Canonical URLs: Full §03 · Raw Markdown · Plain-Text Brief · ← §02 Brief · §04 Brief → · All §01–§09 Master Brief · All Articles · Raw Markdown Executive Synthesis: Problem: Entity resolution needs rich comparison structure, but raw contacts, companies, and products should not be sent to external AI providers. — Proves: The production EISA Four-Gate machine validates bounded acts over sealed evidence; Semantic Switch is an experimental alpha structure-preservation research module. — Evaluation Scope: Use the deterministic EISA core for judgment boundaries and evaluate Semantic Switch only against its declared synthetic-test scope. Lossless Parts (6 total, 100% unabridged): Part 1/6 (.txt) · Part 2/6 (.txt) · Part 3/6 (.txt) · Part 4/6 (.txt) · Part 5/6 (.txt) · Part 6/6 (.txt)
- Introduction & Thesis [Part 1/6]: # EISA Four-Gate Verdict Machine & Strict AST Parser The production EISA core determines which bounded acts earn constitutional standing; Semantic Switch is an experimental alpha module for structure-preserving entity-resolution research. **By Blayne Aina · Principal Systems…
- One Constitution, Two Proof Lanes [Part 1/6]: EISA applies the same rule to two very different proposal classes: The recognition spine is
ev_evidence →eisacase_sealed case → boundedEISA_ACT_V1acts → Four-Gate recognition → recognized artifact → immutableeisa_Room submission. - 1. Strict Representation Before Judgment [Part 1/6]: Subtopics: Gate 1 begins with
strict::parse; The exact ten-key act envelope; Record-relative RFC 6901 pointers; The profile bounds — A grammar cannot govern a transition if two parsers can assign different meanings to the same bytes. EISA does not deserialize submissions through an ordinary last-key-wins JSON map. — Spec/Table:version: ExactlyEISA_ACT_V1|stage: A declared stage whose family admits the opcode:FractureorForge|actor: Must match the authenticated caller, case-insensitively |opcode: One of the seven closed opcodes |operands: An object with exactly the opcode's declared keys |evidenceRefs: At most 16 strings naming Room-issuedev_handles; layout relations require exactly… — Key Rules/Claims:visit_mapchecks whether each incoming key already exists and rejects duplicates. |visit_f64refuses non-finite numbers. |de.end()rejects trailing non-whitespace matter after the root value. - 2. The Closed Grammar: Seven Opcodes, Four Layers [Part 1/6]: The opcode set is closed. New semantics cannot be introduced by sending a persuasive string. Dependencies point downward. Layer 0 observations feed Layer 1 pairs; Layer 1 pairs feed Layer 2 closure; Layers 1 and 2 support the Layer 3 partition. — Spec/Table: 0; Opcode:
BindEvidenceTag; Stage: Fracture; Exact operand keys:id,tag,coordinates; Constitutional role: Compare values at two distinct sealed-record coordinates using one positive or counter tag | 0; Opcode:RequestHumanReview; Stage: Fracture; Exact operand keys:id,reason,members; Constitutional role: Emit lawful non-operative review matter over at least one sealed record | 0;… - 3. The Evidence Alphabet: 17 Tags and 12 Comparators [Part 2/6]: Subtopics: The 47 registered fault classes — EISA's tags are not natural-language hints. Each positive or counter tag binds to one deterministic comparator. The two inferential tags are reserved for transitive composition and are refused in
BindEvidenceTag. — Spec/Table: 1; Evidence tag:PHONE_STRONG; Class: Positive; Comparator and exact predicate:DigitsEqual: both digit strings have at least seven digits and are identical | 2; Evidence tag:PHONE_ONE_DIGIT_VARIANCE; Class: Positive; Comparator and exact predicate:DigitsOneSubstitution: equal-length digit strings, Hamming distance exactly one | 3; Evidence tag:PHONE_MULTI_DIGIT_VARIANCE; Class: Positive;… - 4. Sealed Cases: Two Substrates, One Artifact Identity [Part 2/6]: Subtopics: File cases; Passage cases — An act may point only into a sealed case. EISA creates
eisacase_<32hex>by hashing the Room name, a NUL separator, and the case's RFC 8785 JCS record digest. TheartifactHashcarried by every act is that record digest; the Room-qualified case ID is the graph address. — Key Rules/Claims: prepares and validates the citation for the caller and task; | refuses handles from another Room or unrevealed collaborative-pass findings; | loads and reverifies the canonical view and live source; | cuts the exact Unicode-scalar passage using the handle's selector; | refuses textless gaps, empty selections, and… - 5. The Four-Gate Verdict Machine [Part 2/6]: Subtopics: Counterevidence is submission-complete; Transitive closure is a path proof, not a slogan; Partition verification is global across the admitted submission; Consensus identity removes narrative, not responsibility — Each act ends in one consequence class: A malformed whole submission is
