UD for Peripheral Mongolian 
Peripheral Mongolian (ISO 639-3: mvf) is a group of Mongolic varieties spoken
mainly in the Inner Mongolia Autonomous Region of China. The variety documented
in the current Universal Dependencies data is Ordos, spoken in the Ordos
region of Inner Mongolia. Like other Mongolic languages, Ordos is an
agglutinative language with predominantly suffixing morphology and basic
subject–object–verb (SOV) word order.
The current UD data for Ordos is presented in Latin transliteration and accompanied by English translations.
Tokenization and Word Segmentation
- Words are delimited by whitespace characters in the transliteration.
- Punctuation marks are tokenized as separate tokens.
- There are currently no fused multi-word tokens and no words with spaces.
Morphology
Tags
- The annotation uses the universal POS tags. Open classes such as NOUN, VERB, ADJ and ADV are well attested, as are closed classes such as PRON, DET, ADP, NUM and PUNCT.
- The copula bai “to be” is annotated as AUX when used as a copula or in periphrastic verb forms.
Features
The annotation prioritizes the universal morphological features, which account for the large majority of feature annotations. A small number of language-specific features are used to capture distinctions that are morphologically overt in Ordos:
Derivation=Collective,Derivation=Iterative,Derivation=Multiplicativeare used to mark derivational morphology on verbs.Focus=Onlymarks focus particles at the clausal level.- Possessive agreement on nominal heads is marked with
Person[psor](Person[psor]=1,Person[psor]=2,Person[psor]=3) andNumber[psor](Number[psor]=Sing,Number[psor]=Plur), encoding possessive suffixes on nouns. - Nominal case is marked with the Case feature, including nominative, accusative, genitive, dative and other values.
- Verbal categories are marked with Mood, Tense and VerbForm.
Finite verbs use
VerbForm=Fin; converbs (verbal adverbs) functioning as predicates of dependent adverbial clauses useVerbForm=Conv.
Syntax
Core arguments
- Subjects are typically in the nominative case and are annotated with nsubj for nominal subjects.
- Direct objects are typically in the accusative case and are annotated with obj.
- Indirect objects are annotated with iobj.
Copula constructions
- Nonverbal predicates (nominal and adjectival) may be combined with the copula bai, which is attached to the nonverbal predicate with the cop relation.
Language-specific relations
The following language-specific subtypes of dependency relations are used in the Ordos data:
- det:poss — possessive determiner
- nmod:tmod — temporal nominal modifier
- obl:agent — oblique agent
- obl:comit — comitative oblique
- obl:instr — instrumental oblique
- obl:than — oblique standard of comparison
- obl:time — temporal oblique
- obl:tmod — temporal oblique modifier
- subj — subject (used where a finer subject subtype is needed)
References
- Georg, Stefan. 2003. Ordos. In Juha Janhunen (ed.), The Mongolic Languages, 193–209. London: Routledge.
Treebanks
There is one Peripheral Mongolian UD treebank: