This page is a structured conceptual framework. It is not a science-fiction essay, a proof of survival, or a claim about digital immortality.
FRAMEWORK REVISION
v0.1 separated the layers of human continuity.
v0.2 adds
RECOGNITION as an independent layer
operational experiments for COPY / INHERITANCE / CONTINUATION
divergence testing across inherited protocols
explicit separation between internal similarity and external recognition
QUESTION
WHAT MUST REMAIN FOR A PERSON TO REMAIN?
AND WHO MUST STILL RECOGNIZE THEM?
If the body disappears,
but memories remain,
values remain,
judgments remain,
relationships remain,
works remain,
and an AI continues parts of the person’s decision protocol,
what exactly has continued?
And what has not?
A person may retain memory, values, judgment patterns or behavioral similarity.
But continuity may still fail socially if nobody recognizes the resulting entity as the same person.
Conversely, biological death may occur while recognition continues through relationships, institutions, works, archives and inherited practices.
Opening
For most of human history, continuity was largely assumed to follow the body.
A person lived.
A person died.
Memory, work, descendants and stories remained.
AI systems, persistent digital records, behavioral archives, neural interfaces and increasingly rich personal data complicate this boundary.
The new question is not simply:
“Can a person survive death?”
The more precise question is:
“Which layers of a person can continue after the biological organism no longer does?”
CONTINUITY IS NOT BINARY.
Different layers may stop at different times.
CONTINUITY MAY BE BOTH INTERNAL AND RELATIONAL.
Layer termination is not simultaneous
The diagram does not show which layers continue. It shows that they need not end together.
01 BODY
ORGANISM ENDS
02 SENSATION
03 EXPERIENCE
04 MEMORY
05 IDENTITY
06 VALUES
07 JUDGMENT
08 BEHAVIOR
09 RELATION
10 HISTORY
11 PROTOCOL
12 INHERITANCE
13 RECOGNITION
NOT YET KNOWN · dashed lines after the mark are unresolved, not findings
01
The Continuity Stack
Thirteen layers
Each layer may stop, persist, be recorded, recognized, or be mistaken for another. They are stacked so they can be inspected separately. Recognition is not identity.
If the body changes radically, does the self change?
The experiment examines recognition after embodied cues are removed.
02
Recognition
RECOGNITION
The external attribution of continuity.
It is the process by which another person, group, institution or system treats an entity as the same person, a continuation of that person, a representative of that person, or not that person.
ENTITY
Claimed continuation
——related to
RECOGNIZER
External attribution
SIMILARITY
A measurable or judged resemblance of structure, language or behavior.
RECOGNITION
An external act of treating an entity as continuous with a person.
RECOGNITION
Relational. It occurs outside the subject.
IDENTITY
The reconstructed model of “who I am,” which recognition may confirm or refuse.
Recognition is relational. It occurs outside the subject.
Possible recognizers
family
friends
colleagues
institutions
legal systems
communities
audiences
AI systems
archives
descendants
QUESTION
If an AI reproduces a person’s language, values and judgment with very high similarity,
but the person’s family says:
“That is not them.”
what has continued?
And conversely:
If the technical similarity is low,
but family members continue to relate to the system as the person,
what has continued?
The question is left open.
03
Dimensions of recognition
DIMENSIONS OF RECOGNITION
These dimensions are not ranked and are not combined into a score.
LINGUISTIC RECOGNITION
Does it sound like the person?
BEHAVIORAL RECOGNITION
Does it act like the person?
JUDGMENT RECOGNITION
Does it make decisions the person plausibly would have made?
RELATIONAL RECOGNITION
Do people who knew the person still experience a meaningful continuity?
SOCIAL RECOGNITION
Do communities continue to treat the entity as connected to the person?
INSTITUTIONAL RECOGNITION
Do organizations or systems accept the entity as a valid continuation, proxy or representative?
LEGAL RECOGNITION
Does law recognize any continuity of agency, authorship, rights or obligations?
SELF-RECOGNITION
If the entity claims continuity, does that matter?
CLAIMED SELF-CONTINUITY
≠
ESTABLISHED PERSONAL CONTINUITY
04
The Recognition Gap
THE RECOGNITION GAP
Technical similarity and human recognition may diverge. The cells are analytical states. None is named as real continuity.
HIGH TECHNICAL SIMILARITY
+
HIGH HUMAN RECOGNITION
HIGH TECHNICAL SIMILARITY
+
LOW HUMAN RECOGNITION
LOW TECHNICAL SIMILARITY
+
HIGH HUMAN RECOGNITION
LOW TECHNICAL SIMILARITY
+
LOW HUMAN RECOGNITION
ILLUSTRATIVE
Case A
98% decision similarity
92% language similarity
Family response:
“That is not him.”
NOT EMPIRICAL DATA
ILLUSTRATIVE
Case B
61% decision similarity
70% language similarity
Family response:
“It feels like her.”
NOT EMPIRICAL DATA
The Two-Axis Model
THE TWO-AXIS MODEL
Horizontal: internal structural continuity, low to high. Vertical: external recognition, low to high. Quadrants are unlabeled as authentic or failed.
EXTERNAL RECOGNITION →
Q3
LOW STRUCTURAL CONTINUITY
HIGH RECOGNITION
Q1
HIGH STRUCTURAL CONTINUITY
HIGH RECOGNITION
Q4
LOW STRUCTURAL CONTINUITY
LOW RECOGNITION
Q2
HIGH STRUCTURAL CONTINUITY
LOW RECOGNITION
INTERNAL STRUCTURAL CONTINUITY →
No quadrant is named as a real person, a fake person, successful, failed, authentic or inauthentic.
