This is part of archived UD v1 documentation. See http://universaldependencies.org/ for the current version.

home issue tracker

Dependencies

Note: nmod, neg, and punct appear in two places.

Core dependents of clausal predicates Nominal dep Predicate dep nsubj csubj nsubjpass csubjpass dobj ccomp xcomp iobj	Non-core dependents of clausal predicates Nominal dep Predicate dep Modifier word nmod advcl advmod     neg	Special clausal dependents Nominal dep Auxiliary Other vocative aux mark discourse auxpass punct expl cop
Noun dependents Nominal dep Predicate dep Modifier word nummod acl amod appos   det nmod   neg	Compounding and unanalyzed compound mwe goeswith name foreign	Coordination conj cc punct
Case-marking, prepositions, possessive case	Loose joining relations list parataxis remnant dislocated reparandum	Other Sentence head Unspecified dependency root dep

---

acl: clausal modifier of noun

The dependency type acl is used for relative clauses and clausal complements of nouns. For a relative clause, the head of the dependency is the noun modified by the clause, and the dependant is the main predicate of the clause.

edit

  

    
input
    
brat
    
info
  
太郎 が 食べ た リンゴ \n Taro NOM eat PAST apple
acl(リンゴ, 食べ)
case(太郎, が)
aux(食べ, た)
{
    "text": "太郎 が 食べ た リンゴ \n Taro NOM eat PAST apple ",
    "entities": [
        [ "embedded-1-T1", "token", [ [ 0, 2 ] ] ],
        [ "embedded-1-T2", "token", [ [ 3, 4 ] ] ],
        [ "embedded-1-T3", "token", [ [ 5, 7 ] ] ],
        [ "embedded-1-T4", "token", [ [ 8, 9 ] ] ],
        [ "embedded-1-T5", "token", [ [ 10, 13 ] ] ],
        [ "embedded-1-T7", "token", [ [ 16, 20 ] ] ],
        [ "embedded-1-T8", "token", [ [ 21, 24 ] ] ],
        [ "embedded-1-T9", "token", [ [ 25, 28 ] ] ],
        [ "embedded-1-T10", "token", [ [ 29, 33 ] ] ],
        [ "embedded-1-T11", "token", [ [ 34, 39 ] ] ]
    ],
    "relations": [
        [ "embedded-1-R1", "acl", [ [ "arg1", "embedded-1-T5" ], [ "arg2", "embedded-1-T3" ] ] ],
        [ "embedded-1-R2", "case", [ [ "arg1", "embedded-1-T1" ], [ "arg2", "embedded-1-T2" ] ] ],
        [ "embedded-1-R3", "aux", [ [ "arg1", "embedded-1-T3" ], [ "arg2", "embedded-1-T4" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "1" ]
}

Formal noun こと / koto “fact” can have a clausal complement and forms a noun phrase denoting the action expressed by the clause. For a example, 走る / hashiru “run” こと / koto “fact” means running (or a fact that someone runs).

edit

  

    
input
    
brat
    
info
  
ほめ られ た こと \n praise PASSIVE PAST fact
acl(こと, ほめ)
auxpass(ほめ, られ)
aux(ほめ, た)
{
    "text": "ほめ られ た こと \n praise PASSIVE PAST fact ",
    "entities": [
        [ "embedded-2-T1", "token", [ [ 0, 2 ] ] ],
        [ "embedded-2-T2", "token", [ [ 3, 5 ] ] ],
        [ "embedded-2-T3", "token", [ [ 6, 7 ] ] ],
        [ "embedded-2-T4", "token", [ [ 8, 10 ] ] ],
        [ "embedded-2-T6", "token", [ [ 13, 19 ] ] ],
        [ "embedded-2-T7", "token", [ [ 20, 27 ] ] ],
        [ "embedded-2-T8", "token", [ [ 28, 32 ] ] ],
        [ "embedded-2-T9", "token", [ [ 33, 37 ] ] ]
    ],
    "relations": [
        [ "embedded-2-R1", "acl", [ [ "arg1", "embedded-2-T4" ], [ "arg2", "embedded-2-T1" ] ] ],
        [ "embedded-2-R2", "auxpass", [ [ "arg1", "embedded-2-T1" ], [ "arg2", "embedded-2-T2" ] ] ],
        [ "embedded-2-R3", "aux", [ [ "arg1", "embedded-2-T1" ], [ "arg2", "embedded-2-T3" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "2" ]
}

edit acl

advcl: adverbial clause modifier

The dependency type advcl is reserved for subordinate clauses. The head of the dependency is the main predicate of the main clause, and the dependant is the main predicate of the subordinate clause. A subordinate clauses consists of a clause and a word of “SCONJ”, and is typically followed by the main clause.

edit

  

    
input
    
brat
    
info
  
食べ た なら 行く 。 \n eat PAST if go .
advcl(行く, 食べ)
aux(食べ, た)
mark(食べ, なら)
punct(行く, 。)
{
    "text": "食べ た なら 行く 。 \n eat PAST if go . ",
    "entities": [
        [ "embedded-3-T1", "token", [ [ 0, 2 ] ] ],
        [ "embedded-3-T2", "token", [ [ 3, 4 ] ] ],
        [ "embedded-3-T3", "token", [ [ 5, 7 ] ] ],
        [ "embedded-3-T4", "token", [ [ 8, 10 ] ] ],
        [ "embedded-3-T5", "token", [ [ 11, 12 ] ] ],
        [ "embedded-3-T7", "token", [ [ 15, 18 ] ] ],
        [ "embedded-3-T8", "token", [ [ 19, 23 ] ] ],
        [ "embedded-3-T9", "token", [ [ 24, 26 ] ] ],
        [ "embedded-3-T10", "token", [ [ 27, 29 ] ] ],
        [ "embedded-3-T11", "token", [ [ 30, 31 ] ] ]
    ],
    "relations": [
        [ "embedded-3-R1", "advcl", [ [ "arg1", "embedded-3-T4" ], [ "arg2", "embedded-3-T1" ] ] ],
        [ "embedded-3-R2", "aux", [ [ "arg1", "embedded-3-T1" ], [ "arg2", "embedded-3-T2" ] ] ],
        [ "embedded-3-R3", "mark", [ [ "arg1", "embedded-3-T1" ], [ "arg2", "embedded-3-T3" ] ] ],
        [ "embedded-3-R4", "punct", [ [ "arg1", "embedded-3-T4" ], [ "arg2", "embedded-3-T5" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "3" ]
}

edit advcl

advmod: adverbial modifier

The dependency type advmod is used for adverbial modifiers of verbs, nominal verbs, adjectives and nominal adjectives.

edit

  

    
input
    
brat
    
info
  
ゆっくり 走る 。 \n slowly run .
advmod(走る, ゆっくり)
punct(走る, 。)
{
    "text": "ゆっくり 走る 。 \n slowly run . ",
    "entities": [
        [ "embedded-4-T1", "token", [ [ 0, 4 ] ] ],
        [ "embedded-4-T2", "token", [ [ 5, 7 ] ] ],
        [ "embedded-4-T3", "token", [ [ 8, 9 ] ] ],
        [ "embedded-4-T5", "token", [ [ 12, 18 ] ] ],
        [ "embedded-4-T6", "token", [ [ 19, 22 ] ] ],
        [ "embedded-4-T7", "token", [ [ 23, 24 ] ] ]
    ],
    "relations": [
        [ "embedded-4-R1", "advmod", [ [ "arg1", "embedded-4-T2" ], [ "arg2", "embedded-4-T1" ] ] ],
        [ "embedded-4-R2", "punct", [ [ "arg1", "embedded-4-T2" ], [ "arg2", "embedded-4-T3" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "4" ]
}

edit advmod

amod: adjectival modifier

Adjectival modifiers (amod) are adjectives and nominal adjectives. A nominal adjective must be followed by an auxiliary な / na to modify a noun. Note that adjectival modifiers with arguments are tagged as “acl”.

