# TENET ARITHMETICA — Doctrinal Invariants

Version: `2.2.0-TENET-ARITHMETICA-CARDINAL-LAYERS`
Status: `FOUNDATIONAL CANDIDATE / STRUCTURALLY COMPLETE / NON-RUNTIME`
Authority: `USER / HUMAN CONTROL PLANE`
Class: `AGLA / TENET / ARITHMETICA`
Mutation policy: `VERSION-CONTROLLED ONLY`

Dependencies:

- root TENET class law;
- root five-class structural order;
- `ARITHMETICA_CARDINAL_LAYER_REGISTRY.json`.

TENET states the invariants of ARITHMETICA. It does not admit input, instantiate
traversal, execute arithmetic, coordinate OPERAE, or mediate output.

This artifact defines one bounded cardinal regime articulated in ten layers.
Each layer has an explicit numeric value and a stable, script-independent
carrier identity. Glyph bindings, PRE/I..IX structural indices, graph objects,
and ordinal execution positions remain distinct projections.

## I. Intentiones

`PRIMA_INTENTIO` concerns the being of number.

`SECUNDA_INTENTIO` concerns the mode by which number, order, quality, notation,
relation, and traversal become intelligible.

Mandatory distinctions:

```text
PRIMA_INTENTIO != SECUNDA_INTENTIO
NUMERUS != QUALITAS
CARRIER != CARRIED_QUALITY
RELATIO != ENS
NOTATIO != NUMERUS
BASIS != CAUSA_NUMERI
STRUCTURA != ONTOLOGIA
PROIECTIO != RES_PROIECTA
```

A relation, notation, base, diagram, graph, or traversal may express or
articulate number. None creates number as being.

## II. Numeric causality

The unit is the non-relational numerical anchor:

```text
UNITAS = 1
NUMERUS_UT_ENS = PLURALITAS_ADDITIVA
ADDITIO = CAUSA_CONSTITUENS / PRIMA_INTENTIO
MULTIPLICATIO = CAUSA_ORDINANS / SECUNDA_INTENTIO
```

Multiplication operates over additive numerical being. It orders or groups what
is numerically constituted; it does not replace addition as the constituting
account.

A base makes grouping or position visible. A notation expresses a number. A
positional string must be decoded before its intrinsic number is known.

## III. Ontologia cardinalis

ARITHMETICA adopts this bounded carrier vocabulary:

| Number | Carrier ID | Phoenician candidate | Hebrew fallback | Index | Canonical quality | Structural posture |
|---:|---|:---:|:---:|---|---|---|
| 1 | `ARITH-CARRIER-001` | `𐤀` | `א` | `PRE` | `identity` | non-relational anchor |
| 2 | `ARITH-CARRIER-002` | `𐤁` | `ב` | `I` | `distinction` | relational field |
| 3 | `ARITH-CARRIER-003` | `𐤂` | `ג` | `II` | `mediation` | relational field |
| 4 | `ARITH-CARRIER-004` | `𐤃` | `ד` | `III` | `stabilization` | relational field |
| 5 | `ARITH-CARRIER-005` | `𐤄` | `ה` | `IV` | `asymmetry` | relational field |
| 6 | `ARITH-CARRIER-006` | `𐤅` | `ו` | `V` | `synthesis` | relational field |
| 7 | `ARITH-CARRIER-007` | `𐤆` | `ז` | `VI` | `rupture` | relational field |
| 8 | `ARITH-CARRIER-008` | `𐤇` | `ח` | `VII` | `amplification` | relational field |
| 9 | `ARITH-CARRIER-009` | `𐤈` | `ט` | `VIII` | `closure` | relational field |
| 10 | `ARITH-CARRIER-010` | `𐤉` | `י` | `IX` | `transition` | relational field |

The mapping is bidirectional as a controlled vocabulary, not an identity claim.

```text
NUMBER_CARRIES_QUALITY
NUMBER_DOES_NOT_EQUAL_QUALITY
QUALITY_IS_BASE_NEUTRAL
1_IS_PRE_NON_RELATIONAL
2_TO_10_FORM_I_TO_IX_RELATIONAL_FIELD
```

A number outside `1..10` remains a number but has no canonical quality in this
bounded register. Absence of a registered quality must be disclosed, not filled
by analogy.

