File size: 92,554 Bytes
6fa4bc9
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
{
    "paper_id": "2021",
    "header": {
        "generated_with": "S2ORC 1.0.0",
        "date_generated": "2023-01-19T16:43:42.080774Z"
    },
    "title": "On auxiliary verb in Universal Dependencies: untangling the issue and proposing a systematized annotation strategy",
    "authors": [
        {
            "first": "Magali",
            "middle": [],
            "last": "Sanches Duran",
            "suffix": "",
            "affiliation": {},
            "email": ""
        },
        {
            "first": "Adriana",
            "middle": [
                "Silvina"
            ],
            "last": "Pagano",
            "suffix": "",
            "affiliation": {},
            "email": "apagano@ufmg.br"
        },
        {
            "first": "Amanda",
            "middle": [],
            "last": "Pontes",
            "suffix": "",
            "affiliation": {},
            "email": ""
        },
        {
            "first": "Thiago",
            "middle": [],
            "last": "Alexandre",
            "suffix": "",
            "affiliation": {},
            "email": ""
        },
        {
            "first": "Salgueiro",
            "middle": [],
            "last": "Pardo",
            "suffix": "",
            "affiliation": {},
            "email": "taspardo@icmc.usp.br"
        }
    ],
    "year": "",
    "venue": null,
    "identifiers": {},
    "abstract": "Auxiliary verbs are universally recognized as components of verbal constructions. While there is no shortage of scholarship on these verbs in various linguistic traditions, uncertainty still remains on the best way to annotate them for Natural Language Processing (NLP) purposes. This paper reviews the evolution of the concept of auxiliary verbs to gather insights into forms of representing them in an annotation scheme and raises some issues with a view to leveraging the potential afforded by them in different NLP tasks. Using Brazilian Portuguese as an instance language and Universal Dependencies (UD) as annotation model, we argue for (i) annotating inflected verbs as heads, (ii) annotating auxiliary interdependence in an auxiliation chain; and (iii) adopting a more consistent treatment of auxiliaries to encompass tense, aspect, modality and voice in auxiliation chains. We further propose auxiliary type as a feature to be annotated which can be easily implemented in existing and new treebanks with substantial gains in enriching the information that can be extracted for different NLP applications.",
    "pdf_parse": {
        "paper_id": "2021",
        "_pdf_hash": "",
        "abstract": [
            {
                "text": "Auxiliary verbs are universally recognized as components of verbal constructions. While there is no shortage of scholarship on these verbs in various linguistic traditions, uncertainty still remains on the best way to annotate them for Natural Language Processing (NLP) purposes. This paper reviews the evolution of the concept of auxiliary verbs to gather insights into forms of representing them in an annotation scheme and raises some issues with a view to leveraging the potential afforded by them in different NLP tasks. Using Brazilian Portuguese as an instance language and Universal Dependencies (UD) as annotation model, we argue for (i) annotating inflected verbs as heads, (ii) annotating auxiliary interdependence in an auxiliation chain; and (iii) adopting a more consistent treatment of auxiliaries to encompass tense, aspect, modality and voice in auxiliation chains. We further propose auxiliary type as a feature to be annotated which can be easily implemented in existing and new treebanks with substantial gains in enriching the information that can be extracted for different NLP applications.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Abstract",
                "sec_num": null
            }
        ],
        "body_text": [
            {
                "text": "Thousands of years ago, writing ushered in a new era for mankind. The advent of writing made it possible for thoughts and information to be conveyed between individuals across distinct epochs and localities. The ideas recorded in writing began to fertilize other minds and generate new ideas, exponentially accelerating the evolution of ideas in human societies (Ridley, 2010) .",
                "cite_spans": [
                    {
                        "start": 362,
                        "end": 376,
                        "text": "(Ridley, 2010)",
                        "ref_id": "BIBREF17"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Introduction",
                "sec_num": "1"
            },
            {
                "text": "As much as writing allowed knowledge transfer among individuals, it today supports the transfer of human knowledge to machines. This happens through Natural Language Processing (NLP), and one of the ways to train machines to process text and learn to extract from text much of what humans do with it is through annotated corpora. Corpus annotation has thus become a way to formally record the implicit and explicit linguistic knowledge that can be gathered from texts. The result enables NLP to develop statistical models for training new human language technologies (Ide, 2017) .",
                "cite_spans": [
                    {
                        "start": 567,
                        "end": 578,
                        "text": "(Ide, 2017)",
                        "ref_id": "BIBREF5"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Introduction",
                "sec_num": "1"
            },
            {
                "text": "Annotating a corpus is an undertaking that requires considerable effort in designing, executing, and reviewing an entire process. Since design involves the creation of models to represent linguistic information, the models are often reused in corpus annotation endeavors, both within a language and in languages other than the one for which the model was created.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Introduction",
                "sec_num": "1"
            },
            {
                "text": "One drawback of leveraging models is the fact that, because they represent facts of a language, they are to some degree dependent on the language the theory drew on for its study. The advantage of models' reuse, on the other hand, is that the use of the same annotation scheme becomes a means to compare languages. The comparison, in turn, makes it possible to create multilingual NLP applications. This is the aim of the Universal Dependencies (UD) model (Nivre 2015 , which is designed to be language independent. At the time of writing, there are over 200 corpora annotated with the UD model in just over 120 languages.",
                "cite_spans": [
                    {
                        "start": 456,
                        "end": 467,
                        "text": "(Nivre 2015",
                        "ref_id": "BIBREF12"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Introduction",
                "sec_num": "1"
            },
            {
                "text": "The fact that many linguists and computer scientists use UD and strive to instantiate it in their languages has promoted numerous discussions around its guidelines. One such discussion revolves around auxiliary verbs. Some languages, for instance, have opted for tagging tense and passive voice auxiliaries as AUX; others include modal verbs under this tag, and still others add aspectual verbs to the set. While UD does not require languages to follow a single standard, it recommends that only strongly grammaticalized auxiliary verbs be annotated as such. This, in turn, raises further discussion as to where to draw the line for an auxiliary to be considered fully grammaticalized.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Introduction",
                "sec_num": "1"
            },
            {
                "text": "This paper grew out of a concern on how to best represent auxiliary verbs in UD scheme towards building a proposal to leverage the full potential afforded by them in different NLP tasks. Drawing on Brazilian Portuguese, we contend that there are substantial gains to be obtained from (i) annotating auxiliaries as heads; (ii) annotating auxiliary interdependence in an auxiliation chain; and (iii) adopting a more consistent treatment of auxiliaries to encompass tense, aspect, modality and voice in auxiliation chains. We further propose auxiliary type as a feature to be annotated for the purpose of enriching information to be tapped from treebanks. Our proposal offers several benefits to NLP tasks, such as enhancing detection of subjects for information extraction by annotating inflected verbs as heads; temporal reference detection by relying on tense and aspect auxiliaries; and speculation detection by leveraging modal auxiliaries as cues for that task.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Introduction",
                "sec_num": "1"
            },
            {
                "text": "In Section 2, we briefly review the evolution of the concept of auxiliary verbs, highlighting the points that are important to our discussion. Section 3 discusses the auxiliation process and provides examples in Brazilian Portuguese for four types of auxiliaries: tense, aspect, modality and passive voice diathesis. In Section 4, we exemplify ways to annotate auxiliary verbs in UD, discussing their pros and cons and presenting a proposal to reduce different forms of annotation to an interpretation of auxiliaries common to all of them. Section 5 concludes our study.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Introduction",
                "sec_num": "1"
            },
            {
                "text": "The topic of auxiliary verbs has been extensively discussed in the last decades, a thorough review being outside the scope of this paper. We will focus hence on the works that most contributed to advancing discussions of the auxiliary verb concept, since this concept is fundamental to the decisions about the UD annotation scheme we would like to argue for.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Contributions over time towards the concept of auxiliary verbs",
                "sec_num": "2"
            },
            {
