A reference for identifying when recovery capacity is quietly disappearing
Systems rarely fail because solutions are unknown. They fail when the conditions enabling effective response deteriorate — gradually, without obvious threshold — until the point at which response would have been sufficient has passed.
This document provides a minimal vocabulary for identifying that deterioration before it becomes irreversible. It is not a theory. It is a working reference derived from observed patterns across independent cases. Use it as a lens, not a rulebook.
Recovery capacity deterioration consistently appears along three observable dimensions:
The variables interact. Loss of redundancy elongates recovery time. Suppressed variation prevents redundancy from being recognised as absent. Reductions below threshold in any variable produce disproportionate effects on the others.
The direction of movement along the three variables is more diagnostically significant than the current level. A system with moderate capacity and a stable trajectory is less at risk than a system with higher current capacity and a declining trajectory.
Ask not: where is the system? Ask: which direction is it moving, and how fast?
Recovery does not restore symmetrically. Deterioration occurs incidentally under normal efficiency pressure. Restoration requires deliberate intervention and extended time. Prevention costs less than recovery.
Observable before deterioration reaches critical levels. Each requires domain-specific operationalisation. Treat as hypotheses, not rules.
Working threshold: three or more signals simultaneously present with declining trajectories indicates active deterioration requiring attention. Four or more indicates probable transition toward irreversibility without deliberate intervention. These numbers are indicative, not validated.
The Irreversible zone is the hardest to detect from inside. The system continues to analyse, decide, and allocate resources — but its actions no longer alter the outcome in question. External observation is required.
Provide structured language for trajectory discussion.
Enable comparison across otherwise unrelated systems.
Generate diagnostic questions before formal analysis begins.
Identify where closer investigation is warranted.
Predict system failure.
Replace domain expertise or operational risk assessment.
Justify decisions directly.
Claim validity in domains not yet examined.
This document describes recurring structural relationships, not universal laws. All propositions are subject to falsification. Similarity across domains is not equivalence. Use accordingly.
Use it before formal analysis, not after. It is most valuable in the stage when a situation is being named — when the question is still "what kind of problem is this?" rather than "what is the solution?"
Apply it to the system you are observing, including the institution doing the observing. Institutions themselves exhibit the same three variables, the same five signals, and the same four zones. A deteriorating institution cannot reliably diagnose a deteriorating system.
Do not use it to rank, certify, or assign blame. Use it to generate questions that would otherwise not be asked: Is recovery time elongating? Where has redundancy been consumed? What signals are no longer reaching the centre?
If the framework is useful, the questions it generates will be more valuable than any score it produces.