between traditional 'algorithmic' and a more biologically relevant 'organizational' complexity. We develop the formal properties of the latter with reference to hierarchies of explanation. The utility of both measures of complexity is discussed and examples of practical application given. We discuss relationships between complexity of environment, cognition, behaviour and brain, outline a possible metacognitive function for perception of complexity, and consider the implications for evolutionary psychology of the special nature of social complexity.