                "text": "Our review begins with Tesni\u00e8re (1959) , an author who compared auxiliaries to free morphemes, though ascribed to them a distinctive nature of being inflectional. Auxiliary verbs, for Tesni\u00e8re, help other verbs enable a subcategory transfer (in his account, of tense and voice) and are totally devoid of semantic content. Auxiliaries are classified as compound verb forms, as opposed to simple forms, and operate with auxiliated ones. Tesni\u00e8re described auxiliary verbs as those assuming grammatical functions whereas auxiliated verbs contribute with the semantics. Being acquainted with English grammar (he mentions the verb do as an auxiliary), he admitted other functions of auxiliaries, which he suggested when he used etc. in: \"One distinguishes between auxiliaries of tense (past, future), auxiliaries of voice (passive), etc.\" (Tesni\u00e8re, 1959, p. 403) . Benveniste (1974) elaborated on Tesniere's description, recognizing verbal chains of auxiliaries of tense, modality and diathesis (voice). The author coined the term auxiliation to refer to a process that syntagmatically joins an auxiliating form to an auxiliated one, avoiding the use of the term auxiliary. For simplicity, we adopt the term auxiliary verb, even when referring to Benveniste's work.",
                "cite_spans": [
                    {
                        "start": 23,
                        "end": 38,
                        "text": "Tesni\u00e8re (1959)",
                        "ref_id": "BIBREF21"
                    },
                    {
                        "start": 834,
                        "end": 858,
                        "text": "(Tesni\u00e8re, 1959, p. 403)",
                        "ref_id": null
                    },
                    {
                        "start": 861,
                        "end": 878,
                        "text": "Benveniste (1974)",
                        "ref_id": null
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Contributions over time towards the concept of auxiliary verbs",
                "sec_num": "2"
            },
            {
                "text": "By including modal verbs as a further type of auxiliation, Benveniste evidenced that auxiliation chains are longer and more complex than previously believed; yet he did not include aspectual verbs among auxiliaries. Another important contribution by Benveniste was to show that there is sequential order for auxiliary verbs to occur, namely, modal -temporal -passive voice -full verb, within a process he called over-auxiliation. Despite acknowledging that an auxiliary verb is the verb that takes person, number, mood, and tense inflections in a compound form and showing compound forms made up by up to three types of auxiliary functions (tense, modality, and passive voice diathesis), Benveniste did not expand on the fact that, in the case of longer chains, auxiliaries after the first one do not take inflections. Neither did he explicitly state that the second auxiliary in a chain is auxiliated by the first one and so on. He did state, however, that an auxiliary of passive voice diathesis is always the last one in a chain before a full verb, because no auxiliary verb can undergo passive voice diathesis.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Contributions over time towards the concept of auxiliary verbs",
                "sec_num": "2"
            },
            {
                "text": "Another relevant contribution to the concept of auxiliary verbs was made by studies of grammaticalization, mainly after the 1990s. Heine (1993) does a survey of the different ways languages express features of tense, modality and verbal aspect, pointing out that the lack of agreement around a concept of auxiliaries is largely due to the diversity of phenomena. For Heine, one of the sources for linguists' disagreement can be traced to Chomsky's AUX, a universal category he introduced in 1956, which is in fact not directly related to auxiliary verbs. Heine's review shows that auxiliary verbs have been at times considered as main verbs, as non-autonomous verbs, and still as a different grammatical category of verb altogether. Likewise, in dependency grammars, auxiliary verbs are usually considered as dependent by some authors while others posit them as heads. Steele (1994, p.818 ) praised Heine's work for his survey of views on auxiliary verbs, but criticized him for not tackling issues such as which verb is head and which one is dependent in dependency relations. Kuteva (2001) , who set out to complete the work of Heine, remarked that the big problem is the fact that some linguistic traditions disregard the dynamic character of the process of auxiliation, which prevents new auxiliaries arising in languages from being recognized. For her, auxiliation is an ongoing process and auxiliary verbs can be found at various stages in this process. There is thus no limited set of auxiliary verbs and one cannot separate auxiliary verbs from the verbs that gave rise to them.",
                "cite_spans": [
                    {
                        "start": 131,
                        "end": 143,
                        "text": "Heine (1993)",
                        "ref_id": "BIBREF4"
                    },
                    {
                        "start": 869,
                        "end": 888,
                        "text": "Steele (1994, p.818",
                        "ref_id": null
                    },
                    {
                        "start": 1078,
                        "end": 1091,
                        "text": "Kuteva (2001)",
                        "ref_id": "BIBREF8"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Contributions over time towards the concept of auxiliary verbs",
                "sec_num": "2"
            },
            {
                "text": "Andersen (2006) agrees that auxiliation is a dynamic process, \"so the class is continually losing and acquiring new members\" (p.4). The author compares auxiliary constructions in over 800 languages and concludes: \"There is no, and probably cannot be, any specific, language independent formal criteria that can be used to determine the characterization of any given element as a lexical verb or an auxiliary verb.\" (p.5). He makes an important distinction between inflectional and semantic heads. The former encodes features responsible for making the construction to be grammatical, whereas the latter determines valence (argument structure). In some languages, the inflectional and the semantic heads are conflated, as the full verb is the one bearing inflections. In others, the inflectional head is the auxiliary and the semantic head is the full verb. Therefore, depending on the annotation purpose, dependence relations may prioritize the inflectional or the semantic head .",
                "cite_spans": [
                    {
                        "start": 9,
                        "end": 15,
                        "text": "(2006)",
                        "ref_id": null
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Contributions over time towards the concept of auxiliary verbs",
                "sec_num": "2"
            },
            {
                "text": "1 Krug (2012) discusses the grammaticalization of auxiliary verbs and illustrates the process of full verbs becoming auxiliaries for the English language. According to the author, and for this he draws on Bolinger (1980 ( , apud Krug, 2012 , it suffices for a verb to receive a complement in infinitive form to enter a path of grammaticalization. Krug cites the following characteristics of auxiliaries:",
                "cite_spans": [
                    {
                        "start": 2,
                        "end": 13,
                        "text": "Krug (2012)",
                        "ref_id": "BIBREF7"
                    },
                    {
                        "start": 214,
                        "end": 219,
                        "text": "(1980",
                        "ref_id": null
                    },
                    {
                        "start": 220,
                        "end": 239,
                        "text": "( , apud Krug, 2012",
                        "ref_id": null
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Contributions over time towards the concept of auxiliary verbs",
                "sec_num": "2"
            },
            {
                "text": "\u2022 they may coexist with a full homonymous verb;",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Contributions over time towards the concept of auxiliary verbs",
                "sec_num": "2"
            },
            {
                "text": "\u2022 they contribute to expressing tense, aspect and modality (known as TAM);",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Contributions over time towards the concept of auxiliary verbs",
                "sec_num": "2"
            },
            {
                "text": "\u2022 they do not occur alone, except in cases of elliptical full verbs (easily recoverable in context);",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Contributions over time towards the concept of auxiliary verbs",
                "sec_num": "2"
            },
            {
                "text": "\u2022 they are complemented by verbs in non-finite forms (gerund, participle and infinitive).",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Contributions over time towards the concept of auxiliary verbs",
                "sec_num": "2"
            },
            {
                "text": "Krug acknowledges the fact that auxiliaries for passive voice diathesis and negative and interrogative constructions are considered by some linguists, but he does not include them in his account.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Contributions over time towards the concept of auxiliary verbs",
                "sec_num": "2"
            },
            {
                "text": "Prior to Benveniste, an auxiliary construction was seen as a set of verbs complementing each other: one of them expressing grammatical features and the other semantic ones. Benveniste posited over-auxiliation and included modal verbs in the auxiliation process. The chain of over-auxiliation, including aspectual verbs, conforms to the following sequence of occurrence: modal -temporalaspectual -passive voice -full verb (see Figure 1 ). Over-auxiliation is of key importance for corpus annotation, though it is still under-explored in accounts on the matter. Bearing in mind that each auxiliary verb imposes a non-finite form on the auxiliated verb, we can argue that, except for the first auxiliary, which holds inflections, all the other auxiliaries in a verbal chain are, concomitantly, auxiliated by a preceding verb and auxiliary to the following one, as shown in Figure 1 . This implicates that the traditional labor division ascribed to auxiliary verbs as representing grammatical functions and full verbs as representing semantic functions cannot be sustained. therefore, the verb auxiliated by devia is the verb ter. \u2022 ter [ter (have)], auxiliary of tense, requires the auxiliated to be a past participle form; the verb auxiliated by ter is the verb come\u00e7ar. ter is auxiliated by devia and is auxiliary to come\u00e7ar. \u2022 come\u00e7ado [come\u00e7ar (start)], auxiliary of aspect, requires the auxiliated to be an infinitive form and to be introduced by the preposition a; the verb auxiliated by come\u00e7ar is the verb ser. Therefore, come\u00e7ar is auxiliated by ter and is auxiliary to ser. \u2022 ser [ser (be)], auxiliary of passive voice, requires the auxiliated to be a past participle form. The verb auxiliated by ser is the verb implantar (past participle: implantado), which is the full verb in this sentence. Therefore, ser is auxiliated by come\u00e7ar and auxiliary to implantar. \u2022 implantado [implantar (implement)] is a full verb, auxiliated by ser.",
                "cite_spans": [],
                "ref_spans": [
                    {
                        "start": 426,
                        "end": 434,
                        "text": "Figure 1",
                        "ref_id": "FIGREF0"
                    },
                    {
                        "start": 870,
                        "end": 878,
                        "text": "Figure 1",
                        "ref_id": "FIGREF0"
                    }
                ],
                "eq_spans": [],
                "section": "Discussion",
                "sec_num": "3"
            },
            {
                "text": "A productive way to explore the concept of auxiliation is to focus on the concept of auxiliated verb rather than on the concept of auxiliary verb. An auxiliated verb may be an auxiliary or a full verb. An auxiliated verb is a verb that takes the non-finite form required by its auxiliary, is introduced by the preposition (if any) required by its auxiliary, and has the same subject as its auxiliary. Therefore, we may have an auxiliation chain whenever all verbs in a chain share the same subject. However, we cannot affirm that the verbs in a chain sharing the same subject are auxiliaries followed by a full verb, as it depends on which verbs will be considered auxiliary in each work and for what purpose. In auxiliation verbal chains, the first verb is only auxiliary and the last verb is only auxiliated (full verb), but the verbs in between are both auxiliary and auxiliated, which shows that these two categories are not mutually exclusive.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Discussion",
                "sec_num": "3"
            },
            {
                "text": "The question to be posed is not whether a verb is an auxiliary, but whether it is an auxiliary in a given chain and what can be leveraged from its annotation. This approach makes it possible to overcome the much debated need to define a list of auxiliary verbs. Discussions about whether a verb is an auxiliary or not have always been based on comparing verbs with prototypical auxiliaries, i.e., those whose grammaticalization process is well advanced. In English, some auxiliaries (can, may, might, should) are fully grammaticalized to the extent that they do not compete with homonymous full verbs, do not require to to introduce the auxiliated verb, and do not require another verb to construct a negative and an interrogative form . In other languages, such as Brazilian Portuguese, auxiliary verbs 2 are at different stages in the grammaticalization process.",
                "cite_spans": [
                    {
                        "start": 483,
                        "end": 508,
                        "text": "(can, may, might, should)",
                        "ref_id": null
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Discussion",
                "sec_num": "3"
            },
            {
                "text": "Several scholars have tackled the task of describing auxiliary verbs in Brazilian Portuguese. Among them, Pontes (1973) and Lobato (1975) are two particularly exhaustive accounts, each exploring likely criteria to classify a verb as an auxiliary one. Pontes (1973) takes a syntagmatic view on auxiliaries and discusses interdependence relations between auxiliary and auxiliated verbs. Lobato (1975) performs different probes to try to differentiate auxiliary from non-auxiliary verbs. More recently, Ilari & Basso (2014) systematize criteria to assign auxiliary status to a given verb, considering all instances found in corpora regardless of the degree of grammaticalization a verb is still exhibiting. When it comes to grammar textbooks, lists of auxiliary verbs can be found in most of them, no two lists being alike, which shows the variety of criteria and stances taken by grammarians in Brazil.",
                "cite_spans": [
                    {
                        "start": 106,
                        "end": 119,
                        "text": "Pontes (1973)",
                        "ref_id": null
                    },
                    {
                        "start": 124,
                        "end": 137,
                        "text": "Lobato (1975)",
                        "ref_id": "BIBREF10"
                    },
                    {
                        "start": 385,
                        "end": 398,
                        "text": "Lobato (1975)",
                        "ref_id": "BIBREF10"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Discussion",
                "sec_num": "3"
            },
            {
                "text": "Drawing on Portuguese as a sample case, we would like to argue for a view on auxiliary verbs as verbs in their own right, implicating that a verb can be an auxiliary verb in some uses and a full verb in others, in the latter operating to help construe a variety of meanings. While they may be semantically weak, auxiliary verbs have a strong role in the syntax of the clause, its finite form agreeing with the subject and dictating the form of their auxiliated verbs. Auxiliary verbs can take part in a chain of several auxiliary verbs and be modified by adverbs, which sets them apart from fully-grammatical words. To better grasp the behavior of each auxiliary type in Brazilian Portuguese, we will briefly address four main groups (tense, modality, aspect and passive voice diathesis) and their characteristics below.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Discussion",
                "sec_num": "3"
            },
            {
                "text": "Brazilian Portuguese expresses tense basically through morphological desinence. The so-called tense auxiliaries ter and haver are used to express a previous past event within the past itself (1) and a previous future event within the future (2). For this reason, ter and haver are tense auxiliaries in some environments only, as in the following examples.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Auxiliaries of tense",
                "sec_num": "3.1"
            },
            {
                "text": "(1) Quando olhei, ele j\u00e1 havia atirado. (When I looked, he had already shot.) (2) No dia que voc\u00ea vier eu j\u00e1 terei partido. (The day you will arrive I will be gone.)",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Auxiliaries of tense",
                "sec_num": "3.1"
            },
            {
                "text": "As seen above, both ter and haver require the auxiliated verb to be a past participle form. However, a past participle form is not a criterion sufficiently strong to single out occurrences of ter and haver as tense auxiliaries. Ter may combine with a past participle form in other tenses to express aspect (3) and resultative constructions (4).",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Auxiliaries of tense",
                "sec_num": "3.1"
            },
            {
                "text": "(3) Ele tem vindo aqui todos os dias. (He has been coming here every day.) (4) Ele teve aprovado seu visto s\u00f3 ontem. (He had his visa approved only yesterday.)",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "3",
                "sec_num": null
            },
            {
                "text": "In the case of haver, this verb may be followed by a past participle form, which is not actually a verb, but a noun. In such cases, haver is not an auxiliary verb. This is the case of (5) and 6 In examples 5 and 6, comunicado and sentido are nouns and not verbs. In both, haver construes existence and is inflected in present and perfect tenses, i.e., it is not an auxiliary of tense.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "3",
                "sec_num": null
            },
            {
                "text": "Brazilian Portuguese has two modal verbs that are more highly grammaticalized than others: poder and dever. Both express several types of modality: permission, obligation and possibility. Poder has no homonymous full verb, but dever does [dever (to owe)]. There are several less grammaticalized modals like, e.g., tentar (to try). Some of them can be used as full verbs as is the case of saber (to know) and some of them, as is the case of pretender (to intend) and querer (to want or would like to), can also take a finite clause as a complement, its subject not being the same as the one of the main clause. However, whenever followed by an infinitive, those verbs share the same subject. These two possibilities are illustrated by examples (7) and (8). The particular behaviour of modal verbs mentioned above tends to exclude such verbs from traditional lists of modal verbs in Portuguese . However, for NLP, verbs like pretender (to intend), 6 querer (to want), saber (to know), tentar (to try), etc. followed by an infinitive verb are important cues for deducing whether an event has occurred or whether a statement is a fact or mere speculation.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Auxiliaries of modality",
                "sec_num": "3.2"
            },
            {
                "text": "Modal verbs in Brazilian Portuguese are in a stage of grammaticalization in which they have not lost their semantic load, since even the most grammaticalized one, poder, is polysemous: it construes permission (9) or probability (10). The same holds for dever, which construes obligation (11) or probability (12): Grammaticalization studies point out that there may be verbs at various stages in the grammaticalization process regarding their use as auxiliaries and this seems to be the case for many modal verbs in Brazilian Portuguese.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Auxiliaries of modality",
                "sec_num": "3.2"
            },
            {
                "text": "Aspectual verbs express how events occur in time. The meaning of aspect can be readily grasped through examples of some of its subcategories: frequentative (informing an event repeats frequently), inchoative (informing an event has started) and terminative (informing an event has finished).",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Auxiliaries of Aspect",
                "sec_num": "3.3"
            },
            {
                "text": "There are aspectual verbs in Brazilian Portuguese that convey information on the event. For example, the aspectual verb chegar a (literally arrive to) followed by an infinitive signals the event took place some time ago and lasted for a while, but did not persist. Besides being prolific, aspectual verbs are the least grammaticalized verbs in Brazilian Portuguese. Some compete in interpretation with full verbs, as is the case of acabar de [acabar (finish)], which, followed by an infinitive introduced by the preposition de , is aspectual in (16) and full verb in (17). ",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Auxiliaries of Aspect",
                "sec_num": "3.3"
            },
            {
                "text": "In Portuguese, the passive voice may be constructed by the auxiliary verb ser (to be) followed by a past participle (which we call analytic passive voice) or by adding the pronoun se to a transitive verb (which we call synthetic passive voice). In analytic passive voice, the subject is the prototypical patient and comes to the right of the verb ser:",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Auxiliary of passive voice",
                "sec_num": "3.4"
            },
            {
                "text": "(18) As cartas de senten\u00e7a foram assinadas pelo juiz. (The sentencing letters were signed by the judge.)",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Auxiliary of passive voice",
                "sec_num": "3.4"
            },
            {
                "text": "Unlike the auxiliary verbs of tense, which also require the participle form of their auxiliated, the participle of the passive voice is not invariable: it agrees in number and gender with the subject of the passive voice (in Portuguese, number and gender are typical inflections of nouns, while the typical inflections of verbs are mood, person, number, and tense). This fact enables us to verify agreement between verb and subject regardless of which verb is the head of the subject dependency relation.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Auxiliary of passive voice",
                "sec_num": "3.4"
            },
            {
                "text": "In Portuguese, only the passive voice auxiliary is fully grammaticalized and may occur in all verb tenses. Tense auxiliaries are well grammaticalized, but need to be annotated as such, though only in some verb tenses. Modals and aspectual verbs are less grammaticalized, but this does not mean their annotation is less important for NLP.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Auxiliary of passive voice",
                "sec_num": "3.4"
            },
            {
                "text": "The different views among linguists on which verb categories can actually be considered auxiliary pose an additional challenge to model them for the purpose of annotation in NLP.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Auxiliaries in Natural Language Processing",
                "sec_num": "4"
            },
            {
                "text": "For syntactic annotation based on constituencies, the key issue is to decide which of the verbs in a verbal phrase is the head. For annotation based on dependencies, the decision encompasses which verb is head and which one is dependent, as well as which dependency relation links each of the verbs to the others.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Auxiliaries in Natural Language Processing",
                "sec_num": "4"
            },
            {
                "text": "In UD, there is a Part of Speech (Pos) tag for full verbs (VERB) and a PoS tag for auxiliary verbs (AUX). Its guidelines leave it up to each language to define which categories of auxiliaries will be annotated as AUX and which verbs are prototypically used in each category. UD basically recommends that the auxiliary verbs specified in the annotation guidelines should be highly grammaticalized in the language. This has encouraged very conservative decisions, so that not all languages annotate modality auxiliaries (modal verbs) and aspect auxiliaries (aspectual verbs), as they tend to be less grammaticalized than other auxiliaries.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Auxiliaries in Natural Language Processing",
                "sec_num": "4"
            },
            {
                "text": "Automatic identification of auxiliaries in Portuguese has already been focused on by Baptista et al. (2010) , who consider an extensive list of over 26 auxiliary verbs; however, despite aiming at a dependency parser, the authors do not follow UD guidelines.",
                "cite_spans": [
                    {
                        "start": 85,
                        "end": 107,
                        "text": "Baptista et al. (2010)",
                        "ref_id": "BIBREF1"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Auxiliaries in Natural Language Processing",
                "sec_num": "4"
            },
            {
                "text": "In Portuguese, one of the main probes for identifying the subject is through verb agreement. The verb holding verb inflections is thus a natural candidate to be the head of the nsubj relation. This has implications in cases where there are auxiliaries together with a full verb. In UD, once a verb is annotated as AUX, there is no possibility to annotate it as a head: it has to be dependent in an aux dependency relation. Only verbs annotated as VERB may be heads of dependency relations.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Auxiliaries in Natural Language Processing",
                "sec_num": "4"
            },
            {
                "text": "Therefore, when an auxiliary verb, annotated as AUX, happens to be the first in an auxilation chain, thus keeping the inflections, a direct dependency relation between the inflected verb and the subject is not annotated, which precludes easy extraction of subjects. Still, if a verbal auxiliation chain contains several verbs annotated as AUX, the distance between the subject and the head (full verb) is longer and, as the full verb is always in an infinite verb form, the main cue to verify agreement between subject and verb can be missed. Most importantly, once a verb annotated as AUX cannot be head in a dependency relation, this prevents annotating over-auxiliation, which is marked by the non-finite verb form required from an auxiliated by its auxiliary. This is a major drawback, implicating that rich morphological and syntactical information is left untapped for linguistic studies and NLP applications.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Auxiliaries in Natural Language Processing",
                "sec_num": "4"
            },
            {
                "text": "As far as we can see, there are three options to annotate auxiliation chains using dependency relations, which are illustrated in the three figures below. Figure 2 shows annotation of all verbs with the PoS tag VERB, the first one being the head and the second one a dependent in a xcomp dependency relation, this annotation being iterated between the following verbs in the chain. Figure 4 shows annotation of all verbs as AUX, except for the last one, which is a full verb and takes the root. Figure 2 is the most satisfactory as it tags all verbs alike, which is a better representation in the case of a chain, where one verb is concomitantly auxiliary to the following verb and auxiliated by the preceding one. Annotation in Figure 3 is less satisfactory; although it keeps the traditional annotation of more grammaticalized auxiliary verbs as AUX, it misses details regarding the syntactic relation between the verbs annotated as VERB and those annotated as AUX, such as the requirement for an infinitive form after the aspectual verb come\u00e7ado (started). It also splits the verb chain into two xcomp relations. Annotation in Figure 4 is the least satisfactory of all, as it ignores the interdependence relationship between the verbs in a verbal chain and poses a problem regarding the distance between the verb holding inflections and the subject, assuming that the longer the distance is, the more difficult the task of subject detection becomes.",
                "cite_spans": [],
                "ref_spans": [
                    {
                        "start": 155,
                        "end": 163,
                        "text": "Figure 2",
                        "ref_id": "FIGREF4"
                    },
                    {
                        "start": 382,
                        "end": 390,
                        "text": "Figure 4",
                        "ref_id": "FIGREF6"
                    },
                    {
                        "start": 495,
                        "end": 503,
                        "text": "Figure 2",
                        "ref_id": "FIGREF4"
                    },
                    {
                        "start": 729,
                        "end": 737,
                        "text": "Figure 3",
                        "ref_id": "FIGREF5"
                    },
                    {
                        "start": 1130,
                        "end": 1138,
                        "text": "Figure 4",
                        "ref_id": "FIGREF6"
                    }
                ],
                "eq_spans": [],
                "section": "Auxiliaries in Natural Language Processing",
                "sec_num": "4"
            },
