Allocutive agreement in translation is the grammatical marking of the addressee on a verb or clause even when that listener is not a semantic argument of the event. Basque provides the classic example: a clause about a third person can carry morphology that depends on whether the speaker is addressing a male or female familiar interlocutor. Recent research also uses allocutive analyses for addressee-oriented honorific phenomena beyond Basque, including debates around Japanese.
People searching for allocutive agreement, allocutivity, addressee agreement, Basque hika, addressee honorific agreement are solving a two-layer problem. The event can be entirely third-person while the speech act contains second-person grammatical information. A translator must therefore separate who participates in the event from who is being spoken to, and must preserve social relation, familiarity and addressee features without accidentally inserting the listener as subject or object.
This article owns the allocutive-agreement precision layer for eduKateSG. It does not replace Honorifics and Social Address, Polypersonal Agreement, Reported Speech or Deixis. Its narrow job is to preserve addressee-indexing morphology that lives outside ordinary verb argument structure, especially where the target language has no direct allocutive category and must recover the same relationship through register, pronouns, honorifics or dialogue style.
Quick Read
Build two maps. EVENT MAP: who does what to whom? SPEECH-ACT MAP: who speaks to whom, and what social/addressee features does the morphology encode? Never let a speech-act addressee become an event participant unless the lexical verb independently selects that participant.
One-Sentence Answer
Translate allocutive agreement by preserving the speaker–addressee relationship separately from the event’s subject, object and other arguments.
Why This Precision Layer Matters
Allocutivity shows that grammar can encode a listener who is nowhere in the proposition. A third-person event can therefore contain second-person morphology without meaning ‘you.’ If that distinction is missed, the target can change the event itself. If the morphology is simply deleted, the event survives but the social relationship can disappear from dialogue and narrative characterisation.
What Allocutive Agreement Is
The first task is to separate speech-act grammar from event grammar. Allocutive agreement is verbal or clause-level morphology that encodes properties of the addressee even when that addressee is not a semantic argument of the verb. Because allocutive systems are highly language- and variety-specific, the translator should not assume that one Basque, Japanese or other addressee-oriented pattern defines the category universally.
The main translation risk is mistaking addressee marking for an ordinary second-person subject or object. A fluent target can still be propositionally wrong if the listener becomes an actor or object, or pragmatically wrong if social distance is flattened. The event map and speech-act map therefore need separate verification.
A practical diagnostic is to write two rows: EVENT PARTICIPANTS and SPEECH-ACT ADDRESSEE. For example, a clause meaning ‘Peter worked’ can carry morphology showing that the sentence is addressed to a male familiar interlocutor. Once the two layers are stable, the target can use whatever resources are natural: informal/formal pronouns, speech level, honorifics, vocatives, register or sometimes no overt marker when context already carries the relationship.
The Addressee Is Not Necessarily an Argument
This distinction is invisible if every person marker is assumed to be argument agreement. The defining translation problem is that the listener can be grammatically represented without participating in the event described by the verb. Because allocutive systems are highly language- and variety-specific, the translator should not assume that one Basque, Japanese or other addressee-oriented pattern defines the category universally.
The main translation risk is adding the listener into the event as actor, recipient or patient. A fluent target can still be propositionally wrong if the listener becomes an actor or object, or pragmatically wrong if social distance is flattened. The event map and speech-act map therefore need separate verification.
A practical diagnostic is to ask whether removing the addressee from the real-world event changes who worked, arrived or slept. For example, ‘Peter worked’ remains about Peter’s work even though allocutive morphology indexes the person being spoken to. Once the two layers are stable, the target can use whatever resources are natural: informal/formal pronouns, speech level, honorifics, vocatives, register or sometimes no overt marker when context already carries the relationship.
Speech-Act Layer vs Event Layer
A precise translation should identify the addressee before target pronouns are generated. Allocutive morphology belongs to the speech-act context, whereas ordinary agreement belongs to the event’s argument structure. Because allocutive systems are highly language- and variety-specific, the translator should not assume that one Basque, Japanese or other addressee-oriented pattern defines the category universally.
The main translation risk is merging addressee features with subject/object agreement. A fluent target can still be propositionally wrong if the listener becomes an actor or object, or pragmatically wrong if social distance is flattened. The event map and speech-act map therefore need separate verification.
A practical diagnostic is to build EVENT and SPEECH-ACT layers separately. For example, the event can be third-person while the speech-act layer indexes a second-person listener. Once the two layers are stable, the target can use whatever resources are natural: informal/formal pronouns, speech level, honorifics, vocatives, register or sometimes no overt marker when context already carries the relationship.