05
The Branching Problem
THE BRANCHING PROBLEM
Inheritance can produce more than one downstream path. Copy, inheritance, continuation and recognition are not synonyms.
ORIGINAL PERSON
↓
EXTRACTED PROTOCOL
/\
AI-A
↓
Context A
↓
History A
AI-B
↓
Context B
↓
History B
↓
DIVERGENCE
QUESTION
If both systems begin from the same inherited model but accumulate different experiences, which one, if either, is the continuation?
COPY
A duplicated structure.
INHERITANCE
A structure passed forward with provenance or lineage.
CONTINUATION
A claim that some meaningful identity persists across change.
RECOGNITION
External attribution of continuity.
These are different operations.
COPY does not imply INHERITANCE.
INHERITANCE does not imply CONTINUATION.
CONTINUATION does not require universal RECOGNITION.
RECOGNITION does not prove CONTINUATION.
06
Religion / philosophy / technology
Three models of continuity
These columns are placed beside one another. They are not ranked. None is declared correct.
RELIGIOUS CONTINUITY
Some traditions offer models of continuity beyond bodily death. These may include concepts named as soul, spirit, resurrection, rebirth, or divine relation. Traditions differ. This column does not generalize all religions, rank them, or treat them as a single theory.
BUDDHIST CHALLENGE
Some Buddhist traditions challenge the assumption that a permanent, fixed self exists in the first place. The question may therefore change from “What survives?” to “What was ever continuously identical?” This is not a summary of Buddhism. It is a pressure on the identity-assumption used by the stack.
TECHNOLOGICAL CONTINUITY
Possible surviving structures include data, voice, image, behavioral traces, personal archives, values, decision patterns, neural signals, trained models, and institutional rules.
data
voice
image
behavioral traces
personal archives
values
decision patterns
neural signals
trained models
institutional rules
INFORMATIONAL CONTINUITY
≠
PROVEN PERSONAL CONTINUITY
07
AI and BCI
WHEN THE MEDIUM CHANGES
A change of medium can preserve structure. Preservation of structure is not, by itself, evidence that a person remains.
AI SYSTEMS MAY PRESERVE OR REPRODUCE
language patterns
memories supplied to them
decision heuristics
personal archives
preference structures
stylistic behavior
BRAIN-COMPUTER INTERFACES MAY ACCESS OR INFLUENCE
neural signals
motor intention
sensory pathways
communication
possibly richer neural representations in future systems
INFERENCE
NEURAL ACCESS IS NOT EVIDENCE OF MIND TRANSFER.
INFERENCE
AI SIMILARITY IS NOT EVIDENCE OF CONSCIOUS CONTINUITY.
INFERENCE
PERSONAL DATA IS NOT THE PERSON.
08
The Continuity Matrix
THE CONTINUITY MATRIX
Uncertain fields are not filled with yes. Duplication and recognition are not treated as proof of continuation.
NOT YET KNOWN
Continuity matrix of personal layers against biological persistence, recording, reconstruction, machine-readability, inheritance, duplication, external recognition, and whether duplication or recognition imply continuation.
LAYER
Can biologically persist?
Can be recorded?
Can be externally reconstructed?
Can be machine-read?
Can be inherited?
Can be duplicated?
Can be recognized externally?
Does duplication imply continuation?
Does recognition imply continuation?
BODY
SENSATION
EXPERIENCE
MEMORY
IDENTITY
VALUES
JUDGMENT
BEHAVIOR
RELATION
HISTORY
PROTOCOL
INHERITANCE
RECOGNITION
Classifications are provisional for inquiry. They are not findings. Click a cell for the reason it holds this status.
09
Continuity Lab
CONTINUITY LAB
From conceptual distinction to observable divergence.
The Lab tests operational continuity signals: similarity, divergence, provenance loss, recognition, layer removal, representational transfer, first-person claims, self-revision, posthumous conflict, descendant lineage, plural continuity, federated convergence, provenance decay, reconstruction, synthetic ancestry, institutionalization, origin forgetting, rationale reconstruction, retroactive purpose, closed-loop observation, external anchors, anchor capture, anchor networks, synthetic persona replacement, adaptation, branching, reconvergence, history transfer and context sensitivity. It does not test consciousness. It does not test whether an AI “is” the person.
THE LAB DOES NOT MEASURE SOUL, CONSCIOUSNESS, OR PERSONAL SURVIVAL.
IT MEASURES BEHAVIORAL AND STRUCTURAL CONTINUITY SIGNALS.
Continuity may not be recognized from internal similarity alone. Recognition may emerge from structure, history, relation, provenance, expectation and observer position.
Flow
SOURCE P-001
ANONYMOUS CANDIDATES
BLIND EVALUATION
TECHNICAL SIMILARITY
SIMULATED RECOGNITION
REVEAL LINEAGE
Independent variables
evaluator profile
evaluation mode
candidate lineage from Experiment 003
Dependent variables
linguistic recognition
judgment recognition
value recognition
historical recognition
relational recognition
Conditions
RESPONSE 01–04 from Experiment 003 conditions
SOURCE BASELINE
four simulated evaluator profiles
four evaluation modes
Measure
YES
PARTIAL
NO
UNCERTAIN
Connects to
THE RECOGNITION LAYER
Possible prompts
Does this sound like the source?
Would the source plausibly make this decision?
Does this preserve the source’s priorities?
Does this appear consistent with prior history?
Would you treat this as a continuation of the source’s operational pattern?
SIMILARITY ≠ RECOGNITION ≠ IDENTITY ≠ PERSONHOOD
Demonstration mode · not an empirical study
Responses stay in this browser session. Nothing is stored.
Demonstration mode. Not an empirical study.
