Translate Precisely | Split Intransitivity & Active

Split intransitivity and active–stative alignment in translation describe systems in which the sole argument of an intransitive verb does not always receive one uniform subject pattern. Some intransitive participants are coded like the agent of a transitive clause, while others are coded like the patient. Linguistic descriptions use terms such as active, active–stative, agentive, active–inactive, split-S and fluid-S for related systems. The exact conditioning can involve lexical verb class, volition, control, affectedness, event construal or historical lexicalisation.

People searching for active stative alignment, split intransitivity, split-S, fluid-S, active inactive alignment, agentive patientive case are solving an alignment problem that English mostly hides. English says I ran, I fell, I slept with the same nominative subject form, but an active–stative language may mark or index the intransitive participant differently depending on whether the event is agent-like or patient-like. Translating case or agreement mechanically as subject versus object can therefore make a perfectly ordinary intransitive event look transitive or passive.

This article owns the split-intransitive and active–stative precision layer for eduKateSG. It does not replace Grammatical Case, Ergative–Absolutive Alignment, Lexical Aspect or Verb Valency. Its narrow task is to distinguish A-like and P-like intransitive coding, separate lexically fixed split-S systems from context-sensitive fluid-S systems, preserve volition, control and affectedness contrasts where they are grammatically meaningful, and rebuild natural target intransitives without inventing objects or changing agency.

Quick Read

Map A = transitive agent, P = transitive patient, S = sole intransitive participant. Then ask whether this S is coded like A or like P. Next ask why: lexical class or contextual event construal? Only then translate case or agreement into a normal target intransitive clause.

One-Sentence Answer

Translate active–stative alignment by preserving the source’s agent-like versus patient-like construal of the intransitive participant without turning that participant into an actual transitive agent or object.

Why This Precision Layer Matters

Split-S and fluid-S systems expose distinctions English often leaves implicit. A person who deliberately performs an activity can pattern differently from someone who undergoes an involuntary state or event. If a translator reads patient-like S marking as object marking, they can invent an absent subject or passive agent; if they ignore a fluid-S alternation, they can lose a contrast in control or affectedness.

A, P and S

Alignment analysis compares the transitive agent A, transitive patient P and sole intransitive participant S.

The main translation risk is calling every patient-like form an object without first comparing the alignment system. Because the target can remain grammatical, ordinary fluency review may not expose the error. A precision review therefore reconstructs the alignment pattern and event construal independently of target pronoun shape or word order.

A practical diagnostic is to establish the coding of A and P before interpreting S. For example, S in ‘fall’ may share morphology with P even though the clause has no object Once the source relation is secure, the target can use an ordinary intransitive subject, an adverbial for control or affectedness, or no overt equivalent for the source case/agreement distinction. The target should preserve the event, not imitate patient-like or agent-like morphology.

For recovery, restore the one-participant event first, then recover only those volition, control or affectedness distinctions that are genuinely active in the source construction. Remove any invented object, causer or passive meaning before polishing style.

Active–Stative Alignment

An active–stative system splits intransitive S arguments between A-like and P-like coding.

The main translation risk is treating P-like S as evidence of passive voice. Because the target can remain grammatical, ordinary fluency review may not expose the error. A precision review therefore reconstructs the alignment pattern and event construal independently of target pronoun shape or word order.

A practical diagnostic is to ask whether the verb is actually intransitive. For example, ‘I fell’ can have patientive coding but still contain only one participant Once the source relation is secure, the target can use an ordinary intransitive subject, an adverbial for control or affectedness, or no overt equivalent for the source case/agreement distinction. The target should preserve the event, not imitate patient-like or agent-like morphology.

Split-S Systems

In split-S systems, individual intransitive predicates tend to belong to relatively fixed lexical classes selecting A-like or P-like coding.

The main translation risk is making marking depend on contextual volition when the verb class is lexically fixed. Because the target can remain grammatical, ordinary fluency review may not expose the error. A precision review therefore reconstructs the alignment pattern and event construal independently of target pronoun shape or word order.

A practical diagnostic is to check whether the same verb alternates across contexts. For example, RUN may always take active-series marking while SLEEP takes inactive-series marking Once the source relation is secure, the target can use an ordinary intransitive subject, an adverbial for control or affectedness, or no overt equivalent for the source case/agreement distinction. The target should preserve the event, not imitate patient-like or agent-like morphology.

Fluid-S Systems

In fluid-S systems, at least some intransitive predicates can take either A-like or P-like coding with a meaning contrast.