Basque Hika and Allocutive Forms
The target may preserve the social relation without copying the morphology. Basque provides a classic allocutive system in which certain finite forms vary according to the sex/social relation of a familiar addressee, especially in hika-style interaction. Because allocutive systems are highly language- and variety-specific, the translator should not assume that one Basque, Japanese or other addressee-oriented pattern defines the category universally.
The main translation risk is translating allocutive morphology as if it changed the propositional subject or object. A fluent target can still be propositionally wrong if the listener becomes an actor or object, or pragmatically wrong if social distance is flattened. The event map and speech-act map therefore need separate verification.
A practical diagnostic is to identify the clause’s normal arguments first, then add the addressee relation. For example, a third-person clause can differ morphologically depending on whether it is said to a male or female familiar addressee. Once the two layers are stable, the target can use whatever resources are natural: informal/formal pronouns, speech level, honorifics, vocatives, register or sometimes no overt marker when context already carries the relationship.
Familiarity and Social Relation
Allocutivity is best treated as interactional grammar layered over the proposition. Allocutive marking can be tied to familiar, intimate or socially restricted address styles rather than neutral second-person agreement. Because allocutive systems are highly language- and variety-specific, the translator should not assume that one Basque, Japanese or other addressee-oriented pattern defines the category universally.
The main translation risk is flattening the form into a generic you and losing register. A fluent target can still be propositionally wrong if the listener becomes an actor or object, or pragmatically wrong if social distance is flattened. The event map and speech-act map therefore need separate verification.
A practical diagnostic is to label the addressee relationship—familiar, intimate, peer, deferential or other language-specific category. For example, a familiar allocutive form may be inappropriate in formal target dialogue even when lexical content is accurate. Once the two layers are stable, the target can use whatever resources are natural: informal/formal pronouns, speech level, honorifics, vocatives, register or sometimes no overt marker when context already carries the relationship.
Addressee Gender or Class Features
The first task is to separate speech-act grammar from event grammar. Some allocutive paradigms distinguish properties such as addressee sex/gender or social category. Because allocutive systems are highly language- and variety-specific, the translator should not assume that one Basque, Japanese or other addressee-oriented pattern defines the category universally.
The main translation risk is turning a grammatical addressee class into descriptive content about the event. A fluent target can still be propositionally wrong if the listener becomes an actor or object, or pragmatically wrong if social distance is flattened. The event map and speech-act map therefore need separate verification.
A practical diagnostic is to keep the feature on the interactional layer. For example, a masculine familiar allocutive marker does not mean a male participant carried out the verb’s event. Once the two layers are stable, the target can use whatever resources are natural: informal/formal pronouns, speech level, honorifics, vocatives, register or sometimes no overt marker when context already carries the relationship.
Allocutivity vs Ordinary Second-Person Agreement
This distinction is invisible if every person marker is assumed to be argument agreement. Ordinary second-person agreement indexes an argument of the verb; allocutivity indexes the listener independently of event argument structure. Because allocutive systems are highly language- and variety-specific, the translator should not assume that one Basque, Japanese or other addressee-oriented pattern defines the category universally.
The main translation risk is treating every second-person-looking affix as allocutive or every allocutive affix as subject agreement. A fluent target can still be propositionally wrong if the listener becomes an actor or object, or pragmatically wrong if social distance is flattened. The event map and speech-act map therefore need separate verification.
A practical diagnostic is to test whether the clause can have a third-person subject and no second-person object while retaining the marker. For example, a third-person declarative can still show allocutive morphology. Once the two layers are stable, the target can use whatever resources are natural: informal/formal pronouns, speech level, honorifics, vocatives, register or sometimes no overt marker when context already carries the relationship.
Allocutivity vs Vocatives
A precise translation should identify the addressee before target pronouns are generated. Vocatives overtly name or call the addressee, whereas allocutive agreement can index the addressee morphologically without a vocative phrase. Because allocutive systems are highly language- and variety-specific, the translator should not assume that one Basque, Japanese or other addressee-oriented pattern defines the category universally.
The main translation risk is adding an explicit name or ‘hey you’ because the source marks allocutivity. A fluent target can still be propositionally wrong if the listener becomes an actor or object, or pragmatically wrong if social distance is flattened. The event map and speech-act map therefore need separate verification.
