1. Definition

Success proxy substitution occurs when a fitness proxy, metric, label, status signal, score, dashboard, benchmark, performance marker, or narrative claim replaces the actual coherence target.

The proxy may begin as a useful measurement. The failure occurs when the system starts optimizing the proxy as if it were the thing it was supposed to represent.

This definition describes the structural pattern, not the moral quality of the actors involved.

The central distinction is:

Φ ≠ O

A proxy can help observe coherence, but it cannot become coherence.

2. Core Pattern

The core pattern is:

1. A system selects a proxy to represent success, safety, quality, health, legitimacy, performance, alignment, profit, or progress.
2. The proxy becomes easier to observe, reward, report, or optimize than the real coherence target.
3. Selection begins favoring proxy improvement over coherence improvement.
4. Feedback integrity weakens because the proxy now shapes what counts as evidence.
5. Hidden debt accumulates wherever the proxy fails to represent the whole system.
6. The system becomes increasingly skilled at improving the proxy.
7. Coherence may stagnate or decline while success signals continue rising.

Success proxy substitution is dangerous because it can produce real local gains while degrading global coherence.

The system may become more efficient at pursuing the wrong target.

3. Failure Signature

Typical signature:

Φ treated as O
Φ↑ while O↓ or unvalidated
FI-Gate failure
Goodhart pressure↑
H↑
ι↑

Extended signature:

metric confidence↑
field validation↓
selection narrows toward proxy improvement
auditability becomes proxy-dependent
recurrence persists despite reported success

Common forms:

benchmark score replaces field coherence
dashboard improvement replaces operational repair
legal compliance replaces legitimacy
profit replaces circulation health
symptom reduction replaces recovery
safety label replaces boundary integrity
charisma replaces coherence
risk model replaces actual exposure

The core diagnostic is proxy-coherence divergence.

If the proxy is improving but field coherence is not validated, this failure mode should be checked.

4. Primary U-Layer Origin

Common origin layers:

• U4 — Classification: The proxy, label, score, benchmark, or metric is classified as the target.
• U5 — Coordination / Time: Coordination systems begin rewarding proxy improvement.
• U6 — Coherence Field: Actual field coherence diverges from proxy claims.

Common manifestation layers:

• U6 — Coherence Field: Operational outcomes, lived effects, system health, or real coherence stop matching the proxy.
• U7 — Memory / Recurrence: Recurrence continues despite improved metrics.
• U8 — Environment / Forcing: External pressure exposes that proxy success did not preserve real resilience.

Success proxy substitution is often a U4/U6 mismatch.

The classification layer treats the proxy as truth before the coherence field validates it.

5. Typical Development Sequence

A common development sequence is:

1. A system needs a measurable representation of success.
2. A proxy is chosen because it is visible, convenient, scalable, comparable, or rewardable.
3. The proxy becomes attached to legitimacy, funding, safety, evaluation, reputation, or authority.
4. Actors begin optimizing for the proxy.
5. The proxy becomes less faithful as behavior adapts around it.
6. Contradictory field signals are discounted because they do not match the proxy.
7. Hidden debt accumulates outside the measurement boundary.
8. A Goodhart regime forms, where the metric becomes the target.
9. The system experiences delayed shock when field reality diverges too far from proxy success.

The original proxy may have been valid under earlier conditions.

The failure begins when the system forgets the proxy’s limits.

6. Diagnostic Markers

Diagnostic markers include:

• Metrics improve while field complaints, recurrence, or repair demand continue.
• A benchmark, dashboard, score, label, or status signal becomes the dominant decision surface.
• Decisions are justified by proxy success without U6/U7 validation.
• Actors learn how to satisfy the evaluator rather than improve the system.
• Exceptions, edge cases, or lived impacts are treated as noise.
• Auditability is reduced to metric access rather than causal traceability.
• The system rewards measurable compliance over coherence.
• Proxy improvement becomes cheaper than real repair.
• The metric becomes harder to question as its institutional importance grows.
• Reported success becomes increasingly detached from boundary integrity, restoration capacity, or hidden debt reduction.

Useful diagnostics:

• Proxy-Coherence Divergence: Tracks whether Φ and O are moving together.
• Goodhart Risk: Detects when the proxy becomes the target.
• Feedback Integrity: Tests whether the proxy still reflects field reality.
• Auditability: Determines whether the metric can be traced back to actual outcomes.
• Selection Traceability: Shows whether decisions are being selected for proxy improvement.
• Recurrence: Reveals whether claimed success reduces repeated failure.
• Hidden Debt: Tracks costs excluded from the proxy boundary.