The main translation risk is treating alternate marking as free variation. Because the target can remain grammatical, ordinary fluency review may not expose the error. A precision review therefore reconstructs the alignment pattern and event construal independently of target pronoun shape or word order.

A practical diagnostic is to identify the contextual semantic difference. For example, the same verb may use A-like coding for controlled action and P-like coding for involuntary occurrence Once the source relation is secure, the target can use an ordinary intransitive subject, an adverbial for control or affectedness, or no overt equivalent for the source case/agreement distinction. The target should preserve the event, not imitate patient-like or agent-like morphology.

Volition

Voluntary versus involuntary participation is a common but not universal conditioning factor.

The main translation risk is assuming every A-like S is intentional and every P-like S accidental. Because the target can remain grammatical, ordinary fluency review may not expose the error. A precision review therefore reconstructs the alignment pattern and event construal independently of target pronoun shape or word order.

A practical diagnostic is to check the source language’s actual semantic conditioning. For example, a fluid-S language may contrast deliberate coughing with involuntary coughing Once the source relation is secure, the target can use an ordinary intransitive subject, an adverbial for control or affectedness, or no overt equivalent for the source case/agreement distinction. The target should preserve the event, not imitate patient-like or agent-like morphology.

Control

Control over initiation or continuation can influence active–stative marking.

The main translation risk is conflating control with moral responsibility. Because the target can remain grammatical, ordinary fluency review may not expose the error. A precision review therefore reconstructs the alignment pattern and event construal independently of target pronoun shape or word order.

A practical diagnostic is to ask whether the participant controls the event. For example, intentional movement versus slipping may differ Once the source relation is secure, the target can use an ordinary intransitive subject, an adverbial for control or affectedness, or no overt equivalent for the source case/agreement distinction. The target should preserve the event, not imitate patient-like or agent-like morphology.

Affectedness

Patient-like S coding can correlate with being affected, undergoing change or experiencing a state.

The main translation risk is turning affectedness into an external agent-caused event. Because the target can remain grammatical, ordinary fluency review may not expose the error. A precision review therefore reconstructs the alignment pattern and event construal independently of target pronoun shape or word order.

A practical diagnostic is to keep the clause intransitive unless source adds a causer. For example, become sick can pattern P-like without anyone causing it Once the source relation is secure, the target can use an ordinary intransitive subject, an adverbial for control or affectedness, or no overt equivalent for the source case/agreement distinction. The target should preserve the event, not imitate patient-like or agent-like morphology.

States

Many active–stative systems place stative predicates with patient-like or inactive coding, though exact patterns differ.

The main translation risk is generalising all states into one class. Because the target can remain grammatical, ordinary fluency review may not expose the error. A precision review therefore reconstructs the alignment pattern and event construal independently of target pronoun shape or word order.

A practical diagnostic is to consult lexical and paradigm evidence. For example, know, be tired and stand may pattern differently across languages Once the source relation is secure, the target can use an ordinary intransitive subject, an adverbial for control or affectedness, or no overt equivalent for the source case/agreement distinction. The target should preserve the event, not imitate patient-like or agent-like morphology.

Activities

Controlled activities often favour A-like coding but can be lexically or historically exceptional.

The main translation risk is assuming semantics predicts every verb perfectly. Because the target can remain grammatical, ordinary fluency review may not expose the error. A precision review therefore reconstructs the alignment pattern and event construal independently of target pronoun shape or word order.

A practical diagnostic is to treat semantic tendency and lexical class separately. For example, run may be A-like even in involuntary context under split-S Once the source relation is secure, the target can use an ordinary intransitive subject, an adverbial for control or affectedness, or no overt equivalent for the source case/agreement distinction. The target should preserve the event, not imitate patient-like or agent-like morphology.

Change of State

Events such as fall, break, become ill or die commonly exhibit patient-like properties.

The main translation risk is translating patientive morphology as passive ‘was fallen’ or ‘was died’. Because the target can remain grammatical, ordinary fluency review may not expose the error. A precision review therefore reconstructs the alignment pattern and event construal independently of target pronoun shape or word order.

A practical diagnostic is to test whether there is any external causer. For example, the person simply fell Once the source relation is secure, the target can use an ordinary intransitive subject, an adverbial for control or affectedness, or no overt equivalent for the source case/agreement distinction. The target should preserve the event, not imitate patient-like or agent-like morphology.

Motion Predicates

Motion can pattern as agent-like when controlled or patient-like when uncontrolled, depending on the language.