inadmissible. A recognized submission reportsdecision: partition,review,relation, ornone; it does not acquire a blanket database-execution verdict. — Spec/Table: 1. Syntax and stage isolation; What the recognizer establishes: Strict JSON, exact keys, closed symbols, authenticated actor, unique IDs, bounded sizes, one act family, matchingartifactHash; What it refuses to let the proposal do: Smuggle duplicate fields, foreign opcodes, mixed stages, self-declared actors, or unbounded matter | ****2. — Key Rules/Claims:operative— a recognized act with a consensus identity. |lawful_non_operative— a valid review request, which creates no operative identity claim. |anti_knowledge— a valid rejection of other matter. |typed_fault— preserved residue with a fault code and no standing. - Exhibit A — A Valid Contact Submission [Part 3/6]: The following synthetic submission binds two email relationships, asserts a pair, and proposes a complete partition over three sealed records.
artifactHashis the case's bare digest. Values are resolved from the sealed case, not trusted from the prose. - Exhibit B —
counterevidence_suppressed[Part 3/6]: Suppose the same submission also admittedact_bind_counter_03, bindingDOMAIN_ROOT_MISMATCHon the pair, butact_dup_101_102omitted it fromcounter. Its recognized-act report entry is: The refusal is not a score penalty. It is the absence of standing for that pair. - 6. Proof Lane A: Spatial Relations Without Visual Guesswork [Part 3/6]: Subtopics: From PDF operators to integer evidence; The
CONNECTSproof; Branch labels enter relation identity — A flowchart can have perfect OCR and still produce a false graph. Recognizing the words inside two boxes does not establish which arrow connects them, which direction the arrow points, whether the tip touches a node or an enclosing frame, or whether “Yes” and “No” label the same… — Key Rules/Claims: Provenance. All three records come from one Room, view digest, source revision, parser digest, and page. Each record's committedLayoutFact.nodeis exactly the node its evidence handle cites. - Exhibit C — An Operative Layout Relation [Part 3/6]: This is the layout portion of a synthetic recognized-act report. The frame fields are observations in the current report shape, not an agent-authored
layout_proofobject. - Exhibit D — Two Distinct Partition Failures [Part 3/6]: The first splits records already proven equivalent. The second groups records without proving connectivity. They are different defects and receive different faults.
- 7. Proof Lane B: Semantic Switch First, Then the Entity-Resolution Stack [Part 4/6]: Subtopics: Stage 0 — Semantic Switch: protected computation before external AI; Stage 1 — qBlock: candidate nomination over protected records; Stage 2 — EISA Fracture: identity adjudication with dissent preserved; Stage 3 — EISA Forge: canonical state is a separate decision; Stage 4 — Shield: the baseline moves forward or does not move — Entity resolution adds a different danger: the evidence can contain direct identifiers, and the final merge can be difficult or impossible to reverse through the destination's ordinary product operations. — Spec/Table:
SS01: Privacy-first stage responsibility. Semantic Switch transforms native records before provider-exposed qBlock, Fracture, Forge, or Shield computation. Separate stages own identity judgment and destination mutation. |SS02: Whole blocks remain the unit of work. Disjoint IDs, member existence, non-overlap, and in-block edge evidence are checked; no pair-request fragmentation or cross-block… — Key Rules/Claims: every surviving value names its source record and field pointer; | provenance and validity intervals remain inspectable; | declared field quorums and canonical claims decide support; | unsupported values do not win through confidence alone; | a conditional reversal plan records pre-state and post-state commitments… - 8. From Duplicate Records to Evidence-Linked Relationship Graphs [Part 4/6]: The commercial value is not “AI deduplication” as a slogan. It is a defensible identity substrate on which relationship paths can be inspected. A six-hop query can be a useful exploration horizon. It is not proof of influence, consent, or authority.
