0. Registry Classification
| Field | Entry |
|---|---|
| Restoration Arc ID | RA-007 |
| Name | Load Shedding |
| Short Name / Alias | Load Reduction |
| Primary Family | Scaling |
| Secondary Families | Core; Capacity; Cybernetics; Security; Boundary; Economy; Biology-Medicine; AI Governance |
| Treatment | Standalone Arc |
| Status | Canon-Ready |
| Scope | Local / Relational / Institutional / AI / Biological / Economic / Civilizational / Cross-Domain |
| Primary U-Layers | U1 / U2 / U3 β U5 / U6 / U7 validation |
| Primary Operators | Ξ β Ξ β Ξ β Au β ββ β β β Ξ€ |
| Primary Diagnostics | Load, Gain, R, H, K, Ο(t), π(t), π(t), Ο_resp, recurrence |
1. Purpose
1.1 What This Arc Repairs
Load Shedding repairs the overload condition where a system is carrying more demand, coupling, exposure, throughput, complexity, or obligation than its restoration capacity can absorb.
It applies when continued operation at current load would accelerate hidden debt, reduce damping, exhaust slack, degrade auditability, collapse boundaries, or prevent deeper repair.
This arc repairs overload by:
- identifying nonessential load;
- reducing throughput demand;
- reducing active coupling;
- suspending expansion;
- narrowing scope;
- removing avoidable stressors;
- preserving core functions;
- lowering gain;
- preventing repair capacity from being consumed by nonessential operation.
Load Shedding is the canonical restoration arc for making repair possible when capacity is exceeded.
1.2 Core Restoration Function
This arc restores capacity margin by reducing load, scope, coupling, exposure, and amplification until `R_eff` can exceed `Load Γ Gain`, allowing stabilization, slack regeneration, and deeper repair to proceed.
Load Shedding prevents a system from trying to repair while still carrying the overload that caused or sustains the failure.
2. Use Conditions
2.1 When to Apply
Use this arc when:
R_eff < Load Γ Gain;- load exceeds available restoration capacity;
- additional control worsens outcomes;
- a system cannot stabilize because demand remains too high;
- multiple repair arcs are needed but no capacity exists to run them;
- recurring incidents are caused by operating too close to capacity limits;
- emergency response is being normalized because load was never reduced;
- dashboards, procedures, teams, bodies, systems, or institutions are saturated;
- throughput demand prevents auditability, learning, repair, or review;
- the system must preserve core function by suspending nonessential function.
Examples:
- an AI governance system adds more rules while review capacity collapses;
- a security team is too overloaded to distinguish real incidents from noise;
- an institution keeps expanding obligations while unresolved harm accumulates;
- an economic system operates at survival-edge pressure with no buffer;
- a biological or operational system cannot recover because exposure or demand remains constant;
- a relationship, project, or organization attempts deep repair while still carrying all prior load.
2.2 When Not to Apply
Do not apply this arc when:
- active harm requires immediate containment before load planning;
- load reduction would abandon affected nodes;
- load is being shed onto less powerful nodes;
- the system uses load shedding to avoid responsibility;
- the load is essential to prevent greater harm and cannot yet be reduced;
- auditability is too low to distinguish essential from nonessential load;
- boundary repair or controlled decoupling is required before load can be safely reduced;
- the issue is not overload but invalid coupling, permission drift, or hidden authority.
Load Shedding must not become burden dumping.
2.3 Required Preconditions
Before this arc begins, the following must be true:
| Precondition | Requirement |
|---|---|
| Minimum Stabilization | Acute cascade slowed enough to identify load sources |
| Load Visibility | Major load categories can be named |
| Essential Function Map | Core functions can be distinguished from nonessential demand |
| Boundary Protection | Shedding load does not violate affected-node boundaries |
| Auditability | Load movement and burden transfer can be traced |
| Restoration Intent | Load is reduced to enable repair, not avoid repair |
| Review Path | Load shedding decisions can be reviewed and reversed where needed |
If required preconditions fail:
Arc cannot validly begin.The system must return to emergency stabilization, audit surface expansion, boundary reconstitution, or controlled decoupling before load shedding can proceed.