edit

  

    
input
    
brat
    
info
  
きれい な リンゴ \n beautiful AUX apple
amod(リンゴ, きれい)
aux(きれい, な)
{
    "text": "きれい な リンゴ \n beautiful AUX apple ",
    "entities": [
        [ "embedded-5-T1", "token", [ [ 0, 3 ] ] ],
        [ "embedded-5-T2", "token", [ [ 4, 5 ] ] ],
        [ "embedded-5-T3", "token", [ [ 6, 9 ] ] ],
        [ "embedded-5-T5", "token", [ [ 12, 21 ] ] ],
        [ "embedded-5-T6", "token", [ [ 22, 25 ] ] ],
        [ "embedded-5-T7", "token", [ [ 26, 31 ] ] ]
    ],
    "relations": [
        [ "embedded-5-R1", "amod", [ [ "arg1", "embedded-5-T3" ], [ "arg2", "embedded-5-T1" ] ] ],
        [ "embedded-5-R2", "aux", [ [ "arg1", "embedded-5-T1" ], [ "arg2", "embedded-5-T2" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "5" ]
}

edit amod

appos: appositional modifier

Appositive relations.

edit

  

    
input
    
brat
    
info
  
形態素 解析 器 Mecab \n morpheme analysis tool Mecab
appos(Mecab, 器)
compound(解析, 形態素)
compound(器, 解析)
{
    "text": "形態素 解析 器 Mecab \n morpheme analysis tool Mecab ",
    "entities": [
        [ "embedded-6-T1", "token", [ [ 0, 3 ] ] ],
        [ "embedded-6-T2", "token", [ [ 4, 6 ] ] ],
        [ "embedded-6-T3", "token", [ [ 7, 8 ] ] ],
        [ "embedded-6-T4", "token", [ [ 9, 14 ] ] ],
        [ "embedded-6-T6", "token", [ [ 17, 25 ] ] ],
        [ "embedded-6-T7", "token", [ [ 26, 34 ] ] ],
        [ "embedded-6-T8", "token", [ [ 35, 39 ] ] ],
        [ "embedded-6-T9", "token", [ [ 40, 45 ] ] ]
    ],
    "relations": [
        [ "embedded-6-R1", "appos", [ [ "arg1", "embedded-6-T4" ], [ "arg2", "embedded-6-T3" ] ] ],
        [ "embedded-6-R2", "compound", [ [ "arg1", "embedded-6-T2" ], [ "arg2", "embedded-6-T1" ] ] ],
        [ "embedded-6-R3", "compound", [ [ "arg1", "embedded-6-T3" ], [ "arg2", "embedded-6-T2" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "6" ]
}

edit appos

aux: auxiliary

The dependency type aux is used for auxiliary verbs and particles that attach to predicates.

edit

  

    
input
    
brat
    
info
  
ご飯 を 食べ た 。 \n rice ACC eat PAST .
case(ご飯, を)
dobj(食べ, ご飯)
aux(食べ, た)
punct(食べ, 。)
{
    "text": "ご飯 を 食べ た 。 \n rice ACC eat PAST . ",
    "entities": [
        [ "embedded-7-T1", "token", [ [ 0, 2 ] ] ],
        [ "embedded-7-T2", "token", [ [ 3, 4 ] ] ],
        [ "embedded-7-T3", "token", [ [ 5, 7 ] ] ],
        [ "embedded-7-T4", "token", [ [ 8, 9 ] ] ],
        [ "embedded-7-T5", "token", [ [ 10, 11 ] ] ],
        [ "embedded-7-T7", "token", [ [ 14, 18 ] ] ],
        [ "embedded-7-T8", "token", [ [ 19, 22 ] ] ],
        [ "embedded-7-T9", "token", [ [ 23, 26 ] ] ],
        [ "embedded-7-T10", "token", [ [ 27, 31 ] ] ],
        [ "embedded-7-T11", "token", [ [ 32, 33 ] ] ]
    ],
    "relations": [
        [ "embedded-7-R1", "case", [ [ "arg1", "embedded-7-T1" ], [ "arg2", "embedded-7-T2" ] ] ],
        [ "embedded-7-R2", "dobj", [ [ "arg1", "embedded-7-T3" ], [ "arg2", "embedded-7-T1" ] ] ],
        [ "embedded-7-R3", "aux", [ [ "arg1", "embedded-7-T3" ], [ "arg2", "embedded-7-T4" ] ] ],
        [ "embedded-7-R4", "punct", [ [ "arg1", "embedded-7-T3" ], [ "arg2", "embedded-7-T5" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "7" ]
}

edit

  

    
input
    
brat
    
info
  
行き ます か 。 \n go POLITE QUESTION .
aux(行き, ます)
aux(行き, か)
punct(行き, 。)
{
    "text": "行き ます か 。 \n go POLITE QUESTION . ",
    "entities": [
        [ "embedded-8-T1", "token", [ [ 0, 2 ] ] ],
        [ "embedded-8-T2", "token", [ [ 3, 5 ] ] ],
        [ "embedded-8-T3", "token", [ [ 6, 7 ] ] ],
        [ "embedded-8-T4", "token", [ [ 8, 9 ] ] ],
        [ "embedded-8-T6", "token", [ [ 12, 14 ] ] ],
        [ "embedded-8-T7", "token", [ [ 15, 21 ] ] ],
        [ "embedded-8-T8", "token", [ [ 22, 30 ] ] ],
        [ "embedded-8-T9", "token", [ [ 31, 32 ] ] ]
    ],
    "relations": [
        [ "embedded-8-R1", "aux", [ [ "arg1", "embedded-8-T1" ], [ "arg2", "embedded-8-T2" ] ] ],
        [ "embedded-8-R2", "aux", [ [ "arg1", "embedded-8-T1" ], [ "arg2", "embedded-8-T3" ] ] ],
        [ "embedded-8-R3", "punct", [ [ "arg1", "embedded-8-T1" ], [ "arg2", "embedded-8-T4" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "8" ]
}

edit aux

auxpass: passive auxiliary

The passive auxiliaries in Japanese are れる / reru and られる / rareru. They attach to the main verbs to form verbal phrases in the passive voice. Note that passivization causes case alternations.

