Guardrail and Failure-Mode Register

Document ID

TEM-P3-3A3B3C-SD03-GUARDRAIL

Version / issue

Issue 1 - aligned to the Issue 1 paper sequence

Date

2026-09-24

Status

Controlled Part 3 support document aligned to the eight Issue 1 canonical papers, GOT-V1-3a-1 through GOT-V1-3c-8. Secondary to the canonical papers.

Supersedes

Guardrail and Failure-Mode Register - Issue 1 - Formatted.docx. This content reconciliation preserves the eight-paper sequence and updates source identities, ownership, locators, and support dependencies. Earlier release history remains recoverable in that source copy.

Function

Records Part-level guardrails, failure modes, severity classes, corrections, and diagnostic questions while preserving distinct 3a/3b/3c ownership.

Canonical rule

Primary papers govern technical definitions. Failure-control may expose drift but may not create new technical content.

Internal split rule

Section 3a, Section 3b, and Section 3c remain distinct canonical regions inside one Part-level support control. Downstream papers may inherit and apply upstream architecture but do not retroactively redefine it.

Source-lock rule

Use the eight canonical Part 3 papers and current SD-01/SD-02 as active Part 3 control. Use Parts 1/2 only for inherited terms and guardrails.

Template rule

Retain the established support-document layout and control schema. Earlier support copies retain historical evidence; the eight Issue 1 papers govern current content. Parts 1 and 2 remain inherited-control references only.

GOT-V1-3a-1

GOT-V1-3a-1 - The Gödel Horizon - Issue 1.docx | sha16 44a8b20a68541a55

The Gödel Horizon

Horizon grammar: relational availability; availability/closure boundary conditions; Soft Horizon; Gödel Horizon; Procedural Availability Horizon; coupled loci; abstraction boundaries; comparator and stabilisation discipline.

GOT-V1-3a-2

GOT-V1-3a-2 - The Discipline of the Unknown - Issue 1.docx | sha16 04d3becc2b5a5029

The Discipline of the Unknown

Unknown-statuses; residual; positive exclusion; live formulations; Best Available Formulation; Weighted Residual; action threshold, safeguard, monitoring, stop condition, update, and branch-memory.

GOT-V1-3b-3

GOT-V1-3b-3 - On the Origin of Power - Issue 1.docx | sha16 d6d961fbb277a77c

On the Origin of Power

Primitive Power as IPL-indexed distinction-conditioned branch potential; Counterfactual Inclusion Field; potential/application/consequence separation; Counterfactual Exclusion; thermodynamic grounding.

GOT-V1-3b-4

GOT-V1-3b-4 - On the Origin of Responsibility - Issue 1.docx | sha16 ef17be5480bc49e7

On the Origin of Responsibility

Primitive Responsibility as locus-indexed attribution; Responsibility Provenance; distributed and temporal provenance; Responsibility Gradient; Immediate-Frame Principle; Responsibility Warrant.

GOT-V1-3b-5

GOT-V1-3b-5 - On the Origin of Trust - Issue 1.docx | sha16 d8a68ac35982e5a0

On the Origin of Trust

Trust as reliance closure under residual; primitive audit; branch-space economy; trust thresholds; vicarious trust; Epistemic Authority and the minimum Structural Authority distinction (§8); trust carriers; laundering; betrayal, repair, and formalisation handoff.

GOT-V1-3c-6

GOT-V1-3c-6 - The Case for Formalisation - Issue 1.docx | sha16 b4d11bd226838155

The Case for Formalisation

Load-bearing language; recoverable authority; distinct epistemic grounding and structural authorisation (§1); epistemic debt; institutions and machines; formalisation as preservation; auditable claim packet; proportional formalisation; retrospective review.

GOT-V1-3c-7

GOT-V1-3c-7 - The Constitutional Governance of Formalisation - Issue 1.docx | sha16 a3540c539c730360

The Constitutional Governance of Formalisation

Governance of formalisation; constitutional/technical/application levels; application of Epistemic Authority inherited from 3b-5; custodial and operational applications of Structural Authority (§6); amendment, forks, certification, and capture controls.

GOT-V1-3c-8

GOT-V1-3c-8 - The Epistemic Method - Issue 1.docx | sha16 dc08048b02787be4

The Epistemic Method

Integrated relational audit of claim-strength and reliance-strength; indexed relation; claim and reliance paths; comparator, residual, reliance limit; preferential reliance and carrier-path preservation (§6); conditional Power/Responsibility audits; method-compliance is not closure.

