Crisis Cognitive Degradation and Defense: The SIDA-F Framework v1.4
14 cases bidirectionally anchored — 5 failure cases revealing "where cognition breaks," 9 success cases mapping "what mechanisms intercept degradation"
In One Sentence
Under extreme conditions, the cognitive functions of organizations and individuals do not get stronger — they get weaker. SIDA-F tells you where degradation occurs, how it cascades and amplifies, and what mechanisms can intercept it.
How Degradation Occurs: The SIDA-F Five-Layer Loop
Crisis cognition follows a five-layer loop: Sense → Interpret → Decide → Act → Feedback. Degradation at any layer cascades and amplifies degradation at the next layer.
Sense-layer degradation: Signals are drowned out by noise (Fukushima — 200+ information streams with no prioritization), information becomes asynchronous across hierarchical levels (Katrina — information vacuum after communications infrastructure was physically destroyed), and instruments systematically deceive when operating beyond their design range (Fukushima — radiation meters saturated and displayed "normal").
Interpret-layer degradation: Premature judgment lock-in (Challenger — "O-ring erosion is acceptable"), burden of proof is inverted (Deepwater Horizon — "cannot prove it is unsafe" replaced "needs to prove it is safe"), and the entire interpretive framework becomes unmodifiable (Katrina — the federalist response framework fails to switch under catastrophic conditions).
Decide-layer degradation: Authority suppresses dissent (all three nuclear/aerospace cases — Fukushima's TEPCO headquarters → Prime Minister's Office filtering chain), sunk costs lock in decisions (Challenger — "we've already delayed so many times, we can't delay again because of one Thiokol engineer"), and decisions disappear within bureaucratic processes (Katrina — Mission Assignment approvals required "innumerable signatures").
Act-layer degradation: Protocols are rigidly executed in situations they do not cover (all five failure cases — fire brigades entering nuclear radiation zones following routine fire response procedures, the NRP framework not being broken under catastrophic conditions).
Feedback-layer degradation: Feedback arrives too late (Fukushima — the true state of the reactors was not confirmed for weeks), feedback is forever unattainable (all five failure cases — "what if we had chosen B" can never be verified), and successes are erroneously coded as evidence of "system safety" (Challenger — 7 instances of O-ring erosion without explosion = 7 instances of "safety" evidence → cumulative explosion risk).
How Defense Works: Four Principles Revealed by Success Cases
Principle 1: Cognitive Offloading — Individual → Structure. 8/9 success cases do not defend by "training stronger individuals" — but by transferring cognitive burden from individual working memory to external structures. The colored jerseys on an aircraft carrier deck let you know who is doing what without language. The trauma team's differential diagnosis exhaustion ("at least 3 differential diagnoses") means you don't need to deliberately remember not to miss something — the rule remembers for you.
Principle 2: Format Determines Executability. Whether dissent can penetrate authority gradients depends on the format in which it is expressed. Technical parameter format ("sensor reading X, deviation Z%, exceeds threshold") → received as protocol invocation. Personal opinion format ("I have a bad feeling about this") → received as authority challenge. The former does not require courage. The latter does — and in high power-distance organizations, courage is often insufficient.
Principle 3: Appropriate Preloading Outperforms Forced Maintenance. Under certain conditions, the optimal strategy is not to protect cognitive function — but to ensure cognitive function does not run in the crisis at all. Japan's Sanriku region "tsunami tendenko" — "earthquake → run. Don't look back. Don't wait for family." — this decision was not made when the tsunami warning sounded. It is a sentence a 6-year-old hears repeatedly at school. In the 2011 Great East Japan Earthquake, 2,900 primary and secondary school students in Kamaishi City — almost all survived.
Principle 4: Cluster Determines Defense Philosophy. Seconds-to-minutes-scale operations (trauma teams, flight decks) → defense philosophy = preloading and automation. Hour-to-day-scale organizations (command centers, multi-agency responses) → defense philosophy = structure and institutions. The same principle requires completely different implementation forms across these two temporal scales.
Honest Boundaries of This Framework
Only human organizational crisis cognition is examined — AI is not covered. Not because AI does not exist, but because JEPA Theorem 5.2 has mathematically proven: under non-Gaussian latent variables (heavy tails, truncation, phase transitions — the mathematical signatures of crisis scenarios), AI cannot learn identifiable world models. The judgment responsibility that cannot be outsourced in crisis cognition — ultimately rests with humans.
Among the 14 cases, only Cluster 2 (hour-to-day-scale organizations) contains failure cases. The degradation patterns of Cluster 1 (second-grade operations) and Cluster 3 (cultural-embodied evolution) are based on theoretical inference, without independent validation by failure cases.
This is a diagnostic framework — not a predictive framework. It can tell you "which cognitive link is most vulnerable in this organization" — but it cannot tell you "what the next crisis will be."
🔄 Belief Update Log
v1.0 (prior): Crisis cognition is a unified meta-domain; 8 fields share underlying structure ↓ 2026-06-03 New evidence: HRO and disaster science share degradation structure (ICS is a direct product of HRO), but anthropology and traditional communities may use fundamentally different cognitive mechanisms ↓ v1.1 (posterior): 3 cognitive clusters, parametric differences rather than architectural rupture ↓ 2026-06-03 New evidence: 6 degradation patterns had zero occurrence across 5 cases — false patterns manufactured by framework symmetry ↓ v1.2 (posterior): 6 false patterns demoted to L1 appendix ↓ 2026-06-04 New evidence: 9 success cases reveal cross-cluster commonalities in defense mechanisms — cognitive offloading independently appeared in 8/9 cases ↓ v1.3–1.4 (posterior): Bidirectional framework — 5 degradation-side cases + 9 defense-side cases → 14-case bidirectional anchoring