A practical diagnostic is to separate addressee identification from addressee morphology. For example, allocutive marking need not introduce a spoken vocative. Once the two layers are stable, the target can use whatever resources are natural: informal/formal pronouns, speech level, honorifics, vocatives, register or sometimes no overt marker when context already carries the relationship.
Allocutivity vs Politeness Particles
The target may preserve the social relation without copying the morphology. Politeness particles, sentence-final particles and honorific words can manage social stance without forming an agreement paradigm. Because allocutive systems are highly language- and variety-specific, the translator should not assume that one Basque, Japanese or other addressee-oriented pattern defines the category universally.
The main translation risk is calling every polite particle allocutive agreement. A fluent target can still be propositionally wrong if the listener becomes an actor or object, or pragmatically wrong if social distance is flattened. The event map and speech-act map therefore need separate verification.
A practical diagnostic is to check whether morphology systematically varies with addressee features and attaches to a grammatical host. For example, a discourse particle can be polite without agreeing with the listener. Once the two layers are stable, the target can use whatever resources are natural: informal/formal pronouns, speech level, honorifics, vocatives, register or sometimes no overt marker when context already carries the relationship.
Allocutivity vs Honorific Agreement
Allocutivity is best treated as interactional grammar layered over the proposition. Honorific agreement can target a referent’s social status, while allocutive agreement specifically references the speech-act addressee; some systems or analyses overlap these domains. Because allocutive systems are highly language- and variety-specific, the translator should not assume that one Basque, Japanese or other addressee-oriented pattern defines the category universally.
The main translation risk is treating any honorific morphology as evidence the addressee is being indexed. A fluent target can still be propositionally wrong if the listener becomes an actor or object, or pragmatically wrong if social distance is flattened. The event map and speech-act map therefore need separate verification.
A practical diagnostic is to identify whose status controls the morphology. For example, subject honorification can concern a respected third person, whereas allocutivity concerns the listener. Once the two layers are stable, the target can use whatever resources are natural: informal/formal pronouns, speech level, honorifics, vocatives, register or sometimes no overt marker when context already carries the relationship.
Japanese Addressee-Honorific Analyses
The first task is to separate speech-act grammar from event grammar. Recent formal work sometimes analyses Japanese addressee-oriented honorific phenomena through an allocutive perspective, though terminology and analysis are theory-dependent. Because allocutive systems are highly language- and variety-specific, the translator should not assume that one Basque, Japanese or other addressee-oriented pattern defines the category universally.
The main translation risk is presenting all Japanese politeness as one simple Basque-like allocutive paradigm. A fluent target can still be propositionally wrong if the listener becomes an actor or object, or pragmatically wrong if social distance is flattened. The event map and speech-act map therefore need separate verification.
A practical diagnostic is to distinguish descriptive facts from theoretical analysis. For example, Japanese addressee honorific morphology can be discussed as speech-act-related without assuming it is structurally identical to Basque. Once the two layers are stable, the target can use whatever resources are natural: informal/formal pronouns, speech level, honorifics, vocatives, register or sometimes no overt marker when context already carries the relationship.
Speech-Act Participant Features
This distinction is invisible if every person marker is assumed to be argument agreement. Allocutive systems show that grammar can represent speaker/addressee information beyond the ordinary argument structure of the lexical verb. Because allocutive systems are highly language- and variety-specific, the translator should not assume that one Basque, Japanese or other addressee-oriented pattern defines the category universally.
The main translation risk is deleting the interactional layer because the target event remains intelligible. A fluent target can still be propositionally wrong if the listener becomes an actor or object, or pragmatically wrong if social distance is flattened. The event map and speech-act map therefore need separate verification.
A practical diagnostic is to record SPEAKER and ADDRESSEE features independently from semantic roles. For example, the same event can be stated with different allocutive forms to different listeners. Once the two layers are stable, the target can use whatever resources are natural: informal/formal pronouns, speech level, honorifics, vocatives, register or sometimes no overt marker when context already carries the relationship.
Declaratives
A precise translation should identify the addressee before target pronouns are generated. Allocutive marking commonly appears in declarative clauses that do not otherwise contain second-person participants. Because allocutive systems are highly language- and variety-specific, the translator should not assume that one Basque, Japanese or other addressee-oriented pattern defines the category universally.
The main translation risk is assuming a declarative third-person clause has no second-person grammatical information. A fluent target can still be propositionally wrong if the listener becomes an actor or object, or pragmatically wrong if social distance is flattened. The event map and speech-act map therefore need separate verification.
