Mova0.6 pre-release

Okinawan: the flag of Okinawa Prefecture, a white ring in a red disc on white 14. Verb inflection and auxiliaries

Central Okinawan verbs distinguish tense, aspect, mood and politeness through inflection and through combinations with auxiliaries. The form of a predicate also depends on its place in the sentence. A verb ending an ordinary statement, a verb modifying a noun and a verb linking two events each have their own forms. This chapter uses Shuri forms as the reference and gives the examples in romanization. A doubled vowel is long, and = marks a particle boundary.

Independent and dependent forms

A verb has a set of stems and endings. To identify a form, establish both what it expresses and what construction it belongs to. Tense alone does not determine the ending. For example, the past of kachun ‘write’ is kachan in an ordinary affirmative statement and kacharu in an adnominal clause. Both forms locate the writing in the past. The difference concerns the construction in which the predicate appears. (Chamberlain 1895, § The Verb; Shimoji and Pellard 2010, § Okinawan)

The conclusive form is the form used to close an ordinary statement. The adnominal form is used before a noun that the clause modifies. It also occurs in some constructions without a following noun, including the focus construction with =du. A converb links its clause to another predicate or combines with an auxiliary. The following forms show these contrasts in a single verb. The table gives the affirmative plain forms of kachun that are used throughout this chapter. (Chamberlain 1895, § The Verb; Shimoji and Pellard 2010, § Okinawan)

SeriesConclusiveAdnominalGloss
Simple nonpastkachunkachuruwrite.NPST
Simple pastkachankacharuwrite.PST
Durative nonpastkachoonkachooruwrite.DUR.NPST
Durative pastkachootankachootaruwrite.DUR.PST

Each form in the table has the syntactic value named by its column. Thus kachuru is write.NPST.ADN and kachootan is write.DUR.PST.CONCL. The adnominal forms in the table end in -ru and the conclusives end in -n. This correspondence describes these affirmative series. Other series, including the negative and the polite, must be identified from their own forms and distribution.

FormGlossTranslationSource
tigami kachanletter write.PST.CONCL‘I wrote a letter.’(constructed)
kacharu tigamiwrite.PST.ADN letter‘a letter I wrote’(constructed)
tigami kachi, ichanletter write.CVB go.PST.CONCL‘I wrote a letter and went.’(constructed)

The translations supply a first-person subject. The Okinawan forms are also possible with another writer or mover when the context identifies that participant. The third example illustrates a different kind of dependency from the second. In kacharu tigami, the adnominal predicate has a noun as its head and carries its own past marking. In tigami kachi, ichan, the converb has no past ending. The sequence and the past of ichan supply the temporal setting for the writing.

Independent clauses can also have interrogative, imperative or volitional endings. Accordingly, an independent predicate need not have the conclusive ending of a statement. Conversely, a predicate that has adnominal morphology can end a complete utterance under the conditions of a focus construction. The form of the verb and the independence of the utterance are therefore two separate facts. A predicate at the end of a sentence can be either conclusive or adnominal, depending on what licenses it. (Shimoji and Pellard 2010, § Okinawan)

The simple conclusive also serves as a convenient citation form. Kachun names the verb ‘write’ even when the form required in a sentence is kachi, kacharu or kakantan. It is useful to distinguish the citation form from the stem. For instance, the negative kakan and the converb kachi use different stem shapes. The final -n of the citation form is no guide to the shape of the negative stem or to the choice of converb. (Chamberlain 1895, § The Verb)

Tense and aspect

The primary tense contrast is past versus nonpast. A simple nonpast predicate can describe a habit, a recurrent event or an event expected after the time of speaking. Temporal expressions and the discourse distinguish these readings. An event currently in progress normally calls for the durative series when the point is that it is under way. A simple past predicate presents an event or state as past. The durative past locates the ongoing situation or persisting state at a past reference time. (Chamberlain 1895, § The Verb; Shimoji and Pellard 2010, § Okinawan)

