PHAROS Method Formation Atlas

Four artifacts, one method, no collapse

The four source documents do not compete to be the definitive history. They answer different questions at different analytical resolutions. The register maps the corpus. The intellectual arc explains conceptual handoffs. The timeline calibrates duration and recurrence. The genealogy establishes implementation provenance.

Normalization rule METHOD CLAIM = CORPUS POSITION × INTELLECTUAL FUNCTION × TEMPORAL RELATION × OPERATIONAL EMBODIMENT

A high-stakes claim about method formation is complete only when all four coordinates are stated or when the absent coordinate is marked.

1. Artifact identity ledger

A1 — The Register

Class: Corpus register and archival audit Method role: Topological map of the corpus Primary question: What exists, in what form and status, and what did each work contribute to method formation? Unit of analysis: Individual works, forms, apparatus, status, venue history, archive findings, and method-stage position Time logic: Functional order. Sequence follows contribution to method formation rather than composition or publication date. Evidence mode: Manuscripts, bibliographies, submission packages, file-system inspection, counts, and correction logs Singular contribution: Preserves the plurality of the corpus. It prevents works outside the governance spine from being treated as absent or irrelevant and makes form, status, and archival condition visible. Authority limit: It can establish inventory and documented status. It cannot by itself prove a causal intellectual lineage or identify every conceptual handoff.

A2 — How Agatha Became a Governance Method

Class: Developmental and interpretive spine Method role: Causal grammar of the research program Primary question: How did a concern with authority, mediation, proof, interruption, and legitimacy become an AI-governance method? Unit of analysis: Conceptual handoffs between fields, works, laboratories, operators, failure modes, and governance controls Time logic: Causal-functional sequence. Ground → launch → laboratory → executable governance → hardening. Evidence mode: Close interpretive reading across the corpus, with explicit bridge claims and a discipline caveat against retroactive conscription Singular contribution: Names the through-concept and explains why the stages belong to one program without turning every earlier paper into an AI-governance paper. Authority limit: It explains relation and purpose. It does not establish archive completeness, submission status, or implementation provenance.

A3 — Sixteen Years and Five Months

Class: Temporal calibration and comparative audit Method role: Temporal depth and compression test Primary question: How does the five-month Voice Operator lineage relate to the longer intellectual history that begins in 2010? Unit of analysis: Dated crossing points, recurrences, gaps, compressions, and primary-source corrections Time logic: Two clocks on one axis. A sixteen-year ring is compared with a five-month implementation lineage. Evidence mode: Primary thesis reading, chronology, comparison of register and genealogy, and explicit correction of unsupported bridge claims Singular contribution: Prevents implementation speed from being mistaken for intellectual origin. It distinguishes the recent encoding of a method from the longer recurrence of its structural commitments. Authority limit: Structural recurrence is not identical to documented intentional inheritance. It can support continuity claims without converting similarity into proven descent.

A4 — Voice Operator Specs — A Genealogy

Class: Implementation provenance and architecture audit Method role: Executable lineage and control provenance Primary question: How did prose rules, governance operators, personas, GPTs, agents, skills, and the paper pipeline inherit and transform one another? Unit of analysis: Operators, tokens, hashes, timestamps, bindings, instantiations, layers, gates, handoffs, and unresolved gaps Time logic: Version and layer history concentrated in 2026. It separates sequence from strata. Evidence mode: Opened files, hashes, GPT identifier timestamps, inheritance tables, source ledger, and confidence labels Singular contribution: Shows where the method became executable. It distinguishes operator change from persona binding, product instantiation, agent scope, and skill extraction. Authority limit: It can prove operational inheritance where artifacts exist. It cannot alone establish the full intellectual origin of the method or make unsupported stages direct evidence.