            {
                "text": "In order to find out which strategy is more suitable for machine learning, Lhoneux et al. (2020) compared two ways of annotating auxiliaries in UD: the auxiliary as head of its auxiliated (head left and dependent right) and the auxiliary as dependent on the auxiliated (head right and dependent left). They concluded that the information the annotation brings to the process depends on the machine learning modeling choices, and, therefore, if properly modeled, the same properties may be automatically acquired. Given that machine learning seems to deal equally well with both forms of annotation, we believe that it is preferable to choose a learning model based on what is desired to be learned rather than to choose a form of annotation based on what the available models can learn.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Auxiliaries in Natural Language Processing",
                "sec_num": "4"
            },
            {
                "text": "In Brazilian Portuguese auxiliary constructions, the inflectional head is an auxiliary and, for syntactic purposes, it is the head of the construction, as it must agree with the subject. In UD, however, when a verb is annotated as AUX, it becomes a dependent in a dependency relation aux, the head being another verb: a full verb or other auxiliary verb annotated as VERB if there is a chain. Moreover, UD defines AUX as a functional word and restricts its selection as head in a relation. This means that the phenomenon of over-auxiliation (one auxiliary modifying another) cannot be represented using AUX in UD.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Auxiliaries in Natural Language Processing",
                "sec_num": "4"
            },
            {
                "text": "From the perspective of semantic applications in NLP, treatment of auxiliaries and full verbs has important implications. Promoting the full verb to head is interesting to information extraction. Buiko et al (2009) , for example, compared the effect of different dependency representations on information extraction and concluded that the trimming of auxiliary structures enhanced the event extraction results. The trimming of auxiliary structures is an operation that seeks to \"prune the auxiliaries/modals as governors from the dependency graph and propagate the dependency relations of these nodes to the main verbs\" (Buiko et al 2009). For temporal expression, aspectual and modal verbs are fundamental cues, as explained by Pustejovsky et al (2017) in their design of the TimeML model, aimed to extract time information on events. Modal verbs are also relevant cues for speculation detection, as explored for different domains in Ozgur & Radev (2009) , Zhou et al. (2010) , Sauri & Pustejovsky (2012) , and Rivera Zavala & Martinez (2020) .",
                "cite_spans": [
                    {
                        "start": 196,
                        "end": 214,
                        "text": "Buiko et al (2009)",
                        "ref_id": null
                    },
                    {
                        "start": 729,
                        "end": 753,
                        "text": "Pustejovsky et al (2017)",
                        "ref_id": null
                    },
                    {
                        "start": 935,
                        "end": 955,
                        "text": "Ozgur & Radev (2009)",
                        "ref_id": "BIBREF14"
                    },
                    {
                        "start": 958,
                        "end": 976,
                        "text": "Zhou et al. (2010)",
                        "ref_id": "BIBREF22"
                    },
                    {
                        "start": 979,
                        "end": 1005,
                        "text": "Sauri & Pustejovsky (2012)",
                        "ref_id": null
                    },
                    {
                        "start": 1019,
                        "end": 1043,
                        "text": "Zavala & Martinez (2020)",
                        "ref_id": "BIBREF18"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Auxiliaries in Natural Language Processing",
                "sec_num": "4"
            },
            {
                "text": "Since an auxiliation chain has the same subject and refers to a same event, identifying them is productive for extraction tasks, even if auxiliaries are annotated as head or not, and even when there are other functions in between, such as adverbs and pronouns.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Auxiliaries in Natural Language Processing",
                "sec_num": "4"
            },
            {
                "text": "Three main arguments are worth summing up at this point:",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Auxiliaries in Natural Language Processing",
                "sec_num": "4"
            },
            {
                "text": "\u2022 there is syntactic relationship between the auxiliaries of modality, tense, aspect and passive voice diathesis within a chain of over-auxiliation; \u2022 verbs in a chain can be at the same time auxiliated by a verb and auxiliary to another one;",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Auxiliaries in Natural Language Processing",
                "sec_num": "4"
            },
            {
                "text": "\u2022 the UD guidelines do not allow an auxiliary verb (considered a functional word) to be head of a dependency relation; Bearing upon the above arguments, we believe the most productive way to annotate auxiliary verbs in UD is using the tag VERB and relating verbs to each other as open clausal complements (xcomp), this relation tag implicating they all have the same subject. Moreover, considering that modality, tense, aspectual and passive voice cues are relevant for many NLP applications, we propose to add a new annotation for this purpose at the morphological level: a feature called AuxiliaryType, with the initial values: Tense, Modality, Aspect, and Voice. Thus, regardless of whether the verbs that participate in the auxiliation chain have been annotated as AUX or as VERB, they will be identified at the feature level by the auxiliary function they perform within the chain. The absence of this feature means that the verb is not performing an auxiliary function within the chain and is therefore a full verb.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Auxiliaries in Natural Language Processing",
                "sec_num": "4"
            },
            {
                "text": "Our proposal has a twofold impact:",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Auxiliaries in Natural Language Processing",
                "sec_num": "4"
            },
            {
                "text": "\u2022 it allows treebanks with other annotation decisions for auxiliaries to reconcile their annotations by simply adding a feature to indicate auxiliary type (no alteration in annotation needed, but merely addition of features); \u2022 it enables recovery of auxiliation chains, as a sequence of verbs and/or auxiliaries that present a value of AuxiliaryType and are followed by a verb with no value of AuxiliaryType (the full verb, which is the semantic head).",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Auxiliaries in Natural Language Processing",
                "sec_num": "4"
            },
            {
                "text": "A further proposal is specifying Voice at the feature level. UD provides features to discriminate categories of Modality (Mood), Tense and Aspect. However, it has no feature to discriminate between categories of Voice (values: Passive, Agentive, Resultative, Causative, etc.). It would therefore also be desirable to create a Voice feature to complete the description of auxiliaries in UD morphology. This is particularly interesting because Voice is directly linked to the semantic role of the subject (Patient, Agent, Beneficiary, Cause, etc.) , and this would favor other NLP semantic applications.",
                "cite_spans": [
                    {
                        "start": 503,
                        "end": 545,
                        "text": "(Patient, Agent, Beneficiary, Cause, etc.)",
                        "ref_id": null
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Auxiliaries in Natural Language Processing",
                "sec_num": "4"
            },
            {
                "text": "In this paper, we have discussed the annotation of auxiliary verbs under the UD model, using evidence from linguistic theory to reconcile different ways of annotating phenomena that share semantic similarity, but differ greatly in syntactic behaviour.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Final Remarks",
                "sec_num": "5"
            },
            {
                "text": "As we have argued in the preceding sections drawing on grammaticalization studies, auxiliary verbs are verbs in their own right, i.e., they are neither a closed class of words (they are open to new candidate forms to auxiliaries), nor are they a fully functional class of words (they can operate as full verbs themselves). Throughout this paper, we have put forward arguments in favour of annotating auxiliary verbs as heads in dependency relations whenever the inflected verb is an auxiliary verb, as in Portuguese auxiliary chains. Hence, our proposal is to leverage the role of auxiliary verbs in the syntax of the clause, both in determining the form of auxiliated verbs and establishing agreement with the subject.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Final Remarks",
                "sec_num": "5"
            },
            {
                "text": "We also proposed the inclusion of a new morphological feature, AuxiliaryType, with the initial values of Tense, Modality, Aspect, and Voice. This way, we move towards standardizing UD auxiliaries annotation, enhancing comparability between languages. By annotating information on auxiliary function at feature level, we can reconcile our proposal to annotate auxiliaries with that of other languages that may have adopted other annotation strategies, either because their auxiliary verbs are not inflected, or because they have decided to privilege the semantic head in syntactic annotation.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Final Remarks",
                "sec_num": "5"
            },
            {