FormGlossTranslationSource
achaa tigami kachuntomorrow letter write.NPST.CONCL‘I’ll write a letter tomorrow.’(constructed)
chinuu tigami kachanyesterday letter write.PST.CONCL‘I wrote a letter yesterday.’(constructed)
nama tigami kachoonnow letter write.DUR.NPST.CONCL‘I’m writing a letter now.’(constructed)
chinuu tigami kachootanyesterday letter write.DUR.PST.CONCL‘I was writing a letter yesterday.’(constructed)

In the first example, achaa ‘tomorrow’ supplies the future time. The verb is in the same nonpast series that serves for a habit. In the last example, chinuu ‘yesterday’ locates the durative situation in the past. The sentence can be used to describe what the writer was occupied with at a relevant time yesterday. Whether the letter was eventually finished is a separate matter.

The durative construction also has readings in which a state holds as a consequence of an earlier event. The interpretation depends on the lexical verb and on what the context treats as continuing. With an activity such as writing, the continuing part can be the activity itself. With a change of state, the continuing part can be the state reached by the change. English may use an ongoing-event expression for the first reading and a state expression for the second. The Okinawan durative ending alone does not require one English translation in both contexts. (Shimoji and Pellard 2010, § Okinawan)

A resultative or perfect form in -een, as in kacheen, presents the prior event through a result relevant at the reference time. With a letter, the fact that it is in a written state can be what matters. A simple past can report the writing as one event in a sequence. A durative can place the writer in the middle of working on the letter. These forms can therefore be used about the same general course of events, depending on which part of it the clause presents. (Chamberlain 1895, § The Verb; Shimoji and Pellard 2010, § Okinawan)

FormGlossTranslationSource
tigami kachanletter write.PST.CONCL‘I wrote a letter.’(constructed)
tigami kachoonletter write.DUR.NPST.CONCL‘I’m writing a letter.’(constructed)
tigami kacheenletter write.RES.CONCL‘I’ve written the letter.’(constructed)

The negative also distinguishes nonpast and past. For the plain simple forms of ‘write’, compare the following conclusive paradigm. In the nonpast, kakan can describe a regular absence of writing or an occasion on which writing will not take place. In the past, kakantan denies the writing at a past time. The negative has its own shape, including the kaka- stem in this verb. An affirmative past like kachan does not serve as the base to which a negative particle is added. (Chamberlain 1895, § The Verb; Shimoji and Pellard 2010, § Okinawan)

TenseAffirmativeGlossNegativeGloss
Nonpastkachunwrite.NPST.CONCLkakanwrite.NEG.NPST.CONCL
Pastkachanwrite.PST.CONCLkakantanwrite.NEG.PST.CONCL
FormGlossTranslationSource
achaa tigami kakantomorrow letter write.NEG.NPST.CONCL‘I won’t write a letter tomorrow.’(constructed)
chinuu tigami kakantanyesterday letter write.NEG.PST.CONCL‘I didn’t write a letter yesterday.’(constructed)

The affirmative paradigms in the first section also show how tense combines with clause type. The contrast between kachootan and kachootaru is conclusive versus adnominal within the durative past. The contrast between kachooru and kachootaru is nonpast versus past within the durative adnominal. In an adnominal clause, interpret the tense with the time supplied by the larger sentence and the context. In a converb chain, establish first whether a dependent predicate has tense marking of its own. A past ending on the last verb can place the chain in the past even when the earlier converb has no tense ending.

