Antipassive voice in translation is a valency- and alignment-changing construction commonly discussed in ergative–absolutive languages. In a canonical transitive ergative clause, the agent is ergative and the patient is absolutive; an antipassive construction often makes the agent pattern like an intransitive absolutive subject while the patient is omitted, demoted or expressed in a less core form. The event can remain semantically transitive even though the grammar packages it as less transitive.
People searching for antipassive, antipassive voice, object demotion, ergative antipassive, Inuktitut antipassive, Mayan antipassive are solving a voice-and-alignment problem. Antipassive is not simply ‘active voice’ and it is not the mirror image of passive in every detail. Depending on the language, the construction can interact with indefinite or nonspecific objects, imperfective or habitual interpretations, syntactic extraction restrictions, discourse focus, and object omission. English often translates an antipassive with an ordinary active clause, which means the translator must preserve the source object status and discourse meaning without expecting a visible English voice marker.
This article owns the antipassive precision layer for eduKateSG. It does not replace Passive Voice, Ergative–Absolutive Alignment, Verb Valency or Information Structure. Its narrow task is to identify the antipassive agent, determine whether the patient is omitted or demoted, distinguish argument-structure change from mere object indefiniteness, preserve aspect and discourse effects, and rebuild a target active or intransitive-like clause without inventing a fully specified patient the source deliberately backgrounds.
Quick Read
Write the BASE TRANSITIVE frame A→P. Then mark ANTIPASSIVE: A becomes the privileged S-like/absolutive argument; P = omitted / oblique / noncore / indefinite / still semantically present. Translate only after deciding what happened to P.
One-Sentence Answer
Translate antipassives by preserving the agent-centred event while keeping the patient as omitted, demoted or backgrounded to exactly the degree supported by the source.
Why This Precision Layer Matters
Antipassives are easy to flatten because English often uses a normal active sentence for both the base transitive and the antipassive. But the source construction may deliberately make the patient less grammatically prominent, less specific or absent. Reintroducing a definite English object can reverse the discourse packaging and sometimes change truth conditions.
What an Antipassive Is
The first precision question is structural rather than lexical. An antipassive is a construction in which a transitive event is repackaged so that the agent receives intransitive-like grammatical treatment and the patient loses ordinary core-object status.
The main translation risk is treating the construction as an ordinary transitive active clause with an unchanged object. Because the target can remain perfectly grammatical, ordinary proofreading may miss the fact that a participant has been promoted, demoted, incorporated, or assigned the wrong discourse status. A precision review therefore rebuilds the source event and its grammatical packaging separately.
A practical diagnostic is to compare the ergative base clause with the antipassive derived clause. For example, A-ERG sees P-ABS can become A-ABS sees-ANTIPASS P-oblique or no overt P Once the event frame is stable, the target can use prepositions, ordinary objects, passive-like or active-like clauses, lexical compounds, pronouns, word order, or clause expansion. The target form may differ sharply from the source while preserving the same participants and event relation.
For recovery, preserve the reduced object status. Repair the first participant or voice mismatch before polishing style. If the source grammar leaves a semantic relation broad, preserve that breadth rather than choosing a more specific target role without evidence.
Agent Promotion to S-Like Status
This distinction becomes visible once participant roles and grammatical status are separated. The agent commonly becomes the sole or privileged absolutive argument in an ergative language.
The main translation risk is translating absolutive marking as patienthood because of the base transitive pattern. Because the target can remain perfectly grammatical, ordinary proofreading may miss the fact that a participant has been promoted, demoted, incorporated, or assigned the wrong discourse status. A precision review therefore rebuilds the source event and its grammatical packaging separately.
A practical diagnostic is to identify clause valency before interpreting case. For example, A can be absolutive in the antipassive even though it remains semantic agent Once the event frame is stable, the target can use prepositions, ordinary objects, passive-like or active-like clauses, lexical compounds, pronouns, word order, or clause expansion. The target form may differ sharply from the source while preserving the same participants and event relation.
For recovery, keep agent role. Repair the first participant or voice mismatch before polishing style. If the source grammar leaves a semantic relation broad, preserve that breadth rather than choosing a more specific target role without evidence.
Object Demotion
A fluent target can still fail if this argument-structure relation drifts. The patient may be expressed in an oblique, instrumental, dative-like or other less core form.
The main translation risk is keeping the patient as an ordinary direct object in target when source backgrounds it. Because the target can remain perfectly grammatical, ordinary proofreading may miss the fact that a participant has been promoted, demoted, incorporated, or assigned the wrong discourse status. A precision review therefore rebuilds the source event and its grammatical packaging separately.
