You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
another major documentation effort; new aspect hierarchy; cleaned up some spurious ambiguity; some work on number NPs; initial effort to take full advantage of REPP, and make the grammar more useful with out-of-the-box preprocessors;
; This is a dummy type that should not be used directly.
198
+
; This type is to allow MEI to be used alongside GUO, and also as the directly negation of LE.
199
+
; """.
200
+
201
+
; le+pos := le
202
+
; """
203
+
; Restriction for verbs that can take, at least, 了 as aspectual particle.
204
+
; """.
205
+
206
+
; mei := guo & le+neg
207
+
; """
208
+
; 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)
209
+
; """.
210
+
211
+
; zhe+guo := zhe & guo
212
+
; """
213
+
; LZG (LE-ZHE-GUO) restriction: can, at least, take ZHE and GUO.
214
+
; """.
215
+
216
+
; le+zhe+guo := mei & le+neg & le+pos & zhe & guo
217
+
; """
218
+
; LZG (LE-ZHE-GUO) restriction: can take LE, ZHE and GUO.
This essentially blocks predicative adjectives from taking any kind of aspect.
228
+
This essentially restricts adjectives to interactions with 不 (and excludes all other aspect-related words).
71
229
This is being done here (and not in zhong-lextypes.tdl) since this is seen as
72
230
a Mandarin only restriction.
73
231
[FIXME] Need to ask Joanna whether or not this restriction would also happen
74
232
in Cantonese and, if so, more this higher. It is confusing to be split.
75
-
""".
233
+
[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.
234
+
""".
76
235
77
236
78
237
;; SSH 2015-04-28: verbal particles
@@ -84,17 +243,45 @@ Aspect markers such as 了 take a complement (currently +vj)
84
243
85
244
"""
86
245
[ SYNSEM.LOCAL.CAT.VAL [ COMPS < [ BOUND na-or--,
87
-
LOCAL [CONT.HOOK.INDEX.E.ASPECT perfective,
88
-
CAT.HEAD +vj] ] > ] ].
246
+
LOCAL [ CONT.HOOK.INDEX.E.ASPECT perfective,
247
+
CAT.HEAD +vj ] ] > ] ].
248
+
249
+
pfv-marker_le := pfv-marker &
250
+
"""
251
+
TEST FOR NEW HIERARCHY
252
+
This required that the COMP for perfective markers are compatible only with verbs that can take 'LE' as a perfective marker.
0 commit comments