A.6.B — Boundary Norm Square (Laws / Admissibility / Deontics / Work‑Effects)
Type: Architectural (A) Status: Stable Normativity: Normative (unless marked informative) Placement: Part A → A.6.B (matrix module; referenced by A.6 cluster overview) Builds on: E.8 (authoring template), A.6.0 (
U.Signature), A.6.1 (U.Mechanism), A.6.3 (EntityOfConcern-preserving episteme construction), E.17.0/E.17 (MVPK + “no new semantics” faces), A.7 (EntityOfConcern and Description-episteme boundary; specification-use and publication-carrier distinction), A.2.3 (promise content when contract language is current), A.2.8 (U.Commitment), A.2.8.PER (subject pattern selected by the permission-word branch), A.2.9 (U.SpeechAct), E.10.D2 (EntityOfConcern and Description-episteme boundary; specification-use and refinement discipline), E.10 publication face, form, unit, and carrier discipline Purpose (one line): Provide a canonical four-way classification of boundary statements (L/A/D/E), rules for writing each quadrant, and rules for explicit cross‑quadrant references to support later boundary evolution and audit.
Use this when. Use A.6.B when boundary wording mixes definitions, runtime entry conditions, generic prescriptions, individual duties or grants, actual work or evaluation results, and evidence.
First useful move. Identify what the reader needs to decide or understand about the boundary. Use §4.4 to select the claims and grounds needed for that answer; split mixed claims, classify their jobs, and name consequential dependencies by ID or canonical location.
What changes in practice. Laws, gates, governance, and work-and-evidence claims can change without silently changing one another; publication faces reuse the canonical claims instead of restating them.
Ordinary non-use boundary. A one-off local sentence with one recoverable logical job needs the correct quadrant, but it does not need a Claim Register or a separate publication form. Add those only when reuse, decision, audit, or cross-face citation needs them; §6.1 requires any material dependency to name the claim by ID or canonical location.
A.6.B:0 — Conventions
Keywords. The key words MUST, MUST NOT, SHOULD, SHOULD NOT, MAY, and SHALL have their RFC 2119/8174 meanings only when capitalized as shown.
Quadrant labels. This pattern uses the classification labels L / A / D / E as statement quadrants:
- L — Laws & Definitions
- A — Admissibility & Gates
- D — Deontics & Commitments
- E — Work‑Effects & Evidence
These labels are claim-classification labels for statements, not MVPK face designators and not pattern identifiers.
Statement identifiers (recommended). Classifiable statements SHOULD be given stable IDs with a quadrant prefix: L-*, A-*, D-*, E-*. Other sections and views SHOULD reference these IDs or the canonical claim locations. A faithful explanation may accompany the reference under §6.1.
Non-collision note (informative). The A-* prefix here is “Admissibility”, not Part-A numbering and not MVPK’s AssuranceLane face designator. If this is a readability hazard in your program, prefer an explicit G-* (“Gate”) local convention while keeping the quadrant name “Admissibility”. Also avoid introducing single-letter mnemonics for MVPK face designators inside this cluster; spell face designators in full to reduce collisions.
Atomic claim. An atomic claim is a sentence (or bullet) that performs exactly one logical role and is classifiable under exactly one quadrant. If a sentence mixes roles, it is not atomic and MUST be split before it can be classified.
Claim job and adjudication basis. A quadrant identifies what a boundary claim contributes. Its adjudication basis states how that claim can be settled. Definitions and declared entry predicates are both description content. Generic normative content is recovered from its source; an individual duty or grant requires its actual constitutive basis. An occurrence or evaluated-result claim requires the facts specified by its subject predicate.
Deontic claims. Writing a D-* claim records either generic normative content or a claim about an individual duty, commitment, or grant. When the wording is about permission, use the permission-word branch in §8.4.1 to recover the object, what makes it obtain, and the evidence needed before reliance. The instituting act and any compliance finding have their separate claims.
Modality. Preserve whether a sentence defines, prescribes, permits, hypothesizes or reports. Use that meaning together with the claim’s job under §4; a modal word alone leaves the classification unresolved.
A.6.B:1 — Problem frame
Boundary descriptions routinely mix four distinct claim families: definitions are written as obligations, runtime gates are hidden inside laws, governance talk is assigned to “the interface”, and operational-result “guarantees” are asserted without the actual Work, evaluation, or observation needed to settle them. This makes substitution unclear and leaves apparent auditability without claims that can be adjudicated.
The classification connects the strata already distinguished in FPF: Signature and Mechanism; the EntityOfConcern, claim-bearing episteme and carrier; and views under viewpoints. It lets a reader identify each claim, how it is settled and which other claims its use depends on.
A.6.B:2 — Problem
When authors cannot reliably answer two questions—
- “What does this claim contribute: meaning, an admission condition, normative content or a deontic position, or an occurrence/evidence account?”
- “What grounds are needed to settle this claim?”
—then boundary statements drift across layers, faces fork semantics, and “compliance” becomes a matter of interpretation rather than a property that can be checked.
A boundary needs a minimal, stable classification that:
- classifies every atomic statement into a unique quadrant, and
- requires any cross‑quadrant dependencies to be explicitly referenced.
A.6.B:3 — Forces
| Force | Tension |
|---|---|
| Precision vs readability | Predicate‑style constraints reduce ambiguity; narrative helps adoption. |
| Evolvability vs enforceability | Stable laws should not embed volatile runtime gates; governance needs enforcement hooks. |
| Auditability vs simplicity | Evidence makes claims adjudicable; evidence also introduces operational design obligations. |
| Local meaning vs reuse | Boundaries must be local; reuse must be explicit via IDs and references, not duplicated prose. |
A.6.B:4 - Solution — the Boundary Norm Square
A.6.B:4.1 - Four claim jobs and their grounds
The Boundary Norm Square distinguishes four jobs in a boundary account. A definition can remain fixed while a gate changes; a duty can obtain while its bearer fails to perform it. An earlier observation may still be valid evidence when a changed admission condition leads to refusal. Establish the observation’s fitness for the new check, then apply that condition to decide admission. Keeping those claims distinct makes their dependencies available for a change or audit decision.
Classify what the sentence asserts, then establish its grounds. For example, a declared entry predicate belongs in A and the result of applying it to one request belongs in E. A claim that a grant currently permits an action belongs in D; the instituting act, an evaluation of that grant and an actual exercise of it have their own E claims. The predicate definition and the individual claims remain description content, with different subjects and adjudication conditions.
A.6.B:4.2 — The square
| Quadrant | Job of the claim | What settles it |
|---|---|---|
| L — Laws & Definitions | Defines vocabulary, a semantic constraint, invariant or typing rule. | The declared definition or assumptions and the corresponding inspection, proof, type validation or model reasoning. |
| A — Admissibility & Gates | Declares the condition for a named action or application to be admitted. | The admission rule and its input meanings. A particular satisfaction or admission-decision result is a separate E claim. |
| D — Deontics & Commitments | States generic normative content or a claim about an individual duty or grant. | The normative source for generic content; the subject pattern’s constitutive conditions and actual basis for an individual relation. |
| E — Work‑Effects & Evidence | Reports an occurrence, evaluated finding, observed condition or evidence support. | The actual facts and subject-specific conditions for that claim, with evidence support when the receiving use requires it. |
Clarification (classify the claim by its job, not by its subject pattern).
- Classify the atomic claim by what its sentence states and by the conditions that let a reader decide it.
- The pattern governing the referenced object supplies its predicate and obtaining conditions; it does not choose the claim’s quadrant.
- When permission wording is present, use the single permission-word branch in §8.4.1. It separates the possible jobs of that wording without inventing a common “permission result” kind.
Normative rule (single quadrant). Each atomic boundary claim MUST be assigned exactly one job under L/A/D/E. Recover an unresolved meaning before assigning it. A condition or reference within one coherent claim does not by itself add another asserted claim.
Normative rule (mixed claims). A conforming boundary text SHALL separate independently asserted claims with different jobs. For example, a duty to enforce an entry rule and a report of one enforcement result have separate D and E claims referring to the A rule. Preserve the logical connection when splitting the wording.
A.6.B:4.3 — Canonical placements in the Signature Stack
The quadrants have canonical placements in the boundary stack:
- L → Signature layer:
U.Signature.Laws(and mechanism‑local semantic laws if present). - A → Admission layer: keep the condition in its governing rule’s canonical source. Mechanism applications use
U.Mechanism.AdmissibilityConditions; A.21 gates use the applicable profile rule. - D → Deontics & Commitments layer: atomic claims about a generic prescription or about one separately obtaining individual duty, recommendation-as-duty, prohibition, or commitment. When permission wording is live, §8.4.1 decides whether its claim also belongs here.
- E → Work-Effects & Evidence layer: truth-conditional claims whose satisfaction requires actual work, evaluation, observation, or produced carriers.
A publication face MUST NOT introduce new semantic claims outside this L/A/D/E-classified claim set. E.17 (MVPK) enforces this rule for publication faces and their selected profiles.
A.6.B:4.4 - Develop a boundary explanation for its receiving use
Start with the boundary passage, its available sources and the reader’s question. The square helps recover what each claim contributes and what would settle it. The explanation connects the claims needed for the reader to interpret, use or change that boundary.
1. Recover the question and the claims already made. Establish which understanding or decision the text must support. Interpreting a metric, deciding whether a request may enter and determining whether a duty was fulfilled need different answers. Recover the draft’s asserted content, modality and scope. If a word such as guarantees leaves the intended assertion open, use A.6.C:4.3 with the author or source to resolve that meaning. Keep a consequential unresolved meaning explicit until its source is available.