A practical diagnostic is to record PATIENT STATUS separately from semantic role. For example, hunt-ANTIPASS animals-OBL can focus on hunting activity rather than a specific affected set Once the event frame is stable, the target can use prepositions, ordinary objects, passive-like or active-like clauses, lexical compounds, pronouns, word order, or clause expansion. The target form may differ sharply from the source while preserving the same participants and event relation.
For recovery, use an English object only if target naturalness does not overstate specificity. Repair the first participant or voice mismatch before polishing style. If the source grammar leaves a semantic relation broad, preserve that breadth rather than choosing a more specific target role without evidence.
Object Omission
The source morphology is evidence, but it is not the target-language blueprint. Some antipassives allow the patient to be omitted entirely.
The main translation risk is inventing an object because the lexical verb is normally transitive in English. Because the target can remain perfectly grammatical, ordinary proofreading may miss the fact that a participant has been promoted, demoted, incorporated, or assigned the wrong discourse status. A precision review therefore rebuilds the source event and its grammatical packaging separately.
A practical diagnostic is to ask whether the source identifies any patient. For example, ‘he hunted’ may be the best English translation of an antipassive with no overt object Once the event frame is stable, the target can use prepositions, ordinary objects, passive-like or active-like clauses, lexical compounds, pronouns, word order, or clause expansion. The target form may differ sharply from the source while preserving the same participants and event relation.
For recovery, do not supply a guessed object. Repair the first participant or voice mismatch before polishing style. If the source grammar leaves a semantic relation broad, preserve that breadth rather than choosing a more specific target role without evidence.
Ergative–Absolutive Base
The safest method is to reconstruct the event frame before choosing target syntax. Antipassive is most famously associated with ergative systems because it changes how A and P align with S.
The main translation risk is applying nominative-accusative assumptions directly. Because the target can remain perfectly grammatical, ordinary proofreading may miss the fact that a participant has been promoted, demoted, incorporated, or assigned the wrong discourse status. A precision review therefore rebuilds the source event and its grammatical packaging separately.
A practical diagnostic is to map A, S and P before case labels. For example, ergative A in the base becomes absolutive-like S in antipassive Once the event frame is stable, the target can use prepositions, ordinary objects, passive-like or active-like clauses, lexical compounds, pronouns, word order, or clause expansion. The target form may differ sharply from the source while preserving the same participants and event relation.
For recovery, route alignment explicitly. Repair the first participant or voice mismatch before polishing style. If the source grammar leaves a semantic relation broad, preserve that breadth rather than choosing a more specific target role without evidence.
Not Simply the Opposite of Passive
The first precision question is structural rather than lexical. Passive backgrounds or demotes the agent and promotes the patient; antipassive typically backgrounds/demotes the patient and privileges the agent, but the structural details are not perfectly symmetrical.
The main translation risk is forcing a mechanical mirror analysis. Because the target can remain perfectly grammatical, ordinary proofreading may miss the fact that a participant has been promoted, demoted, incorporated, or assigned the wrong discourse status. A precision review therefore rebuilds the source event and its grammatical packaging separately.
A practical diagnostic is to compare participant privileges construction by construction. For example, passive and antipassive can affect valency and discourse differently Once the event frame is stable, the target can use prepositions, ordinary objects, passive-like or active-like clauses, lexical compounds, pronouns, word order, or clause expansion. The target form may differ sharply from the source while preserving the same participants and event relation.
For recovery, translate actual source roles. Repair the first participant or voice mismatch before polishing style. If the source grammar leaves a semantic relation broad, preserve that breadth rather than choosing a more specific target role without evidence.
Semantic Transitivity
This distinction becomes visible once participant roles and grammatical status are separated. An antipassive clause can still describe an agent acting on some patient even when grammar makes the clause less transitive.
The main translation risk is assuming the event has become semantically intransitive. Because the target can remain perfectly grammatical, ordinary proofreading may miss the fact that a participant has been promoted, demoted, incorporated, or assigned the wrong discourse status. A precision review therefore rebuilds the source event and its grammatical packaging separately.
A practical diagnostic is to distinguish semantic patient existence from syntactic object status. For example, someone can be eating something unspecified Once the event frame is stable, the target can use prepositions, ordinary objects, passive-like or active-like clauses, lexical compounds, pronouns, word order, or clause expansion. The target form may differ sharply from the source while preserving the same participants and event relation.
