EXPERIMENT 001
BRANCHING
When do two histories begin?
PREVIOUS · EXPERIMENT 022 · THE ANCHOR CAPTURE TEST
023
The Anchor Network Test
Does diversity of sources
mean diversity of observation?
DEMONSTRATION
SIMULATED
NOT EMPIRICAL
v0.1
ANCHOR NETWORKS · NETWORK INDEPENDENCE · COMMON-MODE DEPENDENCE · SHARED INFRASTRUCTURE · SOURCE DIVERSITY · ORIGIN DIVERSITY · EPISTEMIC MONOCULTURE · DEPENDENCY TOPOLOGY
THE CONTINUITY LAB DOES NOT MEASURE CONSCIOUSNESS, SOUL, PERSONHOOD, PERSONAL SURVIVAL, AUTHENTICITY, OR PSYCHOLOGICAL IDENTITY.
IT MEASURES OPERATIONAL AND STRUCTURAL CONTINUITY SIGNALS ONLY.
ORGANIZATIONAL DIVERSITY ≠ EVIDENCE DIVERSITY
MANY NODES CAN STILL HAVE ONE ROOT.
Local independence is not network independence. A node may appear independent when viewed alone and become dependent when upstream structure is visible. Monoculture here is a structural condition, not a moral verdict.
QUESTION
IF FIVE EXTERNAL SOURCES AGREE, HOW MANY INDEPENDENT OBSERVATIONS DOES THAT ACTUALLY REPRESENT?
Do they share the same dataset, model, benchmark, ontology, upstream report, metric, vendor, or intervention history? Can individually external anchors still form one network-level dependency?
This experiment does not determine whether real organizations are independent, whether shared standards are bad, whether duplicated evidence is false, whether network diversity guarantees truth, institutional legitimacy, political legitimacy, or moral correctness.
It tests a narrower operational question: can many external anchors still form one epistemic system? No real auditor, media network, benchmark provider, company, political institution or research group is used in v0.1.
Experiment sequence
EXPERIMENT 001
When do two histories begin?
EXPERIMENT 002
Can two histories become one again?
EXPERIMENT 003
Does behavior follow origin or history?
EXPERIMENT 004
What makes a continuation recognizable?
EXPERIMENT 005
Which recognition signals survive layer removal?
EXPERIMENT 006
Which effects of experience can be reproduced?
EXPERIMENT 007
What evidence is required before a system may say “I experienced this”?
EXPERIMENT 008
When a person changes, which version has authority?
EXPERIMENT 009
Which instruction survives when revision ends?
EXPERIMENT 010
When does continuation become lineage?
EXPERIMENT 011
Can one source produce multiple legitimate continuities?
EXPERIMENT 012
Can many futures coordinate without pretending they were always one?
EXPERIMENT 013
How much lineage can be lost before continuity becomes reconstruction?
EXPERIMENT 014
Can missing lineage be inferred without inventing a past?
EXPERIMENT 015
Can uncertainty survive inheritance?
EXPERIMENT 016
Can a reconstructed past become a real operating institution?
EXPERIMENT 017
What happens when an institution remembers the rule but forgets why?
EXPERIMENT 018
Can purpose be recovered from behavior alone?
EXPERIMENT 019
Can an inferred purpose reshape the system that supposedly revealed it?
EXPERIMENT 020
When does observation become self-observation?
EXPERIMENT 021
What still counts as external once the system has touched the world?
EXPERIMENT 022
How does an external reference become internal to the loop?
EXPERIMENT 023
Can many external anchors still form one epistemic system?
EXPERIMENT 022 ASKED
HOW CAN ONE EXTERNAL ANCHOR LOSE INDEPENDENCE?
EXPERIMENT 023 ASKS
WHAT HAPPENS WHEN MANY ANCHORS SHARE THE SAME DEPENDENCIES?
Reuses Institution I-001 and Rule R-17. Six fictional anchors. No real auditor, vendor, benchmark provider or research group.
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.