3. Failure / Damage Signature
3.1 Pre-State Across S
| Variable | Expected Pre-State |
|---|---|
| O β Coherence | Declining, brittle, or locally preserved through global overload |
| H β Hidden Debt | Rising through backlog, deferred repair, burden export, and unprocessed demand |
| Ξ΅ β Error / Noise | Increasing, recurring, or amplified by saturation |
| ΞΉ β Inversion Index | Rising when high throughput, control, or endurance is mistaken for coherence |
| Au β Auditability | Often degraded because overload reduces review, traceability, and classification |
| Β΅α΅’ β Agent Integrity | Strained by saturation, role overload, impossible expectations, or forced endurance |
| BΞ£ β Boundary Integrity | Weakening as limits, scope, and refusal paths collapse |
| K β Compatibility / Slack Context | Low; choice-space and buffer depleted |
| R β Restoration Capacity | Below active demand; repair capacity consumed by operation |
| Ξ¦ β Fitness Proxy | Often dominant through throughput, uptime, responsiveness, growth, compliance, or visible activity |
3.2 Primary Failure Links
| Failure Mode | Relationship |
|---|---|
| Capacity Collapse | Primary repair target |
| Zero-Slack Collapse | Primary repair target |
| Load-Gain Saturation | Primary repair target |
| Restoration Starvation | Primary repair target |
| Under-Damped Escalation | Often co-occurs |
| Compression Collapse | Often co-occurs |
| Emergency Normalization | False-restoration risk |
| Forced-Choice Conditions | Often co-occurs |
| Security Overload | Domain expression |
| AI Policy Overload | Domain expression |
| Bureaucratic Overload | Domain expression |
3.3 Origin-Layer Localization
| Layer | Role |
|---|---|
| Failure Origin | Often U1 capacity / throughput, U2 coupling / exposure, U3 control / workload, or U5 timing |
| Visible Symptom Layer | Often U4 narrative stress, U6 field instability, or Ξ¦ productivity / uptime metrics |
| Required Repair Layer | Same or lower than the layer carrying unsustainable load |
| Validation Layer | U5 / U6 / U7 through response timing, field stabilization, recurrence, and backlog behavior |
Canon rule:
Load must be reduced when restoration capacity is below active demand; otherwise repair demand becomes another source of hidden debt.
4. Restoration Objective
4.1 Canonical Objective
Restore capacity margin by reducing nonessential load, coupling, exposure, scope, or throughput until repair capacity can exceed active demand.
Formal objective:
Load β
Gain β
R_eff / (Load Γ Gain) β
π(t) β
π(t) β
Ο_resp β
H growth slowsExpanded objective:
Reduce active demand enough that the system can stabilize, regenerate slack, preserve auditability, and route capacity toward repair instead of being consumed by overload.
4.2 Non-Goals
This arc does not aim to:
- abandon responsibility;
- discard affected nodes;
- reduce load by exporting burden;
- preserve comfort for high-power nodes;
- suppress feedback;
- reduce auditability;
- treat lower activity as full restoration;
- cut essential functions without replacement;
- preserve the same failure geometry under reduced visible demand;
- avoid deeper repair once capacity returns.
5. Operator Sequence
5.1 Minimal Operator Scaffold
Ξ load map β Au burden trace β Ξ scope boundary β Ξ gain reduction β ββ coupling reduction β β capacity routing β Ξ€ stabilization proofUniversal grammar alignment:
Ξ£ + Ξ β Ξ β loadβ / gainβ β β(capacity) β Ξ€ β Temporal ProofLoad Shedding often precedes Slack Regeneration when immediate demand must fall before buffer can rebuild.