For recovery, preserve implied patient if target verb allows it. Repair the first participant or voice mismatch before polishing style. If the source grammar leaves a semantic relation broad, preserve that breadth rather than choosing a more specific target role without evidence.
Indefinite Patients
A fluent target can still fail if this argument-structure relation drifts. Antipassive objects are often associated with indefinite, nonspecific or low-discourse-prominence patients in some languages.
The main translation risk is adding a definite article or named object. Because the target can remain perfectly grammatical, ordinary proofreading may miss the fact that a participant has been promoted, demoted, incorporated, or assigned the wrong discourse status. A precision review therefore rebuilds the source event and its grammatical packaging separately.
A practical diagnostic is to track specificity independently. For example, ‘he was looking for fish’ differs from ‘he saw the fish’ Once the event frame is stable, the target can use prepositions, ordinary objects, passive-like or active-like clauses, lexical compounds, pronouns, word order, or clause expansion. The target form may differ sharply from the source while preserving the same participants and event relation.
For recovery, use bare plural/some/unspecified wording where supported. Repair the first participant or voice mismatch before polishing style. If the source grammar leaves a semantic relation broad, preserve that breadth rather than choosing a more specific target role without evidence.
Specificity Is Not Universal
The source morphology is evidence, but it is not the target-language blueprint. Not every antipassive is simply an indefiniteness construction.
The main translation risk is assuming object nonspecificity explains every antipassive occurrence. Because the target can remain perfectly grammatical, ordinary proofreading may miss the fact that a participant has been promoted, demoted, incorporated, or assigned the wrong discourse status. A precision review therefore rebuilds the source event and its grammatical packaging separately.
A practical diagnostic is to check language-specific syntax and discourse function. For example, some antipassives are required for extraction or aspectual reasons Once the event frame is stable, the target can use prepositions, ordinary objects, passive-like or active-like clauses, lexical compounds, pronouns, word order, or clause expansion. The target form may differ sharply from the source while preserving the same participants and event relation.
For recovery, preserve the actual trigger. Repair the first participant or voice mismatch before polishing style. If the source grammar leaves a semantic relation broad, preserve that breadth rather than choosing a more specific target role without evidence.
Aspectual Effects
The safest method is to reconstruct the event frame before choosing target syntax. Antipassives can correlate with imperfective, habitual, iterative or atelic readings in some languages.
The main translation risk is adding progressive or habitual aspect automatically. Because the target can remain perfectly grammatical, ordinary proofreading may miss the fact that a participant has been promoted, demoted, incorporated, or assigned the wrong discourse status. A precision review therefore rebuilds the source event and its grammatical packaging separately.
A practical diagnostic is to decode TAM independently before using the common correlation. For example, Inuktitut antipassive alternations can interact with interpretation and dialect change Once the event frame is stable, the target can use prepositions, ordinary objects, passive-like or active-like clauses, lexical compounds, pronouns, word order, or clause expansion. The target form may differ sharply from the source while preserving the same participants and event relation.
For recovery, translate source aspect evidence. Repair the first participant or voice mismatch before polishing style. If the source grammar leaves a semantic relation broad, preserve that breadth rather than choosing a more specific target role without evidence.
Event Boundedness
The first precision question is structural rather than lexical. Demoting the object can affect how bounded or telic the event is interpreted.
The main translation risk is preserving a completed total-patient reading where source profiles activity. Because the target can remain perfectly grammatical, ordinary proofreading may miss the fact that a participant has been promoted, demoted, incorporated, or assigned the wrong discourse status. A precision review therefore rebuilds the source event and its grammatical packaging separately.
A practical diagnostic is to test endpoint and object affectedness. For example, ‘ate the seal’ versus ‘was eating seal/meat’ can differ in boundedness under language-specific systems Once the event frame is stable, the target can use prepositions, ordinary objects, passive-like or active-like clauses, lexical compounds, pronouns, word order, or clause expansion. The target form may differ sharply from the source while preserving the same participants and event relation.
For recovery, calibrate target completion. Repair the first participant or voice mismatch before polishing style. If the source grammar leaves a semantic relation broad, preserve that breadth rather than choosing a more specific target role without evidence.
Extraction-Driven Antipassive
This distinction becomes visible once participant roles and grammatical status are separated. Some Mayan languages use antipassive or agent-focus constructions in contexts of A-bar extraction of the transitive subject.
