Decisions in complex systems: finding a balance between competing forces

Disastrous accidents have shown that small deficiencies and shortcomings can lead to large losses in terms of human life, environmental destruction and economic values. Risk research has identified human and organisational factors as the most important set of underlying causes to such accidents. This has made it increasingly evident that decisions taken by managers at different levels in the organisations have a very large influence on the initiation and outcome of such events. The paper discusses models of decision making and complex systems in a search for some kind of "superrational" model of decision making, which would provide a better support in decision making situations that occur in the control of high-risk technologies. The concept of interconnected systems of intelligent autonomous agents may be suggested as a candidate for such a model. The drawback is however that such a model can be used only in a qualitative and metaphoric sense for understanding safety-oriented organisations and not for an accurate prediction of their behaviour.