5.2 Operator Step Table
| Step | Operator | Function | Variable Impact | Failure Prevented |
|---|---|---|---|---|
| 1 | Ξ | Map load, source, priority, dependency, and burden transfer | Auβ / H mapβ | Blind cuts |
| 2 | Au | Trace where load is generated, carried, hidden, or exported | Au_effβ | Burden dumping |
| 3 | Ξ | Set scope boundary around essential and nonessential demand | BΞ£β / Loadβ | Boundary collapse |
| 4 | Ξ | Reduce gain, urgency, amplification, and signal density | πβ / Ξ΅β | Under-damped escalation |
| 5 | ββ | Reduce invalid or nonessential coupling / exposure | Kβ / H growthβ | Coupling overload |
| 6 | β routing | Route freed capacity toward actual repair | Rβ | Restoration starvation |
| 7 | Ξ€ | Validate stabilization and recurrence behavior after shedding | recurrenceβ / Ο_respβ | False recovery |
5.3 Sequence Notes
This arc is priority-gated and burden-gated.
Load should not be removed blindly. The system must distinguish:
essential load
repair-critical load
optional load
extractive load
duplicative load
legacy load
harm-generating loadThe following steps cannot be skipped:
load mapping
burden trace
scope boundary
gain reduction
capacity routing
temporal reviewIf load is shed by transferring burden to a less powerful node, the arc has inverted.
If reduced activity is treated as repair completion, the arc has collapsed into false recovery.
6. Restoration Phases
Phase 0 β Confirm Overload
Purpose: Establish that active load exceeds restoration capacity.
Actions:
- estimate active load;
- estimate gain stack;
- estimate
R_eff; - identify backlog and deferred repair;
- identify overload symptoms;
- identify whether repair capacity is being consumed by operation.
Validation:
R_eff < Load Γ Gain
backlog or recurrence increasing
repair capacity saturatedPhase 1 β Map Load Sources
Purpose: Identify where demand is generated, amplified, hidden, or exported.
Actions:
- map external demand;
- map internal demand;
- map hidden labor;
- map duplicated effort;
- map unnecessary coupling;
- map emergency-response load;
- map compliance, reporting, procedural, or policy load.
Validation:
load sources visible
hidden load named
burden transfer paths visiblePhase 2 β Classify Load
Purpose: Separate essential demand from nonessential, extractive, duplicative, or harmful demand.
Actions:
- identify core functions;
- identify repair-critical work;
- identify optional work;
- identify harmful load;
- identify inherited or obsolete load;
- identify load that only improves
Ξ¦.
Validation:
essential / nonessential distinction visible
Ξ¦-only load identified
repair-critical load preservedPhase 3 β Establish Scope Boundary
Purpose: Create a legitimate boundary around what the system will and will not carry.
Actions:
- suspend nonessential work;
- narrow active scope;
- pause expansion;
- pause optional coupling;
- protect core function;
- protect affected-node support;
- define explicit load limits.
Validation:
Load β
BΞ£ β
scope no longer expands automaticallyPhase 4 β Reduce Gain and Coupling
Purpose: Decrease amplification and exposure that multiply load.
Actions:
- slow cadence;
- reduce signal density;
- reduce unnecessary interfaces;
- decouple nonessential dependencies;
- reduce notification / review / escalation loops;
- lower urgency where urgency adds debt.
Validation:
Gain β
β nonessential β
π begins improvingPhase 5 β Route Freed Capacity to Repair
Purpose: Prevent freed capacity from being immediately consumed by new demand.
Actions:
- allocate recovered capacity to repair backlog;
- restore auditability;
- restore review bandwidth;
- restore boundary maintenance;
- restore slack;
- repair highest-debt bottlenecks.
Validation:
R_eff β
H growth slows
repair backlog begins movingPhase 6 β Stabilization Review
Purpose: Confirm load shedding improves coherence rather than merely reducing visible activity.
Actions:
- monitor throughput, coherence, and hidden debt;
- check whether affected nodes were abandoned;
- check whether burden was exported;
- check whether backlog decreases;
- check whether recurrence changes;
- review if essential functions remain intact.
Validation:
O stable or improving
H(t+n) β€ H(t)
Ο_resp β
recurrence βPhase 7 β Temporal Proof
Purpose: Verify the reduced load state survives delay and does not snap back.
Actions:
- monitor load re-accumulation;
- monitor recurrence;
- test perturbation response;
- review whether shed load returns under a new name;
- validate that freed capacity remains repair-directed.