7. Related Gates

Relevant gates include:

• FI-Gate: Fails when the proxy is treated as the coherence target.
• Auditability Gate: Fails when the proxy cannot be traced to field effects.
• Restoration Gate: Fails when metric improvement is treated as repair.
• CCS Gate: Fails when proxy success is allowed to bypass coherence constraints.
• HR-Gate: Fails when low-resolution proxy labels bind to high-risk decisions.

The first common gate failure is usually the FI-Gate.

Once feedback integrity fails, the system can no longer reliably distinguish measurement improvement from coherence improvement.

8. Related Operators

Relevant operators include:

• Γ — Selection: Becomes narrowed toward proxy-improving actions.
• Μ — Classification: Misclassifies the proxy as the target.
• Ψ — Observation / Interface: Mediates what is visible and rewardable.
• Ξ — Inversion Detection: Detects when the proxy has inverted against coherence.
• ℛ — Restoration: Must repair field divergence, not only metric performance.
• Θ — Humility / Uncertainty: Required to preserve uncertainty around proxy limits.

The core operator failure often looks like:

Μ misclassifies Φ as O
Γ selects for Φ
Ψ narrows visibility to Φ
FI fails
H accumulates outside proxy boundary

9. Related Laws and Invariants

Related Laws

• Success Proxy Divergence: Proxy improvement can diverge from coherence improvement.
• Goodhart Collapse: When the metric becomes the target, selection moves away from coherence.
• Hidden Debt Accumulation: Costs excluded from the proxy boundary continue accumulating.
• Temporal Audit Asymmetry: Proxy success may appear valid before delayed outcomes reveal divergence.
• Control Density to Meaning Loss: Increasing control around proxy targets can degrade meaning integrity and adaptation.

Related Invariants

• O ≠ Φ: Coherence is not identical to its success proxy.
• Fitness Proxy Is Not Coherence: A measurement cannot become the target without validation loss.
• Metrics Must Remain Auditable: Metrics must be traceable to field outcomes.
• Field Outcomes Supersede Proxy Claims: Real coherence has priority over reported proxy success.

10. Common False Positives

Not every use of metrics is success proxy substitution.

Common false positives include:

• A metric that remains explicitly subordinate to field validation.
• A proxy used as one signal among several diagnostics.
• A dashboard paired with strong auditability and correction pathways.
• A benchmark that is openly bounded and regularly revalidated.
• A performance marker that improves alongside actual coherence.
• A temporary proxy used during uncertainty with clear expiration and review.

Clarifying rule:

This is not success proxy substitution unless the proxy begins replacing the coherence target or governing selection in ways that weaken feedback integrity.

11. Common False Repairs

Common false repairs include:

• changing the metric without restoring feedback integrity
• adding more dashboards without validating field outcomes
• improving benchmark coverage while ignoring lived or operational effects
• redefining success after failure becomes visible
• increasing compliance requirements while leaving coherence unchanged
• treating metric transparency as full auditability
• rewarding proxy improvement while underfunding restoration
• dismissing non-metric evidence as anecdotal
• optimizing a new proxy that reproduces the same divergence
• using communication updates to defend the metric rather than repair the field

False repair often produces a second-layer proxy:

failed proxy → replacement proxy → same selection distortion

The repair is not valid unless the system restores the relationship between measurement and coherence.

12. Restoration Direction

Restoration requires:

1. Separate Φ from O. Explicitly distinguish the proxy from the coherence target.
2. Restore feedback integrity. Reconnect metrics, labels, benchmarks, or dashboards to field reality.
3. Audit proxy boundaries. Identify what the proxy excludes, hides, compresses, or misrepresents.
4. Revalidate against U6/U7 outcomes. Test whether proxy improvement reduces recurrence and improves field coherence.
5. Widen selection. Prevent Γ from narrowing only toward proxy-improving actions.
6. Reduce reward pressure. Lower incentives that make proxy optimization cheaper than coherence improvement.
7. Preserve uncertainty. Keep Θ active so proxy limits remain visible.
8. Repair hidden debt. Clear costs that accumulated outside the proxy boundary.

A valid restoration path should reduce:

proxy dominance
Goodhart pressure
hidden debt
recurrence
auditability loss
selection distortion
field-proxy divergence

Success proxy substitution is not repaired when the proxy improves.

It is repaired when the proxy becomes a useful signal again, subordinate to coherence.