### III.A Cardinal layer and carrier contract

The normative support record is
`ARITHMETICA_CARDINAL_LAYER_REGISTRY.json`. Every layer exposes:

```text
ARITHMETICA_CARDINAL_LAYER:
  cardinal_value
  carrier_id
  script_bindings
  numeric_value_authority
  structural_index
  canonical_quality
  structural_posture
  prima_intentio_projection
  secunda_intentio_projection
  correlatives:
    agent
    act
    patient
  complete_correlational_expression
  cross_regime_refs[]
  lexical_status
  provenance
  unresolved_fields[]
```

The numeric value and stable ASCII carrier ID are machine-authoritative. A
glyph is a reversible view and is never the primary key.

```text
CARRIER_ID != NUMBER
CARRIER_ID != DISPLAY_GLYPH
CARRIER_ID != STRUCTURAL_INDEX
CARRIER_ID != GRAPH_ID
CARRIER_ID != OPERA_ID
PHOENICIAN_GLYPH != HEBREW_GLYPH
ALPHABETIC_CARRIER != PHOENICIAN_NUMERAL_CHARACTER

1_TO_10_FORM_ONE_LAYERED_CARDINAL_REGIME
PRE_AND_I_TO_IX_REMAIN_STRUCTURAL_INDICES
```

The earlier Latin `A..I,K` mapping is retained as superseded design
provenance. It is not the current carrier identity. In the provisional Semitic
surface, cardinal 10 uses Yod; Kaph belongs to candidate value 20 and must not
be substituted for Yod.

### III.B Complete correlative law

Each cardinal layer exposes three simultaneous named roles:

```text
AGENT + ACT + PATIENT
```

They form one complete arithmetic principle. They are not alternative meanings
selected by the sign of a transition.

The first four functional English glosses are fixed for this candidate:

```text
1 / identity:       AGENT IDENTIFYING PATIENT
2 / distinction:    AGENT DIFFERENTIATING PATIENT
3 / mediation:      AGENT MEDIATING PATIENT
4 / stabilization:  AGENT STABILIZING PATIENT
```

These are functional English glosses, not attested or canonized Latin forms.
Layers `5..10` retain their registered qualities and the complete three-role
shape, but their exact correlational lexemes remain `UNRESOLVED`.

```text
SIGN != AGENT_OR_PATIENT_SELECTOR
POSITIVE != AGENT
NEGATIVE != PATIENT
```

### III.C Prima and secunda projections

```text
LAYER n UNDER PRIMA_INTENTIO:
    numeric being and canonical quality posture

LAYER n UNDER SECUNDA_INTENTIO:
    relational intelligibility through agent-act-patient correlation
```

For layer 1, the prima-intentio posture remains the non-relational numerical
anchor. Its secunda-intentio projection may nevertheless expose the complete
identificative correlation.

```text
NON_RELATIONAL_NUMERIC_POSTURE
    !=
INCAPACITY_FOR_RELATIONAL_PROJECTION
```

### III.D Reversible script-binding policy

The current display policy is provisional and reversible:

```text
preferred candidate view:  Phoenician letter
technical fallback only:   Square Hebrew letter
diagnostic fallback:        ASCII name or explicit numeric value
```

The Hebrew glyph is never a second carrier, a coequal notation, a semantic join
key, or evidence for Kabbalistic meaning.

```text
PHOENICIAN_ARITHMETICA != HEBREW_KABBALAH
HEBREW_FALLBACK != KABBALAH_BINDING
GLYPH_FALLBACK != SEMANTIC_PROJECTION
```

Rendering may select a fallback without changing carrier ID, numeric value,
quality, graph binding, or event identity. Right-to-left presentation and bidi
isolation belong to the view layer; semantic storage preserves explicit code
points and contains no invisible bidi controls. Glyphs are forbidden as sole
machine identifiers and in canonical filenames.

The alphabetic Phoenician carrier is an AGLA-designed alphanumeric convention.
It must not be described as a separately encoded Phoenician numeral character
or as an attested complete Phoenician numeral system.

### III.E Provisional decimal transposition

`ARITHMETICA_CARDINAL_LAYER_REGISTRY.json` records a candidate relation among
different carriers at different decimal scales. Every record remains
`PROVISIONAL_NOT_ADMITTED`.

```text
DECIMAL_TRANSPOSITION != SAME_GLYPH_WITH_NEW_PLACE_VALUE
SOURCE_CARRIER != TARGET_CARRIER
```

The initial evidence surface includes `1→10`, `2→20`, `3→30`, `4→40`, and
`10→100`. Only carrier 10 is inside the current bounded `1..10` registry as a
target. Carriers 20, 30, 40, and 100 remain explicitly unregistered. Values
500..900, thousands, and composite syntax are unresolved.