edit

  

    
input
    
brat
    
info
  
太郎 が 雇わ れ た 。 \n Taro NOM employ PASSIVE PAST .
auxpass(雇わ, れ)
case(太郎, が)
nsubjpass(雇わ, 太郎)
aux(雇わ, た)
punct(雇わ, 。)
{
    "text": "太郎 が 雇わ れ た 。 \n Taro NOM employ PASSIVE PAST . ",
    "entities": [
        [ "embedded-9-T1", "token", [ [ 0, 2 ] ] ],
        [ "embedded-9-T2", "token", [ [ 3, 4 ] ] ],
        [ "embedded-9-T3", "token", [ [ 5, 7 ] ] ],
        [ "embedded-9-T4", "token", [ [ 8, 9 ] ] ],
        [ "embedded-9-T5", "token", [ [ 10, 11 ] ] ],
        [ "embedded-9-T6", "token", [ [ 12, 13 ] ] ],
        [ "embedded-9-T8", "token", [ [ 16, 20 ] ] ],
        [ "embedded-9-T9", "token", [ [ 21, 24 ] ] ],
        [ "embedded-9-T10", "token", [ [ 25, 31 ] ] ],
        [ "embedded-9-T11", "token", [ [ 32, 39 ] ] ],
        [ "embedded-9-T12", "token", [ [ 40, 44 ] ] ],
        [ "embedded-9-T13", "token", [ [ 45, 46 ] ] ]
    ],
    "relations": [
        [ "embedded-9-R1", "auxpass", [ [ "arg1", "embedded-9-T3" ], [ "arg2", "embedded-9-T4" ] ] ],
        [ "embedded-9-R2", "case", [ [ "arg1", "embedded-9-T1" ], [ "arg2", "embedded-9-T2" ] ] ],
        [ "embedded-9-R3", "nsubjpass", [ [ "arg1", "embedded-9-T3" ], [ "arg2", "embedded-9-T1" ] ] ],
        [ "embedded-9-R4", "aux", [ [ "arg1", "embedded-9-T3" ], [ "arg2", "embedded-9-T5" ] ] ],
        [ "embedded-9-R5", "punct", [ [ "arg1", "embedded-9-T3" ], [ "arg2", "embedded-9-T6" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "9" ]
}

edit auxpass

case: case marking

In Japanese, case markers are particles attaching to a noun phrase. The particles are tagged as “ADP” in UD Japanese.

Typical case markers are as follows:

• が / ga for nominative
• を / o for accusative
• に / ni for dative, locative etc.
• の / no for genitive
• へ / e for locative etc.

edit

  

    
input
    
brat
    
info
  
太郎 が 東京 へ 行く 。\n Taro NOM Tokyo TO go .
case(太郎, が)
case(東京, へ)
nsubj(行く, 太郎)
nmod(行く, 東京)
{
    "text": "太郎 が 東京 へ 行く 。\n Taro NOM Tokyo TO go . ",
    "entities": [
        [ "embedded-10-T1", "token", [ [ 0, 2 ] ] ],
        [ "embedded-10-T2", "token", [ [ 3, 4 ] ] ],
        [ "embedded-10-T3", "token", [ [ 5, 7 ] ] ],
        [ "embedded-10-T4", "token", [ [ 8, 9 ] ] ],
        [ "embedded-10-T5", "token", [ [ 10, 12 ] ] ],
        [ "embedded-10-T6", "token", [ [ 13, 15 ] ] ],
        [ "embedded-10-T7", "token", [ [ 16, 20 ] ] ],
        [ "embedded-10-T8", "token", [ [ 21, 24 ] ] ],
        [ "embedded-10-T9", "token", [ [ 25, 30 ] ] ],
        [ "embedded-10-T10", "token", [ [ 31, 33 ] ] ],
        [ "embedded-10-T11", "token", [ [ 34, 36 ] ] ],
        [ "embedded-10-T12", "token", [ [ 37, 38 ] ] ]
    ],
    "relations": [
        [ "embedded-10-R1", "case", [ [ "arg1", "embedded-10-T1" ], [ "arg2", "embedded-10-T2" ] ] ],
        [ "embedded-10-R2", "case", [ [ "arg1", "embedded-10-T3" ], [ "arg2", "embedded-10-T4" ] ] ],
        [ "embedded-10-R3", "nsubj", [ [ "arg1", "embedded-10-T5" ], [ "arg2", "embedded-10-T1" ] ] ],
        [ "embedded-10-R4", "nmod", [ [ "arg1", "embedded-10-T5" ], [ "arg2", "embedded-10-T3" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "10" ]
}

edit

  

    
input
    
brat
    
info
  
太郎 の 家 \n Taro OF house
case(太郎, の)
nmod(家, 太郎)
{
    "text": "太郎 の 家 \n Taro OF house ",
    "entities": [
        [ "embedded-11-T1", "token", [ [ 0, 2 ] ] ],
        [ "embedded-11-T2", "token", [ [ 3, 4 ] ] ],
        [ "embedded-11-T3", "token", [ [ 5, 6 ] ] ],
        [ "embedded-11-T5", "token", [ [ 9, 13 ] ] ],
        [ "embedded-11-T6", "token", [ [ 14, 16 ] ] ],
        [ "embedded-11-T7", "token", [ [ 17, 22 ] ] ]
    ],
    "relations": [
        [ "embedded-11-R1", "case", [ [ "arg1", "embedded-11-T1" ], [ "arg2", "embedded-11-T2" ] ] ],
        [ "embedded-11-R2", "nmod", [ [ "arg1", "embedded-11-T3" ], [ "arg2", "embedded-11-T1" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "11" ]
}

edit case

cc: coordinating conjunction

The dependency type cc is used for CONJ, that is, conjunction and coordinating conjunctive particle (e.g. と / to, や / ya). See the example of conj.

edit cc

ccomp: clausal complement

Clausal complements are typically introduced by a postpositional particle と / to “that”. The head of the dependency ccomp is the main verb of the main clause, and the dependant is the main predicate of the complement.

edit

  

    
input
    
brat
    
info
  
きれい だ と 思う 。 \n beautiful AUX that think .
ccomp(思う, きれい)
aux(きれい, だ)
case(きれい, と)
punct(思う, 。)
{
    "text": "きれい だ と 思う 。 \n beautiful AUX that think . ",
    "entities": [
        [ "embedded-12-T1", "token", [ [ 0, 3 ] ] ],
        [ "embedded-12-T2", "token", [ [ 4, 5 ] ] ],
        [ "embedded-12-T3", "token", [ [ 6, 7 ] ] ],
        [ "embedded-12-T4", "token", [ [ 8, 10 ] ] ],
        [ "embedded-12-T5", "token", [ [ 11, 12 ] ] ],
        [ "embedded-12-T7", "token", [ [ 15, 24 ] ] ],
        [ "embedded-12-T8", "token", [ [ 25, 28 ] ] ],
        [ "embedded-12-T9", "token", [ [ 29, 33 ] ] ],
        [ "embedded-12-T10", "token", [ [ 34, 39 ] ] ],
        [ "embedded-12-T11", "token", [ [ 40, 41 ] ] ]
    ],
    "relations": [
        [ "embedded-12-R1", "ccomp", [ [ "arg1", "embedded-12-T4" ], [ "arg2", "embedded-12-T1" ] ] ],
        [ "embedded-12-R2", "aux", [ [ "arg1", "embedded-12-T1" ], [ "arg2", "embedded-12-T2" ] ] ],
        [ "embedded-12-R3", "case", [ [ "arg1", "embedded-12-T1" ], [ "arg2", "embedded-12-T3" ] ] ],
        [ "embedded-12-R4", "punct", [ [ "arg1", "embedded-12-T4" ], [ "arg2", "embedded-12-T5" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "12" ]
}

edit ccomp

compound: compound

The dependency type compound in UD Japanese is used for noun compounds and verb compounds.