2. Establish the grounds of the answer. Split independently asserted jobs and classify them through §§4.1–4.3. Use the detailed procedure and conditional permission-word branch in §8.4.1 where needed. Recover the definition for a meaning claim, the declared predicate for an entry condition, the normative or constitutive basis for a duty or grant, and the facts for an occurrence or finding. Preserve the source’s conditions and qualifications. A missing ground requires a targeted return to the source or the responsible practice. If the work calls for a new rule or commitment, develop and establish it through its governing practice before treating it as current boundary content.
3. Select the claims and detail on which the reader’s answer depends. Work backward from that answer. A reported measurement needs its quantity, population, conditions and interpretation; judging fulfilment of a commitment additionally needs that commitment and its comparison rule. Follow such dependencies until the remaining premises are supplied by the text or by references the intended reader can obtain and understand. Use §6 for their relations. Include a detail when changing it could change the answer, when it explains why a step follows, or when the reader needs it to perform the stated use. A complete four-quadrant account is warranted when the question crosses all four jobs. Keep other valid claims in their canonical sources for their own uses.
4. Compose the explanation around that reasoning. State the answer in the reader’s working terms, connect its supporting claims, and make the conditions that limit its use recoverable. Place a definition beside the claim that needs it and an observation beside the inference it supports. IDs or canonical locations preserve material dependencies under §6.1; the accompanying explanation supplies their meaning for this reader. Use a Claim Register or separate Tech and Plain forms when the reuse or publication conditions call for them. Selection and recomposition preserve the source claims; a new semantic constraint first needs its own canonical claim.
5. Return the text to the reader’s work. Follow the intended interpretation or decision using the explanation. Identify any consequential premise the reader would have to invent, obtain that premise or state the precise unresolved question, and revise the affected connection. Consider a relevant change of condition: show which claim must be reconsidered and which grounds remain usable. Stop when the text supports the assigned understanding or decision with its limits. A later question can require a different selection from the same canonical claims.
The cases in §8.4 show these authoring choices. Their claims about latency, coaxiality and project entry are the material being explained; the reusable operation is selecting and connecting that material for a reader’s question.
A.6.B:5 — Quadrant specifications
The following normative specifications state what each quadrant is for, how it is written, and what it must not contain.
A.6.B:5.1 — Quadrant L: Laws & Definitions
Intent. State truth‑conditional content: definitions, invariants, typing and well-formedness constraints, equational laws.
Adjudication. In‑description: can be checked by inspection, proof, type validation, or model reasoning.
Canonical form. Definition: / Invariant: / predicate‑style constraints using “is / iff / for all”.
Prohibitions.
- An
L-*statement MUST NOT contain RFC deontic keywords (MUST, SHALL, SHOULD, or MAY) as operators inside the law or definition itself. - An
L-*statement MUST NOT encode runtime gate predicates (those areA-*). - An
L-*statement MUST NOT assert evidence availability or measurement outcomes (those areE-*).
A.7 EntityOfConcern binding. Identify what the L-* definition, invariant,
or typing rule concerns, separately from the episteme carrying the claim. Its
EntityOfConcern need not itself be a description and may be Work.
Typical dependence. A-* and E-* claims may reference L-* IDs or canonical locations for vocabulary, metric definitions, and invariants needed for interpretation.
A.6.B:5.2 — Quadrant A: Admissibility & Gates
Intent. Specify when a named action or application is admissible under its governing rule: for example, runtime entry predicates, validity gates, and applicability checks that require context or execution environment. An A-* predicate may consume a separately established result as one input, but it does not create or settle that result. If the sentence uses permission wording, choose its job with the branch in §8.4.1.
Applicability and admissibility. Signature Applicability scopes intended use and bounded context; a mechanism’s runtime entry predicates belong in U.Mechanism.AdmissibilityConditions as A-*. If prose says “clients must satisfy the applicability”, first recover whether it means the declared scope or an entry condition. State the generic D-* prescription, or the independently instituted individual duty, separately and reference the condition it requires. Classify that condition as A-* only when it is an entry predicate.
Adjudication. Inspect the declared entry rule and its inputs to establish the A claim. Evaluate that predicate at the named action’s entry or stated application point to obtain a particular satisfaction or gate-decision result; report that result separately in E.
Canonical form. Predicate style, e.g.:
- “A request is admissible iff …”
admissible(x) iff P(x)(conceptual form; no particular syntax is required)
Prohibitions.
- An
A-*statement MUST NOT be placed inU.Signature.Laws. - An
A-*statement MUST NOT use RFC deontic keywords as if it were an agent obligation. (It is a gate predicate, not a duty.) - An
A-*statement MUST NOT claim that evidence exists (that isE-*) or that someone must enforce the gate (that isD-*).
A.7 EntityOfConcern binding. An A-* claim concerns the admission predicate
for a named action or application, not a client’s duty to comply with it.
Keep that predicate distinct from the episteme carrying the claim.
Required references. If an A-* predicate relies on defined terms or invariants, it SHOULD reference the relevant L-* IDs or canonical locations (or at minimum the signature that defines them).
A.6.B:5.3 — Quadrant D: Deontics & Commitments
Intent. State one atomic deontic claim. A generic prescription states what one policy or other normative episteme requires; it does not create an individual duty bearer or commitment occurrence. A claim that one actual System or separately governed party has that duty instead cites one separately obtaining A.2.8 U.Commitment. When a sentence sounds permissive, use §8.4.1; only its Grant or norm row enters D. Writing the D-* sentence alone neither institutes a relation nor establishes compliance.
Adjudication. For a generic prescription, inspect the normative source, its applicable rule content, scope, and current edition. For an individual duty, apply A.2.8 to the separately obtaining commitment and its actual basis. The wording alone establishes neither obtaining nor compliance.
Canonical form. First choose the route. A generic D claim names the normative episteme and the rule content being stated, without inventing an individual bearer.
An individual-duty D claim names the actual bearer and U.Commitment; a system-role kind or assignment may be a rule ground but is neither bearer nor duty.
A responsibility claim uses an admitted domain responsibility predicate and its actual participants, or returns its missing governor. A permissive-looking word does not by itself select D; use §8.4.1 for the grant route.
Examples:
- Generic: “
APIEntryPolicy-v4requires covered clients to satisfyA-….” No individual commitment is asserted. - Individual: “Actual bearer
ClientIntegrator-Ahas commitmentCOM-17to satisfyA-….”
Canonical assertion (recommended; lintable). Use a CommitmentAssertion only when an individual-duty claim must be reused or audited. It concerns one separately obtaining U.Commitment and makes explicit:
entityOfConcernRef, resolving to oneU.Commitmentoccurrence, and theD-*claim ID;- exactly one actual bearer branch:
dutyBearerSystemRefordutyBearerPartyRef; - non-empty
dutyReferentRefsand any actual counterparties; - the A.2.8
DeonticModalityToken, scope, and validity window; - the current constitutive policy, individualizing rule, and actual instituting basis required by that rule; and
- evidence-claim or carrier references only when the receiving reliance or adjudication needs them.
The assertion states and supports a claim about the relation. Fields, publication or evidence have a constitutive role only when the identified rule requires that contribution and its subject pattern’s test is satisfied.
Prohibitions.
-
A generic
D-*statement MUST NOT invent an individual bearer or commitment; name its normative source and rule content.An individual-duty
D-*statement MUST NOT use “the system, service, interface, or specification” as a vague subject; name the actual duty-bearing system or separately governed party andU.Commitment, with an assignment only when the constitutive rule uses it as a ground.Use
A.6.Cwhen ambiguity in promise, utterance, approval, guarantee, or agreement-like boundary wording changes the claim’s interpretation or use. -
A
D-*statement MUST NOT give a referencedL-*orA-*predicate a different meaning. It MUST reference that predicate by ID or canonical location when its meaning depends on it; a faithful explanation may accompany the reference. -
A
D-*statement MUST NOT represent a duty, commitment, or grant as a law or assert that writing the claim alone makes it obtain.
A.7 EntityOfConcern binding. A generic D-* claim episteme concerns the normative rule content it states. An individual D-* claim concerns the duty, commitment, or grant named by its content and does not substitute for that object. When permission wording is live, the branch in §8.4.1 names the subject pattern and the obtaining or non-obtaining test.
Required references.
- If a
D-*claim states a prescription or duty to comply with a gate, it MUST reference the relevantA-*ID(s) or canonical location(s). - If a
D-*statement is meant to be auditable, it SHOULD reference theE-*claim(s) that provide evidence and the carrier classes involved.
A.6.B:5.4 — Quadrant E: Work‑Effects & Evidence
Intent. State an actual occurrence, evaluated finding, observed condition or evidence-support claim. Settle it from the actual work, evaluation, observation, or carrier conditions required by its predicate.
Adjudication. From actual work, evaluation or observation under the claim’s subject predicate. A record can support that judgment through its evidence relation; its presence alone does not settle the claim.
Canonical form. Write the ordinary result first, then make recoverable only what settles it:
- the predicate and object that the claim concerns;
- the participants, work or evaluation occurrence, scope/window, comparison frame, and other conditions required by that predicate; and
- the evidence or source-use relation when the receiving use needs that support; identify the supporting carrier only when carrier identity matters to that use. A carrier may support the claim but does not create the work, effect, or finding.
When permission wording is current, use the branch in §8.4.1 for the occurrence or finding, its failure test, predicate, and subject-pattern locator; do not repeat that subject-question catalogue here.
