Chapter 3 Materials and Data
3.2 The Spontaneous Production Data
3.2.4 Discussion
Based on Table 3.14 to Table 3.17, it is observed that there are some obvious changes from the first period of age (1;11 to 2;6) to the last period (after 3;9). The frequency of the use of omission of VP and unbounded VP is the highest before the age 3.03 and declined successively as the child grew up, except for HY’s data in which the frequency of unbounded VP declined firstly but rose at last.
3.2.4 Discussion
The results shown in Table 3.6 to Table 3.17, mostly, satisfy our expectation that the error rate should decline as the subject grew up. In the spontaneous speech data of HY, presented in Table 3.7, it is not difficult to find that the error rate declined
successively from 35.29% in the first period of age to 20.24% in the third period of age.
And in Table 3.13, the error ratio of CC shows the same result that the error ratio
declined successively as well. Although, in Table 3.9 and Table 3.11, the data of JC and AN do not display smoothly declination of the error ratio (the error ratio from stage 1 to stage 3 in JC’s data: 16.67%, 21.05%, 9.09%; the error ratio from stage 1 to stage 4 in AN’s data: 26.66%, 24.24%, 27.27%, 22.72%), the error ratios of these two children, JC and AN, do decline to the lowest point at the last stage. However, we still find some unexpected results. For example, in Table 3.14, HY’s data, the frequency of the wrong use of verb, unbounded VP and the omission of post-ba NP, did not decrease smoothly.
Hence, we try to give a plausible explanation here.
Why does the percentage of the pattern of wrong use of verb increase from the first period of age to the second period but declined at the third period? We can have a speculation that the ratio of wrong use of verb might successively decline as the child got mature, since, as children get old, the usage of verb will be likely to be more accurate and specific. In this case, the error is a consequence of incomplete lexical knowledge rather than lack of linguistic competence. Take sentences (5a) and (6a) for example, the errors of these two ba sentences are attributable to the incomplete knowledge of the properties of the two verbs, 丟 diu ‘throw’ and 飛 fei ‘fly’. The misuse of diu in the verb phrase 丟死 diusi ‘throwing somebody/ something and make them dead’ is a result of semantic failure; usually, the verb 摔 shuai ‘throw down’ used in the verb phrase 摔死 shuaisi is much adopted by adults. On the other hand, the wrong use of fei is due simply to the misuse of intransitive verb as transitive, which happens a lot in the early children language acquisition.
(5) a. Subject’s Sentence
要把它丟死 (HY: 3;5)
yao ba ta diu si will BA it throw dead
‘Throw it to dead.’
b. Adult’s sentence 要把它摔死
yao ba ta shuai si will BA it throw down dead ‘Throw it to dead.’
(6) a. Subject’s Sentence
你可以把直昇機飛到這裡上面 (HY: 3;0)
ni keyi ba zhishengji fei dao zheli shangmian you can BA helicopter fly to here up
‘Yon can fly your helicopter up to here.’
b. Adult’s sentence
你可以把直昇機開到這裡上面
ni keyi ba zhishengji kai dao zheli shangmian you can BA helicopter drive to here up
‘You can fly your helicopter up to here.’
In Table 3.14, at first sight, it seems strange that the error rate of unbounded VP declined slightly during the age of 3;3 to 3;8 but rose during the age of 3;9 to 4;1.
However, under careful observation, we found that is a statistical mistake. There are two tokens of the unbounded VP counted at the third period of age. See sentences (7a) and (8):
(7) a. 把它打 (HY: 4;0)
ba ta da
BA he hit ‘Hit him.’
(8) 把你拆 (HY: 4;1)
ba ni chai
BA you take apart ‘Take you apart.’
Yet, an utterance immediately following (7a) was found. See (7b) below:
(7) b. 把它打扁 (HY: 4;0)
ba ta da bian BA he hit flat ‘Hit him strongly.’
From sentences (7a) and (7b), this leads to the possibility that (7a) is just a sentence fragment of the well-formed sentence (7b). Therefore, just one token might be counted during the age of 3;9 to 4;1, and the percentage should be 1.19%. Accordingly, the ratios from the first period of age to the third period of age are 1.96%, 1.26%, and 1.19%, which conform to our prediction that the frequency of unbounded VP should decline as children grew up.
In the case of the omission of the post-ba NP, it is unexpected for the reason that following many researches the post-ba NP is a must-have constituent in the ba construction. The post-ba NP is assigned case by ba, and no element can intervene between them (See Huang & Li & Li 2009, for instance). However, we do find this kind of error appearing in the children’s acquisition of ba sentence. In the data presented
above, the error ratio of this pattern is quite inconsistent. In Table 3.15, JC’s data, we do not find this kind of error. In Table 3.16 and Table 3.17, the error ratios of these two children, AN and CC, do not decrease smoothly. The error ratio declines at first, but increases in the later stage. Moreover, in Table 3.14, HY’s data, omission of post-ba NP, does not decrease at all; instead, it increases from the first period of age (0.00%) to the second period age (1.89%) and from the second period of age to the third period of age (3.57%). According to Cheung (1992), this phenomenon persists across different age, and Cheung argued that this error is a performance error because it is found in different age. However, from the point of view of Hsieh (2009), she tried to manifest that this is not a performance error. Hsieh (2009) argued that although this kind of error is found to last for a long time in the data in her study, this phenomenon does not necessarily mean that omission of post-ba NP is a performance error. She also claimed that since the child in her study stops marking this error after he reaches at a certain age (4;8), it is
suggested that this might not be a performance error. In our research, although we can not give a general reason to explain the inconsistency of this error type among our subjects, based on the data of four children our finding is more consistent with Cheung’s (1992) claim. Since the low occurrence and the coincidence of the omission of the post-ba NP are found in the three children in our study, this error type is more likely to be a performance error. However, on account of the lack of data which is after 4;8, we can not confirm whether this error type will cease appearing. Hence, as for the cause of the appearance of omission of the post-ba NP in child language, the future research is absolutely needed.