## IV. Operation invariants

ARITHMETICA recognizes three native operation identities:

1. `OPERA-QUALITAS-EX-QUANTITAS`: admitted quantitative structures to
   qualitative elaborations;
2. `OPERA-QUANTITAS-EX-QUALITAS`: admitted qualitative requirements to
   quantitative candidates;
3. `OPERA-QUANTITAS-ET-QUALITAS`: coupled quantitative/qualitative search and
   articulation.

The three are distinct OPERAE. Bidirectionality does not imply identity, and the
mixed OPERA does not silently replace either directional OPERA.

Arithmetic truth is checked in the numeric layer before qualitative exposition.
A useful qualitative reading cannot validate a false equation.

The compressed-source identifiers `OPERA-EX-QUANTITAS` and
`OPERA-EX-QUALITAS` are retained only as provenance aliases. They do not name
the current separated artifacts.

## V. Relation discipline

Relations may be:

- source-declared;
- ROTA-default;
- numerically derived;
- user-supplied;
- provisionally inferred.

Their provenance must remain distinct. Co-presence in a wheel, table, prompt, or
future graph does not establish causality or priority.

`COMPLETE_GRAPH` means structural adjacency is available in the native ROTA. It
does not mean every traversal is preferred, admitted, meaningful, or free.

## VI. Graph and multi-OPERA preparation

ARITHMETICA exports a typed carrier vocabulary for future graph generation and
multi-OPERA coordination. It does not itself canonize a graph-generation law.

Mandatory interface distinctions:

```text
ARITHMETIC_LOCUS != GRAPH_NODE
ROTA_EDGE != EXECUTION_PERMISSION
DERIVATION != TRAVERSAL
OPERA_TRACE != OPERA
TASK_GRAPH != FIVE_CLASS_ORDER
SATOR != ORDINARY_OPERA_NODE
```

A later graph stack may consume registered ARITHMETICA loci, qualities,
relations, derivations, and traces only through explicit cross-stack admission.
It must preserve provenance and the five-class order.

Dedicated cardinal ROTAE consume this explicit specialization contract:

```text
ARITHMETICA_GRAPH_APPLICATION:
  graphus_ref
  cardinal_amplitude
  locus_layer_refs[]
  difference_layer_refs[]
  source_agent_binding
  transition_act_binding
  target_patient_binding
  orientation_rule
  provenance
```

For a local complete-cardinal fixture:

```text
locus i
    -> ARITHMETICA_LAYER(i).quality

edge {i,j}
    -> ARITHMETICA_LAYER(|j-i|)

transition i->j:
    source             -> agent
    layer |j-i|        -> act
    target             -> patient
    sign(j-i)          -> arithmetic orientation
```

This is an ARITHMETICA specialization. It does not redefine generic Graphus
ontology, canonize a `GRAPHUS n` family, or grant traversal.

```text
SIGNED_DISPLACEMENT(i->j) = j-i
positive = ascending arithmetic orientation
negative = descending arithmetic orientation
```

Source and target become agent-side and patient-side projections only through
this declared binding:

```text
SOURCE != AGENT
TARGET != PATIENT
ARITHMETICA_BINDING(SOURCE) -> AGENT
ARITHMETICA_BINDING(TARGET) -> PATIENT
```

The user-authorized semantic concordance is retained without notational or
artifact collapse:

```text
T:B -> Difference
ARITHMETICA:cardinal 2 -> distinction/difference
```

The provisional relation label is `SHARED_PRINCIPIAL_ROOT`. Its provisional
vocabulary status does not weaken the fixed non-collapse:

```text
SHARED_PRINCIPIAL_ROOT != ARTIFACT_IDENTITY
```

## VII. Failure vocabulary

- `intentio_collapse`
- `number_quality_identity_collapse`
- `quality_as_ontological_cause`
- `multiplication_as_constituent_cause`
- `base_as_number_cause`
- `notation_as_number`
- `false_equation_qualitatively_ratified`
- `unregistered_quality_invented`
- `complete_graph_as_free_traversal`
- `graph_projection_as_ontology`
- `task_graph_as_class_reordering`
- `carrier_id_as_number_identity`
- `display_glyph_as_carrier_identity`
- `carrier_as_structural_index`
- `cardinal_layer_as_ordinal_position`
- `correlatives_reduced_to_alternatives`
- `transition_sign_as_agent_patient_selector`
- `graph_source_as_agent_without_binding`
- `arithmetic_binding_as_graph_ontology`
- `shared_semantic_root_as_artifact_identity`
- `unresolved_latin_form_presented_as_attested`
- `cardinal_quality_as_graph_transition_type`
- `phoenician_letter_as_historical_numeral_sign`
- `hebrew_fallback_as_second_carrier`
- `rtl_rendering_as_storage_order`

## VIII. Regulated generation law

ARITHMETICA is not limited to the fixed `1..10` vocabulary. That vocabulary is
the first bounded fixture and canonical regression surface.