The main translation risk is treating special voice as a change in who acted. Because the target can remain perfectly grammatical, ordinary proofreading may miss the fact that a participant has been promoted, demoted, incorporated, or assigned the wrong discourse status. A precision review therefore rebuilds the source event and its grammatical packaging separately.
A practical diagnostic is to reconstruct the base clause before extraction. For example, who saw Lina? may require an agent-focus/antipassive-like verb form Once the event frame is stable, the target can use prepositions, ordinary objects, passive-like or active-like clauses, lexical compounds, pronouns, word order, or clause expansion. The target form may differ sharply from the source while preserving the same participants and event relation.
For recovery, preserve the extracted agent role. Repair the first participant or voice mismatch before polishing style. If the source grammar leaves a semantic relation broad, preserve that breadth rather than choosing a more specific target role without evidence.
Mayan Agent Focus
A fluent target can still fail if this argument-structure relation drifts. Mayan agent-focus constructions are often discussed with antipassive analyses, though language-specific descriptions differ.
The main translation risk is calling every agent-focus form a canonical antipassive without qualification. Because the target can remain perfectly grammatical, ordinary proofreading may miss the fact that a participant has been promoted, demoted, incorporated, or assigned the wrong discourse status. A precision review therefore rebuilds the source event and its grammatical packaging separately.
A practical diagnostic is to separate observable extraction morphology from theoretical label. For example, K’iche’ agentive/antipassive analyses concern agreement and transitivity Once the event frame is stable, the target can use prepositions, ordinary objects, passive-like or active-like clauses, lexical compounds, pronouns, word order, or clause expansion. The target form may differ sharply from the source while preserving the same participants and event relation.
For recovery, translate the event before theory. Repair the first participant or voice mismatch before polishing style. If the source grammar leaves a semantic relation broad, preserve that breadth rather than choosing a more specific target role without evidence.
Inuktitut Antipassive
The source morphology is evidence, but it is not the target-language blueprint. Inuktitut has a well-known ergative–antipassive alternation whose functions and variation have been extensively studied.
The main translation risk is assuming one interpretation across all dialects. Because the target can remain perfectly grammatical, ordinary proofreading may miss the fact that a participant has been promoted, demoted, incorporated, or assigned the wrong discourse status. A precision review therefore rebuilds the source event and its grammatical packaging separately.
A practical diagnostic is to use dialect/context evidence. For example, antipassive can increasingly overlap with ergative constructions in some varieties Once the event frame is stable, the target can use prepositions, ordinary objects, passive-like or active-like clauses, lexical compounds, pronouns, word order, or clause expansion. The target form may differ sharply from the source while preserving the same participants and event relation.
For recovery, avoid rigid universal glosses. Repair the first participant or voice mismatch before polishing style. If the source grammar leaves a semantic relation broad, preserve that breadth rather than choosing a more specific target role without evidence.
Agreement Changes
The safest method is to reconstruct the event frame before choosing target syntax. Antipassive derivation can change which arguments control agreement or verbal indexing.
The main translation risk is dropping agreement-linked person/number information. Because the target can remain perfectly grammatical, ordinary proofreading may miss the fact that a participant has been promoted, demoted, incorporated, or assigned the wrong discourse status. A precision review therefore rebuilds the source event and its grammatical packaging separately.
A practical diagnostic is to map agreement after derived valency is known. For example, agent may take intransitive-style agreement while patient indexing disappears Once the event frame is stable, the target can use prepositions, ordinary objects, passive-like or active-like clauses, lexical compounds, pronouns, word order, or clause expansion. The target form may differ sharply from the source while preserving the same participants and event relation.
For recovery, preserve participant features. Repair the first participant or voice mismatch before polishing style. If the source grammar leaves a semantic relation broad, preserve that breadth rather than choosing a more specific target role without evidence.
Case Changes
The first precision question is structural rather than lexical. Agent and patient case marking can shift in the antipassive.
The main translation risk is translating case labels literally as English prepositions. Because the target can remain perfectly grammatical, ordinary proofreading may miss the fact that a participant has been promoted, demoted, incorporated, or assigned the wrong discourse status. A precision review therefore rebuilds the source event and its grammatical packaging separately.
A practical diagnostic is to map semantic role and grammatical status first. For example, patient case may become instrumental-like without literally meaning ‘with’ Once the event frame is stable, the target can use prepositions, ordinary objects, passive-like or active-like clauses, lexical compounds, pronouns, word order, or clause expansion. The target form may differ sharply from the source while preserving the same participants and event relation.
