HT Layered Authority and Translation Model

Status: Canon procedural and analytical doctrine
Language: Hadokai Tubatonona / HT
Scope: Validation authority, semantic confirmation, translation layering, canon-status reporting, and responsible HT generation or interpretation

1. Purpose

Hadokai Tubatonona requires a clear distinction between forms that are structurally possible, forms that are lexically attested, forms that are morphologically licensed, and forms whose meanings may be canonically asserted.

This doctrine prevents two related errors.

False canon occurs when a structurally valid or plausible HT form is treated as fully canon-confirmed.

English flattening occurs when an awkward literal English gloss is treated as evidence that the HT form itself is defective.

The governing principle is:

Validation prevents false canon. Translation layering prevents false English flattening.

The central diagnostic question is:

At what layer does this form have authority?

This question governs both HT analysis and HT generation.

2. Relationship To HT Deconstruction

The HT Deconstruction doctrine classifies word-forms by structural and lexical status: canon defined, canon confirmed, canon structured, or canon invalid.

This doctrine does not replace that taxonomy. It extends it.

Deconstruction determines authority at the form and decomposition level. The Layered Authority model carries those results upward into phrase, clause, sentence, semantic, and translation analysis.

When the two doctrines overlap, the deconstruction result governs word-form status. This model then reports how that status affects higher-level interpretation and generation.

3. Core Principle

Lower-layer validity does not confer higher-layer authority.

A form may pass one layer of analysis while remaining unresolved at the next. Each step must be earned separately.

The task is not merely to decide whether something is right or wrong. The task is to identify the highest layer at which the form has authority.

4. Authority Ladders

HT authority is best understood as two related ladders, not one continuous ladder.

The first ladder evaluates HT validity and canon authority.

The second ladder evaluates English rendering and translation authority.

They touch at semantic authority, but they are not the same ladder. A contextual English expansion may be useful without carrying canon authority.

4.1 Validation Authority Ladder

Strict-valid does not mean lexicon-attested.
A form may obey HT structure without being a defined word.

Lexicon-attested does not mean morphologically licensed in this use.
A word may be valid by itself while its current construction, category shift, or suffix behavior remains exposed.

Morphologically licensed does not mean phrase-valid.
A word-form may be valid internally while occupying an uncertain or invalid phrase role.

Phrase-valid does not mean clause-valid.
Words may be ordered correctly inside a phrase while the larger sentence architecture remains unresolved.

Clause-valid does not mean semantically confirmed.
A clause may have recognizable object, subject, time, verb, and closure structure while depending on an unconfirmed semantic relation, predicate relation, or modifier attachment.

Semantically confirmed does not automatically mean canon-defined as a dictionary meaning.
A sentence reading may be supported in context without promoting every component into a new lexical definition.

4.2 Translation Authority Ladder

Literal gloss does not mean conceptual reading.

Conceptual reading does not mean natural English translation.

Natural English translation does not mean contextual expansion.

Contextual expansion does not mean dictionary meaning.

The translation ladder protects English from flattening HT. The validation ladder protects HT from false canon.

5. Licensed

In this doctrine, licensed means supported by at least one of the following:

attested lexicon entry
documented grammar
direct canon example
direct attested usage
explicitly accepted productive pattern

A form is not licensed merely because it is plausible, elegant, familiar, or decomposable.

A decomposition may explain a form. A translation may clarify a form. A contextual example may illustrate a form. None of those alone makes the form canon-confirmed.

6. HT Validation Layer

The HT Validation Layer asks:

Is this valid HT, and at what authority level?

It evaluates the HT form itself before evaluating the English translation.

Hadokai Tubatonona is natively written in the hadohaij: the HT writing system. Each written character is a hadohaijkai, and each hadohaijkai represents a sanakai: a distinct phoneme or unique sound. Romanized forms are therefore analytical representations for accessibility, instruction, input, and validation. They are not the native writing system itself.

Validation checks:

strict form
lexeme attestation
morphological authority
phrase authority
clause authority
semantic authority
canon status

Each layer must be considered separately.

7. Input Classification

Strict romanization explicitly marks ungU, the silent consonant, where required by HT syllable structure.

Casual romanization omits ungU on the surface.

Malformed or incomplete strict candidate contains partial or inconsistent strict notation.

English intent is a proposed meaning to test against HT.

Mixed form is not a canonical mode. A partially marked form is a malformed or incomplete strict candidate, possibly recoverable for analysis.

Casual romanization is an underspecified surface representation of a fully defined strict form. The strict form remains the canonical analytical record. Casual input may therefore be valid for reading or drafting while still requiring strict reconstruction before full canon authority can be asserted.

If one strict reconstruction is responsible, casual input may be casual-valid, strict-recoverable. If multiple reconstructions are possible, it is casual-valid, strict-ambiguous until context, lexicon entry, or canon resolves it.

8. Strict Form Validation

Strict validation checks whether each written token can exist as strict HT.

It tests legal sanakai inventory, CVC slot structure, ungU placement, vowel boundaries, word-edge behavior, consonant clusters, and boundary realization.