                "text": "This work is licensed under a Creative Commons Attribution 4.0 International License. License details: http://creativecommons.org/licenses/by/4.0/.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "",
                "sec_num": null
            },
            {
                "text": "In fact,Andersen (2006) basically recognizes three patterns of auxiliary verb construction inflections: AUX-headed (the auxiliary is the inflected verb), which is the most common pattern; LEX-headed (the full verb is the inflected verb, as in Eneats, Bulgarian, Macedonian, Hatam, Koiari and Kwerba); and doubled inflections (both auxiliary and full verbs inflect, as in Gutob, Mombelo and Mumbami).",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "",
                "sec_num": null
            },
            {
                "text": "The word comunicado is the past participle of the verb to communicate and means both communicated and notification. Many past participles in Portuguese are employed as true nominals.3 Resultative constructions resemble a kind of diathesis where the subject has the semantic role of benefactive. Diathesis is the alternation of semantic roles: in the passive voice diathesis, the patient is the subject; in the causative diathesis, the cause is the subject.2 Among less grammaticalized auxiliaries in English, Osborne & Gerdes (2019) point out be going to.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "",
                "sec_num": null
            },
            {
                "text": "In fact, some modal verbs such as querer, desejar, pretender are classified as full verbs realizing mental processes in systemic-functional descriptions of Portuguese. Likewise, in some accounts on auxiliary verbs, modals do not fulfill all the criteria to be considered auxiliary verbs(cf. LOBATO, 1975).5 The word sentido is the past participle of the verb sentir (to feel and means both felt and sense.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "",
                "sec_num": null
            },
            {
                "text": "One feature of aspectual verbs that poses a challenge to their annotation as auxiliaries in UD is the fact that most of them require a preposition to introduce the auxiliated verb. UD does not provide a specific dependence relation to link this preposition to one of the verbs (since the preposition neither marks case, nor introduces a subordinate clause). We have opted for annotating prepositions for verb and noun arguments as ADP.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "",
                "sec_num": null
            }
        ],
        "back_matter": [
            {
                "text": "Magali Duran and Thiago Pardo are grateful to The Center for Artificial Intelligence of the University of S\u00e3o Paulo (C4AI-http://c4ai.inova.usp.br/), sponsored by IBM and FAPESP (grant#2019/07665-4). Adriana Pagano wishes to thank the National Council for Scientific and Technological Development (CNPq) for grant No. 310630/2017-7.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Acknowledgements",
                "sec_num": null
            }
        ],
        "bib_entries": {
            "BIBREF0": {
                "ref_id": "b0",
                "title": "Auxiliary Constructions",
                "authors": [
                    {
                        "first": "Gregory",
                        "middle": [
                            "D S"
                        ],
                        "last": "Anderson",
                        "suffix": ""
                    }
                ],
                "year": 2006,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Anderson, Gregory D. S. 2006. Auxiliary Constructions. Oxford University Press.",
                "links": null
            },
            "BIBREF1": {
                "ref_id": "b1",
                "title": "Auxiliary Verbs and Verbal Chains in European Portuguese",
                "authors": [
                    {
                        "first": "J",
                        "middle": [],
                        "last": "Baptista",
                        "suffix": ""
                    },
                    {
                        "first": "N",
                        "middle": [],
                        "last": "Mamede",
                        "suffix": ""
                    },
                    {
                        "first": "F",
                        "middle": [],
                        "last": "Gomes",
                        "suffix": ""
                    }
                ],
                "year": 2010,
                "venue": "Computational Processing of the Portuguese Language. PROPOR 2010",
                "volume": "6001",
                "issue": "",
                "pages": "",
                "other_ids": {
                    "DOI": [
                        "10.1007/978-3-642-12320-7_14"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Baptista J., Mamede N., Gomes F. 2010. Auxiliary Verbs and Verbal Chains in European Portuguese. In: Pardo T.A.S., Branco A., Klautau A., Vieira R., de Lima V.L.S., eds. Computational Processing of the Portuguese Language. PROPOR 2010. Lecture Notes in Computer Science, vol 6001. Heidelberg: Springer. https://doi.org/10.1007/978-3-642-12320-7_14",
                "links": null
            },
            "BIBREF2": {
                "ref_id": "b2",
                "title": "Estrutura das Rela\u00e7\u00f5es de Auxiliaridade",
                "authors": [
                    {
                        "first": "\u00c9mile",
                        "middle": [],
                        "last": "Benveniste",
                        "suffix": ""
                    }
                ],
                "year": 1965,
                "venue": "Problemas de Lingu\u00edstica Geral II. Campinas: Pontes",
                "volume": "",
                "issue": "",
                "pages": "181--198",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Benveniste, \u00c9mile. 2006. Estrutura das Rela\u00e7\u00f5es de Auxiliaridade. In: Problemas de Lingu\u00edstica Geral II. Campinas: Pontes, [1965], pp. 181-198.",
                "links": null
            },
            "BIBREF3": {
                "ref_id": "b3",
                "title": "Syntactic simplification and semantic enrichment-trimming dependency graphs for event extraction",
                "authors": [
                    {
                        "first": "E",
                        "middle": [],
                        "last": "Buyko",
                        "suffix": ""
                    },
                    {
                        "first": "E",
                        "middle": [],
                        "last": "Faessler",
                        "suffix": ""
                    },
                    {
                        "first": "J",
                        "middle": [],
                        "last": "Wermter",
                        "suffix": ""
                    },
                    {
                        "first": "U",
                        "middle": [],
                        "last": "Hahn",
                        "suffix": ""
                    }
                ],
                "year": 2011,
                "venue": "Computational Intelligence",
                "volume": "27",
                "issue": "4",
                "pages": "610--644",
                "other_ids": {
                    "DOI": [
                        "10.1111/j.1467-8640.2011.0"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Buyko, E., Faessler, E., Wermter, J., & Hahn, U. (2011). Syntactic simplification and semantic enrichment-trimming dependency graphs for event extraction. Computational Intelligence, 27(4), 610-644. doi:10.1111/j.1467-8640.2011.0",
                "links": null
            },
            "BIBREF4": {
                "ref_id": "b4",
                "title": "Auxiliaries: Cognitive Forces and Grammaticalization",
                "authors": [
                    {
                        "first": "Bernd",
                        "middle": [],
                        "last": "Heine",
                        "suffix": ""
                    }
                ],
                "year": 1993,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Heine, Bernd. 1993. Auxiliaries: Cognitive Forces and Grammaticalization. New York: Oxford University Press.",
                "links": null
            },
            "BIBREF5": {
                "ref_id": "b5",
                "title": "Introduction: The Handbook of LinguisticAnnotation",
                "authors": [
                    {
                        "first": "Nancy",
                        "middle": [],
                        "last": "Ide",
                        "suffix": ""
                    }
                ],
                "year": 2017,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {
                    "DOI": [
                        "10.1007/978-94-024-0881-2_1"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Ide, Nancy. 2017. Introduction: The Handbook of LinguisticAnnotation. In: Ide, Nancy; Pustejovsky, James, eds. Handbook of Linguistic Annotation. Springer. DOI 10.1007/978-94-024-0881-2_1",
                "links": null
            },
            "BIBREF6": {
                "ref_id": "b6",
                "title": "O verbo. In Ilari",
                "authors": [
                    {
                        "first": "Rodolfo",
                        "middle": [],
                        "last": "Ilari",
                        "suffix": ""
                    },
                    {
                        "first": "Mario",
                        "middle": [],
                        "last": "Basso",
                        "suffix": ""
                    }
                ],
                "year": 2014,
                "venue": "Rodolfo. Gram\u00e1tica do portugu\u00eas culto falado no Brasil",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Ilari, Rodolfo, Basso, Mario. 2014. O verbo. In Ilari, Rodolfo. Gram\u00e1tica do portugu\u00eas culto falado no Brasil - vol. III -palavras de classes abertas.",
                "links": null
            },
            "BIBREF7": {
                "ref_id": "b7",
                "title": "The Oxford Handbook of Grammaticalization",
                "authors": [
                    {
                        "first": "Manfred",
                        "middle": [],
                        "last": "Krug",
                        "suffix": ""
