29. Reference, dialogue and discourse
Central Okinawan speakers can identify a referent with a noun phrase, a pronoun or a demonstrative, or leave it implicit. The choice depends on who is speaking, what the participants can see, and what has already been established in the conversation. This chapter describes how those choices work across clauses and turns, using Shuri-based forms.
Introducing and tracking referents
A lexical noun phrase can introduce someone or something into the discourse, bring a previously mentioned referent back into consideration, or distinguish it from another candidate. A demonstrative can help identify the intended referent by directing attention to the situation or to an earlier mention. Once a referent is established, a later clause can carry it forward with an overt expression or an unexpressed argument. The descriptive categories for pronouns and demonstratives also apply to their use across clauses (Chamberlain 1895, § Pronouns; Shimoji and Pellard 2010, § Okinawan).
| Form | Romanization | Gloss | Translation |
|---|---|---|---|
| タルーが ちゃん。みじ ぬだん。 | Taruu-ga chan. Miji nudan. | Taruu-NOM come.PST. Water drink.PST. | “Taruu came. He drank water.” (constructed) |
In this reading, the drinker is Taruu. The first sentence identifies him by name, and the second names the drink but leaves the drinker implicit. The English pronoun he is supplied in the translation. A discourse in which someone else is already prominent can change the interpretation of the second sentence.
Person and discourse status give different kinds of information. A first-person pronoun refers to the speaker of its utterance. A second-person pronoun refers to the addressee, with the choice of form also sensitive to the relationship between the participants. A demonstrative used for a person identifies that person by context, which may include an earlier mention, a gesture, or the participants' shared attention. Chamberlain's description treats personal and demonstrative reference together in the account of pronouns (Chamberlain 1895, § Pronouns).
| Form | Romanization | Gloss | Translation |
|---|---|---|---|
| わん | wan | 1SG | “I, me” (constructed) |
| やー | yaa | 2SG.FAMILIAR | “you,” to one familiar addressee (constructed) |
| あれー ちゃん。 | Aree chan. | DIST.DEM.TOP come.PST | “That person came.” (constructed) |
In the last example, aree combines the distal demonstrative ari with the topic particle ya. Topic marking presents the referent as what the clause is about. The source of its identifiability may lie in the preceding turn or in the situation. Case marking serves another purpose. In Taruu-ga chan, the nominative marks Taruu's role in the clause. A hearer may then carry Taruu forward as the person under discussion, even though the introduction used nominative marking. The interpretation of case and topic marking starts with the clause in which the particle occurs (Shimoji and Pellard 2010, § Okinawan).
Coreference is the relation between two expressions or understood arguments that pick out the same individual. A name and a later demonstrative can corefer. Two occurrences of the same demonstrative can pick out different individuals if the speaker redirects attention, and a first-person form picks out a different person when the speaker changes. Following a referent through a conversation therefore requires keeping track of both the expressions and the circumstances in which they are used.
Zero anaphora and competing antecedents
An argument can remain unexpressed when the hearer can identify it from context. With zero anaphora, the interpretation comes from an already established referent. The referent may have been mentioned in the same sentence, in a previous sentence or by another speaker. An implicit argument can also be supplied by the speech situation. In an answer about the respondent's own actions, for instance, the respondent can be the understood subject without ever having been named. Verbal forms such as the past forms below give tense information. The participant is recovered through the predicate and the context (Shimoji and Pellard 2010, § Okinawan).
| Form | Romanization | Gloss | Translation |
|---|---|---|---|
| タルーが マツーとぅ ちゃん。みじ ぬだん。 | Taruu-ga Matsuu-tu chan. Miji nudan. | Taruu-NOM Matsuu-COM come.PST. Water drink.PST. | “Taruu came with Matsuu. Someone drank water.” (constructed) |
| タルーが マツーとぅ ちゃん。マツーが みじ ぬだん。 | Taruu-ga Matsuu-tu chan. Matsuu-ga miji nudan. | Taruu-NOM Matsuu-COM come.PST. Matsuu-NOM water drink.PST. | “Taruu came with Matsuu. Matsuu drank water.” (constructed) |
In the first version, Taruu is a natural candidate because he is the subject of the first sentence. Matsuu is also available, and a reading in which both drank is possible if the situation supports it. The translation leaves the identity open. The second version names Matsuu as the drinker. Repeating a name in this way can resolve a choice between candidates even when the intended referent has just been mentioned.
Three kinds of evidence matter in recovering a referent. The predicate gives the role to be filled, as nudan requires an interpretation of who drank. The surrounding discourse gives candidates, including the current topic and the participants brought into the most recent event. The situation can then favor one of those candidates or supply a participant through the roles of speaker and addressee. These are cues to interpretation. A recently mentioned noun phrase can be incompatible with the role, and a comitative phrase can provide the intended referent despite another participant's being the preceding subject. If two candidates fit equally well, omission leaves the identity open until a later expression or the situation resolves it.
