UD for Yupik
Tokenization and Word Segmentation
- Words are generally delimited by whitespace or punctuation.
- Words consisting of multiple morphemes are further analyzed into morphemes based on the Chen et al. (2020) Yupik morphological analyzer. See Park et al. (2021) for more details.
- In order to enable the use of morphemes as tokens, we adapted the existing “multiword expressions” annotation mechanism.
- Word-level annotation is available in the github repo as a separate file (not-to-release/ess_sli-ud-test.merged.conllu). We used udapy to remove “multiword expressions” to convert the morpheme-level annotations to word-level annotations.
Morphology
* Click on the hyperlinks for Yupik examples.
Tags
- Most Yupik words are nominals (NOUN), verbals (VERB) or particles (PART).
- If a word is further analyzed into morphemes, noun-yielding derivational morphemes are tagged as NOUN and verb-yielding derivational morphemes are tagged as VERB.
- Particles (PART) do not inflect, so they are not further analyzed into morphemes.
- ADV is used for interrogative adverbs, quantifier-qualifier constructions and demonstrative adverbs.
- PRON is used for demonstrative pronouns and interrogative pronouns.
- DET is used for two interrogative words.
- CCONJ is used for an Yupik enclitic (=llu “and”).
- NUM is used for numerals.
- X is used for inflectional suffixes.
- ADJ, ADP, AUX, INTJ, PROPN, SCONJ, and SYM do not occur in the corpus.
Features
- Unpossessed nouns inflect for Case and Number.
- Possessed nouns also inflect for Number[psor] and Person[psor] to encode the possessor’s number and person information.
- Verbs inflect for Mood, Number[obj], Number[subj], Person[obj] and Person[subj]. Depending on the verb’s valency (Subcat), the verb will encode the number and person information about the subject only or both the subject and object.
- Yupik inflections distinguish third person reflexive (also known as fourth person) from third person. In possessed nouns, Reflex is used when the possessor refers to the subject. Reflex is also used for quantifier-qualifier constructions (see ADV). In verbs, Reflex[obj] or Reflex[subj] is used when the verb’s object or subject refers back to the subject.
Syntax
* Click on the hyperlinks for Yupik examples.
- Yupik does not have a fixed word order.
- Yupik use the absolutive-ergative alignment to distinguish subjects (nsubj) and objects (obj). The subject of an intransitive verb is in absolutive case. The subject of a transitive verb is in ergative case. The direct object of a transitive verb is in absolutive case.
- Nouns in ablative-modalis case may be used as an object of an intransitive verb. In such cases, those nouns are understood as “an indefinite object” (Jacobson 2001, p.20). Because such nouns are not encoded in the verb, we tagged them as oblique nominals in ablative-modalis case (obl:mod) to distinguish them from other oblique nominals (obl).
- There are no known auxiliaries or copulas as stand-alone words. However, some derivational suffixes seem to play the role of auxiliaries or copulas and tagged as such.
- We used dep:aux to tag those derivational suffixes that act like auxiliaries (c.f. aux).
- We used dep:cop for a copula-like derviational suffix (c.f. cop).
- dep:mark is used for derivational suffixes that act like a marker (c.f. mark)
- nmod:arg is used when a nominal base is used as the argument of a noun-elaborating (N→N) derivational suffix.
- dep:infl is used for inflectional suffixes.
- dep:ana is used for the anaphoric prefix.
- dep:emo is used for emotional roots.
- dep:pos is used for postural roots.
Treebanks
There is 1 Yupik UD treebank:
References
- Emily Chen, Hyunji Hayley Park, and Lane Schwartz. 2020. Improved finite-state morphological analysis for St. Lawrence Island Yupik using paradigm function morphology. In Proceedings of the 12th Language Resources and Evaluation Conference, pages 2676–2684, Marseille, France. European Language Resources Association.
- Steven A. Jacobson. 2001. A Practical Grammar of the St. Lawrence Island / Siberian Yupik Eskimo Language, Preliminary Edition, 2nd edition. Alaska Native Language Center, Fairbanks, Alaska.
- Hyunji Hayley Park, Lane Schwartz, and Francis M. Tyers. 2021. Expanding Universal Dependencies for Polysynthetic Languages: A Case of St. Lawrence Island Yupik. In Proceedings of the 1st Workshop on NLP for Indigenous Languages of the Americas (AmericasNLP). Online. Association for Computational Linguistics.