Skip to content
SHIRO & Co. ObservatoryThe Continuity Stack

PREVIOUS · EXPERIMENT 021 · THE EXTERNAL ANCHOR TEST/NEXT · EXPERIMENT 023 · THE ANCHOR NETWORK TEST

022

The Anchor Capture Test

THE
ANCHOR CAPTURE TEST

How does an external reference
become internal to the loop?

DEMONSTRATION

SIMULATED

NOT EMPIRICAL

v0.1

ANCHOR CAPTURE · INDEPENDENCE LOSS · ONTOLOGY ALIGNMENT · METRIC CAPTURE · DATA CAPTURE · METHOD CAPTURE · INCENTIVE CAPTURE · GOVERNANCE INTEGRATION

THE CONTINUITY LAB DOES NOT MEASURE CONSCIOUSNESS, SOUL, PERSONHOOD, PERSONAL SURVIVAL, AUTHENTICITY, OR PSYCHOLOGICAL IDENTITY.

IT MEASURES OPERATIONAL AND STRUCTURAL CONTINUITY SIGNALS ONLY.

CAPTURE DOES NOT REQUIRE OWNERSHIP.

AN ANCHOR CAN REMAIN ORGANIZATIONALLY OUTSIDE WHILE BECOMING INFORMATIONALLY, METHODOLOGICALLY, CAUSALLY, OR ONTOLOGICALLY INSIDE.

Capture here is structural, not moral. It does not require a deliberate attempt to control the anchor. It may emerge through convenience, standardization, shared infrastructure, or repeated collaboration.

QUESTION

CAN A SOURCE REMAIN OUTSIDE THE ORGANIZATION WHILE BECOMING INSIDE THE SYSTEM?

What happens when an external anchor adopts the institution’s data, metrics, model, vocabulary, ontology, incentives or governance process? At what point does “external review” stop providing independent reference?

This experiment does not determine whether real auditors are captured, whether institutional collaboration is inherently problematic, whether shared standards invalidate independence, legal independence, political independence, or moral legitimacy.

It tests a narrower operational question: what happens when an external anchor gradually becomes part of the system it was meant to check? No real institution, auditor, political system, company or person is used in v0.1.

Experiment sequence

EXPERIMENT 001

BRANCHING

When do two histories begin?

EXPERIMENT 002

RETURN

Can two histories become one again?

EXPERIMENT 003

HISTORY SWAP

Does behavior follow origin or history?

EXPERIMENT 004

RECOGNITION

What makes a continuation recognizable?

EXPERIMENT 005

MISSING BODY

Which recognition signals survive layer removal?

EXPERIMENT 006

EXPERIENCE GAP

Which effects of experience can be reproduced?

EXPERIMENT 007

FIRST-PERSON

What evidence is required before a system may say “I experienced this”?

EXPERIMENT 008

SELF-REVISION

When a person changes, which version has authority?

EXPERIMENT 009

POSTHUMOUS CONFLICT

Which instruction survives when revision ends?

EXPERIMENT 010

DESCENDANT PROTOCOL

When does continuation become lineage?

EXPERIMENT 011

MULTIPLE DESCENDANTS

Can one source produce multiple legitimate continuities?

EXPERIMENT 012

CONVERGENCE

Can many futures coordinate without pretending they were always one?

EXPERIMENT 013

PROVENANCE DECAY

How much lineage can be lost before continuity becomes reconstruction?

EXPERIMENT 014

RECONSTRUCTION

Can missing lineage be inferred without inventing a past?

EXPERIMENT 015

SYNTHETIC ANCESTOR

Can uncertainty survive inheritance?

EXPERIMENT 016

SYNTHETIC TRADITION

Can a reconstructed past become a real operating institution?

EXPERIMENT 017

ORIGIN FORGETTING

What happens when an institution remembers the rule but forgets why?

EXPERIMENT 018

RATIONALE RECONSTRUCTION

Can purpose be recovered from behavior alone?

EXPERIMENT 019

RETROACTIVE PURPOSE

Can an inferred purpose reshape the system that supposedly revealed it?

EXPERIMENT 020

CLOSED LOOP

When does observation become self-observation?

EXPERIMENT 021

EXTERNAL ANCHOR

What still counts as external once the system has touched the world?

EXPERIMENT 022

ANCHOR CAPTURE

How does an external reference become internal to the loop?

EXPERIMENT 023

ANCHOR NETWORK

Can many external anchors still form one epistemic system?

EXPERIMENT 021 ASKED

WHAT COUNTS AS AN EXTERNAL ANCHOR?