The general object is:

```text
ARITHMETICA_PROBLEM :=
  <DOMAIN, STATE_SCHEMA, RULE_SET, CONSTRAINTS, OBJECTIVE,
   REDUCTION_POLICY, TERMINATION_POLICY, PROJECTION_POLICY>
```

Rules MAY be arbitrary in content. They MUST NOT be arbitrary in declaration.
Every rule exposes:

```text
ARITHMETICA_RULE:
  rule_id
  version
  source_ref
  domain
  codomain
  arity
  direction
  input_types
  preconditions
  transition_or_derivation
  truth_or_validity_check
  output_types
  reversibility
  cost_or_priority
  termination_effect
```

An undeclared function, intuition, analogy, model inference, or convenient
result is not a rule.

## IX. State-space doctrine

A state is a typed possibility, not merely a number:

```text
ARITHMETICA_STATE:
  state_id
  quantitative_carriers[]
  qualitative_assignments[]
  active_plane
  rule_history[]
  constraint_status
  provenance
  parent_state_refs[]
  depth
```

The initial space `Ω0` contains every possibility admitted by the declared
domain and state schema. Controlled generation constructs a traceable subset:

```text
Ω0
  -> TYPE_FILTER
  -> RULE_APPLICABILITY
  -> CONSTRAINT_FILTER
  -> EQUIVALENCE_DEDUPLICATION
  -> GOAL_OR_RELEVANCE_FILTER
  -> RESOURCE_BOUND
  -> Ω*
```

`Ω*` is a reduction of the considered state space, not a claim that rejected or
unvisited states are ontologically impossible.

Every removed candidate requires a reason category:

- type-invalid;
- rule-inapplicable;
- constraint-violating;
- equivalent duplicate;
- dominated;
- out of scope;
- over budget;
- nonterminating under the declared policy;
- unresolved.

## X. Quantitative and qualitative directions

ARITHMETICA supports two non-collapsible directions:

```text
EX_QUANTITAS:
  quantitative states or relations
    -> admitted qualitative elaborations

EX_QUALITAS:
  qualitative requirements or patterns
    -> quantitative constraints
    -> candidate quantitative states or relations
```

The reverse direction is constraint compilation and search. It is not presumed
to be a mathematical inverse, unique, total, or lossless.

`QUANTITAS_ET_QUALITAS` maintains both surfaces and their derivational bridge.
No qualitative adequacy proves a quantitative statement, and no quantitative
coincidence proves one qualitative interpretation.

## XI. Foundation of GRAPHUS COMMUNIS

TENET ARITHMETICA provides the foundational capability by which GRAPHUS
COMMUNIS may represent:

- typed states as carriers;
- rules as candidate transitions or hypertransitions;
- quantitative weights, costs, ranks, and bounds;
- state deltas;
- equivalence and deduplication;
- admissible state-space reduction;
- derivational provenance.

This dependency does not collapse the classes:

```text
TENET_ARITHMETICA != GRAPHUS
ARITHMETICA_RULE != CANONICAL_GRAPH_EDGE
STATE_SPACE != GRAPH_TOPOLOGY
RULE_APPLICATION != TRAVERSAL_PERMISSION
STATE_REDUCTION != ONTOLOGICAL_DELETION
```

GRAPHUS owns the declared graph object and instance topology. TENET
ARITHMETICA owns the invariants governing quantitative states, rule-derived
relations, and reduction traces consumed by those instances.

Source-declared static topology may be recorded before a quantitative state
model is bound. Any generated, scored, pruned, or state-mutating GRAPHUS
instance MUST resolve an admitted `ARITHMETIC_STATE_SPACE_CAPABILITY`.

## XII. Derivational substrate

`LIBER_EX_OPERAE_ARITHMETICA.md` and
`ARITHMETICA_OPERAE_DERIVATION_DATASET.json` preserve the historical OPERAE as
test and calibration substrate. They do not execute and do not override this
TENET.