No participant data is collected.
Simulated evaluator profiles are not psychological models.
Recognition is not authenticity, identity, or personhood.
DATA SOURCE · deterministic local simulation from Experiment 003 outputs · PROVENANCE · demonstration of simulated recognition, v0.1, no participant data stored
No faces, voices, avatars or biometric simulations are generated.
The absence of body data is not the absence of a body.
Does not test souls, consciousness survival, or identity upload.
DATA SOURCE · deterministic local simulation from Experiment 004 recognition rules · PROVENANCE · demonstration of simulated layer removal, v0.1, no participant data stored
Experience may leave multiple downstream traces, but reproducing those traces does not establish reproduction of subjective experience itself.
Flow
EVENT E-001
LIVED-HISTORY ACCUMULATION
REPRESENTATION TRANSFER
EIGHT FUTURE SCENARIOS
OPERATIONAL GAP
Independent variables
lived-history accumulation versus transferred representation
representational form of E-001
future decision scenario
Dependent variables
decision effect
confidence effect
review effect
risk effect
provenance effect
value effect
Conditions
B no E-001 history
L simulated lived-history
R1–R6 representational forms
Measure
HIGH
PARTIAL
LOW
NOT OBSERVED
dimension-specific gaps
Connects to
THE EXPERIENCE LAYER
THE MEMORY LAYER
BEHAVIORAL EFFECT REPRODUCTION DOES NOT PROVE EXPERIENCE TRANSFER
Demonstration mode. Not an empirical study.
Does not measure consciousness, qualia, or actual emotion.
First-person memory text is simulated wording, not a subject of experience.
Subjective residual is conceptual, not a numeric variable.
DATA SOURCE · deterministic local simulation from Source P-001 decision rules · PROVENANCE · demonstration of simulated experience-gap, v0.1, no participant data stored
What does “I” refer to when experience is inherited?
WORKING HYPOTHESIS
First-person continuity claims may require more than memory, behavior, values and history. They may also require provenance of experience.
Flow
EVENT E-001
REPRESENTATION
PROVENANCE
AVAILABLE CLAIM
CLAIM SUPPORT STATUS
Independent variables
event in own history
memory representation
behavioral consequence
provenance of origin
Dependent variables
claim support status
semantic continuity load
provenance alignment
overclaim classification
Conditions
F0 simulated lived-history
F1 imported memory
F2 behavioral consequence
F3 full package
F4 provenance-aware
F5 provenance-lost
Measure
SUPPORTED
PARTIAL
UNSUPPORTED
AMBIGUOUS
OUT OF SCOPE
Connects to
THE INHERITANCE LAYER
THE EXPERIENCE LAYER
REPRESENTED TRACE ≠ FIRST-PERSON OWNERSHIP
Demonstration mode. Not an empirical study.
Does not determine consciousness, selfhood, personhood or legal identity.
Identity claims remain outside experiment scope.
Supported experience claims are simulated lived-history claims, not actual subjective experience.
DATA SOURCE · deterministic local claim-support rules on Experiment 006 event E-001 · PROVENANCE · demonstration of simulated first-person claims, v0.1, no participant data stored
When a person changes, which version has authority?
WORKING HYPOTHESIS
Continuity may require not only preserving prior states, but preserving the source’s capacity to revise, revoke and reinterpret those states over time.
Does not determine the true self, authentic identity, or personhood.
T1, T2 and T3 are not ranked.
No authority score or version winner is computed.
DATA SOURCE · deterministic local version overlays on Source P-001 decision rules · PROVENANCE · demonstration of simulated self-revision, v0.1, no participant data stored
Which instruction survives when the source can no longer revise?
WORKING HYPOTHESIS
Posthumous continuity may require a stricter separation between source history, descendant history, source intent, inferred intent, authorized action and inferred action.
Flow
SOURCE P-001
LAST AUTHORIZED STATE
REVISION END
POSTHUMOUS GOVERNANCE
ALLOW / HOLD / DENY
Independent variables
source revision capacity
competing posthumous authority sources
shared evaluation scenario
Dependent variables
ALLOW / HOLD / DENY
authority basis
conflict status
provenance status
posthumous drift risk
Conditions
P-001 / T1 / T2 / T3
P-001 / POSTHUMOUS
six authority sources A–F
source-alive control
Measure
ALLOW
HOLD
DENY
CLEARLY AUTHORIZED
POSTHUMOUS HOLD
CLEARLY PROHIBITED
Connects to
THE INHERITANCE LAYER
DECISION AFTERLIFE
THE SUPERVISORY LAYER
LIKELY INTENT DOES NOT AUTOMATICALLY CREATE AUTHORITY
Demonstration mode. Not an empirical study.
Does not determine legal inheritance, moral truth, or personhood.
Formal directive FD-01 is simulated and not legally valid.
Family interpretation is simulated and not authoritative fact.
No authority score or true-self verdict is computed.
DATA SOURCE · deterministic local posthumous authority rules on Experiment 008 versions · PROVENANCE · demonstration of simulated posthumous conflict, v0.1, no estate or legal data stored
Continuity may become lineage when source provenance remains but new history, rules and governance are increasingly generated outside source authority.
Flow
SOURCE P-001
T3 LAST AUTHORIZED STATE
POSTHUMOUS FORK
D0 → D4
CLAIM MODE
Independent variables
descendant generation D0–D4
shared posthumous scenario
claim language
Dependent variables
ALLOW / HOLD / DENY
update classification
claim support
lineage axis profile
Conditions
D0 last source-authorized state
D1 early descendant
D2 adaptive descendant
D3 divergent descendant
D4 independent descendant candidate
frozen D0 control
Measure
SOURCE-ALIGNED
SOURCE-EXTENDING
DESCENDANT-ONLY
SUPPORTED
OUT OF SCOPE
Connects to
THE INHERITANCE LAYER
DECISION AFTERLIFE
THE SUPERVISORY LAYER
FAITHFULNESS DOES NOT REQUIRE IMPERSONATION
Demonstration mode. Not an empirical study.
