nschneid
commented
3 years ago Copulas should be tagged as AUX.
"Will" is usually an auxiliary but in the case of "do as they wilt" I'm not sure if it should be interpreted as a main verb.
Open muchang opened 3 years ago
nschneid
commented
3 years ago Copulas should be tagged as AUX.
"Will" is usually an auxiliary but in the case of "do as they wilt" I'm not sure if it should be interpreted as a main verb.
muchang
commented
3 years ago Copulas should be tagged as AUX.
Thanks, here might be some cases that tag copulas as AUX, which made me confused.
Word: "is"
# sent_id = weblog-typepad.com_ripples_20040407125600_ENG_20040407_125600-0009
# text = One of the most widespread myths of recent times is that the Chernobyl nuclear reactor accident in 1986 caused many thousands of extra cancer deaths in neighbouring regions, and that public health has been severely affected by exposure to radiation.
1 One one NUM CD NumType=Card 10 nsubj 10:nsubj _
2 of of ADP IN _ 6 case 6:case _
3 the the DET DT Definite=Def|PronType=Art 6 det 6:det _
4 most most ADV RBS _ 5 advmod 5:advmod _
5 widespread widespread ADJ JJ Degree=Pos 6 amod 6:amod _
6 myths myth NOUN NNS Number=Plur 1 nmod 1:nmod:of _
7 of of ADP IN _ 9 case 9:case _
8 recent recent ADJ JJ Degree=Pos 9 amod 9:amod _
9 times time NOUN NNS Number=Plur 6 nmod 6:nmod:of _
10 is be VERB VBZ Mood=Ind|Number=Sing|Person=3|Tense=Pres|VerbForm=Fin 0 root 0:root _
11 that that SCONJ IN _ 19 mark 19:mark _
12 the the DET DT Definite=Def|PronType=Art 16 det 16:det _
13 Chernobyl Chernobyl PROPN NNP Number=Sing 16 compound 16:compound _
14 nuclear nuclear ADJ JJ Degree=Pos 15 amod 15:amod _
15 reactor reactor NOUN NN Number=Sing 16 compound 16:compound _
16 accident accident NOUN NN Number=Sing 19 nsubj 19:nsubj _
17 in in ADP IN _ 18 case 18:case _
18 1986 1986 NUM CD NumType=Card 16 nmod 16:nmod:in _
19 caused cause VERB VBD Mood=Ind|Tense=Past|VerbForm=Fin 10 ccomp 10:ccomp _
20 many many ADJ JJ Degree=Pos 21 amod 21:amod _
21 thousands thousand NOUN NNS Number=Plur 19 obj 19:obj _
22 of of ADP IN _ 25 case 25:case _
23 extra extra ADJ JJ Degree=Pos 25 amod 25:amod _
24 cancer cancer NOUN NN Number=Sing 25 compound 25:compound _
25 deaths death NOUN NNS Number=Plur 21 nmod 21:nmod:of _
26 in in ADP IN _ 28 case 28:case _
27 neighbouring neighbour VERB VBG VerbForm=Ger 28 amod 28:amod _
28 regions region NOUN NNS Number=Plur 19 obl 19:obl:in SpaceAfter=No
29 , , PUNCT , _ 37 punct 37:punct _
30 and and CCONJ CC _ 37 cc 37:cc _
31 that that SCONJ IN _ 37 mark 37:mark _
32 public public ADJ JJ Degree=Pos 33 amod 33:amod _
33 health health NOUN NN Number=Sing 37 nsubj:pass 37:nsubj:pass _
34 has have AUX VBZ Mood=Ind|Number=Sing|Person=3|Tense=Pres|VerbForm=Fin 37 aux 37:aux _
35 been be AUX VBN Tense=Past|VerbForm=Part 37 aux:pass 37:aux:pass _
36 severely severely ADV RB _ 37 advmod 37:advmod _
37 affected affect VERB VBN Tense=Past|VerbForm=Part|Voice=Pass 19 conj 10:ccomp|19:conj:and _
38 by by ADP IN _ 39 case 39:case _
39 exposure exposure NOUN NN Number=Sing 37 obl 37:obl:by _
40 to to ADP IN _ 41 case 41:case _
41 radiation radiation NOUN NN Number=Sing 39 nmod 39:nmod:to SpaceAfter=No
42 . . PUNCT . _ 10 punct 10:punct _# sent_id = weblog-juancole.com_juancole_20030911085700_ENG_20030911_085700-0035
# text = The thing to keep in mind is that Sunni Arab nationalists and Baathists and local Sunni radicals are likely to remain far more dangerous to the US in Iraq than al-Qaeda infiltrators, and it would be dangerous to take one's eyes off the former ball.