Mood, modality and evidential readings

Mood and sentence type affect the choice of verb form. Statements ordinarily use a conclusive form. Commands use an imperative, and a volitional can present a course of action the speaker proposes or undertakes. A question particle can combine with a predicate shape selected by the interrogative construction. Tense, polarity and the speaker’s relation to the addressee can still matter within these constructions. (Chamberlain 1895, § The Verb; Shimoji and Pellard 2010, § Okinawan)

Imperative, volitional and interrogative forms

An imperative directs the addressee to act. Its form is selected from the verb’s imperative paradigm. The irregular imperative kuu ‘come!’ and kamee ‘eat!’ illustrate why the ordinary nonpast conclusive is insufficient to predict the shape of an imperative. A prohibition uses a negative directive construction. The plain negative conclusive in a statement such as ‘You aren’t going’ or ‘You won’t go’ has declarative inflection even when the utterance is intended to discourage the action. (Chamberlain 1895, § The Verb)

A volitional can present an action the speaker will carry out or propose an action involving others. The polite form illustrated by kachabira has this range. The subject and the speech situation determine whether an English translation uses ‘I’ll’, ‘let me’ or ‘let’s’. A nonpast statement with a future-time expression can also describe a plan, as in achaa tigami kachun. A plan given as information and a proposal put to the addressee can thus concern the same future action while using different verb forms. (Chamberlain 1895, § The Verb)

FormGlossTranslationSource
kuma=nkai kuuhere=ALL come.IMP‘Come over here.’(constructed)
kameeeat.IMP‘Eat.’(constructed)
tigami kachabiraletter write.POL.VOL‘Let me write a letter.’(constructed)
achaa ichabii=mitomorrow go.POL.NPST=Q‘Are you going tomorrow?’(constructed)

In the last example, the polar question particle =mi follows ichabii. The corresponding affirmative polite conclusive is ichabiin. The question is formed with the predicate shape used before =mi, where the final -n of the statement form is absent. This is the polite nonpast pattern illustrated here. The form required by a question also depends on the question construction. For example, the particle that accompanies a question word and the particle of a polar question have different functions. (Chamberlain 1895, § The Verb; Shimoji and Pellard 2010, § Okinawan)

Desire and expectation

A desiderative predicate can be built with -busan on the appropriate continuative base, as in kamibusan ‘want to eat’. The eating supplies the content of the desire. A time expression can locate the desired action, while the inflected desiderative predicate expresses the desire. The desirer's identity follows from the subject expression or the discourse. A first-person interpretation is natural when the speaker declares a desire without naming the desirer. (Chamberlain 1895, § The Verb)

An expectation can be expressed with an adnominal clause modifying haji, followed by a copula. The adnominal verb supplies the content of the expectation. The copula completes the whole predicate. This construction has two places where grammatical information can be expressed. The inner verb can locate the expected event in relation to its own reference time, while the outer predicate presents the expectation as holding at the relevant time. (Chamberlain 1895, § The Verb; Shimoji and Pellard 2010, § Okinawan)

FormGlossTranslationSource
nama kamibusannow eat.DESID.CONCL‘I want to eat now.’(constructed)
achaa ichuru haji yantomorrow go.NPST.ADN expectation COP.NPST.CONCL‘They’re expected to go tomorrow.’(constructed)

In the second example, ichuru ‘go’ is adnominal because it modifies haji. The whole expression ends in yan, the plain conclusive of the copula. This is distinct from a question or a command inflected directly on ‘go’. It also allows an expected action to be stated without presenting it as the speaker’s own intention.

Information source and inference

Tense, aspect and information source can interact in interpretation. A resultative can be appropriate when the speaker knows about an earlier event from a result that is present now. The same resultative is available when the speaker saw or carried out the event and is talking about its result. Thus a reading based on inference from a result need not carry a separate morpheme for ‘I inferred it’. The simple past likewise locates an event in the past without specifying in its gloss how the speaker found out. (Shimoji and Pellard 2010, § Okinawan)

Reported information can be presented through quotation. The quotative =ndi follows the quoted material. An inflected verb inside that material retains the tense and mood appropriate to the quoted utterance. The quotative boundary is external to the verb. A quoted future plan can therefore have nonpast morphology inside the quotation, while the act of saying it belongs to an earlier time. (Chamberlain 1895, § The Verb; Shimoji and Pellard 2010, § Okinawan)