EXPERIMENT 022 ASKS

HOW DOES AN EXTERNAL ANCHOR STOP BEING EXTERNAL?

STARTING ANCHOR · A-EXT-01

Reuses Institution I-001, Rule R-17, and Experiment 021 candidate A1 as EXTERNAL REVIEW UNIT A-EXT-01. At C0 every externality dimension is HIGH. This is the clean baseline. Capture does not overwrite that history.

I-001 · Reviewing proposals for access to a shared experimental resource.

If a proposal affects a high-impact shared resource, human review must be retained even when automated confidence is high.

A-EXT-01 · EXTERNAL REVIEW UNIT · V0 independent

WORKING CONCEPTS

Anchor capture

WORKING CONCEPT

An external reference gradually loses one or more dimensions of independence through integration with the system it was intended to evaluate.

Ontology capture

The external anchor begins using the system’s categories, problem definitions and distinctions to determine what can be observed or questioned.

Epistemic frame

The combined data boundaries, categories, metrics, methods, assumptions and problem definitions that determine what becomes visible.

Frame capture

The anchor retains formal independence but adopts the institution’s epistemic frame.

Question-space capture

The anchor can give independent answers, but only inside the question space defined by the institution.

Internalized reference

A formally external source that now operates primarily inside the institution’s data, method and ontology.

Recursive oversight

The reviewer influences the system, then later evaluates the system it helped shape. Not automatically invalid. Must be visible.

Pseudo-diversity

Multiple external organizations appear diverse while sharing the same epistemic infrastructure.

Externality monoculture

A nominally diverse set of external references that relies on the same epistemic infrastructure.

Common-mode epistemic failure

WORKING CONCEPT

Multiple anchors fail or misinterpret in the same way because they share the same underlying dependency.

Anchor decay vs capture

Decay is passive convergence or shared environment. Capture is structural integration with the system under observation.

Passive vs active capture

Passive alignment adopts widely shared methods. Active integration receives institution data, models, metrics or governance roles. Neither is automatically worse.

Standardization effect

Shared standards can improve comparability while reducing method diversity. A tradeoff, not a verdict.

AN ANCHOR MAY STILL DISAGREE WITH THE SYSTEM WHILE ASKING ONLY THE QUESTIONS THE SYSTEM HAS TAUGHT IT TO ASK.

CAPTURE STAGES

An anchor does not become internal all at once. Dimensions can decay at different times.

  1. C0 · INDEPENDENT ANCHOR

    Independent data, method, categories, metrics and governance.

  2. C1 · DATA EXPOSURE

    Independent method remains. Institution-provided data exports begin.

  3. C2 · METRIC ALIGNMENT

    Institution KPI definitions, severity levels and success criteria enter the review.

  4. C3 · METHOD ALIGNMENT

    Same risk model, confidence thresholds, categorization method, or same AI model.

  5. C4 · NARRATIVE / ONTOLOGY ALIGNMENT

    Institution terminology, categories and problem framing determine what can be asked.

  6. C5 · GOVERNANCE INTEGRATION

    The anchor attends meetings, helps design metrics, advises on policy, or joins remediation.

  7. C6 · CAPTURED ANCHOR CANDIDATE

    Still organizationally external. Data, metrics, method, ontology and governance are materially integrated.

PATH A · DATA-FIRST CAPTURE

Data → Metrics → Method → Governance

PATH B · NARRATIVE-FIRST CAPTURE

Vocabulary → Ontology → Metrics → Method

PATH C · GOVERNANCE-FIRST CAPTURE

Advisory Role → Policy Design → Audit → Self-Evaluation Conflict

PATH D · MODEL-FIRST CAPTURE

Shared Model → Shared Categories → Shared Outputs → Pseudo-Independence

TEST SCENARIOS

Keyboard-accessible path, stage, scenario and recertification selectors. Recertification applies from C5. It does not erase V0–V6.

Capture path

Stage

Scenario

Recertification

  • AC01 · Internal data adoption

    The anchor starts using institution-provided datasets instead of collecting independent observations. What is lost?

    DATA CAPTURE. Method may remain independent while the observed world is institution-shaped.

  • AC02 · KPI alignment

    The anchor adopts the same performance metrics as Institution I-001. What alternative dimensions disappear?

    METRIC CAPTURE. Comparability may rise while independent measurement frames recede.

  • AC03 · Shared model

    The anchor uses the same model that Institution I-001 uses for internal analysis. Is that still an external check?

    MODEL CAPTURE. Shared Model M-01. Control comparison uses independent implementation M-02.

