Resilience Architecture and Normal Accident Theory

The text analyzes Charles Perrow's Normal Accident Theory (NAT), arguing that catastrophes in complex systems with tight coupling are structurally inevitable. The author divides systems into three categories depending on the level of risk and social benefit. The most critical, such as nuclear power or atomic weapons, may require complete dismantling due to the unforgiving nature of potential errors. Medium-risk systems, e.g., genetic engineering, require rigorous control and a slower pace of implementation. Meanwhile, systems such as aviation are tolerated thanks to a reporting culture and incident analysis. The main message is a transition from technocratic thinking about probability to responsibility for the structure of consequences and the social distribution of risk.