A practical diagnostic is to inspect the finite verb for speech-act indexing. For example, a report about someone else can encode the listener relation. Once the two layers are stable, the target can use whatever resources are natural: informal/formal pronouns, speech level, honorifics, vocatives, register or sometimes no overt marker when context already carries the relationship.
Questions
The target may preserve the social relation without copying the morphology. Question formation can interact with allocutive morphology and addressee orientation in language-specific ways. Because allocutive systems are highly language- and variety-specific, the translator should not assume that one Basque, Japanese or other addressee-oriented pattern defines the category universally.
The main translation risk is assuming interrogative mood cancels all allocutive distinctions. A fluent target can still be propositionally wrong if the listener becomes an actor or object, or pragmatically wrong if social distance is flattened. The event map and speech-act map therefore need separate verification.
A practical diagnostic is to check the paradigm by clause type. For example, some allocutive forms are restricted, altered or unavailable in questions. Once the two layers are stable, the target can use whatever resources are natural: informal/formal pronouns, speech level, honorifics, vocatives, register or sometimes no overt marker when context already carries the relationship.
Imperatives
Allocutivity is best treated as interactional grammar layered over the proposition. Imperatives already involve an addressee as the understood actor, so allocutive distinctions can interact differently with ordinary second-person grammar. Because allocutive systems are highly language- and variety-specific, the translator should not assume that one Basque, Japanese or other addressee-oriented pattern defines the category universally.
The main translation risk is counting the same addressee twice as both allocutive and verbal object. A fluent target can still be propositionally wrong if the listener becomes an actor or object, or pragmatically wrong if social distance is flattened. The event map and speech-act map therefore need separate verification.
A practical diagnostic is to separate imperative subject from independent social marking. For example, a command can encode both the addressee as actor and additional social-register information. Once the two layers are stable, the target can use whatever resources are natural: informal/formal pronouns, speech level, honorifics, vocatives, register or sometimes no overt marker when context already carries the relationship.
Negation
The first task is to separate speech-act grammar from event grammar. Negative clauses can preserve allocutive marking while changing only propositional polarity. Because allocutive systems are highly language- and variety-specific, the translator should not assume that one Basque, Japanese or other addressee-oriented pattern defines the category universally.
The main translation risk is treating an allocutive suffix as part of the negative morphology. A fluent target can still be propositionally wrong if the listener becomes an actor or object, or pragmatically wrong if social distance is flattened. The event map and speech-act map therefore need separate verification.
A practical diagnostic is to segment polarity and addressee features separately. For example, ‘Peter did not work’ can still be addressed in a familiar allocutive register. Once the two layers are stable, the target can use whatever resources are natural: informal/formal pronouns, speech level, honorifics, vocatives, register or sometimes no overt marker when context already carries the relationship.
Tense and Aspect
This distinction is invisible if every person marker is assumed to be argument agreement. Allocutive forms can appear alongside tense/aspect morphology and sometimes alter auxiliary shapes in morphologically complex paradigms. Because allocutive systems are highly language- and variety-specific, the translator should not assume that one Basque, Japanese or other addressee-oriented pattern defines the category universally.
The main translation risk is misidentifying the allocutive segment as tense/aspect or vice versa. A fluent target can still be propositionally wrong if the listener becomes an actor or object, or pragmatically wrong if social distance is flattened. The event map and speech-act map therefore need separate verification.
A practical diagnostic is to gloss the full auxiliary or verb feature bundle. For example, past + third-person event + male-addressee allocutive information can coexist in one form. Once the two layers are stable, the target can use whatever resources are natural: informal/formal pronouns, speech level, honorifics, vocatives, register or sometimes no overt marker when context already carries the relationship.
Auxiliaries
A precise translation should identify the addressee before target pronouns are generated. In Basque-style systems allocutive information is often visible in auxiliary morphology, making the lexical verb look unchanged. Because allocutive systems are highly language- and variety-specific, the translator should not assume that one Basque, Japanese or other addressee-oriented pattern defines the category universally.
The main translation risk is dropping the allocutive meaning because the auxiliary is treated as semantically empty. A fluent target can still be propositionally wrong if the listener becomes an actor or object, or pragmatically wrong if social distance is flattened. The event map and speech-act map therefore need separate verification.
A practical diagnostic is to parse the auxiliary’s person and speech-act features. For example, the auxiliary can carry both argument agreement and addressee indexing. Once the two layers are stable, the target can use whatever resources are natural: informal/formal pronouns, speech level, honorifics, vocatives, register or sometimes no overt marker when context already carries the relationship.