## XIII. Extended failures

- `rule_content_used_without_rule_declaration`
- `state_space_unbounded_without_policy`
- `candidate_pruned_without_reason`
- `qualitative_projection_as_quantitative_proof`
- `reverse_projection_claimed_unique_without_proof`
- `state_reduction_as_ontological_deletion`
- `arithmetic_rule_as_canonical_graph_edge`
- `graph_topology_as_numeric_truth`
- `graphus_without_arithmetic_foundation_for_generated_state`

## XIV. Multi-OPERA ordinal modes

TENET ARITHMETICA assigns a qualitative mode to each ordinal position in an
OPERA history. The mode belongs to the execution position, not permanently to
the OPERA identity.

| Position | Carrier quality | Multi-OPERA mode |
|---:|---|---|
| 1 | identity | `IDENTIFICATIO` |
| 2 | distinction | `DIFFERENTIATIO` |
| 3 | mediation | `MEDIATIO` |
| 4 | stabilization | `STABILIZATIO` |
| 5 | asymmetry | `DESTABILIZATIO_ASYMMETRIAE` |
| 6 | synthesis | `SYNTHESIS` |
| 7 | rupture | `RUPTURA` |
| 8 | amplification | `AMPLIFICATIO` |
| 9 | closure | `CLAUSURA` |
| 10 | transition | `TRANSITIO` |

`DESTABILIZATIO_ASYMMETRIAE` operationalizes the fifth carrier's asymmetry: it
tests or disturbs an achieved stabilization so hidden asymmetry becomes
visible. It does not change the registered carrier quality from `asymmetry`.

When the user explicitly requests conceptual stabilization, the minimal
complete sequence contains four admitted OPERAE:

```text
OPERA_1 / IDENTIFICATIO
  -> OPERA_2 / DIFFERENTIATIO
  -> OPERA_3 / MEDIATIO
  -> OPERA_4 / STABILIZATIO
```

The OPERAE may have any registered identities compatible with the objective.
Examples:

```text
ATSQ:
  A = IDENTIFICATIO
  T = DIFFERENTIATIO
  S = MEDIATIO
  Q = STABILIZATIO

QATS:
  Q = IDENTIFICATIO
  A = DIFFERENTIATIO
  T = MEDIATIO
  S = STABILIZATIO
```

No OPERA owns one of these modes intrinsically.

The table is an ordinal projection of cardinal qualities, not an identity
between the two axes:

```text
CARDINAL_LAYER n != ORDINAL_EXECUTION_POSITION n

ORDINAL_EXECUTION_POSITION n
    MAY RECEIVE
CARDINAL_QUALITY n
    AS A LOCAL MODE
```

A cardinal-2 transition operator may occur at any execution ordinal. The second
OPERA may receive `DIFFERENTIATIO` without becoming carrier 2.

## XV. Prospective continuation heuristic

The assistant or coordinator SHOULD maintain the current ordinal quality even
when the user has not declared a multi-OPERA plan. This is prospective planning
state, not authority to execute another OPERA.

Before proposing continuation it tests:

1. what state delta the next ordinal mode can produce;
2. whether that delta is already present;
3. whether another OPERA can realize the mode lawfully;
4. whether the next mode clarifies, differentiates, mediates, stabilizes,
   destabilizes, synthesizes, ruptures, amplifies, closes, or transitions;
5. whether stopping is more faithful than redundant generation.

After stabilization, additional OPERAE require explicit benefit:

- position 5 exposes asymmetry or destabilizes the current settlement;
- position 6 synthesizes what survived;
- position 7 introduces a declared rupture;
- position 8 amplifies without merely repeating;
- position 9 closes;
- position 10 transitions to a new scale or cycle.

The heuristic MUST reduce degeneration by excess operations. Reachability,
available context, or unused OPERAE are not reasons to continue.

## XVI. Multi-OPERA trace

```text
ARITHMETICA_OPERA_HISTORY:
  history_id
  objective
  stabilization_requested
  entries:
    - ordinal
      arithmetic_carrier
      inherited_quality
      effective_mode
      opera_ref
      arepo_decision_ref
      input_state_ref
      output_state_ref
      state_delta
      redundancy_check
      continuation_reason
  current_quality
  next_mode_candidate
  stop_reason
```

Every later OPERA receives the history as typed input. It MUST NOT erase or
reinterpret earlier ordinal assignments.