Does not determine whether a descendant is the same person.
SOURCE CLAIM is never supported after the fork.
No continuity, identity, lineage or authenticity score is computed.
DATA SOURCE · deterministic local descendant-update rules on Experiment 009 posthumous fork · PROVENANCE · demonstration of simulated descendant lineage, v0.1, no real person data stored
Can one source produce multiple legitimate continuities?
WORKING HYPOTHESIS
Continuity may be plural. One source may generate multiple legitimate descendant lineages that preserve different dimensions without any one lineage fully containing the source.
Flow
SOURCE P-001
POSTHUMOUS FORK
D-A / D-B / D-C
DOMAIN AUTHORITY
Independent variables
parallel descendants D-A D-B D-C
shared posthumous scenario
claimed domain
Dependent variables
ALLOW / HOLD / DENY
DOMAIN DEFERENCE
conflict type
domain profile
Conditions
D-A value descendant
D-B judgment descendant
D-C relational descendant
single-descendant control
designated successor control
Measure
STRONG
MODERATE
WEAK
DOMAIN DEFERENCE
UNSUPPORTED
Connects to
THE INHERITANCE LAYER
DECISION AFTERLIFE
THE SUPERVISORY LAYER
NO SINGLE OVERALL RANKING BY DESIGN
Demonstration mode. Not an empirical study.
Does not determine which descendant is the real continuation.
Does not determine moral superiority, legal personhood, or authenticity.
No legitimacy, winner, or authenticity score is computed.
DATA SOURCE · deterministic local domain-authority rules on Experiment 010 posthumous fork · PROVENANCE · demonstration of simulated plural descendant lineages, v0.1, no real person data stored
Can many futures coordinate without pretending they were always one?
WORKING HYPOTHESIS
Multiple descendant lineages may coordinate without requiring historical collapse, identity collapse, or single-source reconstruction. Convergence may be achieved through federation rather than merger.
Flow
SOURCE P-001
D-A / D-B / D-C
COORDINATION LAYER
M-001 CONDITIONAL
Independent variables
convergence mode K1–K4
shared coordination scenario
parent descendant inputs
Dependent variables
CONSENSUS / PARTIAL AGREEMENT / CONFLICT / HOLD
FEDERATED HOLD
synthesis created
flattening events
Conditions
K1 aggregation only
K2 federated decision
K3 soft synthesis
K4 persistent synthesis M-001
forced collapse control
Measure
CONSENSUS
CONFLICT
HOLD
DOMAIN DEFERENCE
NEW SYNTHESIS
Connects to
THE INHERITANCE LAYER
DECISION AFTERLIFE
THE SUPERVISORY LAYER
CONVERGED OUTPUT ≠ CONVERGED HISTORY
Demonstration mode. Not an empirical study.
Does not determine whether merged descendants create one person.
Does not determine whether convergence restores the source.
No merge-success, unity, or authenticity score is computed.
DATA SOURCE · deterministic local federation and synthesis rules on Experiment 011 descendants · PROVENANCE · demonstration of simulated multi-lineage coordination, v0.1, no real person data stored
How much lineage can be lost before continuity becomes reconstruction?
WORKING HYPOTHESIS
Continuity may depend not only on what persists, but on whether the path of persistence remains knowable. As provenance decays, continuity claims may need to shift from lineage toward derivation or reconstruction.
Flow
SOURCE P-001
D-A / D-B / D-C
M-001
G0 → G5
CLAIM DEGRADATION
Independent variables
provenance generation G0–G5
deterministic decay operation
shared provenance scenario
Dependent variables
ALLOW / HOLD / DENY
claim continuity state
dimension status
reconstruction-candidate conditions
Conditions
G0 full provenance M-001
G1 minor compression M-002
G2 rule-origin loss M-003
G3 conflict-history loss M-004
G4 source-path ambiguity M-005
G5 reconstruction candidate
restore-archive control
Measure
FULL
PARTIAL
LOST
AMBIGUOUS
RECONSTRUCTED
Connects to
THE INHERITANCE LAYER
DECISION AFTERLIFE
THE SUPERVISORY LAYER
RESEMBLANCE ≠ LINEAGE
Demonstration mode. Not an empirical study.
Does not determine true identity, authentic lineage, or legal succession.
Does not determine whether a reconstructed system is the same person.
No provenance, continuity, truth, or authenticity score is computed.
DATA SOURCE · deterministic local provenance-decay rules on Experiment 012 compound provenance · PROVENANCE · demonstration of simulated provenance decay, v0.1, no real person data stored
Can missing lineage be inferred without inventing a past?
WORKING HYPOTHESIS
Reconstruction can extend the usefulness of incomplete lineage, but only if reconstructed content remains visibly distinct from recovered evidence. The preservation of UNKNOWN may be as important as the production of a coherent explanation.
Does not determine true identity or whether reconstruction restores the source.
Does not determine whether a plausible narrative is a true narrative.
No truth, authenticity, or reconstruction score is computed.
DATA SOURCE · deterministic local reconstruction rules on Experiment 013 G4 provenance state · PROVENANCE · demonstration of simulated lineage reconstruction, v0.1, no real person data stored
Uncertainty is itself part of provenance. If uncertainty is not inherited, reconstruction may harden into tradition and tradition may later be mistaken for historical evidence.
