home edit page issue tracker

UD for Cuicatec

Cuicatec (ISO 639-3 cux) is an Oto-Manguean language of the Mixtecan branch, spoken in the Cañada region of Oaxaca, Mexico. The UD data is focused on Tepeuxila Cuicatec (dbaku), spoken around San Juan Tepeuxila.

Tokenization and Word Segmentation

• Words are delimited by whitespace. There are no multiword tokens and no words with spaces.
• Punctuation is split off as separate tokens, including the Spanish-style opening marks ¿ and ¡ (e.g. ¿ A ichad ? “¿Tienes frío?”).
• Possessive, person and other affixes and enclitics are written together with their host and are not split off (e.g. bakud “tu casa”, kuchis “vino (él)”).
• A small number of compound words are written with a hyphen in the source and are kept as one token (e.g. naíd-yádi).
• The glottal stop is written with the modifier letter apostrophe ʼ (U+02BC). The original text, which often uses the ASCII apostrophe, is kept in the text[orig] sentence attribute, and a Spanish translation in text[spa].
• Spelling follows each author’s own orthographic choices, so the same word can appear with different diacritics. Non-standard spellings of a word are marked Typo=Yes, with the standard form given in MISC as Normal (e.g. Un → Normal=ú “yo”).

Morphology

Tags

• SYM is not used.
• NOUN: Nouns take possessive suffixes agreeing with the possessor (e.g. baʼa “casa”, bakud “tu casa”). Relational nouns (NounType=Relat) are tagged NOUN, including the possessive base ñeʼe “de (él)”, which is inflected for its possessor (ñeʼë “de mí”, ñeʼed “de ti”). Nominal classifiers such as ta “señora”, used before women’s names, are also NOUN, with NounType=Clf.
• VERB: Verbs inflect for the person of the subject by suffixes, and in the third person distinguish masculine, feminine, human and animal referents (e.g. kuchis “vino (él)”, jeakuta “llora (ella)”). Aspect and mood are marked by prefixes, such as completive, potential and progressive ji- (e.g. jiyadu “duermo”).
• ADJ: Property words such as cheatea “grande”, lï “pequeño” and kuä “negro” are tagged ADJ. They are frequently used as predicates without a copula, and are then marked VerbForm=Stat.
• AUX: The verb kane/kanu “estar” is tagged AUX when it is the copula of a locative predicate (see Syntax). The modal jikadi “poder” (chikadi “pude”) is tagged AUX when it is followed by the main verb (e.g. Gua jikadi jika dime. “No puedo caminar solo.”).
• DET and PRON: Demonstratives that modify a noun are tagged DET and attached as det. They usually follow the noun: kon/ko/kó “ese”, kun/ku “este”, mi/min “esa/esta” (e.g. Baʼa kun cheatea. “Esta casa es grande.”). Demonstratives used on their own, personal pronouns and interrogatives are tagged PRON. The personal pronouns are ú “yo”, dí “tú”, sakó “él”, takó “ella”, insú “nosotros”, insë “ustedes” and iñsé “ellos”; the interrogatives include duʼu “quién”, dae “qué” and tabea “cuánto”.
• PART: The polar question particle a is tagged PART and attached as discourse (e.g. ¿A ichad? “¿Tienes frío?”). The emphatic particle ne is also PART (e.g. Úu ne chidi. “Yo estornudé.”).
• ADV: Besides manner, time and place adverbs (iku “ayer”, muʼu “aquí”, yeabean “muy”), the negation gua “no” is tagged ADV and attached as advmod.
• ADP: Prepositions such as no “en” and nuku “con” are tagged ADP and attached as case (e.g. No camión kuchi. “Vine en camión.”).
• SCONJ: The complementizer chi “que”, and tochi “cuando”, kat “porque” and tumin “para que”.
• CCONJ: Coordinating conjunctions are rare. They include o “o” and the Spanish loan per “pero”.
• NUM: Cardinal numerals (NumType=Card) such as ama “uno”, obe “dos” and inu “tres” precede the noun.

Features

• Nominal features: NounType distinguishes relational nouns (Relat) and classifiers (Clf). Possession is marked by suffixes on the possessed noun, but it is not yet annotated with features.
• Verbal features: Finite verbs are marked VerbForm=Fin. Where agreement is annotated, verbs carry Person and Number, and third-person forms carry Gender=Masc/Fem or Animacy=Hum/Aml, following the masculine, feminine, human and animal categories of third-person agreement. Aspect (Aspect=Perf completive, Aspect=Prog progressive) and mood (Mood=Pot potential) are also marked, as is transitivity (Subcat=Tran/Intr).
• Stative: Predicative property words are marked VerbForm=Stat.
• Pronouns and determiners: PronType=Prs (personal), Dem (demonstrative) and Int (interrogative).
• Spelling: Typo=Yes marks non-standard spellings (see Tokenization).
• Morphological feature annotation is not yet complete. Only some finite verbs are annotated for person, number, aspect and mood, and lemmas are given for only part of the corpus.

Syntax

• Core arguments: The subject is indexed on the verb by person suffixes, and an overt subject is often omitted (e.g. Gua deanu. “No sé.”). Overt subjects (nsubj) may precede or follow the verb. Objects (obj) are unmarked and usually follow the verb. When both are present, the most frequent order is subject–verb–object (e.g. Takó chikueta ú. “Ella me enseñó.”). In ditransitives, the recipient or addressee is iobj (e.g. Kakued ú yuni. “¡Enséñame el camino!”).
• Nonverbal clauses: Adjectival, nominal and pronominal predicates take no copula. The predicate is the root and the subject is its nsubj (e.g. Ú yaku. “Yo soy alto.”, Ta Marian ndatata. “María es mujer.”, ¿Duʼu noʼöm? “¿Quién es el maestro?”). Locative predicates take the copula kane/kanu “estar”, tagged AUX and attached as cop to the locative expression (e.g. Chida kaneñ no nʼiba. “Mi papá está en el mercado.”). When kane is used without a locative complement, it is a VERB.
• Existence and possession: The existential verb bea “hay” is the root, and the existing entity is its nsubj (e.g. Muʼu bea nuni. “Aquí hay agua.”). Predicative possession uses the same construction, with the possessor expressed by the possessive base ñeʼe (e.g. Bea ntiñu ñeʼë. “Tengo trabajo.”) or by a fronted noun phrase, attached as dislocated (e.g. Uu bea yuchmee. “Yo tengo cerebro.”).
• Adnominal possession: The possessive base ñeʼe, inflected for the possessor, follows the possessed noun and is attached to it as nmod:poss (e.g. Gua yana ñeʼë. “No tengo perro.”).
• Questions and negation: Polar questions are marked by the particle a (discourse), and negation by gua (advmod).
• Complex predicates: A verb followed by a second inflected verb sharing its subject (motion verbs, neʼe “querer”) takes the second verb as xcomp (e.g. Kued kamad. “Ve a leer.”). Complement clauses introduced by chi “que” are ccomp (e.g. Gua deanu chi kuchid. “No sabía que venías.”), and adverbial clauses introduced by tochi “cuando” are advcl.
• Relation subtypes: The only subtype used is nmod:poss, for possessors.

Treebanks

There is 1 Cuicatec UD treebank:

• Cuicatec-MesoTree

© 2014–2024 Universal Dependencies contributors. Site powered by Annodoc and brat

.