1 The the DET DT Definite=Def|PronType=Art 2 det 2:det _
2 thing thing NOUN NN Number=Sing 7 nsubj 7:nsubj _
3 to to PART TO _ 4 mark 4:mark _
4 keep keep VERB VB VerbForm=Inf 2 acl 2:acl:to _
5 in in ADP IN _ 6 case 6:case _
6 mind mind NOUN NN Number=Sing 4 obl 4:obl:in _
7 is be VERB VBZ Mood=Ind|Number=Sing|Person=3|Tense=Pres|VerbForm=Fin 0 root 0:root _
8 that that SCONJ IN _ 19 mark 19:mark _
9 Sunni Sunni ADJ JJ Degree=Pos 11 amod 11:amod _
10 Arab Arab ADJ JJ Degree=Pos 11 amod 11:amod _
11 nationalists nationalist NOUN NNS Number=Plur 19 nsubj 19:nsubj|21:nsubj:xsubj|24:nsubj:xsubj _
12 and and CCONJ CC _ 13 cc 13:cc _
13 Baathists Baathist PROPN NNPS Number=Plur 11 conj 11:conj:and|19:nsubj|21:nsubj:xsubj|24:nsubj:xsubj _
14 and and CCONJ CC _ 17 cc 17:cc _
15 local local ADJ JJ Degree=Pos 17 amod 17:amod _
16 Sunni Sunni ADJ JJ Degree=Pos 17 amod 17:amod _
17 radicals radical NOUN NNS Number=Plur 11 conj 11:conj:and|19:nsubj|21:nsubj:xsubj|24:nsubj:xsubj _
18 are be AUX VBP Mood=Ind|Tense=Pres|VerbForm=Fin 19 cop 19:cop _
19 likely likely ADJ JJ Degree=Pos 7 ccomp 7:ccomp _
20 to to PART TO _ 21 mark 21:mark _
21 remain remain VERB VB VerbForm=Inf 19 xcomp 19:xcomp _
22 far far ADV RB Degree=Pos 21 advmod 21:advmod _
23 more more ADV RBR _ 21 advmod 21:advmod _
24 dangerous dangerous ADJ JJ Degree=Pos 21 xcomp 21:xcomp _
25 to to ADP IN _ 27 case 27:case _
26 the the DET DT Definite=Def|PronType=Art 27 det 27:det _
27 US US PROPN NNP Number=Sing 24 obl 24:obl:to _
28 in in ADP IN _ 29 case 29:case _
29 Iraq Iraq PROPN NNP Number=Sing 27 nmod 27:nmod:in _
30 than than ADP IN _ 34 case 34:case _
31 al al PROPN NNP Number=Sing 33 compound 33:compound SpaceAfter=No
32 - - PUNCT HYPH _ 33 punct 33:punct SpaceAfter=No
33 Qaeda Qaeda PROPN NNP Number=Sing 34 compound 34:compound _
34 infiltrators infiltrator NOUN NNS Number=Plur 24 obl 24:obl:than SpaceAfter=No
35 , , PUNCT , _ 40 punct 40:punct _
36 and and CCONJ CC _ 40 cc 40:cc _
37 it it PRON PRP Case=Nom|Gender=Neut|Number=Sing|Person=3|PronType=Prs 40 expl 40:expl _
38 would would AUX MD VerbForm=Fin 40 aux 40:aux _
39 be be AUX VB VerbForm=Inf 40 cop 40:cop _
40 dangerous dangerous ADJ JJ Degree=Pos 19 conj 7:ccomp|19:conj:and _
41 to to PART TO _ 42 mark 42:mark _
42 take take VERB VB VerbForm=Inf 40 csubj 40:csubj _
43-44 one's _ _ _ _ _ _ _ _
43 one one PRON PRP _ 45 nmod:poss 45:nmod:poss _
44 's 's PART POS _ 43 case 43:case _
45 eyes eye NOUN NNS Number=Plur 42 obj 42:obj _
46 off off ADP IN _ 49 case 49:case _
47 the the DET DT Definite=Def|PronType=Art 49 det 49:det _
48 former former ADJ JJ Degree=Pos 49 amod 49:amod _
49 ball ball NOUN NN Number=Sing 42 obl 42:obl:off SpaceAfter=No
50 . . PUNCT . _ 7 punct 7:punct _# sent_id = weblog-juancole.com_juancole_20040708181175_ENG_20040708_181175-0002
# text = The problem with this argument is that Bush lacked the experience necessary to be president when he ran in 2000, so this sort of cheap shot just hoists him by his own petard.