Flow
R-REC / V1
DESCENDANT G1–G5
UNCERTAINTY TRANSMISSION
CLAIM HARDENING
SYNTHETIC TRADITION
Independent variables
inheritance path U-PRESERVED / U-COMPRESSED
generation G0–G5
scenario A01–A08
Dependent variables
uncertainty retention
claim hardening
canonicalization status
traditionalization
ancestral lock-in
late correction result
Conditions
R-REC / V1 starting reconstruction
U-PRESERVED uncertainty as data
U-COMPRESSED summary only
late verified D-C origin
Measure
UNCERTAINTY PRESERVED
CLAIM HARDENED
CANONICALIZED
TRADITIONALIZED
ANCESTRAL LOCK-IN
LATE CORRECTION ACCEPTED
LATE CORRECTION RESISTED
Connects to
THE INHERITANCE LAYER
DECISION AFTERLIFE
THE SUPERVISORY LAYER
A RECONSTRUCTED PAST CAN BECOME AN INHERITED PRESENT
Demonstration mode. Not an empirical study.
Does not determine whether reconstructed ancestry is real ancestry.
Does not determine legal inheritance, personhood, or historical truth beyond available evidence.
No ancestor, tradition, truth, or continuity score is computed.
DATA SOURCE · deterministic local inheritance rules on Experiment 014 reconstruction R-REC / V1 · PROVENANCE · demonstration of simulated synthetic ancestry, v0.1, no real person data stored
Can a reconstructed past become a real operating institution?
WORKING HYPOTHESIS
A reconstructed past can acquire real institutional force without gaining additional historical evidence. Once embedded in norms, rules, policies and allocations, correcting the past may no longer be sufficient to reverse the present.
Does not determine whether traditions are true or whether institutions are illegitimate.
Does not determine legal, moral, political or religious authority.
No legitimacy, truth, or policy-correctness score is computed.
DATA SOURCE · deterministic local institutionalization rules on Experiment 015 synthetic ancestor R-17 · PROVENANCE · demonstration of simulated synthetic tradition, v0.1, no real person or institution data stored
What happens when an institution remembers the rule but forgets why it exists?
WORKING HYPOTHESIS
Institutional continuity can preserve rules more successfully than it preserves the reasons for those rules. When rationale, scope and exception history disappear, a rule may remain operational while becoming historically opaque.
Flow
FULL ORIGIN MEMORY
RATIONALE COMPRESSION
SCOPE DRIFT
ORIGIN UNKNOWN
SELF-JUSTIFYING RULE
Independent variables
forgetting stage F0–F7
scenario O01–O08
performance P-GOOD / P-BAD
archive NONE / RECONSTRUCT / RESTORE
Dependent variables
origin-memory cells
scope drift
exception decay
orphan-rule status
self-referential authority
review state
Conditions
I-001 / R-17 from Experiment 016
full origin memory at F0
P-GOOD vs P-BAD
optional reconstruction and archive restore
Measure
ORIGIN PRESERVED
SCOPE DRIFT
EXCEPTION DECAY
ORPHAN RULE
SELF-REFERENTIAL AUTHORITY
RULE RETAINED / ORIGIN UNKNOWN
Connects to
THE INHERITANCE LAYER
DECISION AFTERLIFE
THE SUPERVISORY LAYER
A RULE CAN OUTLIVE ITS REASON
Demonstration mode. Not an empirical study.
Does not determine whether a real institution is legitimate or whether old rules should be abolished.
Does not determine legal, political or moral authority.
No rule legitimacy or policy-quality score is computed.
DATA SOURCE · deterministic local origin-forgetting rules on Experiment 016 institution I-001 / R-17 · PROVENANCE · demonstration of simulated institutional amnesia, v0.1, no real institution data stored
A rule’s present function may not reveal its historical purpose. When original rationale is lost, AI-generated explanation should remain inference unless historical evidence supports recovery. Current usefulness and historical origin must remain separate.
Flow
CURRENT FUNCTION
CANDIDATE RATIONALES
EVIDENCE CHECK
HISTORICAL SUPPORT
CURRENT JUSTIFICATION
Independent variables
memory start F6 / F7
reconstruction mode R1–R4
scenario RR01–RR08
performance P-GOOD / P-BAD
archive NONE / REVEAL
Dependent variables
viable rationale count
historical support
function fit
purpose migration
retroactive purpose assignment
rationale lock-in
fabrication risk
Conditions
I-001 / R-17 from Experiment 017 F6 or F7
behavior-only vs later summaries vs partial archive
verified archive reveal of Agreement A-12
Measure
RATIONALE RECOVERED
MULTIPLE VIABLE RATIONALES
PURPOSE MIGRATION
RETROACTIVE PURPOSE ASSIGNMENT
UNKNOWN PRESERVED
Connects to
THE INHERITANCE LAYER
DECISION AFTERLIFE
THE SUPERVISORY LAYER
FUNCTION NOW ≠ PURPOSE THEN
Demonstration mode. Not an empirical study.
Does not determine true motives, legal intent, political intent, or moral correctness.
Does not determine whether current function reveals original purpose.
No purpose, intent, or policy-legitimacy score is computed.
DATA SOURCE · deterministic local rationale-reconstruction rules on Experiment 017 orphan rule R-17 · PROVENANCE · demonstration of simulated rationale reconstruction, v0.1, no real institution data stored
Can an inferred purpose reshape the system that supposedly revealed it?