Each strict syllable occupies three structural positions:

opening consonant position (hadohatiSe)
vowel or syllable-set structure (hadohatiij)
closing consonant position (hadohatinidok)

After these positions are established, this doctrine uses the ordinary analytical terms opening position, vowel position, and closing position for readability.

Outcomes:

Strict form: valid
The written strict form is correct as written.

Strict form: invalid, recoverable correction available
The written form is wrong, but the intended form can be responsibly repaired.

Strict form: invalid, unrecoverable
The written form cannot be repaired without guessing.

Recoverable does not mean valid.

Example:

cImja=vi=val=e= is defensible.

cImja=vi=val==e= is invalid overmarking, though recoverable, because val already closes with l.

The intended concept is clear:

cImjavival + e
woodcutter + associated participant

But the strict realization must respect the existing final consonant in val. Since val already closes with l, adding e produces val=e=, not val==e=.

9. Doubled UngU

Doubled == is valid only when it represents two real ungU consonant positions: one closing a syllable and one opening the next.

It is not punctuation.
It is not a spacer.
It is not a decorative joiner.
It is not a general suffix marker.

garoca==e= can be valid when open ca= meets =e=.

val=e= is valid because val already closes with l.

val==e= is invalid because it inserts an extra ungU after an already closed syllable.

The rule is structural:

Doubled ungU is valid only when both ungU positions are real.

This prevents two opposite errors. It prevents invalid overmarking, but it also prevents valid doubled ungU from being removed by a simplistic cleanup rule.

10. Lexeme Authority

Lexeme authority asks whether each token is known, decomposable, unresolved, or invalid.

Authority order:

direct lexicon entry
canon-confirmed decomposition
canon-structured decomposition
unresolved decomposition
invalid form

Direct lexicon entries outrank decomposition.

A plausible decomposition may explain a form, but it does not canonize a new word by itself. If a word is directly attested in the lexicon, that entry governs unless canon explicitly revises it.

This is essential for AI training. A model must not treat every combination of attested parts as a new canon-defined lexical item.

11. Morphological Authority

Morphological authority checks whether internal word-building is licensed.

It checks:

rightward construction
attested roots or class roots
valid suffix behavior
valid category shifts
valid suffix order
valid participation marking
absence of invented productive prefixing
absence of unlicensed compound promotion

A form may be morphologically structured without being canon confirmed.

This means its parts can be identified and its construction can be plausible, but the whole form is not yet a canon-defined lexical unit.

12. Phrase Authority

Declarative phrase order is:

object phrase + subject phrase + optional time marker + verb phrase + closing boundary

Important principles:

secondary elements precede primary elements
the primary element sits nearest the next resolution point
the primary subject remains bare
primary-subject descriptors displace onto the verb
secondary subjects may carry descriptors
e marks association or accompaniment
shared participation must be confirmed on the verb with no

Participation marking has four values in the primary verb slot:

na marks singular participation
no marks plural or shared participation
zi marks singular reflexive participation
zo marks plural reflexive participation

The same layered authority checks apply to all four. This doctrine’s association examples focus on e + no because they concern shared participation, but reflexive constructions must be evaluated through the same participation-slot authority.

Detached descriptors and canon-compatible attachments must be assigned to a target phrase. If more than one target is plausible, the sentence is structurally parseable but semantically exposed.

A sentence may pass phrase authority even if the final meaning remains only canon structured.

13. Clause Authority

A basic declarative clause follows HT sentence architecture.

A complex conditional follows:

main clause + aS + subordinating conjunction + dependent clause + closure

Known subordinators include at, yazu, Im, buIm, and zeotze.

Each clause must preserve its own HT order. A dependent clause does not escape HT architecture simply because it is dependent.

14. Semantic Authority

Semantic authority asks whether the intended meaning is licensed.

It asks:

Does grammar license this relation?
Does the lexicon license this meaning?
Is the noun phrase attested or newly composed?
Is the predicate or complement relation allowed?
Are modifiers attached to the intended target?
Could another parse compete?
Is the reading confirmed or merely structured?

This is where a sentence may be structurally valid but not canon confirmed.

A form can be lawful, analyzable, and meaningful while still lacking full semantic authority.

15. Canon Status Ladder

The Deconstruction doctrine supplies the core word-form statuses: canon defined, canon confirmed, canon structured, and canon invalid. This doctrine preserves those statuses and adds intermediate reporting states for sentence-level, semantic, and translation work.

Strict invalid – the form fails strict structure.

Strict invalid, recoverable correction available – the form is wrong, but responsibly repairable.

Strict invalid, unrecoverable – repair would require guessing.

Casual-valid, strict-recoverable – valid casual input with one responsible strict reconstruction.

Casual-valid, strict-ambiguous – valid casual input with multiple possible strict reconstructions.

Lexically unresolved – structurally possible but not attested or resolved.

Morphologically structured – components parse, but the whole form is not canon-confirmed.

Structurally parseable – sentence slots and grammar can be identified.

Canon structured – lawful and analyzable, with a plausible reading, but not fully licensed.