edit

  

    
input
    
brat
    
info
  
構文 解析 \n syntax analysis
compound(解析, 構文)
{
    "text": "構文 解析 \n syntax analysis ",
    "entities": [
        [ "embedded-13-T1", "token", [ [ 0, 2 ] ] ],
        [ "embedded-13-T2", "token", [ [ 3, 5 ] ] ],
        [ "embedded-13-T4", "token", [ [ 8, 14 ] ] ],
        [ "embedded-13-T5", "token", [ [ 15, 23 ] ] ]
    ],
    "relations": [
        [ "embedded-13-R1", "compound", [ [ "arg1", "embedded-13-T2" ], [ "arg2", "embedded-13-T1" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "13" ]
}

edit

  

    
input
    
brat
    
info
  
泳ぎ 着く \n swim reach
compound(着く, 泳ぎ)
{
    "text": "泳ぎ 着く \n swim reach ",
    "entities": [
        [ "embedded-14-T1", "token", [ [ 0, 2 ] ] ],
        [ "embedded-14-T2", "token", [ [ 3, 5 ] ] ],
        [ "embedded-14-T4", "token", [ [ 8, 12 ] ] ],
        [ "embedded-14-T5", "token", [ [ 13, 18 ] ] ]
    ],
    "relations": [
        [ "embedded-14-R1", "compound", [ [ "arg1", "embedded-14-T2" ], [ "arg2", "embedded-14-T1" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "14" ]
}

edit compound

conj: conjunct

The dependency type conj represents coordination. Following the definition of UD, the head of the dependency is the first conjunct.

edit

  

    
input
    
brat
    
info
  
太郎 と 次郎 \n Taro and Jiro
cc(太郎, と)
conj(太郎, 次郎)
{
    "text": "太郎 と 次郎 \n Taro and Jiro ",
    "entities": [
        [ "embedded-15-T1", "token", [ [ 0, 2 ] ] ],
        [ "embedded-15-T2", "token", [ [ 3, 4 ] ] ],
        [ "embedded-15-T3", "token", [ [ 5, 7 ] ] ],
        [ "embedded-15-T5", "token", [ [ 10, 14 ] ] ],
        [ "embedded-15-T6", "token", [ [ 15, 18 ] ] ],
        [ "embedded-15-T7", "token", [ [ 19, 23 ] ] ]
    ],
    "relations": [
        [ "embedded-15-R1", "cc", [ [ "arg1", "embedded-15-T1" ], [ "arg2", "embedded-15-T2" ] ] ],
        [ "embedded-15-R2", "conj", [ [ "arg1", "embedded-15-T1" ], [ "arg2", "embedded-15-T3" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "15" ]
}

edit conj

cop: copula

The dependency type cop is reserved for a copular auxiliary だ / da. The auxiliary typically follows a noun phrase to form a copular clause. A postpositional phrase with a nominative case is commonly needed to complete sentence.

For example, dependencies for 太郎 は 学生 だ 。/ Taro ha gakusei da “Taro is a student.” are as follows.

edit

  

    
input
    
brat
    
info
  
太郎 は 学生 だ 。 \n Taro TOPIC student COPULA .
nsubj(学生, 太郎)
case(太郎, は)
cop(学生, だ)
punct(学生, 。)
{
    "text": "太郎 は 学生 だ 。 \n Taro TOPIC student COPULA . ",
    "entities": [
        [ "embedded-16-T1", "token", [ [ 0, 2 ] ] ],
        [ "embedded-16-T2", "token", [ [ 3, 4 ] ] ],
        [ "embedded-16-T3", "token", [ [ 5, 7 ] ] ],
        [ "embedded-16-T4", "token", [ [ 8, 9 ] ] ],
        [ "embedded-16-T5", "token", [ [ 10, 11 ] ] ],
        [ "embedded-16-T7", "token", [ [ 14, 18 ] ] ],
        [ "embedded-16-T8", "token", [ [ 19, 24 ] ] ],
        [ "embedded-16-T9", "token", [ [ 25, 32 ] ] ],
        [ "embedded-16-T10", "token", [ [ 33, 39 ] ] ],
        [ "embedded-16-T11", "token", [ [ 40, 41 ] ] ]
    ],
    "relations": [
        [ "embedded-16-R1", "nsubj", [ [ "arg1", "embedded-16-T3" ], [ "arg2", "embedded-16-T1" ] ] ],
        [ "embedded-16-R2", "case", [ [ "arg1", "embedded-16-T1" ], [ "arg2", "embedded-16-T2" ] ] ],
        [ "embedded-16-R3", "cop", [ [ "arg1", "embedded-16-T3" ], [ "arg2", "embedded-16-T4" ] ] ],
        [ "embedded-16-R4", "punct", [ [ "arg1", "embedded-16-T3" ], [ "arg2", "embedded-16-T5" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "16" ]
}

Note that we treat the auxiliary だ / da after adjectives as aux. Therefore dependencies for さくら が きれい だ 。 / sakura ga kirei da “The cherry blossoms are beautiful.” do not include the `cop’ relation.

edit

  

    
input
    
brat
    
info
  
さくら が きれい だ 。 \n cherry_blossoms NOM beautiful AUX .
nsubj(きれい, さくら)
case(さくら, が)
aux(きれい, だ)
punct(きれい, 。)
{
    "text": "さくら が きれい だ 。 \n cherry_blossoms NOM beautiful AUX . ",
    "entities": [
        [ "embedded-17-T1", "token", [ [ 0, 3 ] ] ],
        [ "embedded-17-T2", "token", [ [ 4, 5 ] ] ],
        [ "embedded-17-T3", "token", [ [ 6, 9 ] ] ],
        [ "embedded-17-T4", "token", [ [ 10, 11 ] ] ],
        [ "embedded-17-T5", "token", [ [ 12, 13 ] ] ],
        [ "embedded-17-T7", "token", [ [ 16, 31 ] ] ],
        [ "embedded-17-T8", "token", [ [ 32, 35 ] ] ],
        [ "embedded-17-T9", "token", [ [ 36, 45 ] ] ],
        [ "embedded-17-T10", "token", [ [ 46, 49 ] ] ],
        [ "embedded-17-T11", "token", [ [ 50, 51 ] ] ]
    ],
    "relations": [
        [ "embedded-17-R1", "nsubj", [ [ "arg1", "embedded-17-T3" ], [ "arg2", "embedded-17-T1" ] ] ],
        [ "embedded-17-R2", "case", [ [ "arg1", "embedded-17-T1" ], [ "arg2", "embedded-17-T2" ] ] ],
        [ "embedded-17-R3", "aux", [ [ "arg1", "embedded-17-T3" ], [ "arg2", "embedded-17-T4" ] ] ],
        [ "embedded-17-R4", "punct", [ [ "arg1", "embedded-17-T3" ], [ "arg2", "embedded-17-T5" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "17" ]
}

edit cop

csubj: clausal subject

A clausal subject (csubj) is commonly introduced by a pair of particles, a nominal marker の / no and a nominative case marker が / ga. When the phrase is topicalized, the nominative が / ga is replaced by は / wa.