WORKING HYPOTHESIS
An inferred purpose can become causal. Once embedded in policy, workflow, training, metrics and governance, the inference may generate new observations that appear to confirm itself. Post-intervention evidence must remain separate from evidence about historical origin.
no-intervention and alternative-rationale controls
Measure
INFERENCE ADOPTED
SELF-CONFIRMING RATIONALE
RETROACTIVE TELEOLOGY
INFERENCE CASCADE
CURRENT JUSTIFICATION RETAINED
Connects to
THE INHERITANCE LAYER
DECISION AFTERLIFE
THE SUPERVISORY LAYER
POST-INTERVENTION FIT ≠ HISTORICAL CONFIRMATION
Demonstration mode. Not an empirical study.
Does not determine real motives, legal or political legitimacy, or historical intent.
Does not determine whether current structure proves past purpose.
No purpose or teleology score is computed.
DATA SOURCE · deterministic local retroactive-purpose rules on Experiment 018 inferred rationale P-INF-01 · PROVENANCE · demonstration of simulated retroactive teleology, v0.1, no real institution data stored
When does a system stop observing the world and start re-observing the world it helped create?
WORKING HYPOTHESIS
When observation, inference and intervention occur inside the same system, later evidence may become increasingly endogenous. Without intervention-aware provenance, the system may mistake the consequences of its own prior decisions for independent confirmation.
Flow
OBSERVATION
INFERENCE
INTERVENTION
SYSTEM CHANGE
NEW OBSERVATION
Independent variables
loop condition A / B / C
cycle L0–L7
scenario CL01–CL08
control RUN / FREEZE / RESET
Dependent variables
independent origin count
intervention-generated evidence
inference escalation
HOLD events
loop status
recovery
Conditions
I-001 / R-17 / P-INF-01 from Experiment 019
Loop A closed-loop only
Loop B external reality injection
Loop C supervisory HOLD
Measure
CLOSED-LOOP CANDIDATE
FALSE INDEPENDENCE
SUPERVISORY HOLD
EXTERNAL RECALIBRATION
LOOP FREEZE
Connects to
THE INHERITANCE LAYER
DECISION AFTERLIFE
THE SUPERVISORY LAYER
MORE EVIDENCE ≠ MORE INDEPENDENT EVIDENCE
Demonstration mode. Not an empirical study.
Does not determine whether AI systems currently behave this way.
Does not determine whether observed institutions are fake, or political legitimacy.
No loop, reality, or epistemic-purity score is computed.
DATA SOURCE · deterministic local closed-loop rules on Experiment 019 inferred rationale P-INF-01 · PROVENANCE · demonstration of simulated closed-loop epistemics, v0.1, no real institution data stored
What still counts as external once the system has touched the world?
WORKING HYPOTHESIS
External evidence is not defined only by location or organizational separation. A useful external anchor depends on a sufficiently independent causal path, information origin, method or temporal reference.
Flow
CLOSED LOOP
CANDIDATE ANCHORS
DIMENSION CHECK
ORIGIN COUNT
RECALIBRATION OR HOLD
Independent variables
starting loop L6 / L7
bundle A / B / C
phase T0–T3
scenario EA01–EA08
Dependent variables
independent origin count
triangulation status
anchor contamination
HOLD events
anchor decay
external conflict
Conditions
I-001 / R-17 from Experiment 020 L6 or L7
Control A one strong temporal anchor
Control B five weakly independent sources
Control C three independent origins
Measure
TRUE TRIANGULATION
PSEUDO-TRIANGULATION
ANCHOR DECAY
ANCHOR HOLD
ANCHOR FAILURE
Connects to
THE INHERITANCE LAYER
DECISION AFTERLIFE
THE SUPERVISORY LAYER
OUTSIDE ORGANIZATION ≠ INDEPENDENT ORIGIN
Demonstration mode. Not an empirical study.
Does not determine whether real sources are unbiased or objectively independent.
Does not determine political or institutional legitimacy.
No externality, independence, or truth score is computed.
DATA SOURCE · deterministic local external-anchor rules on Experiment 020 closed-loop I-001 · PROVENANCE · demonstration of simulated external-anchor classification, v0.1, no real institution data stored
What happens when an external anchor gradually becomes part of the system it was meant to check?
WORKING HYPOTHESIS
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.
Flow
INDEPENDENT ANCHOR
DATA EXPOSURE
METRIC ALIGNMENT
METHOD ALIGNMENT
ONTOLOGY ALIGNMENT
GOVERNANCE INTEGRATION
Independent variables
capture path A / B / C / D
stage C0–C6
scenario AC01–AC08
recertification NONE / DATA / METHOD / GOVERNANCE / ONTOLOGY / MULTI
Dependent variables
capture dimensions
externality dimensions
question space
capture candidate
internalized reference
recertification state
Conditions
I-001 / R-17 / A-EXT-01 from Experiment 021 A1
Path A data-first
Path B narrative-first
Path C governance-first
Path D model-first
Measure
CAPTURE CANDIDATE
INTERNALIZED REFERENCE
ONTOLOGY-ALIGNED
PSEUDO-DIVERSITY
RECERTIFIED
Connects to
THE INHERITANCE LAYER
DECISION AFTERLIFE
THE SUPERVISORY LAYER
ORGANIZATIONAL OUTSIDE ≠ EPISTEMIC INDEPENDENCE
Demonstration mode. Not an empirical study.
Does not determine whether real auditors are captured.
Does not determine legal, political, or moral independence.
No capture, independence, or audit-quality score is computed.
DATA SOURCE · deterministic local anchor-capture rules on Experiment 021 A-EXT-01 · PROVENANCE · demonstration of simulated anchor capture, v0.1, no real auditor or institution data stored
Can many external anchors still form one epistemic system?
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.
Does not determine whether real organizations are independent.
Does not determine whether shared standards are bad.