P1-SD01

Part 1 - Concept Compression Map - Issue 1.docx | sha16 a256c5229849f993

Upstream architecture compression for NTC, BD, IPL, DTS, life-threshold and localisation discipline.

P1-SD02

Part 1 - Glossary - Issue 1.docx | sha16 e97712ced4e2b095

Upstream controlled vocabulary for inherited Part 1 terms.

P1-SD03

Part 1 - Guardrail and Failure-Mode Register - Issue 1.docx | sha16 794360cc1ff16519

Upstream comparator, localisation, source-lock, and anti-inflation guardrails.

P1-SD04

Part 1 - Bibliography and Relation Register - Issue 1.docx | sha16 7b0b194c1552979b

Upstream bibliographic relation-control reference only; no citation transfer into Part 3.

P2-SD01

Part 2 - Concept Compression Map - Issue 1.docx | sha16 eee99287e2cfaa10

Upstream control for RRP, signal, OP/OA/OR/RAS/RSD, knowledge, intelligence, observation, abstraction, scale, and observation-program architecture.

P2-SD02

Part 2 - Glossary - Issue 1.docx | sha16 b144e7dc2eac7133

Upstream controlled vocabulary for inherited informational and observational terms.

P2-SD03

Part 2 - Guardrail and Failure-Mode Register - Issue 1.docx | sha16 f955341321f11d30

Upstream failure-control for local locus, information/signal, observation, abstraction, scale, and descriptor discipline.

P2-SD04

Part 2 - Bibliography and Relation Register - Issue 1.docx | sha16 12dc458d0e940e8c

Upstream bibliographic relation-control reference only; no citation transfer into Part 3.

Protocol

Historical reference: GOT - Bibliographic Identification Classification and Adjacency Discovery Protocol - Issue 1(1).docx | sha16 03dec29bfa52830c

Retained classification context from the prior support release. This protocol file is not in the current supplied control set; its recorded hash is historical and is not represented as a newly verified source. Existing classifications are retained, with no new promotion of inheritance.

P3-SD01

Concept Compression Map - Issue 1.docx | sha16 874e44c76caf3061

Architecture compression and source-discipline control.

P3-SD02

Glossary - Issue 1.docx | sha16 8e0c4e96367405e7

Controlled vocabulary and canonical-home cross-check.

Class A - Mechanical inconsistency

Breaks the architecture.

Redefining upstream terms downstream; treating every cause as Power; role as provenance; method-compliance as closure; carrier as warrant.

Correct before finalisation.

Class B - Terminology drift

Creates ambiguity or future misuse.

Using horizon without availability/closure demand; Power without CIF/locus; Responsibility without attribution target; Trust without closure demand; formalisation without consequence trigger.

Correct before compression or publication.

Class C - Presentation ambiguity

May mislead despite intact architecture.

“Unknown” sounds blank; “power” sounds authority; “responsibility” sounds blame; “trust” sounds moral virtue; “constitutional” sounds immutable; “audit” sounds certification.

Clarify through ownership note, discriminator, residual, or diagnostic question.

Source-lock drift

Superseded papers/support wording or unverified material re-enters as current Part 3 source.

Return to eight current paper hashes and listed controls.

A

Can every technical claim be traced to a current canonical paper or permitted upstream control?

Canonical ownership drift

A support document or later paper silently redefines an earlier canonical term.

Return the definition to the owning paper; use downstream content only as alignment pressure.

A

Which paper owns the discriminator?

Section back-propagation

3b or 3c vocabulary is pre-installed into 3a because it later proves compatible.

Preserve directional inheritance; revise upstream only for actual inconsistency in its owned relation.

A

Would the upstream paper still state the relation without the later vocabulary?

Inheritance by repetition

Repeated citation or use across papers is treated as Indexed Inheritance.

Keep citation, comparator, dependency, and inheritance separate.

A

Has a separate inheritance audit promoted the relation?

Locus drift

Target, carrier, observer, affected locus, applying locus, and responsible locus are silently merged.

Re-index the relation at each material transformation.

A

For whom/what does the claimed relation hold here?

Scale/horizon drift

A claim travels across scale or horizon without preserving its comparator and access conditions.

Restate reference scale/horizon and show a valid bridge.

B

What changed when the claim moved?