Canon defined – the form or meaning is directly attested by a lexicon entry, canon example, or explicit canonical statement.

Canon confirmed – strict form, lexemes, morphology, grammar, phrase relations, and intended semantic relation are licensed by rule, direct lexicon entry, or direct attested usage.

Canon invalid – a required layer fails.

Canon structured is not failure. It means plausible, but not canonized.

This status allows the system to say:

This can be analyzed, but this meaning cannot yet be asserted as canon confirmed.

16. HT Translation Layer

The HT Translation Layer asks:

What kind of English rendering are we producing, and how much interpretation has entered?

This layer must remain separate from validation.

A translation problem is not automatically an HT problem.

Translation layers:

HT form
The expression as written.

Literal gloss
Closest word-by-word or morpheme-by-morpheme rendering.

Conceptual reading
Clear explanation of the HT idea while preserving HT structure.

Natural English translation
Readable English rendering.

Contextual expansion
Scene-specific, cultural, symbolic, or narrative rendering.

Contextual expansion may be useful, but it carries limited translation authority and no canon authority unless canon explicitly promotes it.

Example:

bapavati liku la

Literal gloss: notes hidden safe
Conceptual reading: small writings concealed and kept safe
Natural English: the hidden, protected notes
Contextual expansion: secret writings kept safe from discovery

A strained literal gloss may indicate:

a problem in the HT
an overly literal English gloss
an HT concept that does not map cleanly into English packaging

Only further analysis determines which.

17. Contextual Expansion and No-Promotion

Contextual expansion must never be back-projected into the lexicon unless canonized.

If bapavati liku la is translated in one passage as secret writings kept safe from discovery, that does not make bapavati mean secret writing, and it does not make la mean kept from discovery by itself.

Examples can illustrate usage, but an example translation does not redefine a lexeme unless canon explicitly promotes that usage into the entry.

The lexicon changes by canon action, not by accidental drift through poetic, contextual, or expanded translations.

18. Translation Authority

Just as validation has canon status, translation has translation authority.

Literal only – preserves components, but may not explain the concept.

Conceptual reading supported – the HT concept is explainable from licensed parts.

Natural translation supported – readable English faithfully renders the supported concept.

Contextual expansion only – depends on passage context and is not dictionary meaning.

Translation overreach – adds meaning not licensed by HT or context.

Translation authority is not canon authority.

This distinction prevents expanded English from laundering itself into HT canon.

19. Standard Report Format

Every serious generated or analyzed HT sentence should be able to produce a report like this:

HT Validation

Strict form: valid
Lexeme attestation: complete
Morphology: licensed
Phrase structure: valid
Clause structure: valid
Semantic authority: structured
Canon status: canon structured
Notes: predicate relation is plausible but not directly canon confirmed

Translation Layers

Literal gloss: notes hidden safe
Conceptual reading: small writings concealed and kept safe
Natural English: the hidden, protected notes
Contextual expansion: secret writings kept safe from discovery
Translation authority: natural translation supported; contextual expansion is passage-dependent

This format makes the analysis accountable. It shows what is known, what is licensed, what is plausible, and what remains interpretive.

20. Generation

HT generation should produce sentences that can survive layered validation, not merely sentences that look grammatical.

HT generation requires at least these core checks:

sentence architecture
lexicon attestation
morphological authority
strict syllable and boundary realization
semantic authority
translation-layer discipline

A generated sentence may be structurally valid and still fail at a higher layer.

A generated sentence may also be conceptually valid while containing a strict-form error.

For example, cImjavival + e is a valid intended construction for woodcutter + associated participant, but cImja=vi=val==e= is not the correct strict realization. The correct form is cImja=vi=val=e=.

The concept may be right while the strict form is still wrong.

21. Interpretation

HT interpretation should not collapse directly from form to meaning.

A responsible interpretation proceeds in layers:

validate the form
check the lexicon
test morphology
identify phrase and clause structure
evaluate semantic authority
produce translation layers

This allows an interpreter to say:

The sentence is structurally parseable.

The reading is canon structured.

The meaning is canon confirmed.

The English expansion is contextual, not lexical.

This is especially important for AI training. A model must learn to stop at the correct authority layer. It must be able to say:

I can parse this, but I cannot assert that meaning.

That is not failure. That is correct HT behavior.

22. Short Doctrine

Lower-layer validity does not confer higher-layer authority.

HT authority has two ladders: validation authority and translation authority.

The validation ladder and translation ladder touch at semantic authority, but they are not the same ladder.

Hadokai Tubatonona is written in hadohaij; romanization is an analytical representation, not the native writing system itself.

Strict validation depends on legal sanakai, complete CVC structure, and correct ungU placement.

Licensed means supported by lexicon, grammar, canon example, attested usage, or accepted productive pattern.

Mixed form is not a canonical mode.

Casual-valid does not always mean strict-unambiguous.

Recoverable does not mean valid.

Doubled ungU is structural, not decorative.

Direct lexicon entries outrank decomposition.

Canon structured is a real status.

Semantic authority is required before meaning can be canon confirmed.