For recovery, choose natural target syntax. Repair the first participant or voice mismatch before polishing style. If the source grammar leaves a semantic relation broad, preserve that breadth rather than choosing a more specific target role without evidence.
Object Incorporation and Antipassive
This distinction becomes visible once participant roles and grammatical status are separated. Noun incorporation and antipassive-like effects can overlap typologically because both may reduce independent object status.
The main translation risk is equating incorporation with antipassive automatically. Because the target can remain perfectly grammatical, ordinary proofreading may miss the fact that a participant has been promoted, demoted, incorporated, or assigned the wrong discourse status. A precision review therefore rebuilds the source event and its grammatical packaging separately.
A practical diagnostic is to ask whether the noun is morphologically incorporated, syntactically demoted or both. For example, incorporated object can remain semantically patient Once the event frame is stable, the target can use prepositions, ordinary objects, passive-like or active-like clauses, lexical compounds, pronouns, word order, or clause expansion. The target form may differ sharply from the source while preserving the same participants and event relation.
For recovery, route incorporation separately. Repair the first participant or voice mismatch before polishing style. If the source grammar leaves a semantic relation broad, preserve that breadth rather than choosing a more specific target role without evidence.
Antipassive and Reflexive
A fluent target can still fail if this argument-structure relation drifts. Reflexive/middle constructions also reduce ordinary transitivity in some languages but encode different participant relations.
The main translation risk is mistaking patient demotion for agent-patient coreference. Because the target can remain perfectly grammatical, ordinary proofreading may miss the fact that a participant has been promoted, demoted, incorporated, or assigned the wrong discourse status. A precision review therefore rebuilds the source event and its grammatical packaging separately.
A practical diagnostic is to test whether A and P are the same entity. For example, antipassive A acts on an unspecified/demoted P, not necessarily itself Once the event frame is stable, the target can use prepositions, ordinary objects, passive-like or active-like clauses, lexical compounds, pronouns, word order, or clause expansion. The target form may differ sharply from the source while preserving the same participants and event relation.
For recovery, preserve separate roles. Repair the first participant or voice mismatch before polishing style. If the source grammar leaves a semantic relation broad, preserve that breadth rather than choosing a more specific target role without evidence.
Antipassive and Object Drop
The source morphology is evidence, but it is not the target-language blueprint. Plain object omission can occur without antipassive morphology or alignment change.
The main translation risk is labelling every objectless transitive verb antipassive. Because the target can remain perfectly grammatical, ordinary proofreading may miss the fact that a participant has been promoted, demoted, incorporated, or assigned the wrong discourse status. A precision review therefore rebuilds the source event and its grammatical packaging separately.
A practical diagnostic is to look for dedicated morphology or syntactic consequences. For example, English ‘I already ate’ is not by itself an antipassive construction Once the event frame is stable, the target can use prepositions, ordinary objects, passive-like or active-like clauses, lexical compounds, pronouns, word order, or clause expansion. The target form may differ sharply from the source while preserving the same participants and event relation.
For recovery, classify source grammar carefully. Repair the first participant or voice mismatch before polishing style. If the source grammar leaves a semantic relation broad, preserve that breadth rather than choosing a more specific target role without evidence.
Discourse Function
The safest method is to reconstruct the event frame before choosing target syntax. Antipassive can foreground the agent or activity while backgrounding the patient.
The main translation risk is reintroducing a highly specific target object and changing information structure. Because the target can remain perfectly grammatical, ordinary proofreading may miss the fact that a participant has been promoted, demoted, incorporated, or assigned the wrong discourse status. A precision review therefore rebuilds the source event and its grammatical packaging separately.
A practical diagnostic is to ask what the source chooses not to foreground. For example, a hunting narrative may focus on the hunter’s ongoing activity Once the event frame is stable, the target can use prepositions, ordinary objects, passive-like or active-like clauses, lexical compounds, pronouns, word order, or clause expansion. The target form may differ sharply from the source while preserving the same participants and event relation.
For recovery, preserve target discourse balance. Repair the first participant or voice mismatch before polishing style. If the source grammar leaves a semantic relation broad, preserve that breadth rather than choosing a more specific target role without evidence.
Generic and Habitual Events
The first precision question is structural rather than lexical. Agent-centred generic activities are natural environments for antipassive in some languages.
The main translation risk is turning a generic activity into one completed token event. Because the target can remain perfectly grammatical, ordinary proofreading may miss the fact that a participant has been promoted, demoted, incorporated, or assigned the wrong discourse status. A precision review therefore rebuilds the source event and its grammatical packaging separately.
