Informations and abstract
The aim of this article is to understand from a sociological point of view the different logics of inquiries (and their perverse effects) that may take place after technological accidents in organizations: the logic of the identification of faults and responsibilities and the logic aimed to the improvement of the system. The thesis of this article is that the mere punishment, or else the dismissal, of the presumed responsible of an accident occurred in a complex organisational system may cause the reiteration of accidental events, even when actors are different. Furthermore, the inquiry aimed to identify faults and responsibilities often generates "organizational inertia" and "un-learning" processes, due to the so-called "blame culture", but it does not eliminate the risk-conditions present in the system. Besides, the article underlines that the logic aimed to the improvement of the system, oriented to the prevention of accidents and to the reinforcement/improvement of safety conditions, does not have the same institutional strength and legitimacy of the inquiry aimed to identify faults and responsibilities, and improvementoriented inquiries may be unheeded. It is consequently difficult for complex organizational systems to develop "auto-improvement" processes, alternative to the penal system, in case of technological accidents or organizational failures.