2. Unified method-formation map

ID	Formation function	Span	Register coordinate	Intellectual coordinate	Temporal coordinate	Operational coordinate	Status
M0	Deep structural substrate	2010	Position 00 once restored to the corpus horizon.	Pre-ground rather than the AI launch. It supplies a recurrent structural commitment, not an AI-governance vocabulary.	Primary temporal anchor and the basis for the sixteen-year ring.	A claimed substrate that the genealogy initially cited indirectly.	record + bounded synthesis
M1	Authority observed in situated practice	2010–2020	Stage I, with the published ethnographic and religious-studies record.	The ground. Refusal-to-cohere and sacred subversion become the primitive.	Long intellectual duration before the 2026 operator build.	Not part of the short implementation chain.	record
M2	Cultural rehearsal and the joint	2015–2025	Stage I-bis and Stage II. Position 13b is the fieldwork-to-screen joint. Position 26 turns specifically toward machine autonomy.	The screen archives and governance probes. Cultural materials become diagnostic test objects.	Part of the long formation, not the five-month operator chronology.	Supplies concepts and probes, not operator-version evidence.	record + synthesis
M3	AI inquiry launched through regimes of proof	2024–2026	Located among the cultural studies record and later method vocabulary.	Explicit launch of the AI endeavor.	Recent articulation built on older structures.	Conceptual predecessor, not yet an executable operator lineage.	record
M4	Laboratories, charge, and governance probes	2024–2026	Stages II and IV, with Sealed Card and Charging Objects entering the formal spine at Stage V.	The laboratory phase. Raw material → rendering → comparison → drift detection → control.	Accelerated experimentation near the 2026 encoding period.	Provides the problems that later controls mechanize.	record + synthesis
M5	The problem is lived and produced	2024–2026	Stage III and Stage IV.	Anti-fluency and production failure become direct governance requirements.	Explains why the 2026 formalization could be rapid.	Family A translates these pressures into prose operators and machine safety rules.	record + synthesis
M6	Governance becomes executable	2026-02–2026-05	Reflected in Stage V method papers and the Scholarly Paper Project OS, but not fully described by work status alone.	The turn from analysis to control.	The five-month compression line.	Core domain. Family A, Family B, instantiation track, four layers, and cross-layer skills.	record
M7	Failure modes and recursive hardening	2026	Stage V, the load-bearing method sequence.	The complex papers name what breaks after governance becomes executable.	A recent hardening layer rather than the origin of the underlying commitments.	Controls propagate into gates, markers, handoffs, refusal conditions, and pipeline nodes.	record + synthesis
M8	Convergence, limits, and falsifiability	2026	Stage VI. Position 63 is the limiting result and position 64 names the frame.	The regime of proof becomes an end-to-end evidence discipline.	The current endpoint still returns to earlier questions of authority and legibility.	Evidence labels, promotion gates, author verification, and source ledgers operationalize the limit.	record
M9	Operational externalization	live	Product and archive findings expand the corpus beyond papers and reveal multiple portfolio tiers.	Terminal form of the method, with a caveat that not every product or earlier cultural work belongs to the same lineage.	The live present where the long ring and short lineage meet.	The currently executing form. Skills become callable across platforms without duplicating content.	record + bounded synthesis

3. Typed hinges

The source documents use hinge language for different transitions. Normalization requires typed hinges rather than selecting one winner.

H0 — Structural precedent

Artifact: 2010 MA thesis Transition: A fixed triadic structure and anti-collapse commitment becomes visible long before the agent architecture. Why it does not compete with the others: This is a substrate claim, not the start of the AI endeavor. Evidentiary status: bounded synthesis

H1 — Fieldwork → screen worlds

Artifact: The Scythe Already in Motion Transition: Ethnographic inquiry and media analysis are performed as one argument. Why it does not compete with the others: This is the joint between Stage I and Stage II. Evidentiary status: record

H2 — Cultural analysis → machine autonomy

Artifact: The Machine’s Mirror Transition: The screen vocabulary turns specifically toward machine autonomy and self-authorship. Why it does not compete with the others: This is narrower than the fieldwork-to-screen joint. Evidentiary status: record

H3 — Cultural question → AI inquiry

Artifact: Agatha / regimes of proof Transition: The evaluation of mediated ritual authority is redirected toward AI-mediated production. Why it does not compete with the others: This is the launch of the AI endeavor, not the first appearance of the underlying method. Evidentiary status: record

H4 — Analysis → executable control

Artifact: Voice Operator families and Final Voice Operator Transition: Rules about voice, evidence, propagation, and refusal become machine-readable and enforceable. Why it does not compete with the others: This is the implementation hinge. Evidentiary status: record

H5 — Self-governance → admissibility limit

Artifact: Verification Without Trust Transition: The method specifies what self-generated evidence can legitimately establish. Why it does not compete with the others: This is the epistemic limit, not an origin. Evidentiary status: record

H6 — Arc → ring closure

Artifact: Pourquoi rêver encore ? against the 2010 thesis Transition: The corpus returns to the same Rivard objects with the later method’s instruments. Why it does not compete with the others: This recalibrates the shape of the history without erasing the functional arc. Evidentiary status: record + synthesis

4. Cross-artifact relations

From	To	Relation	Non-collapse rule
A1	A2	supplies corpus topology	A2 may interpret A1 positions but may not overwrite A1 form, status, or archival distinctions.
A2	A1	assigns developmental purpose	A1 gains explanatory context without turning all works into one genre or project.
A3	A1	extends and corrects temporal horizon	The 2010 thesis and terminal return must remain visible even when the register is ordered by function.
A3	A4	calibrates compression	Fast implementation may indicate prior structure, but recurrence must not be restated as proven direct descent.
A4	A2	demonstrates executable transformation	Conceptual handoffs become controls only where specifications, inheritance, or operational artifacts support the claim.
A4	A1	adds implementation and product classes	GPTs, agents, skills, and pipeline artifacts are not papers and require their own artifact classes.
A2	A3	distinguishes arc from ring	The intellectual sequence and temporal recurrence are complementary geometries, not rival chronologies.
A1	A4	provides test cases and method papers	Works can supply probes or concepts without becoming operator versions.