Objects can also be understood from an earlier clause or question. A question that establishes a particular serving of water makes it possible to answer with just the verb. Both the drinker and what was drunk can then be implicit. The same verb in a fresh utterance can simply report drinking, with no particular earlier referent for the object. An analysis of zero anaphora needs evidence that there is an identifiable referent to recover.
In a clause chain, an implicit participant can continue from one clause into the next. A dependent clause describing one event can also lead into a main clause about another participant. An overt noun phrase can mark that shift, as the sequence with Taruu and Matsuu shows across sentence boundaries. The event relation and participant identity are separate questions. A causal or sequential connection can hold between events with different participants.
Demonstratives in interaction and discourse
The basic three-way set distinguishes a proximal series in k-, a medial series in u- and a distal series in a-. The independent forms can stand as referring expressions, the adnominal forms accompany a noun, and the place forms refer to locations. The place interrogative supplies the corresponding question about location (Chamberlain 1895, § Demonstrative Pronouns; Shimoji and Pellard 2010, § Okinawan).
| Series | Independent form | Adnominal form | Place form |
|---|---|---|---|
| Proximal | くり kuri, “this one” | くぬ kunu, “this” | くま kuma, “here” |
| Medial | うり uri, “that one” | うぬ unu, “that” | うま uma, “there” |
| Distal | あり ari, “that one over there” | あぬ anu, “that over there” | あま ama, “over there” |
The place question form is まー maa “where”. The English translations in the table approximate the contrasts. In a face-to-face exchange, a speaker can use the proximal series to draw attention to something on the speaker's side, the medial series for something on the addressee's side, and the distal series for something away from both. The participants' orientation to the object matters alongside physical distance. The same object can be proximal from one speaker's position and medial from the other's.
| Form | Romanization | Gloss | Translation |
|---|---|---|---|
| くぬ みじ | kunu miji | PROX.ADN water | “this water” (constructed) |
| うぬ みじ | unu miji | MED.ADN water | “that water” (constructed) |
| くまんかい ちゅん。 | Kuma-nkai chun. | Here-ALL come.NPST. | “[Someone] will come here.” (constructed) |
One speaker can use kunu miji while holding a cup and the other can answer with unu miji for the same water. The adnominal form takes its interpretation from the perspective of the person speaking. In the place example, kuma identifies a location and -nkai marks it as the goal of motion. Recovery of the understood mover is a separate matter.
In discourse, a medial expression can direct attention to a referent that the previous turn has made available, even when that referent is out of sight. An independent demonstrative can also point back to a proposal or situation just described. A distal expression can call up something the participants recognize from shared experience. In those uses, the hearer identifies the target through mention or shared knowledge. Physical location is relevant when the conversation or gesture makes it relevant. The demonstrative form by itself can leave open whether the intended target is a person, an object or a whole situation.
Discourse relations in clause chains
A clause can advance a sequence of events, provide a reason for another clause, or set up an expectation that the following clause qualifies. Central Okinawan uses connective verb forms and clause-linking particles for these relations. Adjacent independent clauses can also be related by inference. In the earlier example about Taruu, arrival followed by drinking is a natural chronological reading. Punctuation presents two sentences, and the order of events comes from the narrative interpretation.
| Form | Romanization | Gloss | Use |
|---|---|---|---|
| ぬでぃ | nudi | drink.CVB | Connective form of numun. In a sequence of actions, “drinking” or “having drunk.” |
| くとぅ | kutu | CAUSAL | After a predicate in a causal clause, “because, since.” |
| しが | shiga | ADVERSATIVE | At the end of the first clause in a contrast or concession, “but, although.” |
The table gives the forms illustrated here. The connective form of the verb links a dependent predicate to a following predicate. In a sequence of completed actions, the following finite predicate can supply the tense for interpreting the sequence. With kutu, the clause bearing the linker gives the reason for what follows. With shiga, the first clause supplies a fact or expectation against which the following clause is interpreted. A form's function depends on its construction. Causal kutu is read with the clause it links, and shiga connects propositions in the use illustrated here (Shimoji and Pellard 2010, § Okinawan).
| Form | Romanization | Gloss | Translation |
|---|---|---|---|
| みじ ぬでぃ、やーんかい いちゅん。 | Miji nudi, yaa-nkai ichun. | Water drink.CVB house-ALL go.NPST. | “[I'll] drink water and go home.” (constructed) |
| タルーが ちゃんくとぅ、マツーが みじ ぬだん。 | Taruu-ga chan-kutu, Matsuu-ga miji nudan. | Taruu-NOM come.PST-CAUSAL Matsuu-NOM water drink.PST. | “Because Taruu came, Matsuu drank water.” (constructed) |
| タルーが ちゃんしが、マツーが みじ ぬだん。 | Taruu-ga chan-shiga, Matsuu-ga miji nudan. | Taruu-NOM come.PST-ADVERSATIVE Matsuu-NOM water drink.PST. | “Taruu came, but Matsuu drank water.” (constructed) |
The first example has a reading with a single understood participant who drinks and then goes home. In the causal example, the linker makes the arrival the reason for the drinking, while the circumstances explain why it had that effect. In the adversative example, the listener needs a suitable expectation, such as that Taruu was the one who would drink. The linker signals the contrast and the context supplies its basis. Both examples with linkers have a different overt subject in each clause.