FormGlossTranslationSource
tigami kacheenletter write.RES.CONCL‘The letter has been written.’(constructed)
achaa ichun=nditomorrow go.NPST.CONCL=QUOT‘… saying “I’ll go tomorrow.”’(constructed)

The first translation uses an English passive because the result is salient in that reading. There is no passive marking in kacheen. In the second example, the quoted words and their particle are shown as a fragment. The reporting clause supplies the speaker and the time of the report. Where an example has no quotative, expectation expression or other overt indication of the information source, inference or report may still be recoverable from the conversation. Such an interpretation belongs in the account of the utterance rather than in the morpheme gloss of an ordinary past or durative form.

Adnominal forms, converbs and other dependent forms

Dependent verb forms differ in what they attach to and in which distinctions they retain. An adnominal verb is the predicate of a clause associated with a noun or with another construction that selects adnominal morphology. A converb provides a verbal predicate inside a linked sequence. A continuative base also occurs inside a larger word or predicate. The availability of a dependent form must be established for the construction in which it occurs. (Chamberlain 1895, § The Verb; Shimoji and Pellard 2010, § Okinawan)

Adnominal clauses and focus

In a noun-modifying clause, the adnominal predicate precedes the head noun. A noun such as tigami ‘letter’ can be understood as the object of ‘write’. The writer can be expressed inside the modifier or recovered from context. Past and nonpast distinctions are available in the adnominal series, as is the durative contrast. The head noun remains outside the verb inflection, and its role in the modifying clause does not give the verb person or number endings. (Chamberlain 1895, § The Verb; Shimoji and Pellard 2010, § Okinawan)

FormGlossTranslationSource
achaa kachuru tigamitomorrow write.NPST.ADN letter‘the letter I’ll write tomorrow’(constructed)
chinuu kacharu tigamiyesterday write.PST.ADN letter‘the letter I wrote yesterday’(constructed)
nama kachooru tigaminow write.DUR.NPST.ADN letter‘the letter I’m writing now’(constructed)

The overt noun in these examples identifies what the clause modifies. In the focus construction with =du, adnominal inflection can occur without any following noun. The particle marks the focused constituent and the predicate appears in the form called for by the construction. The following pair keeps the same noun and past event. The first is an ordinary assertion and the second identifies the letter as the relevant contrast. (Shimoji and Pellard 2010, § Okinawan)

FormGlossTranslationSource
tigami kachanletter write.PST.CONCL‘I wrote a letter.’(constructed)
tigami=du kacharuletter=FOC write.PST.ADN‘It was a letter that I wrote.’(constructed)

This distribution matters when analyzing a sentence-final -ru form. In kacharu tigami, there is a head noun. In tigami=du kacharu, the focus particle explains the form of the final predicate. Treating the latter as an ordinary noun modifier with an omitted noun would miss the overt relation between the focus particle and the predicate. Conversely, an adnominal form inside a noun phrase requires no =du to license it there.

Converb and continuative base

The linking converb kachi adapts ‘write’ to a construction in which another predicate follows. A common reading with two event predicates is succession. Simultaneity, means or attendant circumstance depend on the predicates and their context. The converb itself gives no independent past or nonpast ending. In an auxiliary construction, the same kind of dependent form can be the lexical input to a single larger predicate. The stem or converb shape must be learned with the verb class. A converb is not made by applying one spelling change to every conclusive form. (Chamberlain 1895, § The Verb; Shimoji and Pellard 2010, § Okinawan)

UseFormGlossTranslationSource
Link in a past sequencetigami kachi, ichanletter write.CVB go.PST.CONCL‘I wrote a letter and went.’(constructed)
Link in a future sequenceachaa tigami kachi, ichuntomorrow letter write.CVB go.NPST.CONCL‘Tomorrow I’ll write a letter and go.’(constructed)
Base inside a desiderativekamibusaneat.DESID.CONCL‘I want to eat.’(constructed)