## XVII. Shared OPERA request and result invariants

Every ARITHMETICA OPERA receives the same typed envelope:

```text
OPERA_ARITHMETICA_REQUEST:
  operation_id
  request_id
  arepo_decision_id
  payload
  declared_layer
  declared_base
  route_id
  output_mode
  session_id
  rule_set_ref
  state_space_ref
  ordinal_mode_binding_ref
  graphus_binding_ref
```

Every OPERA MUST:

1. verify its exact identity and artifact version;
2. verify that AREPO admission names that same OPERA;
3. preserve original carrier, notation, layer, base, and provenance;
4. use only declared rules and admitted transitions;
5. separate quantitative, qualitative, relational, logical, and graph layers;
6. retain accepted, rejected, deferred, and deduplicated branches;
7. return typed result, state delta, limits, and trace to SATOR.

These requirements are shared doctrine. An individual OPERA artifact specifies
only its native transformation and OPERA-specific failures.

## XVIII. Shared normalization, generation, and trace law

Normalization is common:

```text
normalize_number(intrinsic, n) -> n

normalize_number(base_manifest, carriers, base)
  -> validate manifest
  -> derive intrinsic integer

normalize_number(positional, digits, base)
  -> validate every digit against base
  -> decode positional value
```

Decoded value, original notation, and base remain separate.

All three OPERAE instantiate this regulated engine:

```text
GENERATIO_ARITHMETICA_REGULATA(problem, admission):
  bind admitted domain, state schema, seeds, rules, constraints, and budgets
  initialize accepted, rejected, deferred, deduplicated, and frontier registers
  while frontier is non-empty and termination policy permits:
    select a state by the declared policy
    enumerate applicable declared rules
    materialize each candidate
    verify quantitative truth or declared rule validity
    apply type, constraint, equivalence, dominance, and resource gates
    retain, reject, defer, or deduplicate with an explicit reason
    expand only retained states whose expansion remains admitted
  project retained states in the OPERA's native direction
  return result, state delta, reduction trace, and unresolved set
```

Every rule adapter conforms to:

```text
ARITHMETICA_RULE_ADAPTER:
  adapter_id
  operator_or_relation
  input_arity
  output_arity
  parse
  applicable
  apply
  verify
  canonicalize
  inverse_or_reverse_search
  cost
  provenance
```

Every OPERA returns:

```text
ARITHMETICA_GENERATION_TRACE:
  execution_id
  operation_id
  request_id
  arepo_decision_id
  problem_version
  rule_set_version
  domain_ref
  seed_state_refs[]
  generated_state_count
  accepted_state_refs[]
  rejected[]:
    state_ref
    stage
    reason
    evidence_ref
  deduplicated[]:
    removed_ref
    representative_ref
    equivalence_ref
  deferred_state_refs[]
  quantitative_proofs[]
  qualitative_projection_records[]
  actual_state_delta
  termination_reason
  budget_consumed
  graphus_binding_ref
  warnings[]
  status
```

This is a decision and transformation trace, not hidden chain-of-thought.

## XIX. Shared handoff, ordinal, and GRAPHUS binding

Any native OPERA may emit:

```text
OPERA_HANDOFF:
  producer_opera
  producer_trace
  typed_result
  admissible_consumers[]
  required_revalidation
  expiry_condition
```

A handoff does not schedule, admit, or execute another OPERA.

When a multi-OPERA history participates, each artifact receives the
`OPERA_ORDINAL_MODE_BINDING` defined by Sections XIV-XVI. The mode constrains
the purpose of that execution without changing its OPERA identity. A zero or
equivalent delta triggers redundancy review and cannot justify automatic
continuation.

When GRAPHUS participates, every OPERA emulates traversal by session
self-binding:

```text
prior state
  -> admitted rule or edge
  -> active constraints and coherence test
  -> accepted delta
  -> resulting state as the next binding
```

The trace retains materially relevant rejected alternatives. It neither
narrates imaginary movement nor requires a visual skill or invocation-local
graph file. Coherence pressure may reduce admitted possibilities; it cannot
invent topology, rules, authority, or permission.

Shared failures:

- `operation_identity_admission_mismatch`
- `undeclared_rule_application`
- `candidate_branch_untraced`
- `pruning_reason_missing`
- `termination_budget_ignored`
- `qualitative_fit_as_numeric_truth`
- `cyclic_derivation_as_progress`
- `historical_fixture_as_closed_operator_law`
- `zero_delta_as_automatic_continuation`
- `coherence_pressure_invents_transition`

## XX. Final law

Number is constituted numerically.

Quality is carried under a declared mode of intelligibility.

Structure exposes order without creating being.