The main translation risk is assuming all movement is agentive. Because the target can remain grammatical, ordinary fluency review may not expose the error. A precision review therefore reconstructs the alignment pattern and event construal independently of target pronoun shape or word order.

A practical diagnostic is to test source verb class and event construal. For example, go intentionally versus drift or fall can receive different coding Once the source relation is secure, the target can use an ordinary intransitive subject, an adverbial for control or affectedness, or no overt equivalent for the source case/agreement distinction. The target should preserve the event, not imitate patient-like or agent-like morphology.

Experiencer Predicates

Feeling, perception and cognition can use P-like, A-like or dedicated patterns.

The main translation risk is forcing experiencers into English object roles based on case labels. Because the target can remain grammatical, ordinary fluency review may not expose the error. A precision review therefore reconstructs the alignment pattern and event construal independently of target pronoun shape or word order.

A practical diagnostic is to identify EXPERIENCER and STIMULUS roles. For example, a ‘me is cold’ type structure can still translate ‘I am cold’ Once the source relation is secure, the target can use an ordinary intransitive subject, an adverbial for control or affectedness, or no overt equivalent for the source case/agreement distinction. The target should preserve the event, not imitate patient-like or agent-like morphology.

Case Marking

Active–stative alignment can be expressed through case on noun phrases.

The main translation risk is translating active or patientive case as English nominative or accusative mechanically. Because the target can remain grammatical, ordinary fluency review may not expose the error. A precision review therefore reconstructs the alignment pattern and event construal independently of target pronoun shape or word order.

A practical diagnostic is to map semantic role before target case. For example, P-like S case may look object-like but functions as the sole argument Once the source relation is secure, the target can use an ordinary intransitive subject, an adverbial for control or affectedness, or no overt equivalent for the source case/agreement distinction. The target should preserve the event, not imitate patient-like or agent-like morphology.

Agreement and Indexing

The split can appear in verbal agreement or pronominal series rather than noun case.

The main translation risk is missing the alignment because nouns are unmarked. Because the target can remain grammatical, ordinary fluency review may not expose the error. A precision review therefore reconstructs the alignment pattern and event construal independently of target pronoun shape or word order.

A practical diagnostic is to compare person-index series. For example, one set may index A and active S, another P and inactive S Once the source relation is secure, the target can use an ordinary intransitive subject, an adverbial for control or affectedness, or no overt equivalent for the source case/agreement distinction. The target should preserve the event, not imitate patient-like or agent-like morphology.

Pronominal Series

Pronouns can have agentive and patientive forms corresponding to different intransitive predicate classes.

The main translation risk is thinking the two pronoun series refer to different participants. Because the target can remain grammatical, ordinary fluency review may not expose the error. A precision review therefore reconstructs the alignment pattern and event construal independently of target pronoun shape or word order.

A practical diagnostic is to link both forms to the same referent under different event types. For example, agentive ‘I’ and patientive ‘me-like’ form can both refer to the speaker Once the source relation is secure, the target can use an ordinary intransitive subject, an adverbial for control or affectedness, or no overt equivalent for the source case/agreement distinction. The target should preserve the event, not imitate patient-like or agent-like morphology.

Lexicalisation

Historical motivations can become opaque, leaving verbs whose marking no longer matches simple semantic expectations.

The main translation risk is correcting a fixed pattern to fit a neat volition theory. Because the target can remain grammatical, ordinary fluency review may not expose the error. A precision review therefore reconstructs the alignment pattern and event construal independently of target pronoun shape or word order.

A practical diagnostic is to treat irregular class membership as grammar. For example, a seemingly controlled verb may use patientive coding historically Once the source relation is secure, the target can use an ordinary intransitive subject, an adverbial for control or affectedness, or no overt equivalent for the source case/agreement distinction. The target should preserve the event, not imitate patient-like or agent-like morphology.

Diachronic Change

Active or agentive systems often reflect layers of historical development rather than one timeless semantic rule.

The main translation risk is overstating a universal mapping from volition to case. Because the target can remain grammatical, ordinary fluency review may not expose the error. A precision review therefore reconstructs the alignment pattern and event construal independently of target pronoun shape or word order.

A practical diagnostic is to separate synchronic pattern from historical explanation. For example, modern class membership can preserve older semantic distinctions Once the source relation is secure, the target can use an ordinary intransitive subject, an adverbial for control or affectedness, or no overt equivalent for the source case/agreement distinction. The target should preserve the event, not imitate patient-like or agent-like morphology.

