Modelling a behaviour of intelligent agents, it seems quite natural to study such action scenarios.In this paper we consider action scenarios where some actions have typical and/or atypical effects. We propose a preferential strategy adequate for reasoning about such actions. This method, intuitively...