The accumulation of interdependent processes, interfaces, and exceptions that can make the whole system increasingly difficult for any one operator to understand.
Overview
Complexity in critical systems often accumulates through incremental decisions that make sense in isolation: an interface added to reduce friction, an exception created for a special case, or a legacy process preserved to avoid disruption. Over time, the growing number of processes, interfaces, and exceptions can make the whole system increasingly difficult for any one person or team to understand.
Operational complexity is different from technical complexity. A system can be technically simple but difficult to operate, or technically sophisticated but relatively straightforward in practice. What matters for resilience is how well the people responsible for the system can understand its behaviour, particularly under stress.
§01
How complexity accumulates
Adding a new procedure, system, or interface can sometimes be easier than removing or replacing an existing one. The addition may solve an immediate problem, while removal can affect several teams or depend on agreement about what should take its place. Complexity can therefore grow through decisions that are reasonable in their immediate setting.
Mergers, regulatory changes, and technology transitions can each leave another layer behind. Legacy processes coexist with new ones. Parallel systems continue while migration is planned. Exceptions accumulate around standards. Each layer may be defensible in its own terms, while their combined effect becomes difficult for any individual or team to see in full.
§02
Why system growth outpaces comprehension
The people who operate critical systems rotate, retire, and move on. Their successors may inherit more processes, interfaces, and exceptions than earlier teams managed. Documentation can drift out of date, while training tends to focus on standard procedures rather than unusual cases. The system may continue to grow even as shared understanding becomes harder to maintain.
Understanding can also become fragmented. Different teams know different parts of the system, without any one person holding the full operating picture. During routine operation, the gaps between those areas of knowledge may matter little. Under stress, they can make coordination harder and contribute to unexpected behaviour.
§03
What happens under stress
Stress can reveal complexity that was less visible during routine operation. A system that usually runs smoothly may behave differently when several exceptions occur together. Operators who would normally manage one deviation may have to handle several at once without a clear picture of how they interact.
Under these conditions, interactions that were manageable or unnoticed during routine operation can become more important and more difficult to understand. A decision made to protect one part of the system may affect another in an unexpected way when the connection was not documented or modelled. This does not necessarily mean the system was poorly designed; its operation may have changed beyond the shared understanding of the people responsible for it.
§04
What reversibility looks like
Reversibility is the capacity to undo a change and return to a known state. Growing interdependence can make this harder, particularly when teams no longer share the same understanding of the original state. A change that was once simple to reverse may later require coordination across several parts of the system.
Some organisations respond by simplifying parts of the system: retiring interfaces, consolidating exceptions, or rebuilding a shared account of how operations fit together. Such work can be costly and disruptive, and it does not remove every source of complexity. It is one way organisations try to preserve the ability to understand and reverse changes as the system evolves.