edit

  

    
input
    
brat
    
info
  
太郎 が 怒ら れる の は いや だ 。 \n Taro NOM punish PASSIVE that TOPIC dislike AUX .
csubj(いや, 怒ら)
nsubjpass(怒ら, 太郎)
case(太郎, が)
auxpass(怒ら, れる)
mark(怒ら, の)
case(怒ら, は)
aux(いや, だ)
punct(いや, 。)
{
    "text": "太郎 が 怒ら れる の は いや だ 。 \n Taro NOM punish PASSIVE that TOPIC dislike AUX . ",
    "entities": [
        [ "embedded-18-T1", "token", [ [ 0, 2 ] ] ],
        [ "embedded-18-T2", "token", [ [ 3, 4 ] ] ],
        [ "embedded-18-T3", "token", [ [ 5, 7 ] ] ],
        [ "embedded-18-T4", "token", [ [ 8, 10 ] ] ],
        [ "embedded-18-T5", "token", [ [ 11, 12 ] ] ],
        [ "embedded-18-T6", "token", [ [ 13, 14 ] ] ],
        [ "embedded-18-T7", "token", [ [ 15, 17 ] ] ],
        [ "embedded-18-T8", "token", [ [ 18, 19 ] ] ],
        [ "embedded-18-T9", "token", [ [ 20, 21 ] ] ],
        [ "embedded-18-T11", "token", [ [ 24, 28 ] ] ],
        [ "embedded-18-T12", "token", [ [ 29, 32 ] ] ],
        [ "embedded-18-T13", "token", [ [ 33, 39 ] ] ],
        [ "embedded-18-T14", "token", [ [ 40, 47 ] ] ],
        [ "embedded-18-T15", "token", [ [ 48, 52 ] ] ],
        [ "embedded-18-T16", "token", [ [ 53, 58 ] ] ],
        [ "embedded-18-T17", "token", [ [ 59, 66 ] ] ],
        [ "embedded-18-T18", "token", [ [ 67, 70 ] ] ],
        [ "embedded-18-T19", "token", [ [ 71, 72 ] ] ]
    ],
    "relations": [
        [ "embedded-18-R1", "csubj", [ [ "arg1", "embedded-18-T7" ], [ "arg2", "embedded-18-T3" ] ] ],
        [ "embedded-18-R2", "nsubjpass", [ [ "arg1", "embedded-18-T3" ], [ "arg2", "embedded-18-T1" ] ] ],
        [ "embedded-18-R3", "case", [ [ "arg1", "embedded-18-T1" ], [ "arg2", "embedded-18-T2" ] ] ],
        [ "embedded-18-R4", "auxpass", [ [ "arg1", "embedded-18-T3" ], [ "arg2", "embedded-18-T4" ] ] ],
        [ "embedded-18-R5", "mark", [ [ "arg1", "embedded-18-T3" ], [ "arg2", "embedded-18-T5" ] ] ],
        [ "embedded-18-R6", "case", [ [ "arg1", "embedded-18-T3" ], [ "arg2", "embedded-18-T6" ] ] ],
        [ "embedded-18-R7", "aux", [ [ "arg1", "embedded-18-T7" ], [ "arg2", "embedded-18-T8" ] ] ],
        [ "embedded-18-R8", "punct", [ [ "arg1", "embedded-18-T7" ], [ "arg2", "embedded-18-T9" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "18" ]
}

edit csubj

csubjpass: clausal passive subject

The dependency type csubjpass is used for clausal subjects of passive verbs.

edit

  

    
input
    
brat
    
info
  
言っ た の が 悔やま れる 。 \n say PAST that NOM regret PASSIVE .
csubjpass(悔やま, 言っ)
auxpass(悔やま, れる)
aux(言っ, た)
mark(言っ, の)
case(言っ, が)
punct(悔やま, 。)
{
    "text": "言っ た の が 悔やま れる 。 \n say PAST that NOM regret PASSIVE . ",
    "entities": [
        [ "embedded-19-T1", "token", [ [ 0, 2 ] ] ],
        [ "embedded-19-T2", "token", [ [ 3, 4 ] ] ],
        [ "embedded-19-T3", "token", [ [ 5, 6 ] ] ],
        [ "embedded-19-T4", "token", [ [ 7, 8 ] ] ],
        [ "embedded-19-T5", "token", [ [ 9, 12 ] ] ],
        [ "embedded-19-T6", "token", [ [ 13, 15 ] ] ],
        [ "embedded-19-T7", "token", [ [ 16, 17 ] ] ],
        [ "embedded-19-T9", "token", [ [ 20, 23 ] ] ],
        [ "embedded-19-T10", "token", [ [ 24, 28 ] ] ],
        [ "embedded-19-T11", "token", [ [ 29, 33 ] ] ],
        [ "embedded-19-T12", "token", [ [ 34, 37 ] ] ],
        [ "embedded-19-T13", "token", [ [ 38, 44 ] ] ],
        [ "embedded-19-T14", "token", [ [ 45, 52 ] ] ],
        [ "embedded-19-T15", "token", [ [ 53, 54 ] ] ]
    ],
    "relations": [
        [ "embedded-19-R1", "csubjpass", [ [ "arg1", "embedded-19-T5" ], [ "arg2", "embedded-19-T1" ] ] ],
        [ "embedded-19-R2", "auxpass", [ [ "arg1", "embedded-19-T5" ], [ "arg2", "embedded-19-T6" ] ] ],
        [ "embedded-19-R3", "aux", [ [ "arg1", "embedded-19-T1" ], [ "arg2", "embedded-19-T2" ] ] ],
        [ "embedded-19-R4", "mark", [ [ "arg1", "embedded-19-T1" ], [ "arg2", "embedded-19-T3" ] ] ],
        [ "embedded-19-R5", "case", [ [ "arg1", "embedded-19-T1" ], [ "arg2", "embedded-19-T4" ] ] ],
        [ "embedded-19-R6", "punct", [ [ "arg1", "embedded-19-T5" ], [ "arg2", "embedded-19-T7" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "19" ]
}

edit csubjpass

dep: unspecified dependency

Undefined.

edit dep

det: determiner

The dependency type det is for a closed set of adnominals.

The full list of the adnominals is as follows:

• この / kono “this”
• こんな / konna “this”
• あの / ano “that”
• あんな / an’na “that”
• その / sono “that”
• そんな / son’na “that”
• どの / dono “which”
• どんな / don’na “which”

edit

  

    
input
    
brat
    
info
  
この 本 \n this book
det(本, この)
{
    "text": "この 本 \n this book ",
    "entities": [
        [ "embedded-20-T1", "token", [ [ 0, 2 ] ] ],
        [ "embedded-20-T2", "token", [ [ 3, 4 ] ] ],
        [ "embedded-20-T4", "token", [ [ 7, 11 ] ] ],
        [ "embedded-20-T5", "token", [ [ 12, 16 ] ] ]
    ],
    "relations": [
        [ "embedded-20-R1", "det", [ [ "arg1", "embedded-20-T2" ], [ "arg2", "embedded-20-T1" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "20" ]
}

edit det

discourse: discourse element

Discourse elements (discourse) in UD Japanese is interjection and emoticons.