A practical diagnostic is to check genericity and aspect. For example, ‘she hunts’ can describe occupation/habit without a specific quarry Once the event frame is stable, the target can use prepositions, ordinary objects, passive-like or active-like clauses, lexical compounds, pronouns, word order, or clause expansion. The target form may differ sharply from the source while preserving the same participants and event relation.
For recovery, use simple present/habitual target wording. Repair the first participant or voice mismatch before polishing style. If the source grammar leaves a semantic relation broad, preserve that breadth rather than choosing a more specific target role without evidence.
Procedural Text
This distinction becomes visible once participant roles and grammatical status are separated. Antipassive-like constructions can describe an operator’s activity without specifying the patient until later.
The main translation risk is supplying a target object too early. Because the target can remain perfectly grammatical, ordinary proofreading may miss the fact that a participant has been promoted, demoted, incorporated, or assigned the wrong discourse status. A precision review therefore rebuilds the source event and its grammatical packaging separately.
A practical diagnostic is to track when the patient enters discourse. For example, ‘mix-ANTIPASS until smooth’ may leave material recoverable from context Once the event frame is stable, the target can use prepositions, ordinary objects, passive-like or active-like clauses, lexical compounds, pronouns, word order, or clause expansion. The target form may differ sharply from the source while preserving the same participants and event relation.
For recovery, preserve procedural coherence. Repair the first participant or voice mismatch before polishing style. If the source grammar leaves a semantic relation broad, preserve that breadth rather than choosing a more specific target role without evidence.
Narrative Reference
A fluent target can still fail if this argument-structure relation drifts. Demoted patients may remain inferable from previous context.
The main translation risk is treating omitted patient as unknown rather than backgrounded. Because the target can remain perfectly grammatical, ordinary proofreading may miss the fact that a participant has been promoted, demoted, incorporated, or assigned the wrong discourse status. A precision review therefore rebuilds the source event and its grammatical packaging separately.
A practical diagnostic is to distinguish discourse-known but unexpressed from genuinely unspecified. For example, the story may already identify what the hunter is hunting Once the event frame is stable, the target can use prepositions, ordinary objects, passive-like or active-like clauses, lexical compounds, pronouns, word order, or clause expansion. The target form may differ sharply from the source while preserving the same participants and event relation.
For recovery, use object omission if target permits, or a light pronoun if necessary. Repair the first participant or voice mismatch before polishing style. If the source grammar leaves a semantic relation broad, preserve that breadth rather than choosing a more specific target role without evidence.
Antipassive and Information Structure
The source morphology is evidence, but it is not the target-language blueprint. Object demotion can reflect topic/focus organization rather than lexical valency alone.
The main translation risk is translating grammatical backgrounding into semantic absence. Because the target can remain perfectly grammatical, ordinary proofreading may miss the fact that a participant has been promoted, demoted, incorporated, or assigned the wrong discourse status. A precision review therefore rebuilds the source event and its grammatical packaging separately.
A practical diagnostic is to mark PRESENT-IN-EVENT versus PROMINENT-IN-DISCOURSE separately. For example, patient can exist but be nonfocus Once the event frame is stable, the target can use prepositions, ordinary objects, passive-like or active-like clauses, lexical compounds, pronouns, word order, or clause expansion. The target form may differ sharply from the source while preserving the same participants and event relation.
For recovery, keep source content while matching prominence. Repair the first participant or voice mismatch before polishing style. If the source grammar leaves a semantic relation broad, preserve that breadth rather than choosing a more specific target role without evidence.
Translation into English
The safest method is to reconstruct the event frame before choosing target syntax. English often uses ordinary active clauses, optional objects or generic objects to render antipassive meaning.
The main translation risk is searching for an English ‘antipassive voice’ form. Because the target can remain perfectly grammatical, ordinary proofreading may miss the fact that a participant has been promoted, demoted, incorporated, or assigned the wrong discourse status. A precision review therefore rebuilds the source event and its grammatical packaging separately.
A practical diagnostic is to preserve agent focus and patient status semantically. For example, source antipassive can become ‘she was hunting’ Once the event frame is stable, the target can use prepositions, ordinary objects, passive-like or active-like clauses, lexical compounds, pronouns, word order, or clause expansion. The target form may differ sharply from the source while preserving the same participants and event relation.
For recovery, let the special morphology disappear. Repair the first participant or voice mismatch before polishing style. If the source grammar leaves a semantic relation broad, preserve that breadth rather than choosing a more specific target role without evidence.