- 9. The 2026 Research Landscape: Proposal Quality Is Not Transition Standing [Part 5/6]: The current research supports a hybrid division of labor: models can improve nomination, comparison, and review, while deterministic systems must decide which declared predicates actually hold. The sources do not endorse AINA, and their results need their original scopes. — Spec/Table: Teofili et al., Can we trust LLM Self-Explanations for Entity Resolution?, June 2026 preprint; What it establishes: Across three LLMs, ten datasets, and multiple prompting strategies, self-explanations were often unstable, weakly faithful, and poorly aligned with counterfactual evidence.
- 10. The Continuous Wire into the Room [Part 5/6]: EISA recognition and Room admission share one path.
recognizepreviews a verdict;submitruns the same recognition logic and records the result. The caller is authenticated, the Room is admitted, the case is loaded and rechecked, and every evidence handle used by a… — Key Rules/Claims:byand nativesession; |roomand optionaltask; | the case's bareartifactHash; | the complete recognition result; | submission time. - Exhibit E — The Current Room Ledger Envelope [Part 5/6]: This synthetic
showresponse illustrates the current ledger shape with one operative binding. The per-act report remains insideresult. The immutable node is constitutional proof matter for the next boundary. It is not a destination receipt. - Bring One Bounded Protected-Computation Problem [Part 5/6]: The Semantic Switch & Entity Resolution Challenge applies this constitutional layer to one bounded contact, company, product, or other duplicate-record workflow.
- The Article 03 Claim Card [Part 6/6]: Key Rules/Claims: Mechanism: | Semantic Switch Protected Computation + EISA Constitutional Act Recognition | (native-data minimization, relation-isomorphic fictional records, | qBlock candidates, Fracture, Forge, local exact restoration, Shield, | strict parsing, sealed cases, pointer law, deterministic comparators, | Four-Gate recognition, layout proof, and Room ledger) | Failure it governs: | Native identifiers exposed to external AI computation; | comparison structure destroyed by destructive redaction; | ambiguous or schema-valid model prose becoming authority; | fabricated coordinates and geometry; suppressed counterevidence; | broken transitive claims; incomplete or contradictory…
- Before the Room Can Govern, the Act Must Mean One Thing [Part 6/6]: Return to the three agents voting for a merger. The conventional system can report that the vote was unanimous, the JSON matched a schema, and the judge was 97% confident.
- Frequently Asked Questions [Part 6/6]: Subtopics: Why is JSON Schema or constrained decoding insufficient to govern AI actions?; How does EISA prevent an AI agent from suppressing conflicting evidence?; Does
consensusIdentityprove independent agreement?; Why does Semantic Switch run before qBlock?; Is Semantic Switch anonymization?; How does EISA verify flowchart connections without trusting VLM prose?; Does an operative EISA verdict execute the proposed transition? — Constrained decoding establishes shape. It does not resolve record-relative pointers against a sealed case, run deterministic comparators on the values, reject faulted dependencies, force counterevidence into the decision, or prove that a partition is complete, connected, and… - Continue the Series [Part 6/6]: Key Rules/Claims: Series index: From Proposal to Consequence: The AINA Transition-Governance Series | Previous: Article 02 — Zero-Trust Document Ingestion and Server-Cut Evidence | Next: [Article 04 — The Room: Why…
- Primary 2026 Sources [Part 6/6]: Key Rules/Claims: Teofili et al. — Can we trust LLM Self-Explanations for Entity Resolution?,
arXiv:2606.01210| Papadakis et al. — AgenticER: the next frontier in Entity Resolution,arXiv:2607.27435| [Smith et al. - Publication Metadata [Part 6/6]: Title tag: Semantic Switch and EISA: Protected Computation and Entity Resolution Meta description: Semantic Switch protects native data before external AI computation, then qBlock, Fracture, Forge, and Shield turn fictionalized records into governed entity state.