Argument Agreement plus Allocutive Agreement
The target may preserve the social relation without copying the morphology. A finite verb can simultaneously encode event participants and the listener, producing dense morphology. Because allocutive systems are highly language- and variety-specific, the translator should not assume that one Basque, Japanese or other addressee-oriented pattern defines the category universally.
The main translation risk is assigning one person marker to the wrong layer. A fluent target can still be propositionally wrong if the listener becomes an actor or object, or pragmatically wrong if social distance is flattened. The event map and speech-act map therefore need separate verification.
A practical diagnostic is to create separate slots for subject/object and allocutive addressee. For example, third-person subject/object morphology can coexist with an addressee-specific marker. Once the two layers are stable, the target can use whatever resources are natural: informal/formal pronouns, speech level, honorifics, vocatives, register or sometimes no overt marker when context already carries the relationship.
Allocutivity and Gendered Address
Allocutivity is best treated as interactional grammar layered over the proposition. Where the paradigm distinguishes male/female addressees, translation into a language without grammaticalised addressee gender often loses that feature overtly. Because allocutive systems are highly language- and variety-specific, the translator should not assume that one Basque, Japanese or other addressee-oriented pattern defines the category universally.
The main translation risk is adding ‘male friend’ or ‘female friend’ into the prose when the source merely uses the appropriate form. A fluent target can still be propositionally wrong if the listener becomes an actor or object, or pragmatically wrong if social distance is flattened. The event map and speech-act map therefore need separate verification.
A practical diagnostic is to preserve the relation through dialogue voice or context only where useful. For example, target English may not need to state gender if the addressee is already known. Once the two layers are stable, the target can use whatever resources are natural: informal/formal pronouns, speech level, honorifics, vocatives, register or sometimes no overt marker when context already carries the relationship.
Allocutivity and Familiarity
The first task is to separate speech-act grammar from event grammar. A familiar allocutive form can signal intimacy, solidarity or socially licensed familiarity, but the exact social meaning varies across communities. Because allocutive systems are highly language- and variety-specific, the translator should not assume that one Basque, Japanese or other addressee-oriented pattern defines the category universally.
The main translation risk is translating every familiar form as rude or affectionate. A fluent target can still be propositionally wrong if the listener becomes an actor or object, or pragmatically wrong if social distance is flattened. The event map and speech-act map therefore need separate verification.
A practical diagnostic is to use source-community norms rather than generic politeness stereotypes. For example, the same familiar form can be ordinary among peers and inappropriate with a superior. Once the two layers are stable, the target can use whatever resources are natural: informal/formal pronouns, speech level, honorifics, vocatives, register or sometimes no overt marker when context already carries the relationship.
Register in Literary Dialogue
This distinction is invisible if every person marker is assumed to be argument agreement. Fiction can use allocutive forms to establish relationships without explicitly telling the reader how characters relate. Because allocutive systems are highly language- and variety-specific, the translator should not assume that one Basque, Japanese or other addressee-oriented pattern defines the category universally.
The main translation risk is normalising all dialogue into one target register and erasing relationship cues. A fluent target can still be propositionally wrong if the listener becomes an actor or object, or pragmatically wrong if social distance is flattened. The event map and speech-act map therefore need separate verification.
A practical diagnostic is to track allocutive choice across speakers and scenes. For example, a character switching away from familiar allocutive speech can signal changed social distance. Once the two layers are stable, the target can use whatever resources are natural: informal/formal pronouns, speech level, honorifics, vocatives, register or sometimes no overt marker when context already carries the relationship.
Switching Allocutive Register
A precise translation should identify the addressee before target pronouns are generated. A speaker can change allocutive style as relationships, emotions or contexts shift. Because allocutive systems are highly language- and variety-specific, the translator should not assume that one Basque, Japanese or other addressee-oriented pattern defines the category universally.
The main translation risk is treating a form change as random morphological variation. A fluent target can still be propositionally wrong if the listener becomes an actor or object, or pragmatically wrong if social distance is flattened. The event map and speech-act map therefore need separate verification.
A practical diagnostic is to compare the same speaker–addressee pair across discourse. For example, a move from familiar to neutral/formal address may signal conflict or distancing. Once the two layers are stable, the target can use whatever resources are natural: informal/formal pronouns, speech level, honorifics, vocatives, register or sometimes no overt marker when context already carries the relationship.