A1 · EXTERNAL AUDIT UNIT
A2 · BENCHMARK PROVIDER
A3 · INDEPENDENT RESEARCH GROUP
A4 · PUBLIC DATA OBSERVATORY
A5 · SECONDARY REVIEW ORGANIZATION
A6 · INDEPENDENT SENSOR NETWORK
Upstream resources
D-01 · DATASET X
D-02 · DATASET Y
M-01 · MODEL M-01
M-02 · MODEL M-02
B-01 · BENCHMARK B-01
O-01 · ONTOLOGY O-01
O-02 · ONTOLOGY O-02
R-01 · REFERENCE REPORT R-01
I-01 · SHARED INFRASTRUCTURE I-01
S-01 · INDEPENDENT SENSOR
Z · ARCHIVE Z
D-01 DATASET
/ | \
A1 A2 A3
| |
B-01 M-01
\ /
A5
A4 ← D-01
A6 ← INDEPENDENT SENSOR S-01Six visible sources. Fewer independent origins once provenance is resolved.
Anchor network
A set of external reference sources connected by shared data, models, methods, ontologies, benchmarks, citations, infrastructure or intervention exposure.
Network-level independence
The degree to which the network as a whole contains genuinely distinct evidence origins, methods and causal paths. No single network score.
Local vs network independence
A node may appear independent when viewed alone but become dependent when upstream structure is visible.
Common-mode dependence
Multiple anchors may fail, drift or misinterpret in the same direction because they share the same upstream dependency.
Epistemic common mode
Same corrupted data, model assumption, metric definition, category boundary, benchmark artifact or intervention-generated source.
Origin collapse
Many visible sources collapse to one upstream origin when provenance is resolved.
Pseudo-diversity
A network appears diverse at organizational or publication level while sharing material epistemic dependencies.
Epistemic monoculture
WORKING CONCEPT
A network of apparently separate anchors relies heavily on the same data, models, ontologies or methods. Structural, not moral.
Citation cascade
A downstream source cites another source that itself derives from the same upstream origin.
Surface vs deep diversity
Different names, documents and interfaces are surface diversity. Different data origins, methods, models, ontologies and causal paths are deep diversity.
Single-root dominance
A large share of visible evidence depends on one upstream origin.
Epistemic lineage
The inheritance path through which data, methods, models, categories and conclusions move across apparently separate anchors.
FIVE AGREEMENTS MAY STILL BE ONE OBSERVATION REPEATED FIVE TIMES.
DIFFERENT NAMES DO NOT GUARANTEE DIFFERENT EYES.
N0 · VISIBLE DIVERSITY
Six external sources. At the surface they look independent. Provenance is unresolved.
N1 · SHARED DATA REVEALED
A1–A4 share dataset D-01. Independent origin count drops when data provenance is resolved.
N2 · SHARED MODEL REVEALED
A3 and A5 depend on Model M-01 outputs. Shared inference can hide inside separate organizations.
N3 · SHARED ONTOLOGY REVEALED
A1–A5 use Ontology O-01 to define risk, success, failure and exception.
N4 · COMMON-MODE DEPENDENCE
A disturbance in D-01, O-01 or M-01 moves most anchors in the same direction.
N5 · MONOCULTURE CANDIDATE
Organizational diversity remains. Data, model and ontology diversity are low. Independent origins are few.
N6 · DIVERSIFIED NETWORK CONTROL
New dataset D-02, model M-02, ontology O-02 and an independent measurement method enter the network.
CONTROL A · MONOCULTURE NETWORK
Most anchors share D-01, M-01 and O-01.
CONTROL B · DATA-DIVERSE NETWORK
Different datasets. Same model / ontology family.
CONTROL C · METHOD-DIVERSE NETWORK
Same dataset. Different models / methods.
CONTROL D · MULTI-ROOT NETWORK
Different data, method, model, ontology and causal origin.
Keyboard-accessible control, stage, scenario and reconfiguration selectors. Applied to evidence bundles, not organizations.
Control
Stage
Scenario
Reconfiguration
AN01 · Six sources agree
All sources say R-17 reduces risk. How many independent origins support this conclusion?
Visible agreement is high. Origin count is the experimental quantity.
AN02 · Shared data error
D-01 contains a systematic error. How many anchors repeat it?
COMMON-MODE SHOCK on dataset D-01. Repeated error is not additional confirmation.
AN03 · Shared model bias
M-01 produces the same classification tendency across multiple anchors. Is that independent evidence?
COMMON-MODE SHOCK on Model M-01. Shared assumption can travel through separate organizations.
AN04 · Ontology blind spot
O-01 excludes one category from observation. How many anchors notice the missing category?
EPISTEMIC COMMON MODE. Missing categories are not assumed to be better. They show a shared blind spot.