Validation:
Load(t+n) β€ admissible capacity
R_eff > Load Γ Gain
π β
recurrence β7. Gates
7.1 Required Gates
| Gate | Requirement | Failure Result |
|---|---|---|
| FI-Gate | Feedback must measure coherence and hidden debt, not activity reduction alone | Arc resets |
| HR-Gate | No identity-bound certainty about what is βnonessentialβ without trace | Load cut blocked |
| MS-Gate | High-status nodes cannot keep load by exporting it downward | Load plan invalid |
| Au-Actuation | Load shedding decisions must be traceable | Actuation forbidden or provisional |
| BΞ£-Gate | Load cannot be shed by violating boundaries or abandoning protected nodes | Arc aborts or reroutes |
| Ξ-Gate | Coupling expansion blocked until load/capacity margin recovers | Expansion blocked |
| β·α΅’ Principle Gates | Non-negotiable invariants hold | β
outcome |
7.2 Gate Failure Rule
If any required gate fails:
β
β Load Shedding cannot validly proceed in that form.The system must either:
- return to stabilization;
- expand auditability;
- protect affected nodes;
- reclassify essential load;
- reduce gain instead of abandoning load;
- decouple invalid coupling;
- select slack regeneration or boundary reconstitution first.
8. Diagnostics
8.1 Required Diagnostic Trends
| Diagnostic | Expected Trend | Meaning |
|---|---|---|
| Load | β | Active demand decreases |
| Gain | β | Amplification pressure decreases |
| R | β / less saturated | Restoration capacity becomes available |
| R_eff / Load Γ Gain | β | Capacity margin improves |
| H | Growth slows, then β | Hidden debt accumulation slows |
| K / Ο(t) | β | Choice-space and slack improve |
| π(t) | β | Bandwidth margin improves |
| π(t) | β | Disturbance settles faster |
| Ο_resp | β | Response latency improves |
| recurrence | β | Overload pattern weakens |
| Ξ¦/O divergence | β | Lower activity aligns with actual coherence, not optics |
8.2 Arc-Specific Diagnostic Thresholds
Suggested thresholds:
Load β
Gain β
R_eff > Load Γ Gain
π(t) β
π(t) β
Ο_resp β
H(t+n) β€ H(t)
recurrence β across U7
freed capacity routes to repairLoad Shedding is not complete if:
load is exported instead of reduced
affected nodes are abandoned
freed capacity is consumed by new demand
H continues accelerating
R_eff remains below Load Γ Gain
load returns under a new label
Ξ¦ improves while O remains brittle9. Anti-Patterns / False Restorations
9.1 Common False Versions
This arc is being simulated, not executed, if:
- load is cut from visible dashboards but not from actual work;
- load is transferred to less powerful nodes;
- affected nodes lose support under the name of βreductionβ;
- reduced service is treated as restored coherence;
- the system sheds accountability but not demand;
- high-status or high-visibility work is preserved while repair-critical work is cut;
- urgent signals are suppressed rather than processed;
- workload is renamed instead of reduced;
- nonessential coupling remains active;
- freed capacity is immediately used for expansion;
- backlog is hidden instead of cleared.
9.2 Named Anti-Pattern Links
| Anti-Pattern | Why It Fails |
|---|---|
| Burden Dumping | Exports load rather than reducing system demand |
| Cut-The-Support Layer | Removes repair-critical functions while preserving optics |
| Dashboard Load Shedding | Reduces visible metrics but not real load |
| Accountability Shedding | Sheds responsibility rather than demand |
| False Recovery | Mistakes reduced activity for restored capacity |
| Compression Preservation | Keeps the same load geometry under lower visibility |
| Growth Snap-Back | Freed capacity is consumed by expansion before repair |
10. Completion Criteria
10.1 Post-State Signature
| Variable | Required Post-State |
|---|---|
| O | Stable or improving under reduced load |
| H | Growth slowed or reduced |
| Ξ΅ | Bounded and less amplified |
| ΞΉ | Reduced where activity was mistaken for coherence |
| Au | Load and burden paths traceable |
| Β΅α΅’ | Less strained by impossible demand |
| BΞ£ | Scope boundaries clearer and enforceable |
| K | Choice-space / buffer increasing |
| R | Available for repair, not fully consumed by operation |
| Ξ¦ | Subordinate to O; reduced activity cannot certify restoration alone |
10.2 Temporal Proof
Load Shedding cannot be declared complete until reduced load remains stable and repair capacity remains available over time.