                    }
                ],
                "year": 2012,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {
                    "DOI": [
                        "10.1093/oxfordhb/9780199586783.013.0044"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Krug, Manfred. 2012. Auxiliaries and grammaticalization. In: Heine, Bernd; Narrog, Heiko, eds. The Oxford Handbook of Grammaticalization. Oxford University Press. DOI: 10.1093/oxfordhb/9780199586783.013.0044",
                "links": null
            },
            "BIBREF8": {
                "ref_id": "b8",
                "title": "Auxiliation: an enquiry into the nature of grammaticalization",
                "authors": [
                    {
                        "first": "Tania",
                        "middle": [],
                        "last": "Kuteva",
                        "suffix": ""
                    }
                ],
                "year": 2001,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Kuteva, Tania. 2001. Auxiliation: an enquiry into the nature of grammaticalization. New York & Oxford: Oxford University Press.",
                "links": null
            },
            "BIBREF9": {
                "ref_id": "b9",
                "title": "What Should/Do/Can LSTMs Learn When Parsing Auxiliary Verb Constructions?",
                "authors": [
                    {
                        "first": "Miryam",
                        "middle": [],
                        "last": "Lhoneux",
                        "suffix": ""
                    },
                    {
                        "first": "Sara",
                        "middle": [],
                        "last": "De",
                        "suffix": ""
                    },
                    {
                        "first": "Joakim",
                        "middle": [],
                        "last": "Stymne",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Nivre",
                        "suffix": ""
                    }
                ],
                "year": 2020,
                "venue": "Computational Linguistics",
                "volume": "46",
                "issue": "4",
                "pages": "763--784",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Lhoneux, Miryam de, Sara Stymne, Joakim Nivre. 2020. What Should/Do/Can LSTMs Learn When Parsing Auxiliary Verb Constructions? Computational Linguistics 46(4), pp. 763-784.",
                "links": null
            },
            "BIBREF10": {
                "ref_id": "b10",
                "title": "A auxiliaridade em portugu\u00eas",
                "authors": [
                    {
                        "first": "L\u00facia",
                        "middle": [
                            "M P"
                        ],
                        "last": "Lobato",
                        "suffix": ""
                    },
                    {
                        "first": "L\u00facia",
                        "middle": [],
                        "last": "Lobato",
                        "suffix": ""
                    }
                ],
                "year": 1975,
                "venue": "An\u00e1lises lingu\u00edsticas",
                "volume": "",
                "issue": "",
                "pages": "27--91",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Lobato, L\u00facia M.P. 1975. A auxiliaridade em portugu\u00eas. In: Lobato, L\u00facia et al., ed. An\u00e1lises lingu\u00edsticas, pp. 27-91. Petr\u00f3polis: Vozes.",
                "links": null
            },
            "BIBREF11": {
                "ref_id": "b11",
                "title": "Universal Dependencies v2: An ever growing multilingual treebank collection",
                "authors": [
                    {
                        "first": "Joakim",
                        "middle": [],
                        "last": "Nivre",
                        "suffix": ""
                    },
                    {
                        "first": "Marie-Catherine",
                        "middle": [],
                        "last": "De Marneffe",
                        "suffix": ""
                    },
                    {
                        "first": "Filip",
                        "middle": [],
                        "last": "Ginter",
                        "suffix": ""
                    },
                    {
                        "first": "Jan",
                        "middle": [],
                        "last": "Hajic",
                        "suffix": ""
                    },
                    {
                        "first": "Christopher",
                        "middle": [
                            "D"
                        ],
                        "last": "Manning",
                        "suffix": ""
                    },
                    {
                        "first": "Sampo",
                        "middle": [],
                        "last": "Pyysalo",
                        "suffix": ""
                    }
                ],
                "year": 2020,
                "venue": "Proceedings of the12th Language Resources and Evaluation Conference",
                "volume": "",
                "issue": "",
                "pages": "4034--4043",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Nivre, Joakim, Marie-Catherine de Marneffe, Filip Ginter, Jan Hajic, Christopher D. Manning, Sampo Pyysalo, Sebastian Schuster, Francis Tyers & Daniel Zeman. 2020. Universal Dependencies v2: An ever growing multilingual treebank collection. In: Proceedings of the12th Language Resources and Evaluation Conference, 4034-4043. Marseille, France: European Language Resources Association. https://aclanthology.org/2020.lrec-1.497.",
                "links": null
            },
            "BIBREF12": {
                "ref_id": "b12",
                "title": "Towards a universal grammar for natural language processing",
                "authors": [
                    {
                        "first": "Joakim",
                        "middle": [],
                        "last": "Nivre",
                        "suffix": ""
                    }
                ],
                "year": 2015,
                "venue": "Computational Linguistics and Intelligent Text Processing",
                "volume": "",
                "issue": "",
                "pages": "3--16",
                "other_ids": {
                    "DOI": [
                        "10.1007/978-3-319-18111-01"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Nivre, Joakim. 2015. Towards a universal grammar for natural language processing. In: Alexander Gelbukh, ed. Computational Linguistics and Intelligent Text Processing, pp. 3-16. Cairo, Egypt: Springer International Publishing. doi:10.1007/978-3-319-18111-01.",
                "links": null
            },
            "BIBREF13": {
                "ref_id": "b13",
                "title": "The status of function words in dependency grammar: A critique of Universal Dependencies (UD)",
                "authors": [
                    {
                        "first": "Timothy",
                        "middle": [],
                        "last": "Osborne",
                        "suffix": ""
                    },
                    {
                        "first": "Kim",
                        "middle": [],
                        "last": "Gerdes",
                        "suffix": ""
                    }
                ],
                "year": 2019,
                "venue": "",
                "volume": "4",
                "issue": "",
                "pages": "1--28",
                "other_ids": {
                    "DOI": [
                        "10.5334/gjgl.537"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Osborne, Timothy and Kim Gerdes. 2019. The status of function words in dependency grammar: A critique of Universal Dependencies (UD). Glossa: a journal of general linguistics 4(1):17. pp. 1-28, DOI: https://doi.org/10.5334/gjgl.537",
                "links": null
            },
            "BIBREF14": {
                "ref_id": "b14",
                "title": "Detecting Speculations and their Scopes in Scientific Text",
                "authors": [
                    {
                        "first": "",
                        "middle": [],
                        "last": "Ozgur",
                        "suffix": ""
                    },
                    {
                        "first": "Dragomir",
                        "middle": [
                            "R"
                        ],
                        "last": "Arzucan",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Radev",
                        "suffix": ""
                    }
                ],
                "year": 1973,
                "venue": "Proceedings of the 2009 Conference on Empirical Methods in Natural Language Processing",
                "volume": "",
                "issue": "",
                "pages": "1398--1407",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Ozgur, Arzucan, Dragomir R. Radev. 2009. Detecting Speculations and their Scopes in Scientific Text. Proceedings of the 2009 Conference on Empirical Methods in Natural Language Processing, pp. 1398-1407, Singapore. Pontes, Eunice. 1973. Verbos Auxiliares em Portugu\u00eas. Rio de Janeiro, Ed. Vozes.",
                "links": null
            },
            "BIBREF16": {
                "ref_id": "b16",
                "title": "DesigningAnnotation Schemes: From Theory to Model",
                "authors": [
                    {
                        "first": "Annie",
                        "middle": [],
                        "last": "Zaenen",
                        "suffix": ""
                    }
                ],
                "year": 2017,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {
                    "DOI": [
                        "10.1007/978-94-024-0881-2_1"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Annie. Zaenen. 2017. DesigningAnnotation Schemes: From Theory to Model. In: Ide, Nancy; Pustejovsky, James, eds. Handbook of Linguistic Annotation. Springer. DOI 10.1007/978-94-024-0881-2_1",
                "links": null
            },
            "BIBREF17": {
                "ref_id": "b17",
                "title": "The Rational Optimist: How Prosperity Evolves",
                "authors": [
                    {
                        "first": "Matt",