A discourse relation can extend across turns. One participant can ask for a reason and another can supply the causal clause alone, with the event being explained already available from the question. A speaker can also supply a contrasting proposition in response to another person's assertion. The relevant unit for interpretation then includes what both participants have said.
Turn-taking, responses and reported speech
Speaker and addressee roles are assigned afresh on each turn. A question can establish a participant or event that an answer picks up without repeating the relevant words. A short affirmative can confirm what the question asks, and a predicate-only answer can supply the requested information. In an exchange where the question is addressed to the expected visitor, the addressee understood in the question becomes the speaker understood in the answer.
| Turn | Form | Romanization | Gloss | Translation |
|---|---|---|---|---|
| A | あちゃー ちゅみ。 | Achaa chumi. | Tomorrow come.Q. | “Are you coming tomorrow?” (constructed) |
| B | うー、ちゅん。 | Uu, chun. | Yes come.NPST. | “Yes, I am.” (constructed) |
The interrogative form chumi makes the first turn a polar question. In the answer, chun repeats the predicate in its nonpast form. The time and the visit carry over from the question, and the usual interpretation in this exchange assigns both predicates to B. A question already understood to be about a third person could instead receive an answer about that person. An affirmative can also serve as an acknowledgment in a continuing turn. Whether it confirms a proposition or acknowledges the other person depends on what was said and how the exchange proceeds (Shimoji and Pellard 2010, § Okinawan).
A direct quotation presents wording from another speech situation. In quoted wording, a first-person pronoun is interpreted with respect to the quoted speaker, and a time or place expression can take its reference from that speaker's situation. In the surrounding report, those same expressions are interpreted with respect to the reporting utterance. An omitted participant inside a direct quotation can be recovered from the quoted situation. The subjects of the quoted and reporting clauses therefore need to be identified in their respective contexts.
| Form | Romanization | Gloss | Translation |
|---|---|---|---|
| 「あちゃー ちゅん」んでぃ | “Achaa chun” ndi | Tomorrow come.NPST QUOT | “‘I'll come tomorrow,’ [someone said].” (constructed) |
The quotative ndi follows the quoted material. The example is a quotative fragment, with the reporting predicate supplied in the English translation. On the reading shown, the understood visitor is the speaker of the original words, and “tomorrow” is the day after those words were spoken. A report can also give the content in the reporter's wording. Recovering the referents then depends on whether an expression is presented as part of the original utterance or supplied by the reporter. The quotative marks the material being reported. The identity of its source comes from a reporting clause or the discourse (Shimoji and Pellard 2010, § Okinawan).
How Mova handles reference and discourse
Central Okinawan (ryu) has no level-2 model in tools/mova-parse. At level 1, tools/langseg assigns language labels to sentences and spans. Its Latin-script recognition covers English and French, and scripts outside its supported set receive other. It supplies no Central Okinawan identification for kana text. When tools/mova-parse has ryu tokens, it reports the missing model in the # parse comment. The tokens receive _ in LEMMA, UPOS, XPOS, FEATS, HEAD and DEPREL, with Parse=None in MISC.
Mova holds no annotated Central Okinawan treebank, so it has no measured treebank baseline for the language. The baseline uses en xpos-fill on the UD training split, en train to train the parser and tagger, and en parse-eval to measure UAS and LAS on the development split.
Mova's frames, prag and world read English trees and have no Central Okinawan trees to read. The rules in this chapter, including those for demonstrative reference, zero anaphora, clause linking and interpretation across turns, are unimplemented. Squid reads each grammatical point as a claim and queues formalizable claims for implementation. For Central Okinawan, the first implementation needs are the rules required by a tokenizer, a tagger and a UD treebank baseline.
Sources for this chapter
- Chamberlain, Basil Hall. 1895. Essay in Aid of a Grammar and Dictionary of the Luchuan Language. Transactions of the Asiatic Society of Japan 23, supplement. Yokohama. — public domain; not held in the library, to verify
- Shimoji, Michinori, and Thomas Pellard, eds. 2010. An Introduction to Ryukyuan Languages. Tokyo: Research Institute for Languages and Cultures of Asia and Africa, Tokyo University of Foreign Studies. — CC BY-NC-SA; not held in the library, to verify
This page in Markdown → https://md.mova.elamur.ai/free/en/docs/languages/ja/ryu/29-reference-and-discourse/