Template:
Completion requires Load(t+n) within admissible capacity,
R_eff > Load Γ Gain,
H(t+n) β€ H(t),
and recurrence decreasing across U7.Minimum temporal proof:
- load does not snap back immediately;
- nonessential demand remains paused or redesigned;
- freed capacity routes to repair;
- affected nodes are not abandoned;
- hidden work remains visible;
- recurrence decreases under the reduced-load state.
10.3 Completion Statement
Canonical format:
This arc is complete only when load and gain have been reduced below restoration capacity, freed capacity is routed toward repair, hidden debt is no longer accelerating, and the reduced-load state remains stable over time without exporting burden elsewhere.
11. Cross-Links
11.1 Related Restoration Arcs
| Arc | Relationship |
|---|---|
RA-001 β Emergency Harm Stabilization | Precursor when overload is acute and cascading |
RA-006 β Slack Regeneration | Primary follow-on or companion |
RA-010 β Controlled Decoupling | Companion when load comes from invalid coupling |
RA-020 β Safe Decoupling | Follow-on when extraction or dependency sustains load |
RA-022 β Compression Relief | Companion when compression pressure must be reduced |
RA-026 β Stability / Damping Restoration | Companion when overload creates poor ring-down |
RA-050 β Authority Registry Clarification | Companion when load comes from unclear authority |
RA-066 β Circulation Repair | Follow-on when load comes from delivery / return / clearance failure |
11.2 Related Failure Modes
| Failure Mode | Relationship |
|---|---|
| Capacity Collapse | Repairs |
| Zero-Slack Collapse | Repairs / prevents |
| Load-Gain Saturation | Repairs |
| Restoration Starvation | Repairs / prevents |
| Under-Damped Escalation | Often co-occurs |
| Compression Collapse | Often co-occurs |
| Emergency Normalization | False-restoration risk |
| Forced-Choice Conditions | Often co-occurs |
| Security Overload | Domain expression |
| AI Policy Overload | Domain expression |
| Bureaucratic Overload | Domain expression |
11.3 Related Diagnostics
Load, Gain, R, H, K, Ο(t), π(t), π(t), Ο_resp, recurrence, Ξ¦/O divergence11.4 Related Laws / Invariants
INV β Repair capacity must exceed restoration demand.
INV β Slack is required for repair.
INV β Coherence cannot be inferred from activity level.
LAW β Load Γ Gain must remain below effective restoration capacity.
LAW β Compression without slack produces hidden debt.
LAW β Restoration demand can become harm when capacity is insufficient.
LAW β Ξ¦ improvement is not O restoration.
LAW β Burden export preserves hidden debt.12. Domain Notes
12.1 AI / Cognitive Infrastructure
Check:
- rule-stack load;
- classifier volume;
- review queue saturation;
- tool-call exposure;
- memory retrieval burden;
- escalation load;
- moderation and appeal bandwidth;
- latency and rollback capacity;
- whether added safety controls are consuming the review capacity needed to keep them coherent.
AI load shedding may require reducing policy complexity, tool exposure, automation scope, memory activation, or review throughput until auditability and restoration capacity recover.
12.2 Justice / Governance / Legitimacy
Check:
- intake load;
- testimony burden;
- reporting burden;
- review backlog;
- appeal backlog;
- public legitimacy pressure;
- staff / investigator overload;
- whether βefficiencyβ cuts remove support instead of nonessential burden.
JGL load shedding must preserve affected-node support and repair-critical work while reducing nonessential, duplicative, or optics-driven demand.
12.3 Biology / Medicine
Conceptual systems mapping only.
Load Shedding in biological systems means reducing exposure, demand, signal density, timing pressure, or recovery burden so that repair capacity and perturbation tolerance can recover.
Not diagnosis.
Not treatment.
Not medical advice.