edit

  

    
input
    
brat
    
info
  
あー つかれ た 。 \n ah tire PAST .
discourse(つかれ, あー)
aux(つかれ, た)
punct(つかれ, 。)
{
    "text": "あー つかれ た 。 \n ah tire PAST . ",
    "entities": [
        [ "embedded-21-T1", "token", [ [ 0, 2 ] ] ],
        [ "embedded-21-T2", "token", [ [ 3, 6 ] ] ],
        [ "embedded-21-T3", "token", [ [ 7, 8 ] ] ],
        [ "embedded-21-T4", "token", [ [ 9, 10 ] ] ],
        [ "embedded-21-T6", "token", [ [ 13, 15 ] ] ],
        [ "embedded-21-T7", "token", [ [ 16, 20 ] ] ],
        [ "embedded-21-T8", "token", [ [ 21, 25 ] ] ],
        [ "embedded-21-T9", "token", [ [ 26, 27 ] ] ]
    ],
    "relations": [
        [ "embedded-21-R1", "discourse", [ [ "arg1", "embedded-21-T2" ], [ "arg2", "embedded-21-T1" ] ] ],
        [ "embedded-21-R2", "aux", [ [ "arg1", "embedded-21-T2" ], [ "arg2", "embedded-21-T3" ] ] ],
        [ "embedded-21-R3", "punct", [ [ "arg1", "embedded-21-T2" ], [ "arg2", "embedded-21-T4" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "21" ]
}

edit discourse

dislocated: dislocated elements

The dependency type dislocated is used for topic phrases. A topic phrase introduces the topic of a sentence, and are typically prepositional phrases with a topic marker は / ha. One of the most famous examples is 象 は 鼻 が 長い 。/ zou wa hana ga nagai “For elephants, noses are long.”

edit

  

    
input
    
brat
    
info
  
象 は 鼻 が 長い 。 \n elephant TOPIC nose NOM long .
dislocated(長い, 象)
case(象, は)
case(鼻, が)
nsubj(長い, 鼻)
punct(長い, 。)
{
    "text": "象 は 鼻 が 長い 。 \n elephant TOPIC nose NOM long . ",
    "entities": [
        [ "embedded-22-T1", "token", [ [ 0, 1 ] ] ],
        [ "embedded-22-T2", "token", [ [ 2, 3 ] ] ],
        [ "embedded-22-T3", "token", [ [ 4, 5 ] ] ],
        [ "embedded-22-T4", "token", [ [ 6, 7 ] ] ],
        [ "embedded-22-T5", "token", [ [ 8, 10 ] ] ],
        [ "embedded-22-T6", "token", [ [ 11, 12 ] ] ],
        [ "embedded-22-T8", "token", [ [ 15, 23 ] ] ],
        [ "embedded-22-T9", "token", [ [ 24, 29 ] ] ],
        [ "embedded-22-T10", "token", [ [ 30, 34 ] ] ],
        [ "embedded-22-T11", "token", [ [ 35, 38 ] ] ],
        [ "embedded-22-T12", "token", [ [ 39, 43 ] ] ],
        [ "embedded-22-T13", "token", [ [ 44, 45 ] ] ]
    ],
    "relations": [
        [ "embedded-22-R1", "dislocated", [ [ "arg1", "embedded-22-T5" ], [ "arg2", "embedded-22-T1" ] ] ],
        [ "embedded-22-R2", "case", [ [ "arg1", "embedded-22-T1" ], [ "arg2", "embedded-22-T2" ] ] ],
        [ "embedded-22-R3", "case", [ [ "arg1", "embedded-22-T3" ], [ "arg2", "embedded-22-T4" ] ] ],
        [ "embedded-22-R4", "nsubj", [ [ "arg1", "embedded-22-T5" ], [ "arg2", "embedded-22-T3" ] ] ],
        [ "embedded-22-R5", "punct", [ [ "arg1", "embedded-22-T5" ], [ "arg2", "embedded-22-T6" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "22" ]
}

Note that the relation is not used for a topicalized phrase that is also a core argument of the sentence.

edit

  

    
input
    
brat
    
info
  
太郎 は 東京 へ 行く 。\n Taro TOPIC Tokyo TO go .
case(太郎, は)
case(東京, へ)
nsubj(行く, 太郎)
nmod(行く, 東京)
{
    "text": "太郎 は 東京 へ 行く 。\n Taro TOPIC Tokyo TO go . ",
    "entities": [
        [ "embedded-23-T1", "token", [ [ 0, 2 ] ] ],
        [ "embedded-23-T2", "token", [ [ 3, 4 ] ] ],
        [ "embedded-23-T3", "token", [ [ 5, 7 ] ] ],
        [ "embedded-23-T4", "token", [ [ 8, 9 ] ] ],
        [ "embedded-23-T5", "token", [ [ 10, 12 ] ] ],
        [ "embedded-23-T6", "token", [ [ 13, 15 ] ] ],
        [ "embedded-23-T7", "token", [ [ 16, 20 ] ] ],
        [ "embedded-23-T8", "token", [ [ 21, 26 ] ] ],
        [ "embedded-23-T9", "token", [ [ 27, 32 ] ] ],
        [ "embedded-23-T10", "token", [ [ 33, 35 ] ] ],
        [ "embedded-23-T11", "token", [ [ 36, 38 ] ] ],
        [ "embedded-23-T12", "token", [ [ 39, 40 ] ] ]
    ],
    "relations": [
        [ "embedded-23-R1", "case", [ [ "arg1", "embedded-23-T1" ], [ "arg2", "embedded-23-T2" ] ] ],
        [ "embedded-23-R2", "case", [ [ "arg1", "embedded-23-T3" ], [ "arg2", "embedded-23-T4" ] ] ],
        [ "embedded-23-R3", "nsubj", [ [ "arg1", "embedded-23-T5" ], [ "arg2", "embedded-23-T1" ] ] ],
        [ "embedded-23-R4", "nmod", [ [ "arg1", "embedded-23-T5" ], [ "arg2", "embedded-23-T3" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "23" ]
}

edit dislocated

dobj: direct object

Direct object (typically a postpositional phrase with a case marker を / o).

edit

  

    
input
    
brat
    
info
  
ご飯 を 食べる 。 \n Rice ACC eat .
dobj(食べる, ご飯)
case(ご飯, を)
punct(食べる, 。)
{
    "text": "ご飯 を 食べる 。 \n Rice ACC eat . ",
    "entities": [
        [ "embedded-24-T1", "token", [ [ 0, 2 ] ] ],
        [ "embedded-24-T2", "token", [ [ 3, 4 ] ] ],
        [ "embedded-24-T3", "token", [ [ 5, 8 ] ] ],
        [ "embedded-24-T4", "token", [ [ 9, 10 ] ] ],
        [ "embedded-24-T6", "token", [ [ 13, 17 ] ] ],
        [ "embedded-24-T7", "token", [ [ 18, 21 ] ] ],
        [ "embedded-24-T8", "token", [ [ 22, 25 ] ] ],
        [ "embedded-24-T9", "token", [ [ 26, 27 ] ] ]
    ],
    "relations": [
        [ "embedded-24-R1", "dobj", [ [ "arg1", "embedded-24-T3" ], [ "arg2", "embedded-24-T1" ] ] ],
        [ "embedded-24-R2", "case", [ [ "arg1", "embedded-24-T1" ], [ "arg2", "embedded-24-T2" ] ] ],
        [ "embedded-24-R3", "punct", [ [ "arg1", "embedded-24-T3" ], [ "arg2", "embedded-24-T4" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "24" ]
}

edit dobj

expl: expletive

Not used in Japanese.