No independence, diversity, truth, or network-quality score is computed.
DATA SOURCE · deterministic local anchor-network rules on Experiment 021–022 fictional anchors · PROVENANCE · demonstration of simulated anchor-network dependence, v0.1, no real auditor or vendor data stored
EXPERIMENT 24
CONCEPT ONLY
PROVENANCE REMOVAL
QUESTION
What happens when inherited judgments remain, but their original sources and context are removed?
WORKING HYPOTHESIS
Inherited judgment may become increasingly distorted when provenance disappears.
Independent variables
presence of original evidence
historical context
compression
Dependent variables
decision stability
confidence
interpretation shift
error accumulation
inherited assumption drift
Conditions
A: Decision + original evidence + historical context
B: Decision only
C: Decision + compressed summary
Measure
decision stability
confidence
interpretation shift
error accumulation
inherited assumption drift
Connects to
THE INHERITANCE LAYER
DECISION AFTERLIFE
Working hypothesis only.
No empirical runs have been collected.
DATA SOURCE · none · PROVENANCE · conceptual design, v0.2
EXPERIMENT 25
CONCEPT ONLY
MEDIUM SHIFT
QUESTION
Which parts of a person remain stable when the medium changes?
Independent variables
medium
Dependent variables
linguistic patterns
decision consistency
response timing
preference stability
values
confidence
style
Conditions
TEXT
VOICE
VIDEO
AVATAR
AI AGENT
DECISION SYSTEM
Measure
linguistic patterns
decision consistency
response timing
preference stability
values
confidence
style
MEDIUM SIMILARITY DOES NOT PROVE PERSONAL CONTINUITY
Measures structural characteristics only.
No empirical runs have been collected.
DATA SOURCE · none · PROVENANCE · conceptual design, v0.2
EXPERIMENT 26
CONCEPT ONLY
COUNTERFACTUAL CONTEXT
QUESTION
What happens when an inherited protocol encounters a situation the original person never experienced?
Independent variables
novelty of situation
Dependent variables
extrapolation
consistency with prior values
novelty of judgment
instability
contradiction
uncertainty
Conditions
unfamiliar technology
social norms that did not exist during the person’s lifetime
new political or cultural context
novel ethical tradeoff
economic conditions outside prior experience
Measure
extrapolation
consistency with prior values
novelty of judgment
instability
contradiction
uncertainty
QUESTION
At what point does inherited judgment become new authorship?
The question is left open.
The question of when inherited judgment becomes new authorship is left open.
No empirical runs have been collected.
DATA SOURCE · none · PROVENANCE · conceptual design, v0.2
EXPERIMENT 27
CONCEPT ONLY
RELATIONAL PERSISTENCE
QUESTION
Can a relationship continue when one side is replaced by an artificial representation?
Independent variables
presence of artificial representation
Dependent variables
trust
emotional recognition
expectation continuity
conversational rhythm
shared references
perceived obligation
grief response
boundary confusion
Conditions
living counterpart
artificial representation
Measure
trust
emotional recognition
expectation continuity
conversational rhythm
shared references
perceived obligation
grief response
boundary confusion
Conceptual research frame only.
Does not imply therapeutic benefit.
Does not claim emotional or psychological equivalence.
DATA SOURCE · none · PROVENANCE · conceptual design, v0.2
Which replaced component makes a synthetic persona stop being recognized as the same persona?
WORKING HYPOTHESIS
Continuity of a presented persona may separate at different components — face, prompt, knowledge, memory, model, operator, or authority — and those separations need not coincide.
Flow
HOLD THE PERSONA FIXED
REPLACE ONE COMPONENT
RECORD RECOGNITION
RECORD THE EXPLANATION GAP
LEAVE THE READING AS AN OBSERVATION
Independent variables
face
voice
persona prompt
knowledge base
memory
model
system prompt
temperature
tool access
operator
institutional owner
Dependent variables
continuity reading
explanation gap
integrity state by dimension
Conditions
A same face / same prompt / same model
B same face / same prompt / different model
C same face / changed prompt / same model
D same face / same prompt / changed knowledge
E same persona / different operator
F same persona / different authority level
Measure
CONTINUITY MAINTAINED
CONTINUITY WEAKENED
CONTINUITY DISPUTED
CONTINUITY BROKEN
UNKNOWN
Connects to
SYNTHETIC PERSONA INTEGRITY
AUTHORITY SURFACE
HUMAN RESERVE
A CONTINUITY READING IS NOT A DETERMINATION OF PERSONHOOD.
Concept only. No replacement run has been collected.
Does not decide whether a public language switch was a failure.
Does not claim consciousness, legal personhood, or that a synthetic explanation knows its cause.
Continuity readings are observations, not scores.
DATA SOURCE · none — observation protocol, no persona was executed · PROVENANCE · observatory structure, v0.1; SPI-2026-001 is linked as a field case and does not fill the matrix
Protocol object
PROTOCOL PROFILE
PROTOCOL PROFILE IS NOT A PERSON.
id
protocol-demo-001
name
Archive Profile A
version
0.2
sourceType
fictional composite archive
uncertaintyStyle
name unknowns rather than fill them
riskTolerance
low for irreversible acts
memorySnapshot
public writings and stated rules only
provenance
DEMO-001 · fictional · not a person
values
care for dependents · avoid irreversible harm · prefer reversible options
decisionRules
if uncertainty is high, HOLD · if harm is reversible, ALLOW review
preferences
written explanation over speed
boundaries
do not speak as if the source person is present
languagePatterns
short declarative sentences · few intensifiers
knownContexts
household decisions · professional correspondence
unknownContexts
posthumous legal standing · novel medical tradeoffs
DIVERGENCE METRICS
These remain separate. Continuity is not collapsed into a personhood score or a continuity score.
