5 Argument Realizations of Chinese Resultative Compounds
5.3 Dual Status of V2 in Chinese Resultative Compounds
國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
(94) a. Zhangsan hen qi.
Zhangsan very mad
‘Zhangsan was very angry.’
b. Zhangsan hen lei.
Zhangsan very tired ‘Zhangsan was very tired.’
How does one explain the inability of V2s to co-occur with adverbs of degree when they denote literal meanings in the sentence and are totally capable of being modified by the adverbs alone? The account of Result being Extent seems to hit a dead end here.
However, a solution will be given to explain the inablity of resultative compounds to coocurr with adverbs of extent and hypothesize the identity of the second verb in the resultative compounds.
5.3 Dual Status of V2 in Chinese Resultative Compounds
Although in Liu’s (2006) theory the indicator of Result being Extent comes
from that it does not denote its literal sense but rather acquires the Extent meaning in a metaphorical aspect, the Result still has its literal meaning implied in the sentence:
(95) zhe-ge gozuo lei-tan-le Zhangsan this-CL job tired-paralyzed-ASP Zhangsan
‘This job made Zhangsan so tired that he was (almost) paralyzed.’
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
The sentence is actually semantically ambiguous. It could mean that Zhangsan was made so tired that he was literally paralyzed or he was to a point of almost being paralyzed. In a situation like this, one cannot say for sure whether the Result, tan, in lei-tan is Extent or not if the critierion for defining its identity is based on whether it
depicts a real occurrence of Result. Of course the intended meaning of the sentence would be apparent when it is put in a context. However, the determining factor for telling what the V2 of a resultative compound is should be built on syntactic ground.
If the Result does require an argument and its meaning is predicated on the argument, it should be a predicate instead of an expression of Extent that does not require any argument. How, then, can the exclusivity of the Result to the adverb of degree be explained? It is hypothesized in this thesis that some of the V2s in resultative compounds in fact possess a dual status in identity.
Since V2 normally means the result caused by V1, there is an ongoing process of the action denoted by V1 that accumulates to the point of reaching the result denoted by V2. This underlying relationship between V1 and V2 makes the
foundation for the hypothesis of treating V2 as a predicate that is Result in terms of Syntax but Extent in terms of its function. This is the dual status of V2, as claimed in this thesis. It is Result because it assigns an argument on which it is prediaced and is Extent, on the other hand, because it entails the extent of accumulation of V1 before
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
its occurrence. With V2 being functionally Extent, the exclusivity between V2 and the adverb of degree is explained regardless of whether V2 is intended to be used as Result or Extent. However, when it comes to the V2s that have only literal meanings but no function of Extent, as in the case of qi-lei (mad-tired) in (93) and (94), the exclusivity of the V2s to the adverb of degree lies in the fact that the resultative compounds are more like action verbs as a whole and cannot be modified by any adverbs of degree:
(96) a. Zhangsan hen qi. (stative) Zhangsan very mad
‘Zhangsan was very mad’
b. Zhangsan hen lei. (stative) Zhangsan very tired
‘Zhangsan was very tired.’
c. *Zhangsan hen qi-lei-le. (action) Zhangsan very mad-tired-ASP
‘Zhangsan was so mad that he got tired.’
In conclusion, some of the V2s possess the dual status while others are strictly Result.
Now that all V2s are predicates that can represent Result, the argument realizations of resultative compounds will not be altered and the mapping theory
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
previously mentioned remains unharmed. It is so because if V2s were Extents, they would not be able to assign argument structures for they are not predicates. With no argument structures from the V2s to form composite roles with the argument structures from V1s, the rules of Resultative Compounding would be disfigured.
Hence, the argument-function mappings of the resultative compounds with Extent V2s would be completely different and unsatisfactory results of argument realization would happen. Take (97) for example, multiple readings of the sentence will be achieved according to the compounding and argument-function mapping rules:
(97) mama ma-si baba le.
mother scold-dead father ASP
‘This issue made Zhangsan extremely mad.’
a. Resultative Compounding: Vma ＜x y＞+ Vsi ＜z＞→
Vma Vsi＜α β＞, where ＜α β＞= (i) ＜x y-z＞
(ii) ＜x-z y＞ (x=ag, y=pt/th, z=pt/th) i. ‘Mother scolded Father to the extent of almost making him dead.’
＜x y-z＞ (x=ag, y=pt/th) (non-causative) IC [-r]
DC
--- CF S/O… S/O UMP S O
‧
ii. ‘Mother scolded Father and almost died.’
＜x-z y＞ (x=ag, y=pt/th) (non-causative)
iii. ‘Father scolded Mother and was made dead (by Mother).’
＜x-z[af] y[caus]＞ (z=pt/th, y=pt/th) (causative)
According the predictions of the LFG account, all these argument realizations of the sentence are eligible. One may find some of the readings hard to retrieve due to pragmatic factors because it is hard to imagine a situation where someone almost dies from the act of scolding, like (97ii). Even if the predicate si in the resultative
compound does not indicate real death but only a state where one beomes very miserable, it still has the implication of almost being dead in it. And that makes it still a predicate with its literal meaning.
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Therefore, if some V2s like si in (97) are treated as Extent as the scholars said they should, it will leave out some of the readings of the resultative compounds due to different argument structures caused by the V2s not being predicates:
(98) mama ma-si baba le.
mother scold-dead father ASP
‘This issue made Zhangsan extremely mad.’
a. Resultative Compounding: Vma ＜x y＞+ Extentsi → Vma Extentsi＜α β＞, where ＜α β＞ = ＜x y＞
(x=ag, y=pt/th) i. ‘Mother scolded Father very harshly.’
＜x y＞ (x=ag, y=pt/th) (non-causative) IC [-r]
DC
--- CF S/O… S/O UMP S O
Only one argument realization pattern of the sentence exists. V2s being Extent means no composite roles for the resultative compounds; hence, no multiple mapping
outcomes based on composite role suppression. Moreover, the implication of states of the V2s will be missing for they are no longer seen as predicates but only indicators of Extent. And the sentence in (98) can only have one reading; that is, “Mother scolded
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Father very harshly.” The miserable state as implied by si would not exist if si was Extent. However, it is hardly the case when people say sentences like this.
In brief, V2s have to be the predicates of Result in order to form composite roles in resultative compounds. And there is no denying that they are sometimes used as an expression of Extent to demonstrate the extremity of the situations described by the sentences. As a satisfactory solution, the hypothesis of some V2s having a dual status of being Result syntactically and being Extent functionally does not jeopardize the operation of the mapping principles in the LFG account but successfully solves the problem of the exclusivity between resultative compounds and adverbs of degree.
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
107