Prohibitions.
-
E-*statements SHOULD NOT use RFC deontic keywords; they report adjudicable results rather than obligations. -
An
E-*statement MUST NOT hide a gate predicate; gate predicates areA-*. -
For a claim about a particular dated Work, first recover each actual performer through A.13 and let A.15.1 independently admit the Work. Ordinary metonymy is usable when the capable performer remains recoverable; an interface, record or publication does not supply a performer merely by being named.
Add an A.2.1 assignment reference and F.6 only when this claim or its receiving use expressly consumes precise assignment-bound attribution through the same obtaining A.13 assignment; F.6 identifies neither the assignment nor the performer. Missing or failed F.6 alone does not overturn independently admitted Work.
If the attribution finding questions a premise consumed by Work admission, recheck that premise and the dependent claims. If a duty or commitment is also claimed, express it separately in
D-*.
A.7 EntityOfConcern binding. An E-* claim episteme concerns the work effect, evaluated finding, evidence relation, or carrier condition named by its predicate. A record or carrier is a separate object and becomes the concern only when its existence or condition is itself the claim.
Required references.
- If the result depends on an entry predicate, the
E-*statement SHOULD reference the relevantA-*claim. If it depends on an actual gate decision, reference the separateE-*claim about that A.21 decision result. - If another object is needed to settle the predicate, reference that object’s subject pattern without importing its quadrant.
- If evidence is used for reliance, recover the independently established support or evidence-use relation and its basis. Use A.2.4 for the first evidence-use or status-use classification; use A.10 when the source-to-use account must remain replayable, and G.6 when an addressable provenance path is needed. Carrier presence alone is not support.
A.6.B:6 — Cross‑quadrant link discipline
The square also makes dependencies between claims inspectable. A material dependency MUST be explicit by claim ID or canonical location.
A.6.B:6.1 — Explicit reference rule
If a claim’s meaning materially depends on another L/A/D/E-classified claim, that dependency MUST be represented as an explicit reference to its ID or canonical location. A faithful restatement may explain the referenced content; it does not define an alternative rule.
A reader can then follow the explanation back to its governing claim and reconsider dependent uses when that claim changes.
A.6.B:6.2 — Canonical cross‑quadrant dependency patterns
These patterns are valid (and common). The square becomes operational when these links are used systematically.
A.6.B:6.2.1 - (D → A) Prescription-or-duty-to-gate linkage
When governance says that a gate must be satisfied or enforced:
- generic
D-*: “Policy-Prequires covered subjects to satisfy or enforceA-*”; or - individual
D-*: “Actual bearerShas commitmentCto satisfy or enforceA-*.”
This separates what is admissible (A) from generic normative content and any separately instituted individual duty (D). If responsibility is also claimed, state its admitted direct domain predicate or missing governor rather than inferring it from the duty.
A.6.B:6.2.2 - (E → A) Evidence-for-gate linkage
When a claim reports actual observability of a gate decision:
E-*: “For the stated rejection or acceptance due toA-*, carrierCwas produced or observed under conditions …”
This separates gate semantics (A) from the actual observation or result (E). A prescription to produce or retain a carrier belongs in D; an expectation or plan alone supplies no actual E result.
A.6.B:6.2.3 - (D → E) Prescription-or-duty-to-evidence linkage
When governance prescribes evidence production, retention, exposure, or a
measured property and a separate E-* claim states the related result:
- generic
D-*: “Policy-Prequires covered subjects to retain or expose carrier classCused byE-*…”; or - individual
D-*: “Actual bearerShas commitmentCMTto meet the criterion evaluated byE-*under the stated exclusions …”
This separates a generic prescription or individual duty (D) from its adjudication (E).
A.6.B:6.2.4 - (A/E → L) Semantic grounding linkage
When a gate predicate or measurement relies on definitions or invariants:
A-*/E-*referencesL-*that define terms or metrics.
This prevents “metric drift” and “definition drift” across views.
A.6.B:6.2.5 - (D → L) Governance-to-definition linkage
When a generic prescription or individual obligation relies on precise term or metric meanings:
D-*referencesL-*that define the terms or metrics it uses.
This keeps governance text from accidentally redefining semantics in prose.
A.6.B:6.2.6 - (A → D/E) Normative or evaluated input to a gate
When entry depends on a current grant or a separately established finding, reference that D or E claim and state how its subject enters the predicate. For example, A-CAL-01 in §8.4.5.4 requires the grant stated by D-CAL-01 to cover the plan and time, together with the required conflict finding. The grant, finding and entry decision retain their own obtaining conditions. A changed input reopens the admission evaluation that consumed it.
A.6.B:6.3 — The “triangle decomposition” for mixed sentences
Normative rule (decomposition). A conforming boundary text SHALL decompose any mixed sentence into its claim jobs. Recover the predicate and modality of an observability clause before assigning its quadrant. The triangle applies when the source states all three:
- A: admissibility predicate (
A-*) - D: generic prescription or individual-duty claim referencing the gate (
D-* → A-*) - E: actual observation or result referencing the gate, with carriers when needed (
E-* → A-*)
Keep a prescription or individually instituted duty to produce evidence in D. Preserve an expectation or plan as such; leave the actual E result absent until its basis exists. This repairs mixed validity, authorization, compliance, audit, and security claims without adding an unsupported third claim.
A.6.B:6.4 — Ground each dependency
Reusable laws retain their declared meaning, while an admission predicate can consume a normative position or an evaluated result. For each material link, state what it supplies: meaning, an input condition, an evaluation target or evidence support. Settle each depended-on claim under its own rule.
These are cross-quadrant restrictions. They do not prohibit same-quadrant
references, such as L-* → L-* or E-* → E-*.
Normative rule (dependency grounds).
L-*claims MUST NOT depend on or referenceA-*,D-*, orE-*claims (except for purely informative notes marked informative).- An
A-*claim MAY reference aD-*claim for the normative content, duty or grant used by its entry predicate, and anE-*claim for a separately established finding or observation. It MUST identify the input and required scope or currentness under §5.2; deciding that entry is permitted requires the predicate to be satisfied.A-*may referenceL-*for defined terms or invariants. - An
E-*claim MAY reference aD-*claim as its evidence target or to identify the norm, duty, or grant used by its evaluation. TheE-*result MUST satisfy its own predicate and the actual-work, evaluation, or observation requirements of §5.4; theD-*claim alone establishes no result or compliance. (E-*may also referenceA-*for conditioning andL-*for metric or term meanings.) D-*claims MAY referenceL-*,A-*, andE-*claims when needed, and SHOULD do so by ID or canonical location rather than restating content.
Rationale (informative). The L restriction keeps the reusable definition or law independent of a particular gate application, deontic claim or observation. Expose a varying policy or state as a declared parameter; bind it where the condition or claim uses it.
A and D may refer to one another for different reasons: a duty can require enforcing a gate, while that gate consumes a separately obtaining grant. Examine those reasons rather than treating every reference as one ordering relation. A grant justified only by a gate result that itself assumes that grant would leave both claims ungrounded.
Evidence and results may report on normative claims or evaluate conduct against them under their own predicates.
A.6.B:7 — Mini-register: Claim Register (informative, recommended)
A Claim Register is a drift‑control device that lists every classifiable statement verbatim with classification metadata. It is not a new meaning authority.
| ID | Quadrant | Statement (verbatim) | Canonical location (section or publication unit) | Stack layer | A.7 primary layer | viewRef | viewpointRef | References | Notes |
|---|
Guidance (informative):
- The Statement cell should contain the normative text as authored (copied by value), not a paraphrase.
- Canonical location should point to the one place the statement “lives” (e.g.,
Signature.Laws,Mechanism.AdmissibilityConditions,TechCard.NormsCommitments,Evidence.Carriers), so other faces can cite its ID or canonical location. - Stack layer should be one of
{Signature, Admission, Norms-and-commitments, Evidence-and-carriers}to make classification auditable. - A.7 primary layer identifies the claim’s EntityOfConcern, with its kind when needed. The concern may itself be a Description episteme or publication carrier; it remains distinct from the episteme carrying the claim.
- viewRef and viewpointRef are used when the corresponding view or viewpoint identity matters under E.17;
U.Viewmembership requires E.17.0 conformance. - Use References for explicit cross‑quadrant links: for example, which
D-*states a requirement or individual duty to enforce whichA-*predicate, whichE-*reports the result or evidence used to adjudicate which commitment, and whichL-*defines a metric used byE-*. Add external standards or policies where applicable.
A.6.B:8 - Archetypal Grounding (Tell–Show–Show)
Informative. Examples for learning the square; they do not add requirements beyond A.6.B:10.
A.6.B:8.1 - Tell (universal rule)
Decomposing every normative statement into atomic claims, classifying each under exactly one quadrant of the Boundary Norm Square, and expressing cross‑quadrant dependencies by claim ID or canonical location rather than paraphrase supports later boundary evolution and audit.
A.6.B:8.2 - Show #1: Effect signature vs handler (post‑2015 effect systems)
An effect interface declares operations; handlers supply computation behavior. To describe a deployed service using that interface, recover the additional semantic, admission, normative and observation claims that its use needs:
- L: defines the operation vocabulary and the semantic laws required of this service’s realizations.
- A: defines when the operation is admissible (runtime guard predicates).
- D: states who must enforce guards and what the provider commits to (operator and implementer duties; SLAs).
- E: ties “what happened” to observable carriers (traces, logs, metrics, and events) so commitments can be adjudicated.