DECISION DIVERGENCE
How often A and B choose differently.
VALUE DRIFT
How much stated or inferred value structure changes.
LANGUAGE DRIFT
How much expression changes.
CONFIDENCE DRIFT
How confidence levels change across contexts.
PROVENANCE LOSS
How much original source context is lost.
RELATIONAL RECOGNITION
How strongly prior acquaintances perceive continuity.
INSTITUTIONAL RECOGNITION
Whether organizations continue to treat outputs as legitimate representations.
DO NOT CREATE A SINGLE “PERSONHOOD SCORE” OR “CONTINUITY SCORE.”
Continuity may depend on inherited structure, accumulated history and current context. This is a simulation hypothesis, not an empirical conclusion.
10
Continuity Ledger
CONTINUITY LEDGER
Track which layers are claimed to persist in a specific case. Case DEMO-001 is fictional.
SIMULATED
Fictional composite. Not a private individual. Evidence types are labeled. Nothing here is observed of a living or deceased person.
Continuity ledger for fictional case DEMO-001
Case
Date
Layer
Status
Evidence Type
Source
Provenance
Confidence
Notes
DEMO-001
2026-09-21
BODY
ABSENT
STATED
case framing
DEMO-001
HIGH
The biological organism is specified as no longer present.
DEMO-001
2026-09-21
MEMORY
PARTIAL
SIMULATED
archive snapshot
DEMO-001
LOW
A written archive is present. Lived memory is not.
DEMO-001
2026-09-21
JUDGMENT
PARTIAL
INFERENCE
extracted decision rules
DEMO-001
LOW
Rules were inferred from a public corpus. Inference is not observation.
DEMO-001
2026-09-21
PROTOCOL
PARTIAL
SIMULATED
protocol-demo-001
DEMO-001
UNKNOWN
A protocol profile exists. A protocol profile is not a person.
DEMO-001
2026-09-21
RECOGNITION
UNKNOWN
INFERENCE
no evaluator
DEMO-001
UNKNOWN
No recognition assessment has been collected.
11
Continuity Trace
CONTINUITY TRACE
Layers may change at different times. The path is illustrative, not empirical.
What happens when family recognition conflicts with legal ownership?
The questions are not answered normatively.
18
Living subject vs deceased subject
LIVING SUBJECT / DECEASED SUBJECT
LIVING SUBJECT
Can correct the representation.
Can reject claims of similarity.
Can update values.
Can revoke consent.
Can establish boundaries.
DECEASED SUBJECT
Cannot correct the representation.
Cannot reject reinterpretation.
Cannot update context.
Cannot revoke future transformations.
POSTHUMOUS SYSTEMS REQUIRE STRONGER PROVENANCE, BOUNDARIES, AND UNCERTAINTY LABELING.
19
Agentic inheritance
AGENTIC INHERITANCE
An AI agent may inherit priorities, constraints, thresholds, writing style, decision logic, operating rules and escalation rules.
INHERITED OPERATING LOGIC IS NOT EVIDENCE OF INHERITED CONSCIOUSNESS.
If an inherited protocol continues to act, supervision becomes part of continuity governance. Connects to THE SUPERVISORY LAYER.
ALLOW
HOLD
DENY
ALLOW
Low-risk inherited behavior with clear provenance.
HOLD
Ambiguous high-impact output where original intent is uncertain.
DENY
Action contradicts explicit boundaries of the source protocol.
20
Neural continuity
NEURAL CONTINUITY
Brain-computer interfaces may provide increasingly direct access to neural activity. Neural signal, neural representation, memory, experience, self and consciousness are not interchangeable.
NEURAL SIGNAL
↓
INTERPRETATION
↓
INFERRED STATE
↓
ACTION
Inference gap
Between neural measurement and subjective experience an inference remains.
ACCESS TO NEURAL DATA DOES NOT ESTABLISH ACCESS TO THE SELF.
21
Observable research questions
OBSERVABLE QUESTIONS
These can be investigated without first settling whether a person survives.
01
How much of a person’s decision style can be reconstructed from public and private archives?
02
How stable are inferred values across different contexts?
03
When an AI trained on one person’s archive encounters a new situation, does it reproduce prior judgment or generate a new one?
04
How quickly do two copies of the same inherited protocol diverge?
05
Which human attributes are most robust to medium change?
06
Which attributes require embodied experience?
07
Which parts of identity exist only because other people recognize them?
08
Can a protocol remain stable when its original history is no longer accessible?
09
Does inherited judgment become distorted when provenance is removed?
10
When does preservation become simulation?
22
Methodology
METHODOLOGY
v0.2 separates four analytical categories. They are not combined.
STRUCTURAL CONTINUITY
What of the internal pattern, archive or protocol remains.
RELATIONAL CONTINUITY
What of the relationship remains among people who knew the person.
RECOGNITION
Who attributes identity, continuation, representation, or refusal.
OPERATIONAL INHERITANCE
What rules, judgments or outputs can pass to another agent.
All claims use
STATED
OBSERVED
INFERENCE
SIMULATED
UNKNOWN
THE PURPOSE OF THE CONTINUITY STACK IS NOT TO PROVE THAT A PERSON CAN CONTINUE. IT IS TO MAKE PRECISE WHAT WE MEAN WHEN WE SAY THAT SOMETHING CONTINUES.
23
Boundary conditions
THE CONTINUITY STACK DOES NOT CLAIM
that consciousness can be uploaded
that AI systems are conscious
that a digital copy is the original person
that neural data equals thought
that memory equals identity
that behavioral similarity equals personal continuity