-
Notifications
You must be signed in to change notification settings - Fork 8
/
cmn.tdl
364 lines (272 loc) · 11.3 KB
/
cmn.tdl
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
;;; -*- Mode: TDL; Coding: utf-8 -*-
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;; Grammar of Mandarin Chinese - Core
;;;
;;;
;;; origially created by Sanghoun Song (2014-08-28)
;;;
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;================================================================================
; This file (cmn.tdl) exists to store syntactical information that pertains only
; Mandarin Chinese (i.e. where it can differ from Cantonese and other varieties)
; -- expected to be true for both Traditional (zht) and Simplified (zhs).
;
; Currently, this includes mainly information on ASPECT and CLASSIFIERS.
;================================================================================
;================================================================================
; ASPECT
;================================================================================
; Right now we can assume that BU has a similar aspect/restriction as a verb
; allowing only 着. (need to check if they are in fact compatible). [FIXME]
; We could change this to have more meaningful names in the future.
;
;================================================================================
+vjp :+ [ LZG lzg-asp ].
;================================================================================
; TEST FOR NEW ASPECT HIERARCHY
;================================================================================
; Under this new hierarchy, aspect particles should introduce the highest possible
; disjunction of aspects for which it can be used.
;================================================================================
lzg-asp := avm.
; lzg-asp := lzg.
; these top types are kinds of verbs.
; Most will be le-zhe-guo-mei-bu-zai-asp.
; The interactions are restricted by the existence of certain pairs.
; 门在开 - The door is opening.
; 门开着 - The door is open.
; 在坐 - Taking a seat.
; 坐着 - Sitting.
; 她 看 着 电影 。 OK
; 他 刷 着 牙 。 OK
; 他 在 看 着 电影 。 OK
; 他 在 刷 着 牙 。 OK
; 他 在 刷 牙 。 OK
; 她 没 看 着 电影 。 OK
; 他 没 刷 着 牙 。 OK
; 他 没 在 刷 着 牙 。 OK
; 他 没 在 刷 牙 。 OK
; 他 不 在 家 。 OK (but ZAI is a 'verb')
; 她 看 了 电影 。 OK
; 她 不 看 电影 。 OK
; 她 没 看 电影 。 OK
; 她 看 过 电影 。 OK
; 她 没 看 过 电影 。 OK (didn't + have the experience)
; 他 不 (吃着东西看电影) OK, but BU is for the main verb.
; 她 在 看 过 这 个 电影 NO!!!!!!!!
; 她 BU 看 着 电影 。 NO!!!
; le|zhe NEVER
; 她 看 了 着 电影 。 NO!!
; 她 看 着 了 电影 。 NO!!
; 他 在 看 了 电影 。 NO!!
; 他 在 没 看 着 电影 。 NO!!!!!
; 他 不 在 看 电影 。 NO!!! -- (only if it's a place, e.g. TA BU ZAI JIA)
; 她 不 看 了 电影 。 NO!!!
; 她 没 看 了 电影 。 NO!!! ("did + didn't can't coexist")
; 她 看 了 过 这 个 电影 。 NO!!!
; 她 看 过 了 这 个 电影 。 NO!!!
; 她 不 看 过 这 个 电影 。 NO!!!
no-lzg-asp := lzg-asp
"""
For verbs that cannot interact with any of the aspect-related words.
This should probably be used very sparingly.
""".
le-zhe-guo-mei-bu-zai-asp := lzg-asp
"""
For verbs that can interact with all aspect-related words.
""".
le-guo-mei-bu-asp := le-zhe-guo-mei-bu-zai-asp
"""
For verbs that cannot interact with 在 or 着 (essentially blocking the progressive aspect).
[FIXME] IS 在 and 着 the same in restrictions like psychological verbs?
""".
zhe-guo-mei-bu-zai-asp := le-zhe-guo-mei-bu-zai-asp
"""
For verbs that cannot interact with 了 (essentially blocking the perfective aspect).
However they can interact with 过. This, apparently, includes stative/psychological verbs (i.e. verbs that can take degree specifiers).
[FIXME] should be be LE or MEI??? This would probably pose a problem since 过 and 着 are apparently possible.
""".
le-zhe-mei-bu-zai-asp := le-zhe-guo-mei-bu-zai-asp
"""
For verbs that cannot interact with 过 (essentially blocking the experiential aspect).
""".
le-zhe-guo-mei-zai-asp := le-zhe-guo-mei-bu-zai-asp
"""
For verbs that cannot interact with 不 (e.g. 有).
""".
bu-asp := le-guo-mei-bu-asp &
zhe-guo-mei-bu-zai-asp &
le-zhe-mei-bu-zai-asp
"""
For verbs that only interact with 不 (e.g. 是).
""".
zhe-guo-mei-zai-asp := le-zhe-guo-mei-zai-asp &
zhe-guo-mei-bu-zai-asp
"""
""".
zhe-mei-zai-asp := zhe-guo-mei-zai-asp & le-zhe-mei-bu-zai-asp.
guo-zhe-mei-asp := zhe-guo-mei-zai-asp.
le-guo-asp := le-zhe-guo-mei-zai-asp & le-guo-mei-bu-asp.
zhe-zai-asp := zhe-mei-zai-asp.
guo-zhe-asp := guo-zhe-mei-asp.
guo-mei-asp := guo-zhe-mei-asp & le-guo-mei-bu-asp.
zhe-mei-asp := guo-zhe-mei-asp & zhe-mei-zai-asp.
mei-zai-asp := zhe-mei-zai-asp.
; mei-asp := guo-mei-asp & zhe-mei-asp. ; this can't exist because it breaks things
zhe-asp := zhe-mei-asp & zhe-zai-asp & guo-zhe-asp.
guo-asp := guo-mei-asp & le-guo-asp & guo-zhe-asp.
zai-asp := mei-zai-asp & zhe-zai-asp. ; this should possibly also not exist ?
le-asp := le-guo-asp & le-zhe-mei-bu-zai-asp.
;================================================================================
; lzg := avm.
; le := lzg
; """
; LZG (LE-ZHE-GUO) restriction: can, at least, take LE.
; """.
; zhe := lzg
; """
; LZG (LE-ZHE-GUO) restriction: can, at least, take ZHE.
; """.
; guo := lzg
; """
; LZG (LE-ZHE-GUO) restriction: can, at least, take GUO.
; """.
; no-lzg := lzg
; """
; LZG (LE-ZHE-GUO) restriction: cannot take any aspectual particle (e.g. 是).
; """.
; le+zhe := le+neg & le+pos & zhe
; """
; LZG (LE-ZHE-GUO) restriction: can, at least, take LE and ZHE.
; """.
; le+guo := le+neg & le+pos & guo
; """
; Perfective OR Experiential
; LZG (LE-ZHE-GUO) restriction: can, at least, take LE and GUO.
; Should have all but ZAI/ZHE?
; """.
; le+neg := le
; """
; This is a dummy type that should not be used directly.
; This type is to allow MEI to be used alongside GUO, and also as the directly negation of LE.
; """.
; le+pos := le
; """
; Restriction for verbs that can take, at least, 了 as aspectual particle.
; """.
; mei := guo & le+neg
; """
; Restriction for verbs that can take 过 or verbs that can take a negated equivalent of LE (i.e. 没). This should ensure that 没 is compatible with 过 and with verbs that were compatible with 'le'. However, since 'le+neg' is a subtype of 'le', when a verb is compatible with 'le' it is, in fact, compatible with either the positive form (i.e. 了) or it's negative form (i.e. 没)
; """.
; zhe+guo := zhe & guo
; """
; LZG (LE-ZHE-GUO) restriction: can, at least, take ZHE and GUO.
; """.
; le+zhe+guo := mei & le+neg & le+pos & zhe & guo
; """
; LZG (LE-ZHE-GUO) restriction: can take LE, ZHE and GUO.
; """.
basic-pred-adj-lex :+ [ SYNSEM.LOCAL.CAT.HEAD.LZG bu-asp ]
"""
This essentially restricts adjectives to interactions with 不 (and excludes all other aspect-related words).
This is being done here (and not in zhong-lextypes.tdl) since this is seen as
a Mandarin only restriction.
[FIXME] Need to ask Joanna whether or not this restriction would also happen
in Cantonese and, if so, more this higher. It is confusing to be split.
[FIXME] This is also not exactly true, we know adjectives can, in some circunstances, be equated as verbs when interacting with aspect (e.g. LE). We need to understand how this can be processed.
""".
;; SSH 2015-04-28: verbal particles
pfv-marker := aspect-markers &
"""
This is used for 了, to mark perfective aspect.
Aspect markers such as 了 take a complement (currently +vj)
"""
[ SYNSEM.LOCAL.CAT.VAL [ COMPS < [ BOUND na-or--,
LOCAL [ CONT.HOOK.INDEX.E.ASPECT perfective,
CAT.HEAD +vj ] ] > ] ].
pfv-marker_le := pfv-marker &
"""
TEST FOR NEW HIERARCHY
This required that the COMP for perfective markers are compatible only with verbs that can take 'LE' as a perfective marker.