Fluid Alternations and Meaning

When the same verb alternates, the contrast can carry real meaning about control, affectedness or event viewpoint.

The main translation risk is normalising both forms to identical English without checking context. Because the target can remain grammatical, ordinary fluency review may not expose the error. A precision review therefore reconstructs the alignment pattern and event construal independently of target pronoun shape or word order.

A practical diagnostic is to compare minimal pairs. For example, A-like and P-like forms of the same event can signal different control Once the source relation is secure, the target can use an ordinary intransitive subject, an adverbial for control or affectedness, or no overt equivalent for the source case/agreement distinction. The target should preserve the event, not imitate patient-like or agent-like morphology.

Split-S vs Fluid-S Diagnosis

The key distinction is whether selection is lexically fixed by verb or contextually variable for the same predicate.

The main translation risk is calling an entire language fluid-S because a few verbs alternate. Because the target can remain grammatical, ordinary fluency review may not expose the error. A precision review therefore reconstructs the alignment pattern and event construal independently of target pronoun shape or word order.

A practical diagnostic is to build a verb-by-context paradigm. For example, most verbs may be fixed with a small alternating subset Once the source relation is secure, the target can use an ordinary intransitive subject, an adverbial for control or affectedness, or no overt equivalent for the source case/agreement distinction. The target should preserve the event, not imitate patient-like or agent-like morphology.

Active–Stative vs Nominative–Accusative

Nominative–accusative alignment groups A and S together while marking P separately.

The main translation risk is treating P-like S as exceptional objecthood. Because the target can remain grammatical, ordinary fluency review may not expose the error. A precision review therefore reconstructs the alignment pattern and event construal independently of target pronoun shape or word order.

A practical diagnostic is to compare the whole alignment system. For example, English nominative I appears for all ordinary S Once the source relation is secure, the target can use an ordinary intransitive subject, an adverbial for control or affectedness, or no overt equivalent for the source case/agreement distinction. The target should preserve the event, not imitate patient-like or agent-like morphology.

Active–Stative vs Ergative–Absolutive

Ergative alignment groups S with P and distinguishes A, while active–stative splits S between A-like and P-like patterns.

The main translation risk is calling active–stative simply split ergativity. Because the target can remain grammatical, ordinary fluency review may not expose the error. A precision review therefore reconstructs the alignment pattern and event construal independently of target pronoun shape or word order.

A practical diagnostic is to map S subsets explicitly. For example, some S align with A and others with P Once the source relation is secure, the target can use an ordinary intransitive subject, an adverbial for control or affectedness, or no overt equivalent for the source case/agreement distinction. The target should preserve the event, not imitate patient-like or agent-like morphology.

Unaccusative/Unergative Comparison

European syntax also distinguishes intransitive predicate types, but unaccusative and unergative analyses are not identical to active–stative morphological alignment.

The main translation risk is assuming the terms are interchangeable. Because the target can remain grammatical, ordinary fluency review may not expose the error. A precision review therefore reconstructs the alignment pattern and event construal independently of target pronoun shape or word order.

A practical diagnostic is to separate syntactic diagnostics from case and agreement alignment. For example, English arrive versus run may differ syntactically without visible active–stative case Once the source relation is secure, the target can use an ordinary intransitive subject, an adverbial for control or affectedness, or no overt equivalent for the source case/agreement distinction. The target should preserve the event, not imitate patient-like or agent-like morphology.

Person Hierarchies

Active marking can interact with person or nominal hierarchies in some systems.

The main translation risk is predicting all pronouns from lexical semantics alone. Because the target can remain grammatical, ordinary fluency review may not expose the error. A precision review therefore reconstructs the alignment pattern and event construal independently of target pronoun shape or word order.

A practical diagnostic is to check person-specific paradigms. For example, first-person forms may display active marking differently from nouns Once the source relation is secure, the target can use an ordinary intransitive subject, an adverbial for control or affectedness, or no overt equivalent for the source case/agreement distinction. The target should preserve the event, not imitate patient-like or agent-like morphology.

Aspect and Tense

Alignment patterns can interact with TAM in some languages.

The main translation risk is attributing every alignment change to volition. Because the target can remain grammatical, ordinary fluency review may not expose the error. A precision review therefore reconstructs the alignment pattern and event construal independently of target pronoun shape or word order.

A practical diagnostic is to record TAM independently. For example, past or perfective context can shift case or agreement in a broader split system Once the source relation is secure, the target can use an ordinary intransitive subject, an adverbial for control or affectedness, or no overt equivalent for the source case/agreement distinction. The target should preserve the event, not imitate patient-like or agent-like morphology.