Horizon objectification

A horizon is treated as a place, object, metaphysical zone, or fixed edge.

Restore the availability/closure boundary relation.

A

What relation, locus, availability condition, and closure demand produce the boundary?

Gödel inflation

Any uncertainty, difficulty, ignorance, or limitation is labelled a Gödel Horizon.

Require available/formulable relation plus disclosed failure of closure from the present regime’s own resources.

A

Is the closure failure actually internal to the stated regime and demand?

Procedural overclaim

A describable experiment, proof path, evidence path, or process is treated as if realised.

Mark Procedural Availability Horizon and withhold closure until the path exists.

A

Can the required path actually be executed by the locus?

Soft-Horizon reification

Provisional closure travels as final warrant.

Carry residual, scope, comparator/tolerance, purpose, and update with the closure.

A

What remains unresolved and what would force update?

Unknown flattening

All unresolved relation becomes one generic unknown.

Type unknown-status by availability and closure condition.

B

Is the relation unavailable, unrecognised, misclosed, residual, live, weighted, or action-bound?

Residual laundering

Residual is mentioned but allowed to do no limiting work.

State what stronger claim is blocked and what update would change it.

A

Does the residual change claim-strength or action limits?

Live-field equality

All live formulations are treated as equally supported because final closure is absent.

Weight heterogeneously and retain positive exclusions.

B

What evidence/comparator/scale support differentiates them?

BAF closure inflation

Best Available Formulation is treated as final truth.

Retain its purpose type, residual, alternatives, and update conditions.

A

What stronger closure demand remains unmet?

Action-certainty conversion

Action taken under threshold is later narrated as proof or truth.

Separate action threshold from truth/explanatory closure; preserve branch-memory.

A

Would the claim have been equally strong if no action had been required?

Stabilisation hindsight

Realised outcome erases what alternatives were live or knowable beforehand.

Preserve contemporaneous live field, exclusions, and update history.

B

What was actually available to the locus at the time?

Cause-to-Power collapse

Any causal contribution is called Power.

Require IPL-localised distinction-conditioned branch potential under comparator pressure.

A

Would the branch field differ because this organisation can make a discrimination operative?

Power-as-force

Energetic magnitude or visible effect substitutes for the discriminator.

Separate causal magnitude from distinction-conditioned potential.

A

Is organised discrimination doing the branch-changing work?

Power-as-authority

Role, office, command, status, permission, or possession is treated as primitive Power.

Recover the physically supported CIF of the specified locus.

A

What branches are actually included by the locus’s organisation?

Imagined-CIF inflation

Exterior analyst imagination populates the CIF with unrealised but physically unsupported futures.

Include only branches supported by the IPL under stated conditions.

A

What physical path makes each claimed branch available?

Potential/application collapse

Possessing a power potential is treated as equivalent to present application or consequence.

Keep potential, application, and consequence distinct.

B

Which part of the relation is currently established?

CE agency inflation

Counterfactual Exclusion is read as conscious choice, rejection, or metaphysical deletion.

Restore it as the negative side of a realised distinction-conditioned branch event.

B

Was a relevant included alternative simply not equivalently realised?

Thermodynamic escape

Branch-changing power is treated as abstract information unconstrained by physical work/entropy.

Retain physical carriers and lawful thermodynamic accounting.

A

What substrate and energetic pathway carry the distinction and transition?

Causation-to-Responsibility collapse

Every cause becomes Responsibility.

Follow only the distinction-conditioned subset of causal provenance.

A

What local discrimination/organisation did consequential work?

Blame inflation

Primitive Responsibility is treated as blame, guilt, desert, duty, or legal liability.

Keep mature warrant regimes downstream of primitive attribution.

A

Is the claim causal-attributional or normative/allocative?

Provenance transfer

Responsibility is treated as a token passed from one locus to another.

Localise each material transformation; provenance composes but does not simply transfer.

A

What did each locus actually transform, retain, apply, maintain, or update?

Proximity shortcut

Nearest physical actor is assigned the whole responsibility path.

Recover upstream and downstream distinction-conditioned transformations.

A

Which earlier conditions remained materially operative?

Upstream absorption

An early contributor is made responsible for all later consequences.

Test whether later loci corrected, replaced, ignored, or transformed the relation.

A

Did the upstream distinction remain operative in the later branch?

Exterior hindsight attribution

Evaluator uses information unavailable to the locus at the time.