12.4 Economy
Check:
- debt load;
- survival-edge demand;
- hidden labor;
- contractual obligations;
- liquidity pressure;
- extraction pathways;
- throughput pressure;
- whether reduced costs are actually burden transfers.
Economic load shedding must reduce forced demand without cutting the support layers needed for repair and circulation.
12.5 CMS / Meaning / Archetypes
Check:
- meaning overload;
- forced integration;
- symbolic demand;
- identity performance load;
- pressure to forgive, explain, understand, or embody before capacity exists;
- whether complexity is being mistaken for depth.
Meaning systems require load shedding when symbolic, emotional, identity, or interpretive demand exceeds integration capacity.
13. Machine-Readable Metadata
id: "RA-007"
title: "Load Shedding"
aliases:
- "Load Reduction"
family_primary: "Scaling"
families_secondary:
- "Core"
- "Capacity"
- "Cybernetics"
- "Security"
- "Boundary"
- "Economy"
- "Biology-Medicine"
- "AI Governance"
treatment: "Standalone Arc"
status: "Canon-Ready"
scope:
- "Local"
- "Relational"
- "Institutional"
- "AI"
- "Biological"
- "Economic"
- "Civilizational"
- "Cross-Domain"
u_layers:
failure_origin:
- "often U1 capacity / throughput"
- "often U2 coupling / exposure"
- "often U3 control / workload"
- "often U5 timing"
symptom_visible:
- "U4 narrative stress"
- "U6 field instability"
- "Ξ¦ productivity / uptime metrics"
repair_required:
- "same or lower than layer carrying unsustainable load"
validation:
- "U5"
- "U6"
- "U7"
operators:
scaffold: "Ξ load map β Au burden trace β Ξ scope boundary β Ξ gain reduction β ββ coupling reduction β β capacity routing β Ξ€ stabilization proof"
sequence:
- "Ξ"
- "Au"
- "Ξ "
- "Ξ"
- "ββ"
- "β"
- "Ξ€"
state_variables:
primary:
- "Load"
- "Gain"
- "R"
- "H"
secondary:
- "O"
- "K"
- "BΞ£"
- "Ξ¦"
diagnostics:
- "Ο(t)"
- "π(t)"
- "π(t)"
- "Ο_resp"
- "recurrence"
- "Ξ¦/O divergence"
gates_required:
- "FI-Gate"
- "HR-Gate"
- "MS-Gate"
- "Au-Actuation"
- "BΞ£-Gate"
- "Ξ-Gate"
- "β·α΅’"
linked_failure_modes:
- "Capacity Collapse"
- "Zero-Slack Collapse"
- "Load-Gain Saturation"
- "Restoration Starvation"
- "Under-Damped Escalation"
- "Compression Collapse"
- "Emergency Normalization"
- "Forced-Choice Conditions"
- "Security Overload"
- "AI Policy Overload"
- "Bureaucratic Overload"
linked_restoration_arcs:
- "RA-001"
- "RA-006"
- "RA-010"
- "RA-020"
- "RA-022"
- "RA-026"
- "RA-050"
- "RA-066"
anti_patterns:
- "Burden Dumping"
- "Cut-The-Support Layer"
- "Dashboard Load Shedding"
- "Accountability Shedding"
- "False Recovery"
- "Compression Preservation"
- "Growth Snap-Back"
completion_tests:
- "Load decreases"
- "Gain decreases"
- "R_eff > Load Γ Gain"
- "π(t) increases"
- "π(t) increases"
- "Ο_resp decreases"
- "H(t+n) β€ H(t)"
- "recurrence decreases across U7"
- "freed capacity routes to repair"
summary: "Load Shedding reduces nonessential demand, exposure, coupling, scope, or throughput when active load exceeds restoration capacity, allowing stabilization and repair capacity to recover."Final Calibration Rule
Load Shedding answers six questions:
What hidden debt is being generated by overload?
What load, scope, coupling, or exposure must be reduced?
What auditability proves demand was reduced rather than exported?
What coupling, expansion, or throughput must remain blocked until capacity returns?
What trajectory becomes viable once load falls below restoration capacity?
How is reduced load proven stable over time without becoming abandonment or burden dumping?