ABSTRACT

Retrospective analysis of situations in which systems cross the boundary of safe operation is one of the most powerful, and commonly used, approaches for improving safety. Not only do they highlight specific cases in which the system has become unsafe at a component level, they can also expose underlying weaknesses at a bureaucratic level. Reason’s (1990) ‘Swiss cheese model’ serves as a powerful metaphor for introducing this concept. The model highlights how latent conditions at different organisational levels can set the conditions for an accident to occur. While the concept of considering accidents at multiple levels of the organisation is becoming prevalent in the scientific community, a culture of individual culpability, and blame, is still prevalent in our society. In the follow-up to accidents, particularly in cases where human life is lost, calls are frequently made to identify individuals involved and hold them accountable. As Dekker (2007) discusses, the Newtonian view of causality, which assumes symmetry between cause and effect, can lead to the last person performing an action being held responsible for the catastrophic

consequences. As this paper will show, the relationship between cause and effect within complex sociotechnical systems is rarely straightforward.