Apply the Immediate-Frame Principle.

A

Was the relevant relation materially available or recoverable then?

Interior frame laundering

A capable locus retrospectively narrows its stated frame to exclude warnings or alternatives that were actually available.

Reconstruct frame from recoverable paths, not self-description alone.

A

What channels, records, warnings, or correction paths existed?

Warrant-as-provenance

Role, duty, liability, answerability, or institutional allocation is treated as causal provenance.

Keep Responsibility Warrant separate and explicitly related to provenance.

A

What does the warrant allocate, and what causal path does it not establish?

Gradient moralisation

Responsibility Gradient becomes a universal moral score.

Keep it ordinal/relational and tied to capacity plus actual provenance.

B

Is the gradient describing capacity or allocating blame?

Trust-as-certainty

Reliance closure is narrated as proof, confidence without residual, or total transparency.

Restore residual, threshold, and update.

A

What remains unaudited while reliance proceeds?

Trust-carrier reification

Carrier is treated as trust ground because it bears a credential, role, record, or institutional mark.

Recover the closure process the carrier is standing in for.

A

What work did the carrier actually earn?

Trust laundering

Reliance-strength earned in one relation is spent in another.

Re-index relation, carrier, closure demand, residual, and scope.

A

What changed between the relation that earned trust and the relation now relying on it?

Borrowed-closure warrant transfer

Vicarious trust is treated as automatic transfer of warrant or Responsibility Provenance.

Keep outsourced closure, reliance, and provenance distinct.

A

What is merely relied upon versus causally attributable?

Audit/trust opposition

Audit is treated as evidence that trust has failed, or trust as elimination of audit.

Treat audit as the reopening function that preserves corrigibility.

B

What condition should foreground the backgrounded relation again?

Repair by reassurance

Trust repair is asserted by apology, branding, certification, or reassurance without changing the failed relation.

Recover and alter the failure-generating conditions; narrow/refuse trust if needed.

A

What changed in the carrier, process, safeguard, or update path?

Unshared-comparator authority

A locus is preferred for one relation using success on another relation, scale, horizon, or closure demand.

Recover the shared target relation and comparable closure demand, availability, abstraction, scale, and horizon (3b-5 §8).

A

Does the demonstrated advantage concern the relation and use now claimed?

Confidence as epistemic standing

Confidence, precision, quick closure, or reduced expressed doubt is taken as demonstrated epistemic advantage.

Test relation-preservation and the treatment of residual; no single numerical uncertainty measure is required (3b-5 §8).

A

Could the apparently weaker contribution be exposing residual that the confident one conceals?

Relative advantage as sufficient closure

The stronger of two contributions is treated as adequate for a use that neither can support.

Retain the closure demand and residual limits; actionability cannot replace demonstration (3b-5 §8).

A

Is the advantage sufficient for this reliance, or only for further inquiry?

Carrier-path loss of advantage

Standing at the contributing locus is assumed to survive a stale record, altered abstraction, or misleading interface.

Recover both the target relation and the carrier relation for present use (3b-5 §8; 3c-8 §6).

A

What epistemic advantage actually reaches the relying locus?

Recognition as epistemic ground

A contribution acquires supposed standing merely because it is trusted, adopted, or used.

Separate earned standing from actual reliance. Trust makes standing operative but does not create it (3b-5 §8).

A

Would the demonstrated standing remain if the contribution were untrusted or unused?

Authority-type substitution

Structural permission is treated as epistemic standing, or epistemic standing as permission to act or decide.

Apply the two definitions in 3b-5 §8 independently; retain scoped authorisation and demonstrated relation-preservation.

A

Which authority is established, and what would establish the other?

Consequence as epistemic warrant

The occurrence or magnitude of an effect retrospectively establishes the epistemic standing of the relied-upon contribution.

Apply the Power discriminator to the branch difference and the Epistemic Authority discriminator to standing (3b-5 §8).

A

What demonstrated the relation-preserving advantage before consequence was used to validate it?

Universal formalisation

All language is forced into rigid formal structure.

Apply formalisation only where language/output becomes load-bearing and proportionate to consequence.

B

What consequential work triggers the formalisation burden?

Form-completion closure

Completed fields/forms are treated as proof that the relation is preserved.

Audit the relation and paths, not packet completion.

A

Could every field be filled while the target relation remains wrong?