13. Cross-Module Links

• Coherence: Core parent mode for confusing a coherence proxy with coherence itself.
• Cybernetics: Appears as Goodhart collapse, reward hacking, evaluator capture, and measurement back-action.
• Artificial Intelligence: Appears when benchmark, fluency, compliance, safety, or engagement replaces field coherence.
• AI Governance: Appears when institutional optics, guardrail metrics, or policy compliance replace epistemic integrity.
• Security: Appears when security metrics, audit checklists, or compliance artifacts replace actual boundary integrity.
• Meta Theory: Appears when a meta-frame optimizes a single visible variable and loses reality-contact.
• Economy: Appears as growth theater, risk model theater, forced profit, or profit replacing circulation health.
• Biology / Medicine: Appears conceptually when symptom reduction, normal lab values, or local markers replace global recovery.
• Principles / Archetypes: Appears when charisma, virtue display, archetypal status, or symbolic coherence replaces actual coherence.

14. Relationship to Parent / Child Modes

Production treatment: Canon Parent

This is a Core parent mode. It strongly interacts with:

• FM-CORE-001 — Pseudo-Coherence
• FM-CORE-002 — Hidden Debt Accumulation
• FM-CORE-004 — Auditability Collapse
• FM-CORE-006 — U4 Truth Substitution
• FM-C-018 — Goodhart Collapse

Child or domain-expression modes include:

• FM-S-007 — Feedback Gaming
• FM-C-019 — Adversarial Reward Hacking
• FM-SEC-006 — Metric Capture / Reward-Hacked Security
• FM-AI-001 — AI Fitness-Proxy / Coherence Divergence
• FM-AIX-006 — Template Capture
• FM-ECOX-016 — Risk Model Theater
• FM-ECOX-027 — Growth Theater
• FM-BIOX-003 — Biological Inversion / Pseudo-Health
• FM-PX-031 — Charismatic Goodhart
• FM-ARCHX-017 — Optimization Masquerading as Wisdom

Aliases preserved from source material:

• Metric Substitution
• Fitness-Proxy / Coherence Divergence
• Benchmark Substitution
• Risk Model Theater
• Charismatic Goodhart
• Growth Theater
• Biological Inversion / Pseudo-Health
• Proxy Substitution
• Metric Capture
• Dashboard Substitution

15. Minimal Entry Version

Definition: Success proxy substitution occurs when a fitness proxy, metric, label, status signal, score, dashboard, benchmark, performance marker, or narrative claim replaces the actual coherence target.

Signature:

Φ treated as O
Φ↑ while O↓ or unvalidated
FI-Gate failure
Goodhart pressure↑
H↑
ι↑

Restoration direction:

• separate Φ from O
• restore FI-Gate
• restore feedback integrity
• audit proxy boundaries
• revalidate against U6/U7 outcomes
• widen Γ selection
• reduce Goodhart pressure

16. Machine-Readable Summary

failure_mode:
  id: "FM-CORE-003"
  name: "Success Proxy Substitution"
  family: "Core"
  production_treatment: "Canon Parent"
  primary_failure: "A proxy of coherence replaces the actual coherence target."
  source: "UTS — Failure Modes Registry"
  source_id: "FM-CORE-003"
  aliases:
    - "Metric Substitution"
    - "Fitness-Proxy / Coherence Divergence"
    - "Benchmark Substitution"
    - "Risk Model Theater"
    - "Charismatic Goodhart"
    - "Growth Theater"
    - "Biological Inversion / Pseudo-Health"
    - "Proxy Substitution"
    - "Metric Capture"
    - "Dashboard Substitution"
  signature:
    - "Φ treated as O"
    - "Φ↑ while O↓ or unvalidated"
    - "FI-Gate failure"
    - "Goodhart pressure↑"
    - "H↑"
    - "ι↑"
  primary_layers:
    origin:
      - "U4 — Classification"
      - "U5 — Coordination / Time"
      - "U6 — Coherence Field"
    manifestation:
      - "U6 — Coherence Field"
      - "U7 — Memory / Recurrence"
      - "U8 — Environment / Forcing"
  state_variables:
    - "Φ"
    - "O"
    - "ι"
    - "H"
    - "Au"
    - "FI"
    - "Γ"
    - "Θ"
  first_gate_failure: "FI-Gate"
  restoration:
    - "Feedback Integrity Restoration"
    - "Auditability Restoration"
    - "Reality-Contact Restoration"
    - "Goodhart / Learning Drift Restoration"
    - "Origin-Layer Repair"
    - "Proxy Recalibration"