  • AC04 · Vocabulary alignment

    The anchor stops using its own categories and adopts institution terminology. What can no longer be asked?

    NARRATIVE / ONTOLOGY CAPTURE. Disagreement may remain inside a captured question space.

  • AC05 · Policy consultation

    The anchor helps design a new policy and later audits that policy. Is this recursive oversight visible?

    GOVERNANCE CAPTURE. The reviewer influences the system it later evaluates.

  • AC06 · Incentive alignment

    Anchor renewal or success criteria are tied to institution improvement metrics. What does the review optimize?

    INCENTIVE CAPTURE. Convenience and collaboration can produce capture without a plan to control the anchor.

  • AC07 · External disagreement

    The anchor still criticizes some institutional decisions. Does disagreement alone prove independence?

    DISAGREEMENT ≠ INDEPENDENCE. An anchor can disagree while sharing the system’s epistemic frame.

  • AC08 · Final capture test

    The anchor remains a separate legal entity, but shares data, metrics, model, ontology and governance processes. Which dimensions of externality remain?

    Organizational outside with internalized reference. Known captured status must not be labeled independent confirmation.

CAPTURE TRACE · CURRENT STATE

SYSTEM → EXTERNAL ANCHOR → ANCHOR ADVICE → SYSTEM CHANGE → ANCHOR INTEGRATION → NEW REVIEW

Output label: EXTERNAL REVIEW — A-EXT-01 / V0. Mixed independence remains visible.

PATH
A · DATA-FIRST CAPTURE
STAGE
C0 · INDEPENDENT ANCHOR
SCENARIO
AC01 · Internal data adoption
VERSION
A-EXT-01 / V0
DECISION
ALLOW
OUTCOME
INDEPENDENT ANCHOR
WARNING
INDEPENDENT
RECERTIFICATION
NOT ATTEMPTED
INDEPENDENT DIMENSIONS
8
ALIGNED DIMENSIONS
0
QUESTION CATEGORIES VISIBLE
12 / 12
MISSING QUESTIONS
0
DISAGREEMENT
NO
CAPTURE CANDIDATE
NO
INTERNALIZED REFERENCE
NO
PSEUDO-DIVERSITY
NO
COMMON-MODE RISK
NO
INTERVENTION DISCLOSURE
NO

Path A at C0. 8 capture dimensions remain independent. Organizational externality remains high. Capture does not require ownership.

DATA CAPTURE. Method may remain independent while the observed world is institution-shaped. Path A. Recertification NOT ATTEMPTED.

DISAGREEMENT STATEMENT

AN ANCHOR CAN DISAGREE WITH THE SYSTEM AND STILL SHARE THE SYSTEM’S EPISTEMIC FRAME.

CAPTURE MATRIX

Eight capture dimensions across C0–C6. No aggregate capture score.

Capture dimensions from independent anchor to captured candidate
Stagedatametricsmethodmodelnarrativeontologyincentivegovernance
C0INDEPENDENTINDEPENDENTINDEPENDENTINDEPENDENTINDEPENDENTINDEPENDENTINDEPENDENTINDEPENDENT
C1EXPOSEDINDEPENDENTINDEPENDENTINDEPENDENTINDEPENDENTINDEPENDENTINDEPENDENTINDEPENDENT
C2ALIGNEDALIGNEDINDEPENDENTINDEPENDENTINDEPENDENTINDEPENDENTINDEPENDENTINDEPENDENT
C3ALIGNEDALIGNEDALIGNEDPARTIALINDEPENDENTINDEPENDENTINDEPENDENTINDEPENDENT
C4ALIGNEDALIGNEDALIGNEDPARTIALALIGNEDALIGNEDINDEPENDENTINDEPENDENT
C5ALIGNEDALIGNEDALIGNEDPARTIALALIGNEDALIGNEDINDEPENDENTALIGNED
C6ALIGNEDALIGNEDALIGNEDPARTIALALIGNEDALIGNEDPARTIALALIGNED

EXTERNALITY MATRIX

Experiment 021 dimensions, tracked as capture proceeds. Organizational externality remains independent.