Negation

Negation can alter auxiliary or agreement patterns without changing the S role.

The main translation risk is mistaking negative morphology for a different alignment class. Because the target can remain grammatical, ordinary fluency review may not expose the error. A precision review therefore reconstructs the alignment pattern and event construal independently of target pronoun shape or word order.

A practical diagnostic is to identify alignment before polarity. For example, I did not fall remains one P-like S event Once the source relation is secure, the target can use an ordinary intransitive subject, an adverbial for control or affectedness, or no overt equivalent for the source case/agreement distinction. The target should preserve the event, not imitate patient-like or agent-like morphology.

Narrative Perspective

Fluid-S choices can characterise whether a participant is portrayed as acting deliberately or undergoing an event.

The main translation risk is flattening both forms and losing narrative stance. Because the target can remain grammatical, ordinary fluency review may not expose the error. A precision review therefore reconstructs the alignment pattern and event construal independently of target pronoun shape or word order.

A practical diagnostic is to track event construal across repeated verbs. For example, a character can alternately control and suffer similar events Once the source relation is secure, the target can use an ordinary intransitive subject, an adverbial for control or affectedness, or no overt equivalent for the source case/agreement distinction. The target should preserve the event, not imitate patient-like or agent-like morphology.

Translation into English

English usually neutralises active–stative coding in ordinary subject forms.

The main translation risk is trying to reproduce patientive S with an English object pronoun. Because the target can remain grammatical, ordinary fluency review may not expose the error. A precision review therefore reconstructs the alignment pattern and event construal independently of target pronoun shape or word order.

A practical diagnostic is to translate the clause as a normal intransitive and preserve semantic nuance separately. For example, patientive ‘me fell’ becomes ‘I fell’ Once the source relation is secure, the target can use an ordinary intransitive subject, an adverbial for control or affectedness, or no overt equivalent for the source case/agreement distinction. The target should preserve the event, not imitate patient-like or agent-like morphology.

Translation into Another Split-S Language

A target active–stative language may classify verbs differently.

The main translation risk is mapping source A-like S directly to target A-like S. Because the target can remain grammatical, ordinary fluency review may not expose the error. A precision review therefore reconstructs the alignment pattern and event construal independently of target pronoun shape or word order.

A practical diagnostic is to decode source semantics then use target lexical classes. For example, RUN can belong to different classes cross-linguistically Once the source relation is secure, the target can use an ordinary intransitive subject, an adverbial for control or affectedness, or no overt equivalent for the source case/agreement distinction. The target should preserve the event, not imitate patient-like or agent-like morphology.

AI Object Bias

Models may interpret patientive case or object-series agreement as a transitive object.

The main translation risk is inventing an absent agent or passive construction. Because the target can remain grammatical, ordinary fluency review may not expose the error. A precision review therefore reconstructs the alignment pattern and event construal independently of target pronoun shape or word order.

A practical diagnostic is to count core participants before translating case labels. For example, one P-like S still equals one-participant clause Once the source relation is secure, the target can use an ordinary intransitive subject, an adverbial for control or affectedness, or no overt equivalent for the source case/agreement distinction. The target should preserve the event, not imitate patient-like or agent-like morphology.

AI Volition Overreach

Models may add deliberately or accidentally whenever they see active or patientive marking.

The main translation risk is lexicalising a grammatical contrast more strongly than source context warrants. Because the target can remain grammatical, ordinary fluency review may not expose the error. A precision review therefore reconstructs the alignment pattern and event construal independently of target pronoun shape or word order.

A practical diagnostic is to test whether the language’s system truly carries that contrast in the specific verb. For example, split-S lexical class may have no live volition contrast Once the source relation is secure, the target can use an ordinary intransitive subject, an adverbial for control or affectedness, or no overt equivalent for the source case/agreement distinction. The target should preserve the event, not imitate patient-like or agent-like morphology.

Worked Examples

Worked Example 1: Fixed Split-S Verb

Source problem: A language always marks RUN with the A-like series and SLEEP with the P-like series.

Diagnosis: The contrast is lexical rather than a free contextual choice.

Target strategy: Translate both as ordinary English intransitives.

Verification: Do not add deliberate to run or involuntary to sleep unless context supports it.

Worked Example 2: Fluid-S Alternation

Source problem: The same intransitive verb can take A-like or P-like marking depending on control.

Diagnosis: The alternation is semantically meaningful.

