Beyond "bounded rationality" : behaviours and learning in complex evolving worlds

This work challenges the very notion of bounded rationality as dangerously too near to some "unbounded rationality" used as a benchmark. Should we assume that there is an "unbounded" rationality as a benchmark? Should one start, in order to describe and interpret human behaviour, from a model which assumes that we, human beings, have complete and well-defined knowledge of our preferences, all possible states of the world, all possible actions (our "technologies"), the mappings among them, and then look for possible "bounds" and "biases"? Our answer is negative. Rather, the question should be: how do human agents and organizations thereof actually behave in complex and changing environments? Answering this question, we suggest, entails also a significant departure from what is now accepted as behavioural economics, often meant as the analysis of more or less significant deviations from the "Olympic rationality". On the contrary, we suggest, human beings and human organizations behave quite distinctively from the prescriptive model derived from the axioms of rationality.