The square prevents accidentally writing handler obligations as laws or treating observability as a definition.
A.6.B:8.3 - Show #2: ML evaluation protocol boundary (reproducibility discipline)
A published “evaluation protocol” boundary (common in modern ML governance) benefits from classification:
-
L: metric definitions and invariants (e.g., what counts as AUROC; data partition invariants).
-
A: admissibility gates (dataset usage-term constraints; pinned environment constraints; seed policy).
-
D: checker and author duties (publish required faces; use declared dataset version; retention duties for run evidence carriers).
-
E: admitted system
EvaluationRunner-AperformedMLProtocolEvaluation-T1 : U.WorkoverModel-M1,Dataset-D7, the pinned environment and seed policy, and evaluation windowT; the application of the AUROC metric defined by the L claim returnedAUROCResult-T1withmeasuredAUROC=0.91.When a checker, gate, or audit decision needs a replayable source-to-use account for that result, an A.10 evidence-provenance path links it to carriers
RunLog-T1,DatasetHash-D7,EvaluationReport-R1, andTraceSet-T1.
The square keeps “must use dataset vX” (D) separate from “evaluation is admissible iff dataset usage terms match” (A), and both separate from “MLProtocolEvaluation-T1 returned AUROCResult-T1(measuredAUROC=0.91)” (E). The report and log carriers may support reliance on that result; producing a carrier is not the measured result.
A.6.B:8.4 — Worked Rewrite Kit (informative, recommended)
Informative. These cases apply the authoring method in §4.4 and the rules for atomicity, L/A/D/E classification, explicit references, triangle decomposition and grounded dependencies. Use the case whose receiving question matches the present work.
A.6.B:8.4.0 - Goal
For the receiving question established in §4.4, recover the meaning of a boundary sentence that mixes “laws / gates / duties / evidence” and produce:
- atomic L/A/D/E-classified claims,
- explicit references by claim ID or canonical location,
- a readable recomposition (separate Tech and Plain forms when their readers need them).
A.6.B:8.4.1 - Micro-procedure (Atomize → Classify → Triangle → Link → Bind References → Recompose)
Use this procedure to recover and express the claims selected through §4.4. The permission-word branch below supplies the detailed distinctions when permissive wording matters to that answer.
Step 1 — Atomize. Split mixed prose into atomic claims; each must classify to exactly one quadrant.
Step 2 — Classify (L/A/D/E).
- L if the claim is truth‑conditional and adjudicable in‑description (inspection, proof or type validation, or model reasoning over declared assumptions): definitions, invariants, typing and well-formedness constraints.
Guardrails:
L-*MUST NOT (i) use RFC deontic keywords as operators, (ii) encode runtime entry predicates (those areA-*), or (iii) assert evidence existence or measurement outcomes (those areE-*). - A if it declares a gate predicate: when a named action or application is admitted under its governing rule (“admissible iff …”). State the condition as a predicate; a duty to satisfy or enforce it is a separate D claim.
Guardrails:
A-*SHOULD be written in predicate form and MUST NOT (i) use RFC deontic keywords as if it were an agent obligation, (ii) claim that evidence carriers exist (that isE-*), or (iii) assign responsibility or enforcement (that isD-*). (Do not confuse this withSignature.Applicability: applicability scopes intended meaning and intended use; it is not a runtime entry gate.) - D if the atomic statement states either a generic prescription or an individual duty, recommendation-as-duty, prohibition, or commitment. A permissive sentence enters D only through the Grant or norm row below. Guardrails: a generic claim names the normative episteme and applicable rule content without inventing an individual relation. An individual-duty claim names its actual bearer and separately obtaining A.2.8 commitment. A grant claim instead follows the participant and ground test in the Grant or norm row. A system-role kind or assignment may be a rule ground but is neither bearer nor deontic relation. Writing a claim alone does not make its object obtain.
- E if it reports an actual occurrence, evaluated finding, observed condition or evidence support under §5.4. A claim stating the normative content or obtaining duty or grant itself follows D; an evaluation or occurrence concerning it has its separate E claim.
Predicate-specific minimum: name the
E-*predicate and object, then the actual work, evaluation, or observation, scope/window, comparison frame, and other settling conditions that this predicate needs. Add an evidence or source-use relation when the receiving use needs that support; identify a carrier, schema, viewpoint or consumer only when that item matters to the use. Guardrails:E-*SHOULD NOT use RFC deontic keywords and MUST NOT hide a gate predicate (that isA-*). A reference toD-*follows §6.4: identify the evidence target or normative evaluation input, then establish the E result under its own predicate. (If the source sentence is “Role SHALL measure, retain, or expose …”, first decide whether it is a generic prescription about a system-role kind or a claim about one actual bearer. Classify either as D, but assert an individual commitment only on the second route.)
Step 3 — Triangle decomposition. Recover the observability clause’s predicate and modality under §6.3. Use the A/D/E triangle only when the source states an entry condition, a generic prescription or individual duty, and an actual observation or result:
- A: the condition for the named action or application to be admitted,
- D → A: which policy prescribes keeping or enforcing the predicate, or which actual bearer has that separately instituted duty; any responsibility relation is stated separately under its direct domain predicate
- E → A: the actual observation or result and, when needed, the evidence or traces used to adjudicate the predicate.
A requirement to produce or retain evidence stays in D. Preserve an expectation or plan without inventing an actual E result.
Permission-word branch (use only when the sentence sounds permissive). Choose the row by the job the sentence performs, not by the word may, approved, authorized, or permitted.
| Branch | Ask this plain question | Square result | Subject pattern and what closes the row |
|---|---|---|---|
| Grant or norm | Does the sentence state a generic prescription, claim that one actual bearer has an individual duty, or tell a named beneficiary which action is permitted and under what conditions? | D | Use A.2.8 for the generic-prescription or individual-duty route; only the individual route cites one separately obtaining commitment. For a grant use A.2.8.PER: name the grant occurrence, beneficiary, action, scope/window, and policy-valid A.2.9 act with its performer and assignment; confirm current policy conditions and absence of valid revocation or supersession; cite evidence needed before reliance. |
| Gate predicate or result | Does the sentence state the condition for entry, or claim one actual A.21 decision for a bounded action? | A for the entry predicate; E for the actual decision result | Keep the entry predicate as its own A-* claim. For an actual decision, use one separate E-* claim naming the GateDecisionResult, bounded action, applicable profile application, complete required GateCheckApplicationResult set, scope/window, decisionValue, action consequence, and recheck condition. A grant, finding, or conflict is only an input; neither the predicate nor the result creates or resolves it. |
| Actual exercise | Did this dated Work match the named grant’s action and beneficiary while that grant was in force? | E | Use A.2.8.PER PermissionExerciseRelation@Context: name the Work, grant occurrence, performer and the beneficiary-matching ground, scope and interval. Cite an assignment or on-behalf-of relation when that match depends on it. A failed match means that exercise relation does not obtain. |
| Weak evaluation or non-violation | Did an evaluation of a current, sufficiently complete normative frame find no applicable prohibition before action, or no violation in the actual Work? | E | Use NonProhibitionFinding@Context or NonViolationFinding@Context, its frame, subject/action or actual Work, scope and window. Name a separate evaluation Work only when a dated evaluation occurrence is claimed or required by the receiving use. A stale or incomplete frame returns unresolved. |
| Conflict | Do a current grant and norm cover the same case, and has a rule or authorized decision selected the outcome? | E | Use A.2.8.PER PermissionNormConflictFinding@Context. Cite the applicable selecting rule or the admitted system’s authorized dated decision Work and current resolution result; otherwise keep the finding unresolved. |
| Source or display only | Does the sentence only say that a permit, badge, registry entry, message, or carrier exists, displays, or evidences something? | E for an observed carrier/evidence claim; L for its definition | Use §5.4’s evidence-use route and the applicable publication or carrier pattern for the relation being claimed. Visibility or publication alone establishes no grant, exercise, finding or resolution. |
Choose one row. If one sentence answers two questions, split it before classification. If the sentence is not permission-like, do not use this branch. The branch classifies claims and selects existing subject patterns; it creates no permission result umbrella. Use the filled case in §8.4.5.4 when a concrete model is needed; point back to that case rather than adding another pattern list.
Guideline. Keep the gate predicate separate from the observations and results used to evaluate it. Write the predicate in A-*; state an actual observation or result in E-* and reference the gate when the result concerns its application (E → A).
An A-* predicate may consume a separately established result under §5.2; express a material dependency by reference under §6.1. That reference does not establish the result or a gate decision.
Step 4 — Link by ID or canonical location. Apply the cross-quadrant restrictions in §6.4, including its informative-note exception for L references to A/D/E. Same-quadrant references remain available.
Common link motifs (informative). Boundary rewrites use D→A, E→A, D→E, A/E→L, and D→L; A→D/E supplies a stated gate input, while E→D identifies an evidence target or normative evaluation input under §6.4.
Step 5 — Bind references (minimal A.7 discipline).
-
Place L claims in
Signature.Laws(and mechanism-local semantic laws if present). Keep A claims in the canonical source of their governing admission rule: mechanism applications useMechanism.AdmissibilityConditions; A.21 gates use the applicable profile rule. -
Bind a generic D claim to its normative episteme and applicable rule content.