Machine execution laundering

Automation or consistency is treated as repair of upstream under-specification.

Recover training/data/model/interface/decision paths and unresolved abstractions.

A

What relation did the machine actually preserve?

Record-as-warrant

A retained record or documented provenance is treated as epistemic warrant by default.

Separate record/provenance from adequacy and closure.

A

What comparator or evidence shows the record preserves the claimed relation?

Epistemic-debt concealment

Deferred grounding burden is hidden behind stable terminology or process.

Make debt visible as missing relation/path/update obligations.

B

What grounding work has been postponed?

Retrospective outcome bias

Good outcome is treated as evidence the original decision path was strong.

Audit decision quality under the information and conditions available at decision time.

A

Would the path still be judged strong under a different outcome?

Authorisation-path substitution

The mandate governing a claim’s use substitutes for the evidence path that supports its epistemic standing.

Keep structural authorisation and epistemic grounding separately recoverable (3c-6 §1; 3b-5 §8).

A

Does permission to use the claim establish why its target relation is preserved?

Adoption-as-validation

Institutional uptake or implementation scale is treated as evidence the method is correct.

Keep operational adoption distinct from epistemic authority.

A

What relation-preservation was demonstrated independently of adoption?

Custodial-authority inflation

Stewardship of releases/archives is treated as ownership of epistemic validity.

Treat custody as an application of Structural Authority; recover release and correction records without promoting stewardship into epistemic standing (3c-7 §6; 3b-5 §8).

A

Could a steward be wrong while remaining the steward?

Operational-authority inflation

Institutional authority to apply the method becomes authority to redefine it epistemically.

Treat domain authority as an application of Structural Authority; distinguish scoped permission from the epistemic amendment burden (3c-7 §§6–7).

A

Is this an application decision or a technical/constitutional revision?

Preference amendment

Method changes because of convenience, vote, politics, adoption pressure, or author preference.

Require demonstrated relational failure and comparative improvement.

A

What relation does the current formulation fail to preserve?

Founding-text ownership

Originating author/text is treated as immune from correction.

Preserve provenance but deny epistemic ownership.

A

Would a demonstrated counterexample force revision regardless of authorship?

Constitutional-label capture

The highest-level label is used to block challenge.

Raise the burden of revision, not immunity to revision.

A

What would count as demonstrated failure of the proposed constitutional element?

Fork-as-truth

Existence or popularity of a competing fork is treated as evidence of superiority.

Compare relation-preservation, residual, scope, and failure conditions.

B

What does the fork preserve better and at what cost?

Certification closure

Certificate becomes general epistemic ground beyond what was assessed.

Bound certification to assessed relation, scope, residual, and re-audit conditions.

A

What exactly was certified, and what was not?

Method universalisation

Epistemic Method is used as a substitute for science, law, ethics, judgement, or domain expertise.

Use it as relational audit discipline only.

A

What domain-specific burden remains external to the method?

Method-compliance closure

Method-use or completion is treated as epistemic certificate.

Return strength to demonstrated relation-preservation.

A

What claim would remain weak even after perfect procedural compliance?

Path flattening

Claim path, evidence path, record path, carrier path, and reliance path are collapsed.

Recover each load-bearing path separately.

A

Where did content, evidence, record, authority, and reliance each enter?

Claim/reliance-strength collapse

A strong claim is assumed equally safe to rely upon for every locus/purpose, or vice versa.

Index reliance to the relying locus’s closure demand and limit.

A

Does the reliance purpose require more than the claim itself establishes?

Conditional-path universalisation

CIF/CE or Responsibility Provenance fields are required for claims that do not engage Power/Responsibility.

Activate relation-specific audits only when the claim triggers them.

B

What relation-specific claim is being made?

Audit paralysis

Residual is used to prevent all finite action.

Permit action under threshold with proportionate safeguards and update where justified.

B

What action burden exists despite incomplete closure?

Audit theatre

Visible review process substitutes for discriminatory comparator pressure or correction.

Require actual potential to expose error, weaken claims, refuse closure, or force update.

A

Can the audit produce a materially adverse result for the claim?

Preferential-reliance shortcut

Comparison favours another locus without establishing the advantage for the shared relation and closure demand or its survival through the carrier.

Apply the comparator burden where preferential reliance is claimed (3c-8 §6), inheriting 3b-5 §8.

A

What advantage supports this reliance at the receiving locus?