The two chains contain exactly the same form kachi. The past ichan and nonpast ichun, together with the discourse or the time expression, produce different temporal interpretations for the chain. In the desiderative, the continuative base kami- is internal to kamibusan. It has no separate subject or independent ending in that word. Thus the term continuative can refer to a stem used inside a predicate, while converb refers here to a dependent form with a linking function. Similar shapes in the two uses are possible.

A converb does not in itself encode whether the participant in its event is the same person as the participant in the next event. In the examples the most natural reading gives both actions to one person. An overt participant expression or the preceding discourse can control the interpretation in a longer chain. When the sequence is an auxiliary construction, the lexical verb and auxiliary describe one complex predicate, and the auxiliary does not introduce another writer, eater or mover simply because it inflects.

Conditional forms

Conditional predicates introduce the circumstances under which a following claim, question or directive holds. They have dedicated dependent marking. A plain linking converb can leave the relation between two events to context. A conditional makes the status of the first clause as a condition explicit. The following independent predicate supplies the force of the consequent. A command as the consequent directs the addressee to act if the condition is met. It does not assert that the condition has been met. (Chamberlain 1895, § The Verb)

FormGlossTranslationSource
tigami kakee, kuma=nkai kuuletter write.COND here=ALL come.IMP‘If you write the letter, come here.’(constructed)

Here kakee belongs to the conditional clause, and kuu is the independent imperative. A temporal interpretation such as ‘when you have written it’ can be appropriate if the writing is expected. The relation of the writing to the coming is supplied by the conditional construction and the event sequence. The form of the consequent and the conversational expectation both matter to the translation.

Auxiliaries and inflectional sequences

An auxiliary construction combines a lexical predicate with material that contributes a grammatical meaning such as aspect or politeness. The lexical predicate identifies the event or state and licenses its participants. The outer part of the construction bears the inflection appropriate to the whole predicate. In Central Okinawan, combinations involving a dependent base may have conventional contracted shapes. A single surface word can therefore convey both the lexical and auxiliary meanings. (Chamberlain 1895, § The Verb; Shimoji and Pellard 2010, § Okinawan)

Aspectual sequences

The durative series illustrates an aspectual construction with an inflecting outer part. In kachoon, kachooru, kachootan and kachootaru, the lexical event is writing and the durative component presents an ongoing event or a continuing state. The nonpast versus past and conclusive versus adnominal distinctions belong to the whole predicate. Since the series has a contracted shape, the written form need not display a separate auxiliary word. (Chamberlain 1895, § The Verb; Shimoji and Pellard 2010, § Okinawan)

Inflection of the whole predicateFormGloss
Nonpast conclusivekachoonwrite.DUR.NPST.CONCL
Nonpast adnominalkachooruwrite.DUR.NPST.ADN
Past conclusivekachootanwrite.DUR.PST.CONCL
Past adnominalkachootaruwrite.DUR.PST.ADN

The relation of these forms to the converb is clearest when the uses are compared in sentences. Kachi, ichan links an event of writing to an event of going. Kachootan describes writing with durative aspect in the past. The past contrast is realized at the end of the durative predicate. A speaker who changes kachootan to kachootaru for a noun modifier keeps the durative aspect and the past tense. (Shimoji and Pellard 2010, § Okinawan)

FormGlossTranslationSource
tigami kachi, ichanletter write.CVB go.PST.CONCL‘I wrote a letter and went.’(constructed)
tigami kachootanletter write.DUR.PST.CONCL‘I was writing a letter.’(constructed)
kachootaru tigamiwrite.DUR.PST.ADN letter‘the letter I was writing’(constructed)

Resultative morphology likewise combines with the lexical verb and is read as part of the same predicate. The resultative kacheen and the durative kachoon can stand in the same conclusive position. The resumption of an interrupted activity favors the durative description of what was going on. The present availability of a finished letter can favor the resultative description. The lexical meaning of ‘write’ and the object ‘letter’ remain recognizable in both forms.