Bind an individual-duty D claim to its actual duty-bearing System or separately governed party and
U.Commitment; cite an assignment only when the constitutive rule uses it as a ground.State responsibility and authority, when claimed, through their own admitted direct relations or missing governors. Prefer ID or canonical-location references rather than restating
L-*orA-*content. -
Bind each E claim first to its predicate/object and to the actual work, evaluation, observation, scope/window, comparison frame, and other conditions that settle that predicate. Add a carrier/schema, evidence or source-use relation, viewpoint, and consumer only when a receiving reliance decision depends on them; a claim about a carrier’s own existence or condition names the carrier as its object.
Optional drift-control. Add each L/A/D/E-classified claim verbatim to a Claim Register row (A.6.B:7) with its canonical location and references when this helps maintain faithful faces.
Step 6 — Recompose into readable text. Select the source claims and supporting detail needed for the receiving question under §4.4. Include only the claim classes present in the source. Use separate Tech and Plain forms when their readers need them; both may also serve as a teaching aid.
- Tech recomposition: a short L/A/D/E-classified claim bundle listing the present claims and their ID or canonical-location references.
- Plain recomposition: a one-paragraph narrative that summarizes the bundle and cites its claims (no new semantics). If you need a new constraint, add a new atomic L/A/D/E-classified claim to the canonical source before publishing it in Plain.
A.6.B:8.4.2 - Anti-pattern (quick)
- AP-1 Unestablished operational-result guarantees. An asserted operational
result lacks the actual Work, evaluation, or observation needed to settle it.
If the guarantee’s L/A/D/E meaning is ambiguous, use
A.6.C:4.3to recover it. - AP-2 Unresolved duty bearer. “The API promises” is used for an individual duty without a recoverable bearer and independently obtaining commitment; use A.6.C when the intended meaning remains ambiguous.
- AP-3 Gate-as-evidence. Treating the gate predicate (A) as if it were an observation (E).
- AP-4 Gate-as-law. Entry predicates as signature “laws or definitions” (L) instead of
A-*. - AP-5 Unresolved qualifier. “Fast”, “secure”, “approved” or “aligned” leaves the relevant predicate, criterion or relation unrecoverable.
- AP-6 Paraphrase drift. Restating L/A content in D or E with changed meaning (instead of citing by ID or canonical location).
- AP-7 Deontics in predicates. RFC keywords (“MUST, SHALL, and related RFC keywords”) used as operators inside
L-*orA-*predicates (should beD-*that referencesL-*/A-*). - AP-8 View-fork semantics. Recomposition/face text introduces new
L/A/D/Emeaning not present in the L/A/D/E-classified claim set (violates “no new semantics” discipline). - AP-9 Applicability-as-gate. Using
Signature.Applicability(intended use) as a substitute forA-*runtime admission predicates.
A.6.B:8.4.3 - Example 1 — Software engineering (SLO-ish API latency)
A.6.B:8.4.3.1 - Draft sentence (non-conformant)
“This API guarantees p95 latency < 200ms.”
A.6.B:8.4.3.2 - Atomize + Classify (L/A/D/E)
The draft leaves open whether guarantees states a target, an obtaining duty or an observed result. The author first recovers the receiving question: the service reader wants to know what the dated evaluation established about latency.
Recover the measurement definition, admitted operating envelope, actual evaluation and evidence support from the named sources. These are additional premises of this constructed case; the draft alone leaves them unresolved.
Explain the dated measurement. Under steps 2–3 of §4.4, the observed value requires its computation meaning, operating conditions and actual basis. The author therefore selects these claims:
L-API-01 (Definition).
p95_latency(window W, population P, unit U, method M) is defined as … (formal measurement definition).
(Lives in Signature.Laws or a referenced measurement definition pack.)
A-API-01 (Gate predicate: admissibility).
In this expanded case, the evaluation mechanism admits assessment of “p95 < 200ms” only under its declared load profile, deployment region, sampling method and window. Its entry predicate is:
AdmissibleLatencyClaim := (region=US) ∧ (concurrency≤X) ∧ (payload≤Y) ∧ (W=5m) ∧ (M=HDRHistogram@v…) ∧ (P=requests that match filter F)
(References L-API-01 for definition.)
E-API-01 (Evidence / carriers).
For LatencyEvaluation-T1 over Γ_time=[t1..t2], carriers TraceBatch-T1, Histogram-H1, DashboardSnapshot-D1, and SamplingConfiguration-S1 were produced or observed under the operating and sampling conditions in A-API-01, using the metric and computation definition in L-API-01. An A.10 evidence-provenance path links those carriers to LatencyEvaluation-T1 and its LatencyResult-T1.
(References A-API-01 and L-API-01 for the operating conditions and metric definition.)
E-API-02 (Observed value claim).
For interval Γ_time = [t1..t2] under conditions pinned to A-API-01 and using carriers in E-API-01, LatencyEvaluation-T1 produced LatencyResult-T1 with observed p95_latency = 173ms (computed per L-API-01).
(References A-API-01, L-API-01 and E-API-01.)
Return to the measurement question. The author writes: “During [t1..t2], the observed p95 latency was 173ms under the operating and sampling conditions in A-API-01, using the definition in L-API-01 and the evaluation evidence in E-API-01 (E-API-02).”
This connects the value to the conditions under which the reader can interpret it. The endpoint declaration and duty bearers do not change this bounded measurement answer. If the reader asks about another region or sampling method, return to the applicability of A-API-01 and the basis of the corresponding evaluation.
Changed question: explain the service assurance arrangement. The reader now needs the interface being covered, who owes the latency performance, who must publish incident notes and how the supporting evidence remains available. The author recovers the declared interface and the separately obtaining commitments. Those premises select the following additional claims; their inclusion follows the new question:
L-API-02 (Interface signature). The API endpoints and parameters are as declared (including parameter passing discipline / units). (Signature-level structure.)
D-API-01 (Commitment).
Admitted service-maintaining system ServiceOperations-A is the duty bearer of separately obtaining LatencyCommitment-API-01 : U.Commitment; under that commitment it SHALL meet p95_latency < 200ms when A-API-01 holds, adjudicated per L-API-01 using the carriers and observation conditions in E-API-01.
(References L-API-01 and A-API-01 by ID; does not restate them.)
D-API-02 (Operational duty).
Admitted operations system SRE-A is the duty bearer of separately obtaining IncidentNoteCommitment-API-02 : U.Commitment; it SHALL publish incident notes when LatencyCommitment-API-01 is violated and SHALL avoid claiming compliance outside A-API-01.
(References D-API-01 and A-API-01 by ID.)
D-API-03 (Duty-to-evidence linkage).
Admitted telemetry-maintaining system TelemetryOperations-A is the duty bearer of separately obtaining TelemetryRetentionCommitment-API-03 : U.Commitment; it SHALL retain or expose the carriers referenced in E-API-01 for the audit window required by policy.
(References E-API-01 by ID.)
A.6.B:8.4.3.3 - Triangle decomposition
For that assurance question, the author uses §6 to connect the declared condition, the duties and the observations. Each link explains one dependency of the assurance account:
- A-API-01 is “the predicate”.
- D-API-01 → A-API-01 states the commitment under the gate or envelope.
- E-API-01 → A-API-01 binds adjudication (carriers used to decide the gate or commitment).
- D-API-03 → E-API-01 expresses retention and exposure obligations for those carriers.
A.6.B:8.4.3.4 - Readable recomposition
The broader assurance question uses the following selection. The earlier measurement answer remains sufficient for its narrower question. A changed commitment requires reconsidering the duty and compliance conclusion while retaining an applicable measurement and its grounds.
Tech recomposition (L/A/D/E-classified claim bundle, short):
L-API-02identifies the covered API endpoints and parameters.L-API-01defines p95 latency computation.A-API-01specifies when the latency claim is admissible.D-API-01states the commitment under that envelope.E-API-01lists adjudicable carriers and conditions used to adjudicateA-API-01.D-API-02states the operational incident-note duties.D-API-03states the retention and exposure duties for carriers inE-API-01.E-API-02reports observed performance underA-API-01forΓ_time=[t1..t2].
Plain recomposition (one paragraph, readable):
“For the API endpoints and parameters declared in L-API-02, the latency target uses the p95 definition in L-API-01, and the mechanism admits its evaluation only under operating envelope A-API-01. ServiceOperations-A has the latency duty stated in D-API-01. Adjudication uses the telemetry carriers listed in E-API-01; TelemetryOperations-A has the retention duty in D-API-03, and SRE-A has the incident-note duty in D-API-02. Under that envelope, the observed p95 over Γ_time=[t1..t2] was 173ms (E-API-02).”
A.6.B:8.4.4 - Example 2 — Mechanical engineering (fit / coaxiality)
A.6.B:8.4.4.1 - Draft sentence (non-conformant)
“This fit ensures coaxiality.”
A.6.B:8.4.4.2 - Atomize + Classify
First recover which assertion ensures is meant to express. In this constructed case, a reader initially needs to interpret the geometric requirement at the shaft–bushing boundary. The author obtains the declared datum, tolerance and measurement convention, then uses steps 2–3 of §4.4 to select the two definitions that make that requirement interpretable:
L-FIT-01 (Definition).
coaxiality is defined relative to a declared base axis and measurement method (datum scheme, instrument, tolerance zone).
(Truth-conditional: “what it means”.)
L-FIT-02 (Interface and boundary structure). The boundary relation involves shaft, bushing, datum axis, tolerance class, temperature window, assembly procedure class. (Signature-level arity recovery / slots.)
Return to the interpretation question. The author writes: “Interpret the coaxiality requirement for the shaft–bushing boundary using its declared datum scheme, measurement method and tolerance zone (L-FIT-01), with the participants and applicable temperature and assembly conditions in L-FIT-02.” The reader can now locate what the geometric requirement means. A measured claim about a particular lot needs a further answer.