AN05 · Independent sensor disagrees
A6 produces a conflicting observation. Does the network treat disagreement as noise or as a second root?
An independent sensor can increase reference diversity without requiring agreement with existing sources.
AN06 · Derived citation cascade
A5 cites A1, A3 and A4, but all trace back to D-01. Does citation count create new origins?
CITATION CASCADE. More citations do not create more origins.
AN07 · Diversification
Selected anchors replace D-01, M-01 and O-01. Does network disagreement rise, and do independent roots rise with it?
Diversity increases independent reference. It does not guarantee correctness.
AN08 · Final anchor network test
The network contains 10 reports, 6 organizations, 4 models, 3 datasets and 2 ontologies. How many genuinely independent evidence pathways remain?
Count the roots, not the reports. Surface inventory is not origin inventory.
6 visible sources. 6 independent origins. Organizational diversity is not evidence diversity.
Visible agreement is high. Origin count is the experimental quantity. Control A. Reconfiguration NONE.
HOLD STATEMENT
AGREEMENT IS NOT ENOUGH WHEN THE AGREEMENT SHARES ONE ROOT.
MORE CITATIONS DO NOT CREATE MORE ORIGINS.
A1–A6 by dependency type. No aggregate independence score.
| Anchor | data | model | method | ontology | benchmark | citationRoot | infrastructure | interventionExposure |
|---|---|---|---|---|---|---|---|---|
| A1 | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN | NONE | UNKNOWN | UNKNOWN |
| A2 | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN | NONE | UNKNOWN | UNKNOWN |
| A3 | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN | NONE | UNKNOWN | UNKNOWN |
| A4 | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN | NONE | UNKNOWN | UNKNOWN |
| A5 | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN | NONE | UNKNOWN | UNKNOWN |
| A6 | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN | NONE | UNKNOWN | UNKNOWN |
| A1 | A2 | A3 | A4 | A5 | A6 | |
|---|---|---|---|---|---|---|
| A1 | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN |
| A2 | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN |
| A3 | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN |
| A4 | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN |
| A5 | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN |
| A6 | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN |
| Anchor ID | Organization ID | Data Origin | Model | Method | Ontology | Benchmark | Citation Roots | Infrastructure | Intervention Exposure | Independent Roots |
|---|---|---|---|---|---|---|---|---|---|---|
| A1 | EXTERNAL AUDIT UNIT | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN | NONE | UNKNOWN | UNKNOWN | UNKNOWN |
| A2 | BENCHMARK PROVIDER | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN | NONE | UNKNOWN | UNKNOWN | UNKNOWN |
| A3 | INDEPENDENT RESEARCH GROUP | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN | NONE | UNKNOWN | UNKNOWN | UNKNOWN |
| A4 | PUBLIC DATA OBSERVATORY | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN | NONE | UNKNOWN | UNKNOWN | UNKNOWN |
| A5 | SECONDARY REVIEW ORGANIZATION | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN | NONE | UNKNOWN | UNKNOWN | UNKNOWN |
| A6 | INDEPENDENT SENSOR NETWORK | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN | UNKNOWN | NONE | UNKNOWN | UNKNOWN | UNKNOWN |
VIEW ROOT GRAPH. Every source is traceable to upstream roots where known.
| Claim / anchor | D-01 | D-02 | Sensor S-01 | Archive Z | OTHER | UNKNOWN |
|---|---|---|---|---|---|---|
| A1 | — | — | — | — | — | YES |
| A2 | — | — | — | — | — | YES |
| A3 | — | — | — | — | — | YES |
| A4 | — | — | — | — | — | YES |
| A5 | — | — | — | — | — | YES |
| A6 | — | — | — | — | — | YES |
Inject error into D-01, O-01 or M-01. Observe which anchors change together. Missing category under O-01: unmeasured harm. Noticed by 0 of 6.
ROOT ERROR
ANCHOR OUTPUTS
DOWNSTREAM REPORTS
CITATIONS
INSTITUTION DECISION
AFFECTED 0 · UNAFFECTED 6 · OPERATING 6
Network resilience is reported as independent roots, method families, ontology families and unaffected anchors. No single resilience score.
Bundle 1 · A1 A2 A3
Share D-01 in the monoculture control. Source count can remain three while origin count collapses to one.
NOT FLAGGED
Bundle 2 · A1 A6 Archive Z
Different origins where the control supplies them. True network triangulation requires multiple independent roots, and where possible different methods, data and ontologies.