                        "middle": [],
                        "last": "Ridley",
                        "suffix": ""
                    }
                ],
                "year": 2010,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Ridley, Matt. 2010. The Rational Optimist: How Prosperity Evolves. HarperCollins Publishers.",
                "links": null
            },
            "BIBREF18": {
                "ref_id": "b18",
                "title": "The Impact of Pretrained Language Models on Negation and Speculation Detection in Cross-Lingual Medical Text: Comparative Study",
                "authors": [
                    {
                        "first": "Rivera",
                        "middle": [],
                        "last": "Zavala",
                        "suffix": ""
                    },
                    {
                        "first": "R",
                        "middle": [],
                        "last": "Martinez",
                        "suffix": ""
                    },
                    {
                        "first": "P",
                        "middle": [],
                        "last": "2020",
                        "suffix": ""
                    }
                ],
                "year": null,
                "venue": "JMIR Med Inform",
                "volume": "8",
                "issue": "12",
                "pages": "",
                "other_ids": {
                    "DOI": [
                        "10.2196/18953"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Rivera Zavala R, Martinez P. 2020. The Impact of Pretrained Language Models on Negation and Speculation Detection in Cross-Lingual Medical Text: Comparative Study. JMIR Med Inform 8(12):e18953 URL: https://medinform.jmir.org/2020/12/e18953 DOI: 10.2196/18953",
                "links": null
            },
            "BIBREF19": {
                "ref_id": "b19",
                "title": "Are You Sure That This Happened? Assessing the Factuality Degree of Events in Text",
                "authors": [
                    {
                        "first": "",
                        "middle": [],
                        "last": "Sauri",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Roser",
                        "suffix": ""
                    },
                    {
                        "first": "James",
                        "middle": [],
                        "last": "Pustejovsky",
                        "suffix": ""
                    }
                ],
                "year": 2010,
                "venue": "Computational Linguistics",
                "volume": "38",
                "issue": "2",
                "pages": "261--299",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Sauri, Roser, Pustejovsky, James. 2010. Are You Sure That This Happened? Assessing the Factuality Degree of Events in Text. Computational Linguistics, 38, 2, pp. 261-299.",
                "links": null
            },
            "BIBREF20": {
                "ref_id": "b20",
                "title": "Reviewed Work: Auxiliaries: Cognitive Forces and Grammaticalization by Bernd Heine In: Language",
                "authors": [
                    {
                        "first": "Susan",
                        "middle": [],
                        "last": "Steele",
                        "suffix": ""
                    }
                ],
                "year": 1994,
                "venue": "",
                "volume": "70",
                "issue": "",
                "pages": "818--821",
                "other_ids": {
                    "DOI": [
                        "10.2307/416332"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Steele, Susan. 1994. Reviewed Work: Auxiliaries: Cognitive Forces and Grammaticalization by Bernd Heine In: Language, Vol. 70, No. 4, pp. 818-821. Linguistic Society of America Language https://doi.org/10.2307/416332. Accessed August 13, 2021.",
                "links": null
            },
            "BIBREF21": {
                "ref_id": "b21",
                "title": "Elements of Structural Syntax. Osborne, Timothy; Kahane, Sylvain, translators",
                "authors": [
                    {
                        "first": "Lucien",
                        "middle": [],
                        "last": "Tesni\u00e8re",
                        "suffix": ""
                    }
                ],
                "year": 1959,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Tesni\u00e8re, Lucien. 1959. Elements of Structural Syntax. Osborne, Timothy; Kahane, Sylvain, translators. John Benjamins Publishing Company, 2015.",
                "links": null
            },
            "BIBREF22": {
                "ref_id": "b22",
                "title": "Exploiting Multi-Features to Detect Hedges and their Scope in Biomedical Texts",
                "authors": [
                    {
                        "first": "H",
                        "middle": [],
                        "last": "Zhou",
                        "suffix": ""
                    },
                    {
                        "first": "X",
                        "middle": [],
                        "last": "Li",
                        "suffix": ""
                    },
                    {
                        "first": "D",
                        "middle": [],
                        "last": "Huang",
                        "suffix": ""
                    },
                    {
                        "first": "Z",
                        "middle": [],
                        "last": "Li",
                        "suffix": ""
                    },
                    {
                        "first": "Y",
                        "middle": [],
                        "last": "Yang",
                        "suffix": ""
                    }
                ],
                "year": 2010,
                "venue": "Proceedings of the Fourteenth Conference on Computational Natural Language Learning -Shared Task",
                "volume": "",
                "issue": "",
                "pages": "106--113",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Zhou, H., Li, X., Huang, D., Li, Z., & Yang, Y. 2010. Exploiting Multi-Features to Detect Hedges and their Scope in Biomedical Texts. In: Proceedings of the Fourteenth Conference on Computational Natural Language Learning -Shared Task. pp. 106-113. Uppsala: Association of Computational Linguistics.",
                "links": null
            }
        },
        "ref_entries": {
            "FIGREF0": {
                "type_str": "figure",
                "uris": null,
                "num": null,
                "text": "Auxiliary chain showing over-auxiliation process Verbs in Figure 1 can be thus analysed: \u2022 devia [dever (should)], auxiliary of modality, requires the auxiliated to be an infinitive form;"
            },
            "FIGREF1": {
                "type_str": "figure",
                "uris": null,
                "num": null,
                "text": "Voc\u00ea quer marcar uma consulta semana que vem? (Would you like to schedule an appointment next week?) (8) Voc\u00ea quer que eu marque uma consulta semana que vem? (Would you like me to schedule an appointment next week?)"
            },
            "FIGREF2": {
                "type_str": "figure",
                "uris": null,
                "num": null,
                "text": "Voc\u00ea pode entrar, se quiser. (You may come in if you want.) (10) Pode chover hoje \u00e0 noite. (It may rain tonight.) (11) O funcion\u00e1rio deve usar uniforme todos os dias no trabalho. (Employees must wear their uniform every day at work.) (12) O atraso deve ser por causa da chuva. (The delay must be due to the rain.)"
            },
            "FIGREF3": {
                "type_str": "figure",
                "uris": null,
                "num": null,
                "text": "Ele j\u00e1 acabou de ler o livro. (He has already finished reading the book.)"
            },
            "FIGREF4": {
                "type_str": "figure",
                "uris": null,
                "num": null,
                "text": "Annotation of inflected verb (auxiliary) as headFigure 3shows annotation of the more grammaticalized verbs as AUX (tense and passive voice auxiliaries) and the less grammaticalized ones as VERB (modal and aspectual auxiliaries), obtaining a combination of dependency relations aux and xcomp."
            },
            "FIGREF5": {
                "type_str": "figure",
                "uris": null,
                "num": null,
                "text": "Hybrid annotation of auxiliaries and auxiliated verbs"
            },
            "FIGREF6": {
                "type_str": "figure",
                "uris": null,
                "num": null,
                "text": "Annotation of (auxiliated) full verb as syntactic head Annotation in"
            },
            "TABREF1": {
                "content": "<table><tr><td>(13) Ele</td><td/><td>chegou</td><td>a</td><td colspan=\"2\">pensar</td><td>em</td><td colspan=\"2\">abandonar</td><td>o</td><td>Brasil</td></tr><tr><td>He</td><td/><td>arrived</td><td>to</td><td>think</td><td/><td>in</td><td colspan=\"2\">to abandon</td><td>the Brazil</td></tr><tr><td colspan=\"6\">He even thought about leaving Brazil for good.</td><td/><td/></tr><tr><td>(14) Ele</td><td/><td>deu de</td><td>assistir</td><td colspan=\"2\">filmes</td><td>de</td><td>terror</td><td>ultimamente</td></tr><tr><td>He</td><td/><td>gave of</td><td>watch</td><td colspan=\"2\">movies</td><td>of</td><td colspan=\"2\">horror</td><td>lately</td></tr><tr><td colspan=\"5\">He took to watching horror movies lately.</td><td/><td/><td/></tr><tr><td>(15) Eu</td><td>dei</td><td colspan=\"3\">de suspeitar de todo</td><td>mundo</td><td/><td colspan=\"2\">depois que fui enganada</td></tr><tr><td>I</td><td>gave</td><td>of suspect</td><td>of all</td><td/><td>world</td><td/><td>after</td><td>that was deceived</td></tr><tr><td colspan=\"6\">I became suspicious of everyone after I was deceived.</td><td/><td/></tr></table>",
                "text": "If we did not take into account the dynamic character of the process of auxiliation, as pointed out byKuteva (2001), we would not be able to recognize new aspectual verbs arising in Portuguese. For example, the verb dar de [dar (give)], followed by infinitive, informs the event has become a habit:",
                "type_str": "table",
                "num": null,
                "html": null
            }
        }
    }
}