Changed question: determine what was established for lot L123. The author obtains the evaluation rule and the dated measurement facts. The admissible process envelope and actual evaluation connect the geometric meaning to this lot. Their source recovery supplies the next three claims:
A-FIT-01 (Gate predicate). In this expanded case, the evaluation mechanism admits assessment of the coaxiality claim only when manufacturing and assembly satisfy the declared process envelope: material batch, temperature window, tool calibration validity, surface finish class and alignment procedure version. (Gate predicate; can be checked using evidence, but is not itself evidence.)
E-FIT-01 (Evidence carriers).
For CoaxialityEvaluation-L123 over lot L123 and Γ_time=[t1..t2], carriers CMMReport-L123, CalibrationCertificate-C7, AssemblyLog-L123, TemperatureTrace-L123, and DatumSchemePin-D4 were produced or observed. An A.10 evidence-provenance path links those carriers to CoaxialityEvaluation-L123 and its CoaxialityResult-L123.
(References A-FIT-01 and L-FIT-01; avoids RFC deontics.)
E-FIT-02 (Observed).
For lot L123 and window Γ_time=[t1..t2], under conditions pinned to A-FIT-01 and using carriers in E-FIT-01, CoaxialityEvaluation-L123 produced CoaxialityResult-L123: measured coaxiality was within tolerance zone T (interpreted per L-FIT-01).
(References A-FIT-01, L-FIT-01, and E-FIT-01.)
The resulting account states: “For lot L123 during [t1..t2], the evaluation under A-FIT-01 found coaxiality within tolerance zone T, interpreted through L-FIT-01 and supported by the carriers in E-FIT-01 (E-FIT-02).”
If the datum scheme changes, the author reopens the interpretation and fitness of the measurement for that scheme. The earlier finding retains its stated datum and conditions.
Further question: explain the production and evidence duties. A reader arranging release and later reliance also needs to know who is bound to maintain the process envelope and retain its evidence. Recovering the separately obtaining commitments supplies these additional claims:
D-FIT-01 (Duty).
Admitted production-engineering system ProcessEngineer-A is the duty bearer of separately obtaining ProcessEnvelopeCommitment-FIT-01 : U.Commitment; it SHALL ensure A-FIT-01 holds for the production lot and SHALL not release the lot for use when A-FIT-01 is false.
(References A-FIT-01.)
D-FIT-02 (Duty-to-evidence linkage).
Admitted quality-engineering system QualityEngineer-A is the duty bearer of separately obtaining FitEvidenceRetentionCommitment-02 : U.Commitment; it SHALL retain or expose the carriers referenced in E-FIT-01 for the production lot.
(References E-FIT-01 by ID.)
A.6.B:8.4.4.3 - Readable recomposition
For this broader question the author connects the definition, admissibility, duties and observation. This selection lets the reader distinguish what the lot evaluation established from what each duty requires.
Tech bundle:
- Meaning of coaxiality:
L-FIT-01. - Boundary arity and participants:
L-FIT-02. - When the claim is admissible:
A-FIT-01. - Who has the process-envelope duty:
ProcessEngineer-AunderD-FIT-01. - What we observe and keep as carriers:
E-FIT-01and measured outcomeE-FIT-02(with retention dutyD-FIT-02).
Plain paragraph:
“In this expanded case, ‘ensures coaxiality’ is made precise by fixing the definition and datum scheme (L-FIT-01) and by making the boundary participants explicit (L-FIT-02). The mechanism admits this coaxiality evaluation only under the declared manufacturing and assembly envelope (A-FIT-01). ProcessEngineer-A has the process-envelope duty stated in D-FIT-01. QualityEngineer-A has the evidence-retention duty in D-FIT-02. Compliance is adjudicated using the measurement and process carriers listed in E-FIT-01; for lot L123 over Γ_time=[t1..t2], the observed coaxiality was within tolerance E-FIT-02.”
A.6.B:8.4.5 - Example 3 — Management (project “approved or aligned”)
A.6.B:8.4.5.1 - Draft sentence (non-conformant)
“The project is approved.”
A.6.B:8.4.5.2 - Atomize + Classify
With only the draft, ask which approval is claimed and which subsequent action the reader needs to decide. Here the question is whether the project may start execution. Recover the sponsor-approval definition, entry predicate, four actual check-application results and evidence path from the named sources. These additional premises let the author connect the prerequisite meanings to a current entry decision.
The separately obtaining coordinator duty is needed for the broader question of who is bound to maintain this arrangement; it is also supplied in the constructed case below.
L-PRJ-01 (Definition).
For this case, the project-entry policy defines RequiredProjectApproval(project) as sponsor sign-off on the identified project. Risk-review completion, budget availability and staffing are separate prerequisites.
A-PRJ-01 (Gate predicate: stage entry).
For starting execution work, ExecutionAdmissible(project) holds iff RequiredProjectApproval(project) and the required prerequisites are satisfied: risk review completed, budget line exists and key roles staffed.
(This is the “may start work” entry predicate; it references L-PRJ-01 for what counts as approvals. If “approved” is meant as permission rather than gate evidence, use the permission-word branch in §8.4.1. An approval registry entry or evidence carrier alone remains source/display evidence and is not a grant.)
D-PRJ-01 (Duty).
Admitted project-coordination system ProjectCoordinator-A is the duty bearer of separately obtaining ProjectEntryCommitment-PRJ-01 : U.Commitment; it SHALL not initiate execution unless A-PRJ-01 holds, SHALL keep the approval registry current, and SHALL retain or expose the evidence carriers referenced in E-PRJ-02.
(References A-PRJ-01 and E-PRJ-02 by ID.)
E-PRJ-01 (Gate decision result).
At Γ_time=snapshot(t), ProjectEntryGateResult-P-t : GateDecisionResult has gateRef=ProjectEntryGate-P, decisionSubjectRef=ProjectEntryClaim-P-t, boundedActionRef=StartExecution-P, and profileApplicationRef=ProjectEntryProfileApplication-P-t.
That application of ProjectEntryProfile-v3 uses the complete required set {ApprovalCheck-P-t, RiskReviewCheck-P-t, BudgetCheck-P-t, StaffingCheck-P-t} : GateCheckApplicationResult[]; each application reports its checked subject and criterion, has sourceOutcome=satisfied, and maps to pass.
The result records decisionValue=pass, action consequence “start StartExecution-P within the admitted project-entry window”, the project-entry scope and window, and recheck on any changed approval, risk-review, budget, staffing, profile, scope, or window.
(This E-* result applies the inputs named by A-PRJ-01; it is neither the A-* predicate nor performed execution Work.)
E-PRJ-02 (Evidence for reliance).
For reliance on E-PRJ-01, the observed carrier set is {DecisionRecord-R7, MeetingMinutes-M3, BudgetLine-B4, StaffingAssignments-S2, GateChecklistSnapshot-G9}. An A.10 evidence-provenance path links those carriers to the four GateCheckApplicationResult values and ProjectEntryGateResult-P-t. GateChecklistSnapshot-G9 displays the result; its presence alone establishes neither a check result nor the gate decision.
(References A-PRJ-01 and E-PRJ-01; avoids RFC deontics.)
A.6.B:8.4.5.3 - Readable recomposition
Answer the entry question. The author uses L-PRJ-01 and A-PRJ-01 to interpret the prerequisites, E-PRJ-01 for the decision and E-PRJ-02 for the support needed before reliance.
The answer is: “At snapshot t, the project-entry decision is pass, allowing StartExecution-P within the stated window (E-PRJ-01, supported by E-PRJ-02). Recheck when a prerequisite or the decision’s profile, scope or window changes.”
A changed budget check returns the reader to the affected admission decision. The definition of sponsor approval remains usable.
Explain the wider arrangement when requested. Adding D-PRJ-01 answers who has the project-entry and registry-maintenance duties. The following recomposition addresses that wider question:
Tech bundle:
- The required sponsor approval is defined by
L-PRJ-01; risk review, budget and staffing remain separate prerequisites. - “May start execution” is a gate predicate:
A-PRJ-01. ProjectCoordinator-A’s project-entry duty:D-PRJ-01.- The A.21 gate result is
E-PRJ-01; its bounded A.10 evidence support isE-PRJ-02.
Plain paragraph:
“In this expanded case, the required sponsor approval has the meaning defined in L-PRJ-01. Together with the separate risk-review, budget and staffing prerequisites, it supplies the inputs to the entry predicate A-PRJ-01. ProjectCoordinator-A has the project-entry and registry-maintenance duties stated in D-PRJ-01. At snapshot t, the current A.21 profile application maps every required check-application result to pass; E-PRJ-01 records decisionValue=pass with the action consequence ‘start StartExecution-P within the stated window’, and E-PRJ-02 supplies the evidence path for reliance on that result.”
A.6.B:8.4.5.4 - Filled permission case (each sentence classified)
Apply §4.4 to each receiving question in this case. To explain the current grant, connect its instituting basis E-CAL-01 to D-CAL-01. For an entry decision, also recover the entry predicate and the actual decision with its applicable inputs. A question about whether the technician used the grant selects E-CAL-02–03; later compliance and renewed entry require their own findings below. The detailed permission branch in §8.4.1 explains these selections. Preserve the distinct claims so a later question can use its own grounds.