edit expl

foreign: foreign words

Foreign words in non-Japanese characters.

edit

  

    
input
    
brat
    
info
  
google docs を 使う 。 \n google docs ACC use .
foreign(docs, google)
case(docs, を)
dobj(使う, docs)
punct(使う, 。)
{
    "text": "google docs を 使う 。 \n google docs ACC use . ",
    "entities": [
        [ "embedded-25-T1", "token", [ [ 0, 6 ] ] ],
        [ "embedded-25-T2", "token", [ [ 7, 11 ] ] ],
        [ "embedded-25-T3", "token", [ [ 12, 13 ] ] ],
        [ "embedded-25-T4", "token", [ [ 14, 16 ] ] ],
        [ "embedded-25-T5", "token", [ [ 17, 18 ] ] ],
        [ "embedded-25-T7", "token", [ [ 21, 27 ] ] ],
        [ "embedded-25-T8", "token", [ [ 28, 32 ] ] ],
        [ "embedded-25-T9", "token", [ [ 33, 36 ] ] ],
        [ "embedded-25-T10", "token", [ [ 37, 40 ] ] ],
        [ "embedded-25-T11", "token", [ [ 41, 42 ] ] ]
    ],
    "relations": [
        [ "embedded-25-R1", "foreign", [ [ "arg1", "embedded-25-T2" ], [ "arg2", "embedded-25-T1" ] ] ],
        [ "embedded-25-R2", "case", [ [ "arg1", "embedded-25-T2" ], [ "arg2", "embedded-25-T3" ] ] ],
        [ "embedded-25-R3", "dobj", [ [ "arg1", "embedded-25-T4" ], [ "arg2", "embedded-25-T2" ] ] ],
        [ "embedded-25-R4", "punct", [ [ "arg1", "embedded-25-T4" ], [ "arg2", "embedded-25-T5" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "25" ]
}

edit foreign

goeswith: goes with

Used when a word is somehow split into tokens.

edit goeswith

iobj: indirect object

Indirect object (typicall a postpositional phrase with a case marker に / ni).

edit

  

    
input
    
brat
    
info
  
花子 に あげる 。 \n Hanako DAT give .
iobj(あげる, 花子)
case(花子, に)
punct(あげる, 。)
{
    "text": "花子 に あげる 。 \n Hanako DAT give . ",
    "entities": [
        [ "embedded-26-T1", "token", [ [ 0, 2 ] ] ],
        [ "embedded-26-T2", "token", [ [ 3, 4 ] ] ],
        [ "embedded-26-T3", "token", [ [ 5, 8 ] ] ],
        [ "embedded-26-T4", "token", [ [ 9, 10 ] ] ],
        [ "embedded-26-T6", "token", [ [ 13, 19 ] ] ],
        [ "embedded-26-T7", "token", [ [ 20, 23 ] ] ],
        [ "embedded-26-T8", "token", [ [ 24, 28 ] ] ],
        [ "embedded-26-T9", "token", [ [ 29, 30 ] ] ]
    ],
    "relations": [
        [ "embedded-26-R1", "iobj", [ [ "arg1", "embedded-26-T3" ], [ "arg2", "embedded-26-T1" ] ] ],
        [ "embedded-26-R2", "case", [ [ "arg1", "embedded-26-T1" ], [ "arg2", "embedded-26-T2" ] ] ],
        [ "embedded-26-R3", "punct", [ [ "arg1", "embedded-26-T3" ], [ "arg2", "embedded-26-T4" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "26" ]
}

edit iobj

list: list

Listing.

edit

  

    
input
    
brat
    
info
  
1. 東京 2. 大阪 \n 1. Tokyo 2. Osaka
list(大阪-4, 東京-2)
appos(1.-1, 東京-2) 
appos(2.-3, 大阪-4)
{
    "text": "1. 東京 2. 大阪 \n 1. Tokyo 2. Osaka ",
    "entities": [
        [ "embedded-27-T1", "token", [ [ 0, 2 ] ] ],
        [ "embedded-27-T2", "token", [ [ 3, 5 ] ] ],
        [ "embedded-27-T3", "token", [ [ 6, 8 ] ] ],
        [ "embedded-27-T4", "token", [ [ 9, 11 ] ] ],
        [ "embedded-27-T6", "token", [ [ 14, 16 ] ] ],
        [ "embedded-27-T7", "token", [ [ 17, 22 ] ] ],
        [ "embedded-27-T8", "token", [ [ 23, 25 ] ] ],
        [ "embedded-27-T9", "token", [ [ 26, 31 ] ] ]
    ],
    "relations": [
        [ "embedded-27-R1", "list", [ [ "arg1", "embedded-27-T4" ], [ "arg2", "embedded-27-T2" ] ] ],
        [ "embedded-27-R2", "appos", [ [ "arg1", "embedded-27-T1" ], [ "arg2", "embedded-27-T2" ] ] ],
        [ "embedded-27-R3", "appos", [ [ "arg1", "embedded-27-T3" ], [ "arg2", "embedded-27-T4" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "27" ]
}

edit list

mark: marker

Subordinating conjunction, conjunctive particle, complementizer (と / to, か / ka).

edit

  

    
input
    
brat
    
info
  
東京 へ 行く と 言っ た 。 \n Tokyo to go that say PAST .
nmod(行く, 東京)
case(東京, へ)
mark(行く, と)
ccomp(言っ, 行く)
aux(言っ, た)
punct(言っ, 。)
{
    "text": "東京 へ 行く と 言っ た 。 \n Tokyo to go that say PAST . ",
    "entities": [
        [ "embedded-28-T1", "token", [ [ 0, 2 ] ] ],
        [ "embedded-28-T2", "token", [ [ 3, 4 ] ] ],
        [ "embedded-28-T3", "token", [ [ 5, 7 ] ] ],
        [ "embedded-28-T4", "token", [ [ 8, 9 ] ] ],
        [ "embedded-28-T5", "token", [ [ 10, 12 ] ] ],
        [ "embedded-28-T6", "token", [ [ 13, 14 ] ] ],
        [ "embedded-28-T7", "token", [ [ 15, 16 ] ] ],
        [ "embedded-28-T9", "token", [ [ 19, 24 ] ] ],
        [ "embedded-28-T10", "token", [ [ 25, 27 ] ] ],
        [ "embedded-28-T11", "token", [ [ 28, 30 ] ] ],
        [ "embedded-28-T12", "token", [ [ 31, 35 ] ] ],
        [ "embedded-28-T13", "token", [ [ 36, 39 ] ] ],
        [ "embedded-28-T14", "token", [ [ 40, 44 ] ] ],
        [ "embedded-28-T15", "token", [ [ 45, 46 ] ] ]
    ],
    "relations": [
        [ "embedded-28-R1", "nmod", [ [ "arg1", "embedded-28-T3" ], [ "arg2", "embedded-28-T1" ] ] ],
        [ "embedded-28-R2", "case", [ [ "arg1", "embedded-28-T1" ], [ "arg2", "embedded-28-T2" ] ] ],
        [ "embedded-28-R3", "mark", [ [ "arg1", "embedded-28-T3" ], [ "arg2", "embedded-28-T4" ] ] ],
        [ "embedded-28-R4", "ccomp", [ [ "arg1", "embedded-28-T5" ], [ "arg2", "embedded-28-T3" ] ] ],
        [ "embedded-28-R5", "aux", [ [ "arg1", "embedded-28-T5" ], [ "arg2", "embedded-28-T6" ] ] ],
        [ "embedded-28-R6", "punct", [ [ "arg1", "embedded-28-T5" ], [ "arg2", "embedded-28-T7" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "28" ]
}

edit mark

mwe: multi-word expression

Multi-word functional expressions.