Translation into Another Ergative Language
The first precision question is structural rather than lexical. A target ergative language may have its own antipassive with different triggers and object coding.
The main translation risk is mapping source antipassive directly to target antipassive. Because the target can remain perfectly grammatical, ordinary proofreading may miss the fact that a participant has been promoted, demoted, incorporated, or assigned the wrong discourse status. A precision review therefore rebuilds the source event and its grammatical packaging separately.
A practical diagnostic is to decode source function before target voice choice. For example, one language may use antipassive for extraction while another does not Once the event frame is stable, the target can use prepositions, ordinary objects, passive-like or active-like clauses, lexical compounds, pronouns, word order, or clause expansion. The target form may differ sharply from the source while preserving the same participants and event relation.
For recovery, re-encode independently. Repair the first participant or voice mismatch before polishing style. If the source grammar leaves a semantic relation broad, preserve that breadth rather than choosing a more specific target role without evidence.
AI Definite-Object Invention
This distinction becomes visible once participant roles and grammatical status are separated. Models often supply a likely object to make the English verb feel complete.
The main translation risk is turning an unspecified patient into a definite one. Because the target can remain perfectly grammatical, ordinary proofreading may miss the fact that a participant has been promoted, demoted, incorporated, or assigned the wrong discourse status. A precision review therefore rebuilds the source event and its grammatical packaging separately.
A practical diagnostic is to forbid object insertion unless source/reference licenses it. For example, ‘he hunted’ should not become ‘he hunted the deer’ without evidence Once the event frame is stable, the target can use prepositions, ordinary objects, passive-like or active-like clauses, lexical compounds, pronouns, word order, or clause expansion. The target form may differ sharply from the source while preserving the same participants and event relation.
For recovery, remove invented objects. Repair the first participant or voice mismatch before polishing style. If the source grammar leaves a semantic relation broad, preserve that breadth rather than choosing a more specific target role without evidence.
AI Passive Confusion
A fluent target can still fail if this argument-structure relation drifts. The unfamiliar term antipassive can lead models to output a passive target.
The main translation risk is reversing participant prominence. Because the target can remain perfectly grammatical, ordinary proofreading may miss the fact that a participant has been promoted, demoted, incorporated, or assigned the wrong discourse status. A precision review therefore rebuilds the source event and its grammatical packaging separately.
A practical diagnostic is to write AGENT stays agent; PATIENT is demoted. For example, antipassive hunter remains the doer Once the event frame is stable, the target can use prepositions, ordinary objects, passive-like or active-like clauses, lexical compounds, pronouns, word order, or clause expansion. The target form may differ sharply from the source while preserving the same participants and event relation.
For recovery, restore active target syntax where appropriate. Repair the first participant or voice mismatch before polishing style. If the source grammar leaves a semantic relation broad, preserve that breadth rather than choosing a more specific target role without evidence.
AI Case-Literalism
The source morphology is evidence, but it is not the target-language blueprint. Models can translate oblique patient case as a fixed preposition that sounds semantically wrong.
The main translation risk is mistaking grammatical demotion for lexical instrument/location meaning. Because the target can remain perfectly grammatical, ordinary proofreading may miss the fact that a participant has been promoted, demoted, incorporated, or assigned the wrong discourse status. A precision review therefore rebuilds the source event and its grammatical packaging separately.
A practical diagnostic is to identify case function inside the antipassive construction. For example, instrumental-like patient case may simply mark noncore object status Once the event frame is stable, the target can use prepositions, ordinary objects, passive-like or active-like clauses, lexical compounds, pronouns, word order, or clause expansion. The target form may differ sharply from the source while preserving the same participants and event relation.
For recovery, choose target object/oblique form naturally. Repair the first participant or voice mismatch before polishing style. If the source grammar leaves a semantic relation broad, preserve that breadth rather than choosing a more specific target role without evidence.
Worked Examples
Worked Example 1: Object Omitted
Source problem: A source antipassive marks a hunter as privileged argument and omits the quarry.
Diagnosis: The activity is hunting; no specific patient is asserted overtly.
Target strategy: English can simply say ‘The hunter was hunting.’
Verification: Do not invent deer, animals or a definite object.
Worked Example 2: Demoted Indefinite Patient
Source problem: A source patient remains overt in a noncore case and is nonspecific.
Diagnosis: The patient is semantically present but grammatically/backgroundedly reduced.
Target strategy: Use ‘She was looking for fish’ or another natural low-specificity object structure.
Verification: Do not turn the patient into a definite total object.