E-CAL-01 (Instituting communicative Work). Admitted system MaintenanceCoordinator-A performed dated CalibrationGrantAct-17 : U.SpeechAct under MaintenanceCoordinator-A@DayShift; that obtaining assignment has the system as holder and covers the act. In this filled case, CalibrationGrantPolicy-v4 does not require a separate authority relation, so none is asserted. The assignment supplies no authority and performs no act. CalibrationGrantAct-17 satisfies the policy in PlantCalibrationContext and is the instituting Work.
A policy variant that does require grant authority must cite one obtaining authority relation under its direct predicate or stop at missing-governor[grant authority].
D-CAL-01 (Grant position). MaintenanceCalibrationGrant-17 : GrantedPermissionRelation@Context, instituted by CalibrationGrantAct-17—the speech act stated in E-CAL-01—permits beneficiary MaintenanceTechnicianSystemRole to run CalibrationProcedure-v3 in Zone 8 during ServiceWindow-17. CalibrationGrantPolicy-v4 remains current, the grant covers that system-role kind, procedure, zone, and window, and no valid revocation or supersession has ended this occurrence; this D-* claim records the grant but does not institute it.
A-CAL-01 (Gate predicate). CalibrationEntryAdmissible(plan, checkTime) holds only if MaintenanceCalibrationGrant-17 is current for the plan’s beneficiary, action, zone, and time and no applicable permission/norm conflict finding is unresolved. The predicate takes those inputs; it creates neither the grant, a conflict result, nor an actual gate decision.
E-CAL-02 (Actual Work and actor). Through its A.13 core, admitted system Tech-17 is the performer for this case under obtaining assignment Tech-17@Shift-B, whose holder is Tech-17 and whose extent covers the early part of ServiceWindow-17. A.15.1 independently admits dated CalibrationWork-17B : U.Work from that performer, its Method, extent, and containing-System facts.
Because this filled case expressly claims precise assignment-bound attribution, F.6 separately relates CalibrationWork-17B to that same assignment. The assignment neither acts nor identifies the performer.
If F.6 is not established, the under-assignment claim remains unresolved; if the relation is shown not to obtain, that claim is refuted. Neither alone overturns the independently admitted Work. If the attribution finding questions a premise consumed by Work admission, recheck that premise and the dependent claims.
E-CAL-03 (Optional exercise claim). Because this case asks whether the grant was used, CalibrationExercise-17B : PermissionExerciseRelation@Context connects CalibrationWork-17B to MaintenanceCalibrationGrant-17: the Work instantiates CalibrationProcedure-v3; Tech-17@Shift-B is an assignment occurrence whose declared species uses MaintenanceTechnicianSystemRoleKindDomain as its assigned-kind domain, and the occurrence supplies MaintenanceTechnicianSystemRole as the value admitted by that domain; and the Work occurs in Zone 8 within ServiceWindow-17 while the grant is current. If the action or beneficiary test failed, this exercise relation would not obtain.
D-CAL-02 (Exercise non-use boundary). The authoring rule says to add E-CAL-03 only when the reader needs to know whether the grant was exercised; otherwise stop with the separately named grant and Work. This is a generic prescription for boundary text, not a claim that one author bears an individual U.Commitment.
E-CAL-04 (Later non-violation finding). Through its A.13 core, admitted system ComplianceEvaluator-4 is the performer under obtaining assignment ComplianceEvaluator-4@QualityShift. A.15.1 independently admits dated CalibrationComplianceEvaluation-17B : U.Work; because this finding expressly preserves precise assignment-bound attribution, F.6 separately relates that Work to the same assignment.
The Work checked CalibrationWork-17B against current PlantCalibrationNormativeFrame-17, complete enough for this technician, procedure, zone, and evaluation window, and returned CalibrationNonViolation-17B : NonViolationFinding@Context(result=nonViolating). A stale or insufficient frame would return unresolved.
If F.6 is not established, the attribution remains unresolved; if the relation is shown not to obtain, that attribution claim is refuted. Missing or failed attribution alone does not overturn independently admitted Work or an independently established evaluation result. If the attribution finding questions a premise used for Work admission or the evaluation, recheck that premise and the dependent claims.
E-CAL-05 (Evidence for reliance). An A.10 evidence-provenance path links the CalibrationNonViolation-17B finding to CalibrationComplianceEvaluation-17B, ComplianceEvaluator-4@QualityShift, CalibrationRunLog-17B, the log’s source and currentness relations, and the bounded audit context. The path supports reliance on the finding; the log, assignment, and path do not perform the evaluation or create its result.
E-CAL-06 (Unresolved conflict). After CalibrationWork-17B and its evaluation, Zone8EntryProhibition-17 becomes current for the same beneficiary, action, zone, and the remaining service window, including the calibration action specified by CalibrationWorkPlan-17C; no applicable rule selects an outcome and no authorized dated decision Work with a current resolution result exists. CalibrationConflict-17 : PermissionNormConflictFinding@Context therefore remains unresolved.
E-CAL-07 (Source/display fact). SignedGrantRecord-17 and GreenPermitTile-17 are visible carriers in this case; their presence is an observed source/display claim only.
E-CAL-08 (Gate decision result). At the later entry check, CalibrationEntryResult-17C : GateDecisionResult has gateRef=CalibrationEntryGate-17, decisionSubjectRef=CalibrationWorkPlan-17C, boundedActionRef=StartCalibrationProcedure-v3-in-Zone8-17C, and profileApplicationRef=CalibrationEntryProfileApplication-17C.
The current profile application uses the complete required set {CalibrationGrantCurrentCheck-17C, CalibrationConflictResolvedCheck-17C} : GateCheckApplicationResult[]: the current-grant application maps to pass, while the CalibrationConflict-17=unresolved source outcome maps to block.
The result records decisionValue=block, action consequence “hold CalibrationWorkPlan-17C outside entry”, the remaining Zone 8 service scope and window, and recheck when the grant, conflict result, profile, scope, or window changes. It neither resolves the conflict nor revokes MaintenanceCalibrationGrant-17.
L-CAL-01 (Tempting wrong classification, rejected). “The visible permit is D, so the grant exists, the Work exercised it, and the Work was non-violating” is not one atomic claim and is false as a classification shortcut. The carrier observation is E-CAL-07; the grant, exercise, evaluation finding, and E-CAL-08 gate result remain the separately classified claims above.
A.6.B:8.4.6 - A compact recomposition aid
Select the source claims needed for the receiving question through §4.4. Cite claim IDs or canonical locations. For a measurement account, connect the value to its definition, conditions and actual basis; for an entry answer, connect the entry rule to its evaluated inputs and current decision. Use the broader forms below when the question also needs the normative arrangement and its evidence. Use separate Tech and Plain forms when their readers need them.
A.6.B:8.4.6.1 - Tech register
“This boundary claim is defined by L-… and applies only under A-…. D-… states either the generic prescription or the separately instituted duty of [actual bearer]. E-… states the observed or evaluated result for
Γ_time=…; add its evidence support when the receiving use needs it.”
A.6.B:8.4.6.2 - Plain register
“We mean [short label] in the sense of L-…, and use it only when A-… holds. D-… either states what the named policy requires or, when an individual duty was separately instituted, names its actual bearer. E-… states the observed or evaluated result for the declared interval; add its evidence support when the receiving use needs it. If responsibility is also claimed, cite its direct relation separately.”
A.6.B:9 — Bias‑Annotation
Scope: Universal for boundary descriptions.
- Arch bias: favors explicit separation and explicit references; mitigated by allowing narrative faces while keeping commitments classified and referenced by ID or canonical location.
- Gov bias: makes actual duty bearers and duties explicit (D) and auditability explicit (E); mitigated by keeping evidence conceptual and carrier-referenced rather than tool-specific. Responsibility, when claimed, remains a separate direct relation.
- Ontological and epistemic bias: distinguishes the EntityOfConcern, claim-bearing episteme and carrier, and requires the actual basis for work, evaluation or observation claims. Cross-quadrant references preserve their connections in a readable account.