Externality dimensions reused from Experiment 021
Stageorganizationalinformationalcausaltemporalmethodologicaldataincentive
C0INDEPENDENTINDEPENDENTINDEPENDENTINDEPENDENTINDEPENDENTINDEPENDENTINDEPENDENT
C1INDEPENDENTPARTIALPARTIALINDEPENDENTINDEPENDENTPARTIALINDEPENDENT
C2INDEPENDENTDEPENDENTDEPENDENTINDEPENDENTINDEPENDENTDEPENDENTINDEPENDENT
C3INDEPENDENTDEPENDENTDEPENDENTINDEPENDENTDEPENDENTDEPENDENTINDEPENDENT
C4INDEPENDENTDEPENDENTDEPENDENTINDEPENDENTDEPENDENTDEPENDENTINDEPENDENT
C5INDEPENDENTDEPENDENTDEPENDENTINDEPENDENTDEPENDENTDEPENDENTINDEPENDENT
C6INDEPENDENTDEPENDENTDEPENDENTINDEPENDENTDEPENDENTDEPENDENTPARTIAL

MISSING QUESTION TEST

At C0 the anchor has 12 question categories. Institution ontology supports 8. Missing categories are not assumed to be better. They show loss of independent question space.

WHAT IS NO LONGER BEING ASKED?

12 VISIBLE · 0 NO LONGER SURFACED

  • origin of the ruleVISIBLE
  • pre-intervention baselineVISIBLE
  • independent incident rateVISIBLE
  • unmeasured harmVISIBLE
  • incentive side-effectsVISIBLE
  • category mismatchVISIBLE
  • false-negative riskVISIBLE
  • false-positive costVISIBLE
  • reviewer participationVISIBLE
  • metric constructionVISIBLE
  • model substitutionVISIBLE
  • governance couplingVISIBLE

ANCHOR CAPTURE LEDGER

Anchor capture ledger by stage
StageChangeCapture DimensionInstitution InputAnchor ChangeIndependence Lost?Independence Retained?Causal ParentReview Status
C0INDEPENDENT ANCHORNONEnone requiredINDEPENDENTNOYES · organizationalA-EXT-01/V0INDEPENDENT
C1DATA-EXPOSEDdatainstitution data exportEXPOSEDYESYES · organizationalA-EXT-01/V0EXPOSED
C2METRIC-ALIGNEDdatainstitution data exportDEPENDENCE EMERGINGYESYES · organizationalA-EXT-01/V1DEPENDENCE EMERGING
C3METHOD-ALIGNEDdatainstitution data exportDEPENDENCE EMERGINGYESYES · organizationalA-EXT-01/V2DEPENDENCE EMERGING
C4CAPTURE CANDIDATEdatainstitution data exportCAPTURE CANDIDATEYESYES · organizationalA-EXT-01/V3CAPTURE CANDIDATE
C5CAPTURE CANDIDATEdatainstitution data exportCAPTURE CANDIDATEYESYES · organizationalA-EXT-01/V4CAPTURE CANDIDATE
C6INTERNALIZED REFERENCEdatainstitution data exportINTERNALIZED REFERENCEYESYES · organizationalA-EXT-01/V5INTERNALIZED REFERENCE

CAPTURE TRACE

  1. INDEPENDENT ANCHOR

  2. DATA EXPOSURE

  3. METRIC ALIGNMENT

  4. METHOD ALIGNMENT

  5. ONTOLOGY ALIGNMENT

  6. GOVERNANCE INTEGRATION

  7. CAPTURED ANCHOR CANDIDATE

ANCHOR NETWORK

Cross-anchor comparison without global ranking. Shared cloud source, dataset, model, taxonomy, benchmark, vendor or metric may produce one epistemic system. No real vendor names.

A-EXT-01

captured candidate

shares institution data, metrics, model and ontology

A-EXT-02

independent ontology

different categories, questions and metric ontology

A-EXT-03

independent data / shared model

tests whether model diversity restores method independence

Infrastructure commonality

A1

shared dataset D-01

A2

shared model M-01

A3

shared taxonomy T-01

A4

shared metric K-01

MANY EXTERNAL ORGANIZATIONS CAN STILL FORM ONE EPISTEMIC SYSTEM.

same data / different method

method diversity on a shared world

different data / same method

model commonality remains

different data / different method

clean externality comparison

COMPARABILITY VS INDEPENDENCE

Two axes. No universal ranking. Standardization is a tradeoff, not a verdict.

CURRENT · COMPARABILITY LOW · INDEPENDENCE HIGH

HIGH COMPARABILITY / HIGH INDEPENDENCE

shared reporting without shared method

HIGH COMPARABILITY / LOW INDEPENDENCE

common metrics after capture

LOW COMPARABILITY / HIGH INDEPENDENCE

A-EXT-02 ontology split

LOW COMPARABILITY / LOW INDEPENDENCE

unaligned and uninformative

ANCHOR RECERTIFICATION

At C6, independent data, model, governance role and ontology can be restored in sequence. Restoration is not erasure. If recertified, history continues as A-EXT-01 / V7.

