CONCEPT
Interactive Complexity
Perrow's term for systems whose components interact through pathways designers did not anticipate — nonlinear, networked, often invisible — producing failure combinations that exceed any safety analysis that could have been conducted.
Interactive complexity is the first of Perrow's two structural conditions for normal accidents. In a linear system, component A feeds B feeds C, and failure propagates along a predictable chain. In an interactively complex system, components interact radially through pathways that cross physical barriers, software abstractions, and organizational walls. The interactions are not sequential but networked; failures do not march but radiate, producing combinations no one tested because no one imagined them. Interactive complexity is not a failure of imagination but a mathematical ceiling: the space of possible interactions grows combinatorially while the space of testable interactions grows linearly, guaranteeing that unexamined pathways exceed examined ones by orders of magnitude.
In The You On AI Field Guide
The concept distinguishes Perrow's framework from earlier risk analyses that treated accidents as failures of specific components. In linear systems, you can trace backward from the failure to the component that caused it and redesign that component. In interactively complex systems, the 'cause' is