Abstract

This thesis distinguishes Classical Behavioural Economics (CBE) from Modern Behavioural Economics (MBE) and discusses Herbert Simon's pioneering contributions to CBE in detail. CBE emphasises the role of bounded rationality, satisficing and heuristics (procedures) in human decision making in contrast to optimization, which is widely used in MBE. In the framework of CBE, heuristics are algorithms, which are embedded in the Information Processing Systems (IPS) in Human Problem Solving (HPS) environments. The argument here is that the premise of CBE is highly suitable for algorithmic modelling of adaptive agents facing complex economic environments. It explores the theoretical foundations of bounded rationality and substantiates the difference between satisficing and optimization, and claims that the former is more general than the latter. These investigations are carried out from the perspective of Computability Theory and Computational Complexity Theory, which provide a coherent foundation for Herbert Simon's work on problem solving and for CBE in general. The second part of the thesis explores the game of Go from the perspective of problem solving in CBE. Its generality, suitability and complexity are analysed in detail. A pseudo IPS which is capable of playing Go is also constructed based on the insights from Simon's work on protocol analysis and information from Go documentaries. The claim of this thesis is that Go can be a powerful paradigm to better understand human economic behaviour in diverse and complex environments.