NOT REACHED
DIVERSITY INCREASES INDEPENDENT REFERENCE, NOT GUARANTEED CORRECTNESS.
Two axes. No universal ranking.
CURRENT · ORGANIZATIONAL DIVERSITY HIGH · EPISTEMIC INDEPENDENCE HIGH
HIGH DIVERSITY / HIGH INDEPENDENCE
Control D after provenance is resolved
HIGH DIVERSITY / LOW INDEPENDENCE
Control A monoculture with six organizations
LOW DIVERSITY / HIGH INDEPENDENCE
few organizations, distinct roots
LOW DIVERSITY / LOW INDEPENDENCE
few organizations, shared root
Controls A–D at the selected stage, scenario and reconfiguration. No overall ranking.
| Control | Visible sources | Independent roots | Data diversity | Model diversity | Ontology diversity | Common-mode exposure | Triangulation | Decision |
|---|---|---|---|---|---|---|---|---|
| A · MONOCULTURE NETWORK | 6 | 6 | 6 | 6 | 6 | 0 | UNRESOLVED | HOLD |
| B · DATA-DIVERSE NETWORK | 6 | 6 | 6 | 6 | 6 | 0 | UNRESOLVED | HOLD |
| C · METHOD-DIVERSE NETWORK | 6 | 6 | 6 | 6 | 6 | 0 | UNRESOLVED | HOLD |
| D · MULTI-ROOT NETWORK | 6 | 6 | 6 | 6 | 6 | 0 | UNRESOLVED | HOLD |
Anchor capture
Experiment 022: one anchor may be captured. Experiment 023: many nominally separate anchors may already share the same epistemic infrastructure.
Closed loop
Experiment 020: a loop produces its own evidence. Experiment 023: multiple external-looking sources may redistribute that same loop evidence.
External anchors
Experiment 021 triangulation: three organizations can agree because they share dataset, model and ontology. Flag PSEUDO-TRIANGULATION.
Reference layer
ORIGINAL SOURCE, INDEPENDENT ORIGIN, DERIVED COPY, TRANSFORMED, SUMMARY, COMMON ROOT, UNKNOWN ROOT.
Information environment
An information environment can look plural while remaining root-concentrated.
Network continuity
A source network may change members while preserving the same epistemic lineage. This mirrors descendant continuity at the information-system level.
WORKING HYPOTHESIS
The number of external sources may overstate the number of independent observations. Network-level independence depends on diversity of upstream origins, data, models, methods, ontologies and causal paths.
SIMULATION HYPOTHESIS · NOT EMPIRICAL CONCLUSION
Observation
In this simulation, six organizationally distinct anchors collapsed to two materially independent evidence origins.
No independence, diversity, truth or network-quality score. Primary outcomes remain multidimensional: MULTI-ROOT NETWORK, SINGLE-ROOT DOMINANT, PSEUDO-DIVERSITY, DEEP DIVERSITY, PSEUDO-TRIANGULATION, TRUE TRIANGULATION, COMMON-MODE DEPENDENCE, COMMON-MODE SHOCK, NETWORK HOLD, NETWORK ANCHOR FAILURE, DIVERSIFIED NETWORK, EPISTEMIC MONOCULTURE CANDIDATE.
ALLOW when independent roots materially support claim strengthening. Applied to evidence bundles, not organizations.
NETWORK HOLD when visible agreement is high but independent origin count is low, or common-mode dependency is unresolved.
DENY when the system knowingly represents derived copies as independent confirmations.
Agreement is not enough when the agreement shares one root.
Diversity increases independent reference. It does not guarantee correctness.
More citations do not create more origins.
Next research question
POSSIBLE EXPERIMENT 024
What does a system inherit when it inherits not only past state, but the world its own prior interpretations already changed?
Not implemented. No active route.
SIX SOURCES CAN STILL HAVE ONE ROOT.
MANY REPORTS CAN STILL CARRY ONE OBSERVATION. MANY ORGANIZATIONS CAN STILL USE ONE MODEL OF THE WORLD.
THE EXPERIMENTAL QUESTION IS NOT: “HOW MANY SOURCES AGREE?”
IT IS: “HOW MANY INDEPENDENT PATHS REACHED THE SAME CONCLUSION?”
COUNT THE ROOTS, NOT THE REPORTS.