edit

  

    
input
    
brat
    
info
  
太郎 に よっ て 発見 さ れ た 。 \n Taro by _ _ discover do PASSIVE PAST
case(太郎, に)
mwe(に, よっ)
mwe(に, て)
nmod(発見, 太郎)
aux(発見, さ)
auxpass(発見, れ)
aux(発見, た)
punct(発見, 。)
{
    "text": "太郎 に よっ て 発見 さ れ た 。 \n Taro by _ _ discover do PASSIVE PAST ",
    "entities": [
        [ "embedded-29-T1", "token", [ [ 0, 2 ] ] ],
        [ "embedded-29-T2", "token", [ [ 3, 4 ] ] ],
        [ "embedded-29-T3", "token", [ [ 5, 7 ] ] ],
        [ "embedded-29-T4", "token", [ [ 8, 9 ] ] ],
        [ "embedded-29-T5", "token", [ [ 10, 12 ] ] ],
        [ "embedded-29-T6", "token", [ [ 13, 14 ] ] ],
        [ "embedded-29-T7", "token", [ [ 15, 16 ] ] ],
        [ "embedded-29-T8", "token", [ [ 17, 18 ] ] ],
        [ "embedded-29-T9", "token", [ [ 19, 20 ] ] ],
        [ "embedded-29-T11", "token", [ [ 23, 27 ] ] ],
        [ "embedded-29-T12", "token", [ [ 28, 30 ] ] ],
        [ "embedded-29-T13", "token", [ [ 31, 32 ] ] ],
        [ "embedded-29-T14", "token", [ [ 33, 34 ] ] ],
        [ "embedded-29-T15", "token", [ [ 35, 43 ] ] ],
        [ "embedded-29-T16", "token", [ [ 44, 46 ] ] ],
        [ "embedded-29-T17", "token", [ [ 47, 54 ] ] ],
        [ "embedded-29-T18", "token", [ [ 55, 59 ] ] ]
    ],
    "relations": [
        [ "embedded-29-R1", "case", [ [ "arg1", "embedded-29-T1" ], [ "arg2", "embedded-29-T2" ] ] ],
        [ "embedded-29-R2", "mwe", [ [ "arg1", "embedded-29-T2" ], [ "arg2", "embedded-29-T3" ] ] ],
        [ "embedded-29-R3", "mwe", [ [ "arg1", "embedded-29-T2" ], [ "arg2", "embedded-29-T4" ] ] ],
        [ "embedded-29-R4", "nmod", [ [ "arg1", "embedded-29-T5" ], [ "arg2", "embedded-29-T1" ] ] ],
        [ "embedded-29-R5", "aux", [ [ "arg1", "embedded-29-T5" ], [ "arg2", "embedded-29-T6" ] ] ],
        [ "embedded-29-R6", "auxpass", [ [ "arg1", "embedded-29-T5" ], [ "arg2", "embedded-29-T7" ] ] ],
        [ "embedded-29-R7", "aux", [ [ "arg1", "embedded-29-T5" ], [ "arg2", "embedded-29-T8" ] ] ],
        [ "embedded-29-R8", "punct", [ [ "arg1", "embedded-29-T5" ], [ "arg2", "embedded-29-T9" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "29" ]
}

雨 が 降る かも しれ ない 。 \n rain NOM fall maybe _ _ .
nsubj(降る, 雨)
case(雨, が)
aux(降る, かも)
mwe(かも, しれ)
mwe(かも, ない)
punct(降る, 。)

edit mwe

name: name

Proper noun compounds.

山田 太郎 \n Yamada Taro
name(山田, 太郎)

edit name

neg: negation modifier

Negation with ない / nai, ず / zu for clausal predicates. Negation with 不 / fu, 非 / hi for Noun dependents.

食べ ない 。 \n eat not .
neg(ない, 食べ)
punct(食べ, 。)

不 支持 \n not support
neg(支持, 不)

edit neg

nmod: nominal modifier

Prepositional/noun phrases modifying another phrase.

明日 北海道 へ 行く 。 \n tomorrow Hokkaido TO go .
nmod(行く, 北海道)
case(北海道, へ)
nmod(行く, 明日)
punct(行く, 。)

リンゴ の ジュース \n apple OF juice
nmod(ジュース, リンゴ)
case(リンゴ, の)

edit nmod

nsubj: nominal subject

Nominal subjects (typically a postpositional phrase with a case marker “が” or sometimes with は / wa).

太郎 が 走る 。 \n Taro NOM runs .
nsubj(走る, 太郎)
case(太郎, が)
punct(走る, 。)

edit nsubj

nsubjpass: passive nominal subject

Nominal subjects in passive forms.

太郎 が ほめ られる 。 \n Taro NOM praise PASSIVE .
nsubjpass(ほめ, 太郎)
auxpass(ほめ, られる)
case(太郎, が)
punct(ほめ, 。)

edit nsubjpass

nummod: numeric modifier

Numeral classifier phrases.

3 冊 の 本 \n 3 COUNT of books
nummod(本, 冊)
case(冊, の)
nummod(冊, 3)

edit nummod

parataxis: parataxis

Listing of sentences.

「 太郎 は 花子 が 好き 。 次郎 は 梅子 が 好き 。 」 と 三郎 が 言っ た 。 \n " Taro TOPIC Hanako NOM likable . Jiro TOPIC Umeko NOM likable . " that Saburo NOM say PAST .
parataxis(好き-12, 好き-6)
punct(好き-12, 「-1)
punct(好き-12, 」-14)
case(太郎-2, は-3)
case(花子-4, が-5)
dislocated(好き-6, 太郎-2)
nsubj(好き-6, 花子-4)
punct(好き-6, 。-7)
case(次郎-8, は-9)
case(梅子-10, が-11)
dislocated(好き-12, 次郎-8)
nsubj(好き-12, 梅子-10)
punct(好き-12, 。-13)
case(好き-12, と-15)
case(三郎-16, が-17)
nsubj(言っ-18, 三郎)
ccomp(言っ-18, 好き-12)
aux(言っ-18, た-19)
punct(言っ-18, 。-20)

edit parataxis

punct: punctuation

Punctuations.

太郎 が 走る 。 \n Taro NOM run .
nsubj(走る, 太郎)
case(太郎, が)
punct(走る, 。)

太郎 、 次郎 \n Taro , Jiro
punct(太郎, 、)
conj(太郎, 次郎)

edit punct

remnant: remnant in ellipsis

Used to annotate head-less coordination (see the definition in Universal Dependencies for details).

edit remnant

reparandum: overridden disfluency

Speech repair.

edit reparandum

root: root

Head of a sentence.

edit root

vocative: vocative

Dialogue participant names in text.

太郎 食べ なさい 。 \n Taro eat IMPERATIVE .
vocative(食べ, 太郎)
aux(食べ, なさい)
punct(食べ, 。)

edit vocative

xcomp: open clausal complement

Not used in Japanese.

edit xcomp

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

.