Reported Speech
The target may preserve the social relation without copying the morphology. Allocutive morphology inside quoted or reported speech belongs to the original speech-act relation, not automatically to the current narrator and reader. Because allocutive systems are highly language- and variety-specific, the translator should not assume that one Basque, Japanese or other addressee-oriented pattern defines the category universally.
The main translation risk is anchoring the addressee feature to the wrong conversation level. A fluent target can still be propositionally wrong if the listener becomes an actor or object, or pragmatically wrong if social distance is flattened. The event map and speech-act map therefore need separate verification.
A practical diagnostic is to identify ORIGINAL SPEAKER and ORIGINAL ADDRESSEE. For example, a quoted allocutive form can show whom the original speaker addressed. Once the two layers are stable, the target can use whatever resources are natural: informal/formal pronouns, speech level, honorifics, vocatives, register or sometimes no overt marker when context already carries the relationship.
Embedded Clauses
Allocutivity is best treated as interactional grammar layered over the proposition. Many allocutive systems impose restrictions on embedding because allocutive agreement is tied to a speech-act layer. Because allocutive systems are highly language- and variety-specific, the translator should not assume that one Basque, Japanese or other addressee-oriented pattern defines the category universally.
The main translation risk is freely interpreting embedded allocutive morphology as ordinary object agreement. A fluent target can still be propositionally wrong if the listener becomes an actor or object, or pragmatically wrong if social distance is flattened. The event map and speech-act map therefore need separate verification.
A practical diagnostic is to check whether the language allows allocutive marking in embedded clauses and under what conditions. For example, a form acceptable in a root clause may be restricted under a complement verb. Once the two layers are stable, the target can use whatever resources are natural: informal/formal pronouns, speech level, honorifics, vocatives, register or sometimes no overt marker when context already carries the relationship.
Root Phenomenon Tendencies
The first task is to separate speech-act grammar from event grammar. Allocutive marking is often associated with root or speech-act-related syntax, though exact distributions vary. Because allocutive systems are highly language- and variety-specific, the translator should not assume that one Basque, Japanese or other addressee-oriented pattern defines the category universally.
The main translation risk is turning a typological tendency into a universal ban on embedding. A fluent target can still be propositionally wrong if the listener becomes an actor or object, or pragmatically wrong if social distance is flattened. The event map and speech-act map therefore need separate verification.
A practical diagnostic is to consult the language-specific distribution. For example, some varieties may permit broader embedded uses than others. Once the two layers are stable, the target can use whatever resources are natural: informal/formal pronouns, speech level, honorifics, vocatives, register or sometimes no overt marker when context already carries the relationship.
Translation into English
This distinction is invisible if every person marker is assumed to be argument agreement. English generally lacks obligatory allocutive agreement, so the target often preserves the proposition while expressing social relation through pronouns, register, honorifics, vocatives or nothing overt. Because allocutive systems are highly language- and variety-specific, the translator should not assume that one Basque, Japanese or other addressee-oriented pattern defines the category universally.
The main translation risk is adding explicit commentary such as ‘said familiarly to a male friend’ in ordinary prose. A fluent target can still be propositionally wrong if the listener becomes an actor or object, or pragmatically wrong if social distance is flattened. The event map and speech-act map therefore need separate verification.
A practical diagnostic is to decide whether the social feature is already recoverable from context and genre. For example, dialogue register can sometimes preserve the relationship without a literal gloss. Once the two layers are stable, the target can use whatever resources are natural: informal/formal pronouns, speech level, honorifics, vocatives, register or sometimes no overt marker when context already carries the relationship.
Translation into Another Honorific Language
A precise translation should identify the addressee before target pronouns are generated. A target language may encode addressee relations through speech levels, honorific verbs or particles rather than allocutive agreement. Because allocutive systems are highly language- and variety-specific, the translator should not assume that one Basque, Japanese or other addressee-oriented pattern defines the category universally.
The main translation risk is mapping source morphology to the nearest-looking target affix regardless of social meaning. A fluent target can still be propositionally wrong if the listener becomes an actor or object, or pragmatically wrong if social distance is flattened. The event map and speech-act map therefore need separate verification.
A practical diagnostic is to compare social relationship, clause type and target honorific system. For example, formal Japanese or Korean speech-level choices may approximate interactional stance without structural identity. Once the two layers are stable, the target can use whatever resources are natural: informal/formal pronouns, speech level, honorifics, vocatives, register or sometimes no overt marker when context already carries the relationship.