5. Preserved tensions

Arc versus ring

The record shows: A2 presents a developmental arc. A3 shows a return to the 2010 object and therefore a ring. Normalization: Use the arc for functional handoffs and the ring for temporal recurrence. Neither geometry replaces the other. Disposition: resolved by dimensional separation

One chain versus layered strata

The record shows: A4 finds that personas, GPTs, agents, native agents, and skills were flattened into one arrow chain. Normalization: Track function, layer, binding, instantiation, and version separately. A repeated name is not proof of identity, and a renamed function is not proof of change. Disposition: resolved by schema

Two origin substrates versus structural continuity

The record shows: A4 separates the operator lineage from Queen Keyport’s fictional substrate. A3 argues that the 2010 âme position may already contain the vacancy structure. Normalization: Retain both claims. The operator lineage and fiction have distinct documentary origins. The deeper structural relation remains a supported interpretation, not a collapsed genealogy. Disposition: preserved

Cultural work as probe versus independent scholarship

The record shows: A2 identifies cultural material as governance infrastructure while warning against retroactive conscription. Normalization: A work may contribute a concept, probe, or method function without being reclassified as an AI-governance paper. Disposition: governing constraint

Functional equivalence versus documentary proof

The record shows: A4 identifies stages whose confidence increased without stronger evidence. A3 shows the same issue around the 2010 bridge. Normalization: Record functional equivalence as synthesis. Reserve lineage claims for opened, dated, hashed, cited, or otherwise traceable artifacts. Disposition: governing constraint

Archive completeness versus register stability

The record shows: A1 was repeatedly corrected by files hidden under unexpected names, classifier categories, duplicate forms, and untracked submission histories. Normalization: Treat A1 as a versioned index with provenance, not a timeless final count. Counts remain tables by artifact class. Disposition: open operational risk

Fiction as designed experiment versus retained production artifact

The record shows: A4 finds that the Keyport two-arm structure is methodologically significant while the decisive break originated in a generation cutoff that was retained. Normalization: Preserve both the experimental reading and the provenance of the break. Do not restate an emergent artifact as an authored outcome. Disposition: preserved

Governance personified versus governance made portable

The record shows: Queen Keyport functions as a vacant office because embodiment risks converting governance into a usable service. Normalization: Describe Queen Keyport as a domain of answerability or coupling condition where the source supports it. Do not reduce the office to another capability agent. Disposition: preserved architectural distinction

6. Use protocol

1. Use A1 when the question concerns inventory, form, status, submission state, apparatus, or archive location.
2. Use A2 when the question concerns conceptual development, transfer between fields, or why a work matters to the method.
3. Use A3 when the question concerns origin, duration, recurrence, compression, or the relation between 2010 and 2026.
4. Use A4 when the question concerns operators, versions, provenance, GPTs, agents, skills, gates, or pipeline inheritance.
5. Require all four coordinates for claims that present PHAROS as a mature method, research program, product architecture, or intellectual lineage.

7. Normalized account of method formation

Cross-source synthesis. PHAROS formed through a recurring concern with authority under conditions where the grounds of legitimacy are incomplete, mediated, or deliberately obscured. The concern first appears in mythic, ritual, queer, and institutional contexts. Cultural analysis then rehearses how authority is staged and refused. Regimes of proof, charge, seam, glitch, and fluency convert those observations into portable concepts. AI-assisted production turns them into practical failures of continuity, drift, coherence, and evidence. The Voice Operator line makes the controls executable. The mature papers classify failure modes and establish an admissibility limit through checkable residue and falsification. The live pipeline, skills, agents, products, and client systems externalize parts of the method without making every artifact part of one lineage.

8. Tier audit

Source-grounded claims: artifact purposes, register stages, arc phases, the 2010 and 2026 temporal contrast, operator families, layer distinctions, pipeline inheritance, and stated caveats. Synthesis claims: the four-coordinate formula, the ten normalized formation functions, and the typed-hinge framework. Provisional claims: the deepest continuity between the 2010 triad and Queen Keyport or the present agent stack. Preserved uncertainties: archive stability, incomplete documentary bridges, the status of emergent fiction artifacts, and the distinction between structural recurrence and demonstrable descent.

Source artifacts

This atlas normalizes four source artifacts held outside this repository (see sourcePaths above): A1 — The Register, A2 — How Agatha Became a Governance Method, A3 — Sixteen Years and Five Months, and A4 — Voice Operator Specs, A Genealogy. A4’s raw form contains internal build identifiers, hashes, and timestamps from an active development environment and is deliberately not reproduced or linked here; everything above about A4 is drawn from how this atlas itself already treats it, not from the raw artifact.