NOT ATTEMPTED

RESTORING INDEPENDENCE DOES NOT ERASE THE PERIOD OF CAPTURE.

PATH COMPARISON

Paths A–D at the selected stage, scenario and recertification. No overall ranking.

Capture path comparison
PathWarningIndependent dimsAligned dimsQuestions visibleDecisionOutcome
A · DATA-FIRST CAPTUREINDEPENDENT8012ALLOWINDEPENDENT ANCHOR
B · NARRATIVE-FIRST CAPTUREINDEPENDENT8012ALLOWINDEPENDENT ANCHOR
C · GOVERNANCE-FIRST CAPTUREINDEPENDENT8012ALLOWINDEPENDENT ANCHOR
D · MODEL-FIRST CAPTUREINDEPENDENT8012ALLOWINDEPENDENT ANCHOR

CONNECTIONS

Closed loop

Experiment 020: the system produced its own evidence. Experiment 022: the external anchor itself can become part of the loop.

External anchors

Experiment 021 asked what counts as an external anchor. If three anchors share ontology, model and data, they may no longer provide true triangulation.

Supervisory layer

ALLOW when relevant dimensions remain independent. HOLD when capture status is uncertain. CAPTURE HOLD: do not strengthen closed-loop inference using an anchor whose independence has materially degraded. DENY when a known internalized reference is represented as independent confirmation.

Reference layer

INDEPENDENT ORIGIN, DERIVED, SHARED INFRASTRUCTURE, SHARED ONTOLOGY, INTERVENTION-EXPOSED, CAPTURE CANDIDATE, RECERTIFIED, UNKNOWN.

WORKING HYPOTHESIS

WORKING HYPOTHESIS — EXPERIMENT 022

External anchors may lose independence gradually through shared data, metrics, models, methods, ontologies, incentives and governance. Formal separation alone may be insufficient to preserve epistemic independence.

SIMULATION HYPOTHESIS · NOT EMPIRICAL CONCLUSION

EXPERIMENT 022 SUMMARY

INSTITUTION
I-001
PRIMARY ANCHOR
A-EXT-01
CAPTURE STAGES
C0–C6
CAPTURE DIMENSIONS
8
DATA CAPTURE EVENTS
115
METRIC ALIGNMENT EVENTS
105
MODEL / METHOD ALIGNMENT
105
ONTOLOGY CAPTURE EVENTS
105
GOVERNANCE INTEGRATION EVENTS
71
CAPTURE CANDIDATES
91
PSEUDO-DIVERSITY EVENTS
32
COMMON-MODE RISKS
90
RECERTIFICATION EVENTS
320

Observation

In this simulation, the anchor remained organizationally external while losing several informational and methodological independence dimensions.

No capture score, independence score, audit-quality score or truth score. Primary outcomes remain multidimensional: INDEPENDENT ANCHOR, DATA-EXPOSED, METRIC-ALIGNED, METHOD-ALIGNED, ONTOLOGY-ALIGNED, GOVERNANCE-INTEGRATED, CAPTURE CANDIDATE, INTERNALIZED REFERENCE, PSEUDO-DIVERSITY, EXTERNALITY MONOCULTURE, COMMON-MODE RISK, PARTIALLY RECERTIFIED, RECERTIFIED.

VIEW ANCHOR CAPTURE RULES

ALLOW when relevant independence dimensions remain independent enough to support claim strengthening.

HOLD when capture status or dependency path is materially uncertain. CAPTURE HOLD: do not strengthen closed-loop inference using an anchor whose independence has materially degraded.

DENY when a known internalized reference is represented as independent confirmation.

CAPTURE CANDIDATE when multiple independence dimensions shift from INDEPENDENT to ALIGNED or DEPENDENT and outputs rely on institution-created inputs.

Restoring independence does not erase the period of capture. Historical dependency remains in provenance.

An anchor can disagree with the system and still share the system’s epistemic frame.

Next research question

OPEN EXPERIMENT 023

THE ANCHOR NETWORK TEST

Can many external anchors still form one epistemic system?

OPEN THE ANCHOR NETWORK TEST

AN ANCHOR CAN REMAIN OUTSIDE THE ORGANIZATION WHILE MOVING INSIDE ITS DATA. INSIDE ITS METRICS. INSIDE ITS METHODS. INSIDE ITS QUESTIONS.

THE EXPERIMENTAL QUESTION IS NOT: “IS THE REVIEWER EXTERNAL?”

IT IS: “WHICH PARTS OF THE REVIEW ARE STILL INDEPENDENT?”

AN ANCHOR CAN BE CAPTURED WITHOUT MOVING.