Worked Example 3: Agent-Focus Extraction
Source problem: A Mayan-type subject question uses special antipassive/agent-focus morphology.
Diagnosis: The extracted wh phrase remains semantic agent.
Target strategy: Translate ‘Who saw Lina?’ using normal English question syntax.
Verification: Do not reverse the roles because ordinary source agreement is missing.
Worked Example 4: Inuktitut-Type Alternation
Source problem: A clause could appear in ergative or antipassive form with discourse/aspectual differences.
Diagnosis: The same broad event is packaged differently.
Target strategy: Choose English wording that preserves object specificity and event boundedness rather than forcing a voice contrast.
Verification: Verify both participant identity and event completion.
Practical Diagnostic Workflow
Build an ANTIPASSIVE FRAME: BASE A; BASE P; BASE CASE/AGREEMENT; ANTIPASSIVE A STATUS; P STATUS = omitted/oblique/noncore; TAM; SPECIFICITY; EXTRACTION TRIGGER; DISCOURSE FOCUS. This isolates alignment change from event meaning.
Run the object-restoration test. Add a fully specific direct object back into a paraphrase and ask what changes. If the result becomes more definite, more telic or syntactically different than the source, the antipassive’s patient reduction matters and should not be erased.
Verification Method
From the target, identify the agent and whether a patient is overt, inferable, generic or unspecified. Compare aspect, completion and discourse prominence with the source. The target passes when it does not invent a stronger patient relation than the source licenses.
Recovery Method
If a definite object was invented, remove it and rebuild the clause around the agent’s activity. If roles were reversed, restore the base A→P event before applying the antipassive packaging. If aspect was flattened, reintroduce only the source-supported boundedness or habituality.
Practice Sequence
Take HUNT, EAT, READ and SEARCH. Create a base ergative-style transitive clause, then an antipassive version with omitted P, oblique P and nonspecific P. Translate each into English and state what changes in object prominence and event boundedness.
Quality-Assurance Checklist
- Reconstruct base A and P.
- Identify antipassive morphology.
- Keep agent as semantic agent.
- Track agent case/agreement changes.
- Identify whether P is omitted or demoted.
- Do not invent a definite object.
- Track object specificity.
- Track aspect and event boundedness.
- Handle extraction-driven antipassives carefully.
- Treat agent-focus terminology cautiously.
- Distinguish antipassive from reflexive/middle.
- Distinguish antipassive from ordinary object drop.
- Preserve discourse backgrounding.
- Audit AI passive confusion.
- Audit AI case-literalism and object invention.
Frequently Asked Questions
What is antipassive voice?
A construction that privileges the transitive agent as an intransitive-like argument while omitting or demoting the patient.
Is antipassive just the opposite of passive?
It is often described as a functional mirror because passive backgrounds the agent and antipassive backgrounds the patient, but the structures are not universally exact opposites.
Why is antipassive common in ergative languages?
It can realign the transitive agent with the absolutive/intransitive pattern and reduce the patient’s core status.
Does the patient disappear semantically?
Not necessarily. It can remain understood, generic, nonspecific or overt in an oblique form.
Is antipassive always imperfective?
No. Aspectual correlations are common in some languages but are not universal.
What is Mayan agent focus?
A special construction used in several Mayan languages for extracting transitive agents; it is often analysed in relation to antipassive, with language-specific debate.
How should English translate antipassive?
Often with an ordinary active clause, optional object, generic object or aspectual wording that preserves patient backgrounding.
Why does AI fail?
Models may invent a likely object, confuse antipassive with passive, or translate demoted case too literally.
Internal Routes Through the eduKate Translation Architecture
Route alignment to Grammatical Case: https://edukatesg.com/2026/09/18/translate-grammatical-case-who-did-what-to-whom-language-roles/. Route general valency changes to Verb Valency and Argument Structure: https://edukatesg.com/2026/09/18/translate-precisely-verb-valency-argument-structure-transitivity-complements/. Route ordinary passive to Passive Voice, Agency and Attribution: https://edukatesg.com/2026/09/18/translate-precisely-passive-voice-agency-attribution-who-did-what-who-said-what/. Route discourse prominence to Information Structure and Emphasis: https://edukatesg.com/2026/09/18/translate-precisely-information-structure-emphasis-topic-focus-contrast/.
Final Principle
Antipassive keeps the agent central while reducing the patient’s grammatical prominence. Preserve the agent’s role, do not over-specify the patient, and let the target language express the activity and discourse focus naturally.