"""
[ SYNSEM.LOCAL.CAT.VAL [ COMPS < [ LOCAL.CAT.HEAD.LZG le-asp] > ] ].
exp-marker := aspect-markers &
[ SYNSEM.LOCAL.CAT.VAL [ COMPS < [ LOCAL.CONT.HOOK.INDEX.E.ASPECT experiential ] > ] ].
exp-marker_le := exp-marker &
"""
TEST
"""
[ SYNSEM.LOCAL.CAT.VAL [ COMPS < [ LOCAL.CAT.HEAD.LZG guo-mei-asp ] > ] ].
; [ SYNSEM.LOCAL.CAT.VAL [ COMPS < [ LOCAL.CAT.HEAD.LZG guo-asp ] > ] ].
;; SSH 2014-09-02: removing irules.
dur-marker := aspect-markers &
[ SYNSEM.LOCAL.CAT.VAL [ COMPS < [ BOUND na-or--,
LOCAL.CONT.HOOK.INDEX.E.ASPECT durative ] > ] ].
dur-marker_le := dur-marker &
"""
TEST
"""
[ SYNSEM.LOCAL.CAT.VAL [ COMPS < [ LOCAL.CAT.HEAD.LZG zhe-mei-asp ] > ] ].
; [ SYNSEM.LOCAL.CAT.VAL [ COMPS < [ LOCAL.CAT.HEAD.LZG zhe-asp ] > ] ].
;; SSH 2014-09-11: [SPEC <>] is added to block 了 不.
;; WWJ 2015-03-17: Added LOCAL...SF prop to block use with A-not-A
;; ZZF 2016-12-2: removed LOCAL.CONT.HOOK.INDEX.SF prop, as they can go with questions as well
;; SSH 2015-04-28: sentence-final particles
;; LMC changed to le (lexical entry type)
;================================================================================
; LE TYPES
;
;================================================================================
crs_le := spart-lex-item & nonque-item &
"""
CRS stands for 'Currently Relevant State', as proposed by Li & Thompson (1981).
It refers to the uses of 了 (LE) and it broadly encompasses multiple uses:
(1) change of state, (2) corrects a wrong assumption, (3) reporting progress
(4) reports what will happen next, (5) marks the speaker's total contribution
to the conversation at that point.
[FIXME]However 'crs' is being used to refer to sentence final particles in
general this should be changed eventually. E.g. 吧_crs, 啊_crs, 啦_crs, etc.
crs_le is possibly not the right type for all these.
We should probably create more subtypes of spart-lex-item, and make sure
we can encode the possible combinations of sent-final-particles.
"""
[ STYLE robust,
SYNSEM [ R-PERIPH +,
SPART crs,
LOCAL.CAT [ VAL [ COMPS < [ SPART not-crs-not-ma,
LOCAL.CAT [ HEAD +vj,
VAL [ SUBJ < >,
COMPS < > ],
MC #mc ],
NON-LOCAL.SLASH 0-dlist ] >,
SPEC < >],
MC #mc ] ] ].
;; SSH 2015-04-16: In Mandarin Chinese, classifiers must co-occur with
;; numerals. (ex. *(一) 只 猫 走 了).
classifier-lex :+ no-spr-classifier-rule-dtr &
[ SYNSEM.LOCAL.CAT.VAL.SPR < [ OPT - ] > ].
no-spr-classifier-lex-rule := const-val-change-only-lex-rule &
[ SYNSEM.LOCAL.CAT [ NONSUBJ +,
HEAD.MOD < [ LOCAL [ CAT.HEAD.GENERIC -,
CONT.HOOK.INDEX.COG-ST type-id ] ] >,
VAL [ SUBJ <>,
COMPS <>,
SPEC <>,
SPR <> ] ],
DTR no-spr-classifier-rule-dtr,
C-CONT [ RELS <! !>, HCONS <! !>, ICONS <! !> ] ].
no-spr-classifier-rule-dtr := word-or-lexrule.
;; SSH 2015-05-07: Mandarin Specific Predsort.
_de_dem_q_rel := dem_rel.
_zhi_dem_q_rel := dem_rel.
_de_p_assoc_rel := related_p_rel.
_de_nom_rel := nm_rel.
_zhi_nom_rel := nm_rel.
_bu_x_rel := neg_x_rel.
_buqi_x_rel := neg_x_rel.
_bu_polar_x_rel := neg_x_rel.
_mei_x_rel := neg_x_rel.
_mei_polar_x_rel := neg_x_rel.
_hen_x_rel := intensifier_x_rel.
_feichang_x_rel := intensifier_x_rel.
_ting_x_rel := intensifier_x_rel.