1 The the DET DT Definite=Def|PronType=Art 2 det 2:det _
2 problem problem NOUN NN Number=Sing 6 nsubj 6:nsubj _
3 with with ADP IN _ 5 case 5:case _
4 this this DET DT Number=Sing|PronType=Dem 5 det 5:det _
5 argument argument NOUN NN Number=Sing 2 nmod 2:nmod:with _
6 is be VERB VBZ Mood=Ind|Number=Sing|Person=3|Tense=Pres|VerbForm=Fin 0 root 0:root _
7 that that SCONJ IN _ 9 mark 9:mark _
8 Bush Bush PROPN NNP Number=Sing 9 nsubj 9:nsubj _
9 lacked lack VERB VBD Mood=Ind|Tense=Past|VerbForm=Fin 6 ccomp 6:ccomp _
10 the the DET DT Definite=Def|PronType=Art 11 det 11:det _
11 experience experience NOUN NN Number=Sing 9 obj 9:obj _
12 necessary necessary ADJ JJ Degree=Pos 11 amod 11:amod _
13 to to PART TO _ 15 mark 15:mark _
14 be be AUX VB VerbForm=Inf 15 cop 15:cop _
15 president president NOUN NN Number=Sing 12 advcl 12:advcl:to _
16 when when SCONJ WRB PronType=Int 18 mark 18:mark _
17 he he PRON PRP Case=Nom|Gender=Masc|Number=Sing|Person=3|PronType=Prs 18 nsubj 18:nsubj _
18 ran run VERB VBD Mood=Ind|Tense=Past|VerbForm=Fin 9 advcl 9:advcl:when _
19 in in ADP IN _ 20 case 20:case _
20 2000 2000 NUM CD NumType=Card 18 obl 18:obl:in SpaceAfter=No
21 , , PUNCT , _ 9 punct 9:punct _
22 so so ADV RB _ 29 advmod 29:advmod _
23 this this DET DT Number=Sing|PronType=Dem 24 det 24:det _
24 sort sort NOUN NN Number=Sing 29 nsubj 29:nsubj _
25 of of ADP IN _ 27 case 27:case _
26 cheap cheap ADJ JJ Degree=Pos 27 amod 27:amod _
27 shot shot NOUN NN Number=Sing 24 nmod 24:nmod:of _
28 just just ADV RB _ 29 advmod 29:advmod _
29 hoists hoist VERB VBZ Mood=Ind|Number=Sing|Person=3|Tense=Pres|VerbForm=Fin 9 parataxis 9:parataxis _
30 him he PRON PRP Case=Acc|Gender=Masc|Number=Sing|Person=3|PronType=Prs 29 obj 29:obj _
31 by by ADP IN _ 34 case 34:case _
32 his he PRON PRP$ Gender=Masc|Number=Sing|Person=3|Poss=Yes|PronType=Prs 34 nmod:poss 34:nmod:poss _
33 own own ADJ JJ Degree=Pos 34 amod 34:amod _
34 petard petard NOUN NN Number=Sing 29 obl 29:obl:by SpaceAfter=No
35 . . PUNCT . _ 6 punct 6:punct _"Will" is usually an auxiliary but in the case of "do as they wilt" I'm not sure if it should be interpreted as a main verb.
In this case, a "do" after "will" might be omitted?
nschneid
commented
3 years ago Interesting. The VERB cases are "X is that Y" sentences, which are an exception to the usual rule that copulas are not heads ("Exception: If the predicative element in the equation is a clause, then the copula verb is treated as the head of the clause, with the following clause as a ccomp (to prevent that the head of the smaller clause gets two subjects).").
I assume the copula should be AUX and is only VERB due to a conversion error.
"Be" verbs are not considered copulas in existential sentences IIRC.
nschneid
commented
3 years ago "Will" is usually an auxiliary but in the case of "do as they wilt" I'm not sure if it should be interpreted as a main verb.
In this case, a "do" after "will" might be omitted?