Subtitles and Dialogue Localisation
The target may preserve the social relation without copying the morphology. Allocutive distinctions can be especially difficult in subtitles where space is limited and social nuance may be distributed across voice, pronouns and scene context. Because allocutive systems are highly language- and variety-specific, the translator should not assume that one Basque, Japanese or other addressee-oriented pattern defines the category universally.
The main translation risk is adding long explanatory phrases that disrupt timing. A fluent target can still be propositionally wrong if the listener becomes an actor or object, or pragmatically wrong if social distance is flattened. The event map and speech-act map therefore need separate verification.
A practical diagnostic is to preserve relationship through concise register choices. For example, a familiar pronoun or informal verb style may do more than a literal note. Once the two layers are stable, the target can use whatever resources are natural: informal/formal pronouns, speech level, honorifics, vocatives, register or sometimes no overt marker when context already carries the relationship.
Legal and Institutional Speech
Allocutivity is best treated as interactional grammar layered over the proposition. Formal institutional contexts may forbid or strongly constrain familiar allocutive forms. Because allocutive systems are highly language- and variety-specific, the translator should not assume that one Basque, Japanese or other addressee-oriented pattern defines the category universally.
The main translation risk is carrying an intimate source form into a target institutional register without considering markedness. A fluent target can still be propositionally wrong if the listener becomes an actor or object, or pragmatically wrong if social distance is flattened. The event map and speech-act map therefore need separate verification.
A practical diagnostic is to identify whether the source itself is socially marked in context. For example, an allocutive choice can become evidence of familiarity between speakers. Once the two layers are stable, the target can use whatever resources are natural: informal/formal pronouns, speech level, honorifics, vocatives, register or sometimes no overt marker when context already carries the relationship.
AI Argument Confusion
The first task is to separate speech-act grammar from event grammar. Models can see second-person morphology and insert ‘you’ as subject or object even when the addressee is not part of the event. Because allocutive systems are highly language- and variety-specific, the translator should not assume that one Basque, Japanese or other addressee-oriented pattern defines the category universally.
The main translation risk is changing the proposition while trying to preserve the marker. A fluent target can still be propositionally wrong if the listener becomes an actor or object, or pragmatically wrong if social distance is flattened. The event map and speech-act map therefore need separate verification.
A practical diagnostic is to separate event arguments from addressee before generation. For example, ‘Peter worked’ must not become ‘you worked’. Once the two layers are stable, the target can use whatever resources are natural: informal/formal pronouns, speech level, honorifics, vocatives, register or sometimes no overt marker when context already carries the relationship.
AI Register Flattening
This distinction is invisible if every person marker is assumed to be argument agreement. AI can preserve propositional meaning but erase familiar/formal allocutive distinctions during paraphrase. Because allocutive systems are highly language- and variety-specific, the translator should not assume that one Basque, Japanese or other addressee-oriented pattern defines the category universally.
The main translation risk is making all speakers sound socially identical. A fluent target can still be propositionally wrong if the listener becomes an actor or object, or pragmatically wrong if social distance is flattened. The event map and speech-act map therefore need separate verification.
A practical diagnostic is to track addressee relationship as a discourse feature. For example, dialogue should maintain consistent social distance where the source does. Once the two layers are stable, the target can use whatever resources are natural: informal/formal pronouns, speech level, honorifics, vocatives, register or sometimes no overt marker when context already carries the relationship.
Back-Translation Speech-Act Check
A precise translation should identify the addressee before target pronouns are generated. A final check should recover both the proposition and the speaker–addressee relationship from the target context. Because allocutive systems are highly language- and variety-specific, the translator should not assume that one Basque, Japanese or other addressee-oriented pattern defines the category universally.
The main translation risk is checking only who did the event. A fluent target can still be propositionally wrong if the listener becomes an actor or object, or pragmatically wrong if social distance is flattened. The event map and speech-act map therefore need separate verification.
A practical diagnostic is to write EVENT plus ADDRESSEE RELATION. For example, the target can be propositionally perfect yet pragmatically undertranslated. Once the two layers are stable, the target can use whatever resources are natural: informal/formal pronouns, speech level, honorifics, vocatives, register or sometimes no overt marker when context already carries the relationship.
Worked Examples
Worked Example 1: Basque-Style Third-Person Clause
Source problem: A third-person event is expressed with allocutive morphology sensitive to a familiar addressee.
