Normal Accident Theory
Some tightly coupled, complex systems may make certain failures difficult to prevent entirely.
The Mechanism: Charles Perrow's normal accident theory argues that accidents may be inevitable in some systems that combine interactive complexity with tight coupling. Unexpected interactions can move faster than operators can understand or contain them.
The Philosophical Pivot: A disaster need not result from one reckless person. Ordinary components and reasonable actions can interact in a way that no single participant anticipated.
The Horizon: The theory informs debates about nuclear power, aviation, and other high-risk systems. It does not claim that every accident is unavoidable or excuse failures of design, oversight, and maintenance.
Looser coupling, simpler designs, buffers, and transparent reporting can reduce some risks. Different scholars disagree about whether certain technologies can be made acceptably safe.
The framework shifts attention from blaming the final operator to examining system architecture and institutional decisions.
When an organization cannot tolerate failure, it must ask whether the system can be redesigned to make failures visible and containable.
A failure can emerge from ordinary parts whose interactions no one has the power to see in time.
If no individual can understand the whole system, who is responsible for deciding whether the system is safe enough to operate?