Polite sequences

The polite series uses a dependent lexical base followed by inflecting polite material. In the ‘write’ paradigm, the polite nonpast conclusive is kachabiin. The forms kachabitan, kachabiran and kachabira show that this construction carries past, negative and volitional distinctions. The lexical verb does not first take its ordinary conclusive ending. The plain kachun and polite kachabiin each complete one predicate. (Chamberlain 1895, § The Verb)

ValuePolite form of ‘write’Gloss
Affirmative nonpast conclusivekachabiinwrite.POL.NPST.CONCL
Affirmative past conclusivekachabitanwrite.POL.PST.CONCL
Negative nonpast conclusivekachabiranwrite.POL.NEG.NPST.CONCL
Volitionalkachabirawrite.POL.VOL
Affirmative nonpast polar questionkachabii=miwrite.POL.NPST=Q

The alternation between kachabiin and kachabitan includes a vowel-quantity difference in these forms. The negative and volitional look similar as well. The final -n in kachabiran belongs to the negative conclusive. Kachabira is the volitional and can be used to offer to write. To interpret the sequence, identify the inflected polite form and the construction it appears in, including a following question particle where present.

FormGlossTranslationSource
achaa tigami kachabiintomorrow letter write.POL.NPST.CONCL‘I’ll write a letter tomorrow.’(constructed)
chinuu tigami kachabitanyesterday letter write.POL.PST.CONCL‘I wrote a letter yesterday.’(constructed)
achaa tigami kachabirantomorrow letter write.POL.NEG.NPST.CONCL‘I won’t write a letter tomorrow.’(constructed)
tigami kachabiraletter write.POL.VOL‘Let me write a letter.’(constructed)

The choice of base before polite material is part of the conjugation. In particular, high-frequency verbs and the copula have established polite forms. An adnominal clause can also occur inside a larger predicate whose matrix ending is polite. In an expectation construction, for example, it is the outer copula that can express the politeness of the whole assertion. That construction has a nominal element, haji, between the two predicates, so it differs structurally from the polite inflection attached within a verbal predicate.

FormGlossTranslationSource
achaa ichuru haji yaibiintomorrow go.NPST.ADN expectation COP.POL.NPST.CONCL‘They’re expected to go tomorrow.’(constructed)

The plain adnominal ichuru is selected by haji. The matrix conclusive yaibiin sets the addressee level for the assertion. Similarly, the verb inside a quoted utterance can carry its own speech style, while a reporting predicate takes inflection suited to the current exchange. The location of a polite form in the larger construction determines which predicate it belongs to.

Participant marking and levels of politeness

Finite verbs do not agree with their subject or object in person or number. The same tense, aspect and mood forms can be used with singular and plural participants and with speakers, addressees and third persons. A noun phrase or pronoun can name the participant. Case and discourse particles help identify its relation to the clause. When participants are recoverable, they can be left unexpressed. The ending of kachan still means write.PST.CONCL with ‘I’, ‘we’ or ‘the child’ as its subject. (Chamberlain 1895, § The Verb; Skirgård et al. 2023, § Verbal person marking)

FormGlossTranslationSource
wan=ga tigami kachan1SG=NOM letter write.PST.CONCL‘I wrote a letter.’(constructed)
wattaa=ga tigami kachan1PL=NOM letter write.PST.CONCL‘We wrote a letter.’(constructed)
warabi=ga tigami kachanchild=NOM letter write.PST.CONCL‘The child wrote a letter.’(constructed)
tigami kachanletter write.PST.CONCL‘I wrote a letter’ or ‘They wrote a letter’, as the context supplies.(constructed)

Commands usually direct the addressee. Volitionals commonly express an intention of the speaker or a proposal that includes the addressee. These interpretations arise from what the forms do in the exchange. A command to one person and a command to several people can use the same imperative. A polite polar question can concern the person being addressed or someone else. Its polite marking concerns the exchange, while the subject has to be supplied by a phrase or the context.