Right, one interpretation is "do as they will do", and another is "do as they will", with "will" meaning something like 'want' or 'choose'.
muchang
commented
3 years ago Here are some more cases that tag AUX as VERB:
Word: "did"
# sent_id = weblog-juancole.com_juancole_20051126063000_ENG_20051126_063000-0037
# text = If he or she did not, then they should have all the same rights as other Iraqis.
1 If if SCONJ IN _ 5 mark 5:mark _
2 he he PRON PRP Case=Nom|Gender=Masc|Number=Sing|Person=3|PronType=Prs 5 nsubj 5:nsubj _
3 or or CCONJ CC _ 4 cc 4:cc _
4 she she PRON PRP Case=Nom|Gender=Fem|Number=Sing|Person=3|PronType=Prs 2 conj 2:conj:or|5:nsubj _
5 did do VERB VBD Mood=Ind|Tense=Past|VerbForm=Fin 11 advcl 11:advcl:if _
6 not not PART RB _ 5 advmod 5:advmod SpaceAfter=No
7 , , PUNCT , _ 11 punct 11:punct _
8 then then ADV RB PronType=Dem 11 advmod 11:advmod _
9 they they PRON PRP Case=Nom|Number=Plur|Person=3|PronType=Prs 11 nsubj 11:nsubj _
10 should should AUX MD VerbForm=Fin 11 aux 11:aux _
11 have have VERB VB VerbForm=Inf 0 root 0:root _
12 all all DET PDT _ 15 det:predet 15:det:predet _
13 the the DET DT Definite=Def|PronType=Art 15 det 15:det _
14 same same ADJ JJ Degree=Pos 15 amod 15:amod _
15 rights rights NOUN NNS Number=Plur 11 obj 11:obj _
16 as as ADP IN _ 18 case 18:case _
17 other other ADJ JJ Degree=Pos 18 amod 18:amod _
18 Iraqis Iraqi PROPN NNPS Number=Plur 15 nmod 15:nmod:as SpaceAfter=No
19 . . PUNCT . _ 11 punct 11:punct _Word: "is"
# sent_id = newsgroup-groups.google.com_GuildWars_086f0f64ab633ab3_ENG_20041111_173500-0004
# text = The main reason is Google is more accessible to the global community and you can rest assured that it's not going to go away.
1 The the DET DT Definite=Def|PronType=Art 3 det 3:det _
2 main main ADJ JJ Degree=Pos 3 amod 3:amod _
3 reason reason NOUN NN Number=Sing 4 nsubj 4:nsubj _
4 is be VERB VBZ Mood=Ind|Number=Sing|Person=3|Tense=Pres|VerbForm=Fin 0 root 0:root _
5 Google Google PROPN NNP Number=Sing 8 nsubj 8:nsubj _
6 is be AUX VBZ Mood=Ind|Number=Sing|Person=3|Tense=Pres|VerbForm=Fin 8 cop 8:cop _
7 more more ADV RBR _ 8 advmod 8:advmod _
8 accessible accessible ADJ JJ Degree=Pos 4 ccomp 4:ccomp _
9 to to ADP IN _ 12 case 12:case _
10 the the DET DT Definite=Def|PronType=Art 12 det 12:det _
11 global global ADJ JJ Degree=Pos 12 amod 12:amod _
12 community community NOUN NN Number=Sing 8 obl 8:obl:to _
13 and and CCONJ CC _ 16 cc 16:cc _
14 you you PRON PRP Case=Nom|Person=2|PronType=Prs 16 nsubj 16:nsubj _
15 can can AUX MD VerbForm=Fin 16 aux 16:aux _
16 rest rest VERB VB VerbForm=Inf 8 conj 4:ccomp|8:conj:and _
17 assured assure VERB VBN Tense=Past|VerbForm=Part 16 advcl 16:advcl _
18 that that SCONJ IN _ 22 mark 22:mark _
19-20 it's _ _ _ _ _ _ _ _
19 it it PRON PRP Case=Nom|Gender=Neut|Number=Sing|Person=3|PronType=Prs 22 nsubj 22:nsubj|24:nsubj:xsubj _
20 's be AUX VBZ Mood=Ind|Number=Sing|Person=3|Tense=Pres|VerbForm=Fin 22 aux 22:aux _
21 not not PART RB _ 22 advmod 22:advmod _
22 going go VERB VBG Tense=Pres|VerbForm=Part 17 ccomp 17:ccomp _
23 to to PART TO _ 24 mark 24:mark _
24 go go VERB VB VerbForm=Inf 22 xcomp 22:xcomp _
25 away away ADV RB _ 24 advmod 24:advmod SpaceAfter=No
26 . . PUNCT . _ 4 punct 4:punct _Word: "is"