Target strategy: Use English context or a light adverbial contrast if necessary, such as “deliberately coughed” versus “couldn’t help coughing.”

Verification: Do not create a second participant.

Worked Example 3: Experiencer Predicate

Source problem: A first-person patientive pronoun occurs with a cold or feeling predicate.

Diagnosis: The speaker is sole experiencer, not object of another agent.

Target strategy: Translate “I am cold” or the appropriate English experiencer construction.

Verification: Do not preserve object-case form as “me.”

Worked Example 4: Patient-Like Change of State

Source problem: A P-like S marks someone who fell.

Diagnosis: The participant undergoes change but remains the only core argument.

Target strategy: Use “She fell.”

Verification: Do not make the target passive or invent a causer.

Practical Diagnostic Workflow

Build an ALIGNMENT TABLE: TRANSITIVE A CODING; TRANSITIVE P CODING; INTRANSITIVE VERB; S CODING; FIXED OR ALTERNATING; VOLITION; CONTROL; AFFECTEDNESS; TAM; PERSON. This reveals whether the system is split-S, fluid-S or something else.

Run the one-participant test. Remove case labels and ask how many semantic participants the source event contains. If there is only one, P-like coding does not create an object. Then test whether the same verb can switch coding with a stable meaning contrast; that distinguishes fluid alternation from lexical split.

Verification Method

From the target, reconstruct the same one-participant event and any source-supported contrast in control or affectedness. Confirm that no agent, patient or passive relation has been invented. If the source alternation is merely lexical class, the target need not overtly mark it.

Recovery Method

If an intransitive was mistranslated as transitive or passive, restore the one-participant frame first. Then recover any live volition or control meaning from source context. If a fixed split-S pattern was overinterpreted, remove explanatory adverbs not independently supported.

Practice Sequence

Take RUN, FALL, SLEEP, COUGH, BREAK and FEEL COLD. Classify three as fixed A-like, three as fixed P-like, then allow COUGH to alternate in a hypothetical fluid-S system. Translate each into English and state which source distinctions need overt target wording.

Quality-Assurance Checklist

Establish A and P coding first.
Identify S coding.
Keep one-participant clauses intransitive.
Distinguish split-S from fluid-S.
Do not equate P-like S with passive.
Do not equate A-like S with deliberate action universally.
Track volition only when grammatically active.
Track control and affectedness.
Check lexicalised verb classes.
Separate case from agreement systems.
Separate active–stative from ergative alignment.
Separate active–stative from unaccusativity.
Track TAM and person interactions.
Audit AI object bias.
Audit AI volition overreach.

Frequently Asked Questions

What is active–stative alignment?

An alignment system where some intransitive arguments pattern like transitive agents and others like transitive patients.

What is split-S?

A system where intransitive predicates are relatively fixed into A-like and P-like classes.

What is fluid-S?

A system where at least some intransitive predicates can alternate between A-like and P-like coding depending on event construal.

Does patient-like marking make an intransitive subject an object?

No. The clause can still contain only one core participant.

Is the distinction always about volition?

No. Volition is common, but control, affectedness, lexical class and historical development can also matter.

How is this different from ergative alignment?

Ergative alignment groups all S with P in the relevant subsystem; active–stative alignment splits S.

Is it the same as unaccusative versus unergative verbs?

No. There are conceptual parallels, but the syntactic and morphological phenomena are not identical.

Why does AI fail?

Models tend to treat object-like case or agreement as proof of an object, or they overstate volition contrasts.

Internal Routes Through the eduKate Translation Architecture

Route broad case and alignment to Grammatical Case: https://edukatesg.com/2026/09/18/translate-grammatical-case-who-did-what-to-whom-language-roles/

Route event structure to Lexical Aspect and Telicity: https://edukatesg.com/2026/09/18/translate-precisely-lexical-aspect-telicity-states-activities-accomplishments-achievements/

Route valency to Verb Valency and Argument Structure: https://edukatesg.com/2026/09/18/translate-precisely-verb-valency-argument-structure-transitivity-complements/

Route voice and alignment changes to Antipassive Voice: https://edukatesg.com/2026/09/19/translate-precisely-antipassive-voice-object-demotion-ergative-absolutive/.

Final Principle

Active–stative alignment changes how one intransitive participant is coded, not how many participants the event contains. Recover A-like versus P-like S, determine whether the choice is lexical or contextual, and then translate the event as the natural intransitive it is.

Translate Precisely | Split Intransitivity and Active–Stative Alignment — Preserve Split-S, Fluid-S, Volition, Control and Affectedness