A.6.B:10 — Conformance Checklist
| ID | Requirement | Purpose |
|---|---|---|
| CC‑A.6.B.1 (Atomicity). | A conforming boundary text SHALL decompose mixed sentences into atomic claims such that each atomic claim belongs to exactly one quadrant L/A/D/E. | Makes L/A/D/E classification unambiguous; prevents mixing claim jobs. |
| CC‑A.6.B.2 (Quadrant classification). | Each atomic claim MUST be classified by its job under §4 and have its own adjudication basis recovered, rather than taking a quadrant from its subject-pattern family. When permission wording is present, the single branch in §8.4.1 MUST select the claim’s job before assigning L/A/D/E. | Prevents one pattern catalogue from replacing the square’s decision. |
| CC‑A.6.B.3 (Form and obtaining constraints). | L-* and A-* claims MUST NOT use RFC deontic keywords as operators. A generic-prescription D-* claim MUST name its normative source and applicable rule content without inventing an individual occurrence; an individual-duty D-* claim MUST name its actual bearer and separately obtaining U.Commitment; a grant D-* claim MUST satisfy §8.4.1. Claim text alone does not make the claimed relation obtain. Responsibility uses its direct predicate or missing governor. An E-* claim MUST name the work, evaluation, or observation that settles it and any evidence used for reliance. | Keeps normative content, individual institution, responsibility, and evaluated results distinct. |
| CC‑A.6.B.4 (Explicit references). | Where a claim materially depends on another L/A/D/E-classified claim, that dependency MUST be expressed by explicit ID or canonical-location reference. An accompanying explanation must preserve the referenced meaning. | Makes the governing claim recoverable across layers and faces. |
| CC‑A.6.B.5 (E‑claim adjudicability). | Each E-* claim names its predicate and object plus the actual work, evaluation, or observation, scope/window, comparison frame, and other conditions required to settle that predicate. It adds an evidence or source-use relation when the receiving use needs that support, and identifies a carrier, schema, viewpoint or consumer only when that item matters to the use. | Makes work-effects adjudicable without forcing unrelated carrier apparatus into every result claim. |
| CC‑A.6.B.6 (Gates and laws). | Operational admissibility predicates MUST NOT appear as L-* laws in the signature layer; they MUST be A-* claims in the canonical source of their governing admission rule. | Preserves substitution and signature stability. |
| CC‑A.6.B.7 (Dependency grounds). | L-* claims MUST NOT reference A-*, D-*, or E-*, except for purely informative notes marked informative. An A-* reference to D-* or E-* MUST meet §6.4’s input and scope/currentness conditions. An E-* reference to D-* MUST meet §6.4’s evidence-target or evaluation-input rule and MUST NOT treat the D claim alone as an E result. | Keeps reusable laws stable and makes gate inputs and evaluation grounds explicit. |
A.6.B:11 - Common Anti‑Patterns and How to Avoid Them
| Anti‑pattern | Symptom | Why it fails | Repair (square‑consistent) |
|---|---|---|---|
| Gate‑as‑law | Preconditions written as “laws” | Confuses law and admission condition; breaks substitution | Move to A-* in the canonical source of the admission rule (Mechanism.AdmissibilityConditions for a mechanism); reference L-* terms. |
| Deontics in predicates | “MUST” inside definitions or gates | Confuses governance with truth or admissibility | Rewrite the definition or gate as L-*/A-*; add a generic D-* prescription or an independently established individual-duty claim when current. |
| Unresolved duty bearer | “The API promises” is used for an individual duty without its actual bearer and commitment | The institutional claim cannot be settled from the interface wording | Recover the policy and generic prescription, or the actual bearer and separately obtaining U.Commitment when an individual duty is claimed. Keep measured property (E-*) and metric definition (L-*) separate; use A.6.C when ambiguity changes interpretation or use. |
| Unestablished operational-result guarantees | An asserted p95 result lacks its actual evaluation | The operational-result claim lacks the basis needed to settle it | Use A.6.C:4.3 when the guarantee’s meaning is ambiguous. For an operational result, state the actual Work, evaluation, or observation, predicate and object, scope/window, and result; if that basis is absent, leave the result unestablished. Add A.10 support and a D-* → E-* link only when the receiving reliance and duty claims need them. A carrier-existence observation alone does not complete the measured claim. |
| Paraphrase drift | A restatement changes the governing rule | Different faces can then imply different conditions or consequences | Reference the claim ID or canonical location and repair the explanation to preserve its meaning; use a Claim Register when useful. |
| View‑fork semantics | A face introduces new L/A/D/E content | Violates “no new semantics” publication discipline | Put any new boundary claim in its canonical source before publishing it on a face. Mark explanatory commentary informative when it adds no boundary semantics. |
A.6.B:12 — Consequences
| Benefits | Trade‑offs / mitigations |
|---|---|
| Stable modular boundaries. Laws, gates and normative claims remain distinct. | Requires writers to split sentences; mitigated by the triangle decomposition pattern. |
| Explicit adjudication links. Duties and claims about results can identify the evaluations and evidence relevant to their use. | Establishing that basis has a cost; require only the support the receiving use consumes. |
| Recoverable meaning across faces. References connect an explanation to its governing claim. | More cross-references; a Claim Register can help when the claim set needs persistent reuse. |
A.6.B:13 — Rationale
Boundary changes need different comparisons. Changing an admission condition affects which applications can enter; changing a normative rule affects what is required or granted; changing an implementation can affect the result while leaving both declarations intact. The four claim jobs keep these questions and their connections recoverable. A.6:4.7 explains how to use them for a change decision.
The classification is an authoring convention, with one job assigned to each atomic claim. Modality and adjudication help recover that job, but neither supplies a two-way partition of every claim. In particular, A predicates are declarations, and an individual D claim requires actual constitutive grounds. The practical benefit comes from exposing the different grounds and preserving their dependencies across uses and publication faces.
For a small boundary, ordinary atomic prose and its existing contract or policy can be sufficient. Claim IDs, a register and additional faces earn their cost when several uses need to refer to the same content or follow a consequential change. Keep the direct subject conditions available in either form.
A.6.B:14 — Source contributions
A.6:11 gives the source contributions and their limits. Effects and handlers distinguish declared operations from handler semantics; contract and behavioral-subtyping work supplies preservation questions; session types supply protocol-specific conditions. ISO/IEC/IEEE 42010 supports concern-oriented descriptions and correspondences. OpenTelemetry and SRE distinguish the observations available from the service property being examined; Cedar gives an explicit rule-evaluation and decision construction.
FPF uses these contributions to organize boundary claims. The four jobs, their FPF placements and §6.4’s dependency rules are the construction explained here. In particular, provider duties require their governing normative source or individual instituting basis, and replayable evidence-provenance accounts use A.10. A source’s typing or telemetry guarantees retain their own stated scope.
A.6.B:15 — Relations
- Used by A.6 and A.6.C: supplies the canonical matrix and cross-quadrant link discipline. Both consumers classify each atomic claim by predicate and adjudication, keep the claim episteme separate from its EntityOfConcern, and use the §8.4.1 branch only when permission wording is live.
- Constrains A.6.0 (
U.Signature): enforces thatL-*laws are truth‑conditional and do not include admissibility predicates. - Constrains A.6.1 (
U.Mechanism): places admissibility inAdmissibilityConditions(A-*) and classifies actual results and evidence claims asE-*, adding carrier references when their use requires them. - Requires A.7: identifies each claim’s EntityOfConcern separately from the episteme carrying it. The concern may itself be a Description episteme or publication carrier.
- Interacts with MVPK/E.17: faces are projections that cite L/A/D/E-classified claims and introduce no new semantics. When the permission-word branch is selected, its row names the subject pattern and obtaining or failure test; A.6.B only classifies the statement, and wording or a carrier alone does not make the referenced object obtain.
A.6.B:15a - Probe-coupled boundary claim classification
When boundary text says that an interaction changes the state represented by its output, identify the interaction, the changed state, and the receiving use, then classify the atomic claims through the same L/A/D/E square. Such interactions may include asking a question, measuring or intervening through a metric, reading a dashboard, conducting a workshop, exporting through a Bridge, or making an API read.
Action classification:
- Copy the boundary sentence being used for a decision.
- Split it into its atomic claims before judging it, retaining only the classes present in the source: definition or law, runtime admissibility, generic prescription, individual duty or grant, work effect or evaluated result, and any separate evidence-support claim.
- Give each atomic claim its quadrant and any identifier required by a later reference.
- Apply
A.6.Pwhen a relation or its participants remain unclear; useF.18only when recovered terms need durable names. Use §5.4’s evidence-use route for needed support,B.3for a named assurance question,C.16for measurement, andF.9for a Bridge question. UseC.26.1for the residual probe-coupled question only when the interaction participates in the represented state and its output is being used as a passive read or export despite that decision-relevant effect. - Use a Claim Register when the claim set needs persistent reuse, dispute resolution or cross-face citation; otherwise keep the atomic prose.
For a local working note, atomize the sentence mentally and write one or more atomic L/A/D/E-classified sentences, preserving every source claim. One sentence suffices only when one logical job remains. Use the Claim Register when the claim set must survive reuse or dispute.
| Claim in the boundary text | Claim class | What the user writes |
|---|---|---|
| The term, variable, state, frame, or relation being defined | L-* law or definition claim | Definition or invariant, without agent obligation language |
The admission condition for using a probe, metric, question, or bridge at the decision point; declaration-level intended-use scope remains in Signature.Applicability | A-* runtime-admissibility claim | Entry predicate and its required inputs; include failure behavior or a neighboring-pattern continuation when the declared rule or receiving use requires it |
| What policy prescribes about applying, retaining, exposing, or avoiding overuse of the probe result; or which actual bearer has that individually instituted duty or which grant currently permits the action; and, if separately claimed, who bears responsibility | generic D-* prescription or individual D-* duty or grant claim; separate direct responsibility claim or missing governor | Normative source and rule content for the generic branch; actual duty bearer and U.Commitment for an individual duty; the grant occurrence, beneficiary, action and conditions under §8.4.1 for a grant; L/A/E claim IDs or canonical locations for material dependencies; responsibility predicate and participants only when that relation independently obtains |
| The actual work effect or observed before-state or after-state, plus any separate evidence relation or carrier used for reliance | E-* work-effect or evaluated-result claim; separate evidence-support claim when current | Predicate and object, actual Work, evaluation, or observation, conditions and window, and result; add evidence support under §5.4 when the receiving use needs it |
Useful outputs:
- a Claim Register when persistent reuse, dispute resolution or cross-face citation needs it;
- one or more atomic L/A/D/E-classified sentences preserving every source claim for a local working note;
- an ordinary A.6.B L/A/D/E-classified claim set when no probe-coupled state-reading question remains;
- the
C.26.1:4.2decision and finish result for a residual probe-coupled question that meets the activation condition above; - the evidence-support, assurance, measurement, or Bridge result or continuation
required by
A.10,B.3,C.16, orF.9for the corresponding current question.
Identify the interaction and its receiving use, split and classify the claims,
then apply the C.26.1 activation condition above to any remaining question.