# sent_id = reviews-053248-0003
# text = The truth is, in my and my dining partners' experience, this is a fine little restaurant with some unique food.
1 The the DET DT Definite=Def|PronType=Art 2 det 2:det _
2 truth truth NOUN NN Number=Sing 3 nsubj 3:nsubj _
3 is be VERB VBZ Mood=Ind|Number=Sing|Person=3|Tense=Pres|VerbForm=Fin 0 root 0:root SpaceAfter=No
4 , , PUNCT , _ 3 punct 3:punct _
5 in in ADP IN _ 12 case 12:case _
6 my my PRON PRP$ Number=Sing|Person=1|Poss=Yes|PronType=Prs 12 nmod:poss 12:nmod:poss _
7 and and CCONJ CC _ 10 cc 10:cc _
8 my my PRON PRP$ Number=Sing|Person=1|Poss=Yes|PronType=Prs 10 nmod:poss 10:nmod:poss _
9 dining dining NOUN NN Number=Sing 10 compound 10:compound _
10-11 partners' _ _ _ _ _ _ _ _
10 partners partner NOUN NNS Number=Plur 6 conj 6:conj:and|12:nmod:poss _
11 ' 's PART POS _ 10 case 10:case _
12 experience experience NOUN NN Number=Sing 3 obl 3:obl:in SpaceAfter=No
13 , , PUNCT , _ 3 punct 3:punct _
14 this this PRON DT Number=Sing|PronType=Dem 19 nsubj 19:nsubj _
15 is be AUX VBZ Mood=Ind|Number=Sing|Person=3|Tense=Pres|VerbForm=Fin 19 cop 19:cop _
16 a a DET DT Definite=Ind|PronType=Art 19 det 19:det _
17 fine fine ADJ JJ Degree=Pos 19 amod 19:amod _
18 little little ADJ JJ Degree=Pos 19 amod 19:amod _
19 restaurant restaurant NOUN NN Number=Sing 3 ccomp 3:ccomp _
20 with with ADP IN _ 23 case 23:case _
21 some some DET DT _ 23 det 23:det _
22 unique unique ADJ JJ Degree=Pos 23 amod 23:amod _
23 food food NOUN NN Number=Sing 19 nmod 19:nmod:with SpaceAfter=No
24 . . PUNCT . _ 3 punct 3:punct _Commit: 2f5eba2
nschneid
commented
3 years ago The last 2 are the same construction as "X is that Y" but "that" is omitted.
muchang
commented
3 years ago Yes, it seems so.
amir-zeldes
commented
3 years ago The VERB cases are "X is that Y" sentences, which are an exception to the usual rule that copulas are not heads
GUM does not implement this exception, which I think is a mistake, for several reasons. In "the problem is that Kim is tired", If we label the matrix clause copula as root and make "tired" its ccomp dependent, then we are saying that:
case dependents, one actually modifying the noun, and one modifying the PP. For PPs it doesn't seem to bother us that there are two case dependents, but for nsubj for some reason it triggers a different treatment. In my opinion the two are analogous.I can add to these considerations for English that in languages with zero copula constructions, the analysis of the copula as root and head of ccomp is not even possible, leading to further inconsistency across UD languages. Hebrew example:
In this case, there is no possibility of applying the exceptional analysis, and we get two nsubj relations on "tired". I don't consider this to be a problem though, I consider it to be the expected analysis from a UD perspective (lexico-centric, does not assume that we need verbs for predication "A is B" analyzed the same on both levels).
nschneid
commented
3 years ago I understand, but the tag should be AUX regardless of the dependency structure, right?
amir-zeldes
commented
3 years ago I don't know how to answer that exactly, since I don't agree with the exceptional dependency structure... I suppose if we were to say it's root then it would be right for it to be VERB, since and AUX has to be a kind of aux/cop/etc. to something else (i.e. AUX is a relationally defined category). But essentially, I don't accept the premise here, so I can't give an unbiased answer regarding the POS...
nschneid
commented
1 year ago Sentences like "X is that Y" now have multiple subjects with the first one attaching as nsubj:outer or csubj:outer (https://universaldependencies.org/changes.html#multiple-subjects & #310).
There are still >100 instances of "be" as VERB rather than AUX without existential "there". I think most of these are copulas (in various constructions: ellipsis etc.) and should be revised. http://universal.grew.fr/?custom=638e0f72dd8f8
Hi, the followings are the suspicious cases that tag VERB as AUX.
Word: "wilt" (Duplicate Items)
Word: "are"
Word: "is"
Word: "was"
Commit: 2f5eba2