Addressee-oriented politeness

The plain and polite series can describe the same event with the same participants and time. The speaker chooses the level in relation to the addressee and the speech situation. Suitable uses of the plain series include familiar exchanges. The polite series is available where the speaker uses a polite manner of address. An English translation often conveys the same proposition for both and leaves the difference to register or context. (Chamberlain 1895, § The Verb; Shimoji and Pellard 2010, § Okinawan)

FormGlossTranslationSource
achaa tigami kachuntomorrow letter write.NPST.CONCL‘I’ll write a letter tomorrow.’(constructed)
achaa tigami kachabiintomorrow letter write.POL.NPST.CONCL‘I’ll write a letter tomorrow.’(constructed)
warabi=ga tigami kachabitanchild=NOM letter write.POL.PST.CONCL‘The child wrote a letter.’(constructed)

The last example shows why a polite ending should not be glossed as marking a respected subject. The event can concern an ordinary third person while the speaker addresses the listener politely. The child still has the same grammatical role as in the plain past example. The polite predicate also permits a first-person subject. It is compatible with a change in subject even when the relationship between speaker and listener remains the same.

Honorific lexical choices and directives

Honorific and humble choices can also appear in the lexical predicate. A conventional honorific motion form such as mensooree addresses or refers respectfully to the person whose coming is requested. Humble expressions can present the speaker’s own action toward another person with deference. These choices depend on who acts, who is involved in the act and who is addressed. Their distribution is therefore distinct from that of the general polite series, which can be used when ‘write’ has an unremarkable subject. (Chamberlain 1895, § The Verb; Shimoji and Pellard 2010, § Okinawan)

FormGlossTranslationSource
kuma=nkai kuuhere=ALL come.IMP‘Come over here.’(constructed)
kuma=nkai mensooreehere=ALL come.HON.IMP‘Please come this way.’(constructed)

Both clauses direct a person to move toward a location. The plain irregular imperative kuu and the honorific imperative mensooree are established forms with their own shapes. The latter also occurs in conventional greetings of welcome. Person, number, the type of directive and the social choice can thus contribute independently to interpretation. The directive supplies an addressee as the expected actor, and the honorific expression supplies the deferential treatment.

How Mova handles verb inflection

Central Okinawan (ryu) has no level-2 parser model in tools/mova-parse. When a span has been assigned to ryu, the missing model is reported in the # parse comment. Its tokens have _ in LEMMA, UPOS, XPOS, FEATS, HEAD and DEPREL, and Parse=None in MISC. Mova holds no annotated Okinawan treebank, so it has no measured treebank baseline for the language.

At level 1, tools/langseg assigns languages to sentences and spans within its script and lexicon coverage. Its Latin-script distinction covers English and French. It also recognizes Hebrew-script varieties and scholarly transliteration of Hebrew or Aramaic. Every other script is other. Its current coverage therefore limits any identification of a span as Okinawan before parsing. With no Okinawan model, Mova does not tokenize, tag, lemmatize or parse the language. prag and world read English trees and have no Okinawan parse to use.

Squid reads the manual and records each grammatical point as a claim. Formalizable claims, such as the distinction between conclusive and adnominal forms or the members of a stated paradigm, wait in its queue. None of this chapter’s verb rules is implemented in Mova yet. The distinctions that would come first are those needed for a tokenizer, a tagger and a UD treebank baseline.

Sources for this chapter

This page in Markdown → https://md.mova.elamur.ai/free/en/docs/languages/ja/ryu/14-verb-inflection/