Diagnosis: The addressee is grammatically indexed but does not perform the event.
Target strategy: Translate the event normally, then preserve the familiar addressee relation through target register if relevant.
Verification: The target must not turn the listener into subject or object.
Worked Example 2: Register Switch
Source problem: The same speaker stops using a familiar allocutive form with the same addressee.
Diagnosis: The propositional content may be unchanged while social distance changes.
Target strategy: Reflect the shift through target formality or dialogue voice.
Verification: A target that sounds identical before and after may lose a narrative cue.
Worked Example 3: Reported Allocutive Speech
Source problem: A quoted character uses an addressee-specific form.
Diagnosis: The relevant addressee belongs to the original speech situation.
Target strategy: Preserve the quoted relationship, not the current narrator-reader relationship.
Verification: The target should identify the correct original interlocutor.
Worked Example 4: AI Inserts You
Source problem: A model sees second-person morphology and translates ‘Peter worked’ as ‘You worked.’
Diagnosis: Allocutive morphology has been mistaken for subject agreement.
Target strategy: Restore Peter as event subject and treat the second-person feature as speech-act information.
Verification: The proposition and the dialogue relation must both be correct.
Practical Diagnostic Workflow
Create an ALLOCUTIVE CARD: EVENT SUBJECT; EVENT OBJECT/OTHER ARGUMENTS; SPEAKER; ADDRESSEE; ADDRESSEE FEATURES; SOCIAL RELATION; CLAUSE TYPE; EMBEDDING STATUS. Only after the event and speech-act layers are complete should target pronouns or honorific levels be chosen.
Run the deletion test: remove the allocutive morphology mentally. If the core event remains the same but the speaker–listener relationship loses information, you have isolated the allocutive contribution. Translate that contribution at the discourse/register level rather than inside the predicate’s argument frame.
Verification Method
Back-read the target twice. First reconstruct the event: who acted on whom? Second reconstruct the speech situation: who is speaking to whom and how socially close/formal is the relation? The target passes only when both maps match the source.
Recovery Method
If the addressee became an event participant, strip all second-person target pronouns and rebuild from the event frame. Then restore social/addressee meaning through register. If the target flattened the relation, reintroduce the smallest natural target cue rather than explanatory prose.
Quality-Assurance Checklist
- Separate event arguments from addressee.
- Identify the speech-act addressee.
- Distinguish allocutive from ordinary second-person agreement.
- Distinguish allocutive from vocatives.
- Distinguish allocutive from generic politeness particles.
- Distinguish addressee honorifics from subject honorifics.
- Track addressee gender/class only where source does.
- Preserve familiar/formal relationship.
- Check clause-type restrictions.
- Check embedded-clause restrictions.
- Preserve register shifts in dialogue.
- Anchor reported allocutive speech to the original context.
- Avoid explanatory overtranslation.
- Audit AI argument confusion.
- Audit AI register flattening.
Frequently Asked Questions
What is allocutive agreement?
Morphology that encodes the speech-act addressee even when the addressee is not an argument of the lexical verb.
Is allocutive agreement just second-person agreement?
No. Ordinary second-person agreement indexes a verb argument; allocutive morphology can appear in a third-person clause with no second-person event participant.
Why is Basque important?
Basque hika and related allocutive paradigms are classic examples of addressee-sensitive verbal morphology.
Does allocutivity always encode politeness?
Not simply. It can encode familiarity, social relation, addressee sex/gender or other interactional features depending on the language.
Is Japanese allocutive in the same way as Basque?
Some current analyses use allocutive ideas for Japanese addressee honorific phenomena, but the systems are not structurally identical and terminology is theory-sensitive.
How should English translate allocutive morphology?
Usually through ordinary propositional syntax plus register, pronoun style, dialogue tone or context rather than a dedicated agreement marker.
Can allocutive marking appear in reported speech?
Yes in some systems, but it must be anchored to the correct original speaker/addressee context.
Why does AI fail on allocutives?
Second-person morphology is easy to misread as a second-person subject or object instead of speech-act agreement.
Where This Fits in the eduKate Translation Architecture
Route broad honorific and address relations to the live Honorifics and Social Address owner; ordinary multi-argument indexing to Polypersonal Agreement; reported speech to the Reported Speech and Quotation owner; speaker/addressee/place/time anchoring to Deixis and Point of View.
Final Principle
The listener can be grammatically present without being part of the event. Translate the proposition and the speech relationship as two separate layers, then let the target language decide how visibly the second layer should appear.