There are two errors found in AN’s and CC’s data respectively. These two errors are difficult to classify into our categories of the non-adult-like ba constructions. These two errors shown in (9) to (10) are repeated below with the conversation context accompanying.
(9) INVESTIGATOR: 你要打給誰
ni yao da gei shei you will call to whom ‘Whom will you call?’
CHILD: 我要把這個打電話 (AN: 3;1)
wo yao ba zhege da-dianhua I will BA this make a phone call
‘I will make a phone call to this.’
(10) INVESTIGATOR: 喔，那你們家有沒有掃把
o ，na nimen jia you-mei-you saoba SEP then you house have-not-have broom
‘Then, do you have a broom in your house?’
CHILD: 我爸爸，把掃把，我爸爸把掃地喔 (CC: 3:1)
wo baba， ba saoba， wo baba ba saodi o my father BA broom my father BA sweep SEP
‘My father takes a broom to sweep.’
In (9), the investigator tried to ask the child to whom the child wants to make a phone call. From the sentence the child produced, it is likely to image that at this point in time the child points to some entity and answers the investigator that he wants to makes a phone call to this entity. Under this view, we can infer that at this moment the child used the ba sentence has had the idea of the disposal property of ba construction.
He plans to deal with or manipulate 這個 zhege ‘this’. That is, the child intends to do something which is 打電話 da-dianhua ‘make a phone call’ to the entity, 這個 zhege
‘this’. However, the child has not acquired the rule of ba construction completely. Even
though he perceives that ba construction is possessed of the disposal property, he doesn’t figure out that an intransitive verb can not follow the [ba NP]. Take (11) for example:
(11) a. *我把它跑步
wo ba ta paobu I BA it run ‘I run.’
b. *我把它飛
wo ba ta fei I BA it fly ‘I fly it.’
In (11), it is easily to find that an intransitive verb, here 跑步 paobu ‘run’ and 飛 fei
‘fly’, can not be shown in the ba construction. Therefore, we can attribute this error to the child’s incomplete knowledge to ba construction.
In (10), CC’s data, it is very difficult to analyze why this child would make this kind of error in the second ba in her reply to the investigator. She not only uses an intransitive verb saodi ‘sweep’ in the ba sentence, but also omits the post-ba NP.
However in the first ba in (10), we can presume that the child wants to said 拿掃把 na saoba ‘take the broom’. If it is the case, we might have a more reasonable and
well-formed reading to the sentence the child produces. It is shown in (12) below.
(12) 我爸爸，拿掃把，我爸爸拿(來)掃地喔
wo baba， na saoba， wo baba na-(lai) saodi o my father take broom my father take to sweep SEP
‘My father takes a broom. My father takes a broom to sweep.’
However, this is only our conjecture, and in the following chapter we will elaborate on the possibility of the verb property of ba. In our research, we do not endeavor to explain every error of ba sentence in our subjects’ production. This will leave to future studies.
Chpater 4 GENERAL DISCUSSION
From the results in the previous chapter, it can be found that ba construction is not simply a grammatical construction for children to acquire for the reason that children committed errors in 20 to 35 percent of their usage of ba construction. We also find that children begin to use ba quite early as they start to produce multiword utterances.
It is well noted that ba in Modern Chinese has lost standard verbal properties. Ba has become “grammaticalized” and does not behave like a pure lexical verb (Li and Thompson, 1981; Sybesma, 1999; Huang & Li & Li 2009). In this chapter, we will not try to give any new proposal to argue against the original one, but, however, we aim to assume that at the early acquisition stage of ba, ba is firstly acquired as a verb, not a grammaticalized category.
In chapter 3, we can see that from the first period of age to the second period of age the frequencies of most of the ba patterns increase. Moreover, it has been shown that the subjects’ non-adult-like ba construction, the error ratio, dropped as they get more
mature. These results manifest that the syntactic representation of the ba construction in children’s language differs considerably from the ba in adult’s language. In order to examine the status of ba in the very early age of children’s language, we try to analyze those non-adult-like ba constructions found in the children’s language. We intend to provide evidence to support the hypothesis that ba is, firstly, acquired as a verb in young children’s language through the errors children make in producing ba sentence. The investigation of children’s grammatical errors in ba sentence is approached from two angles: the errors whose ratio decrease as the increase of age, and specific patterns of errors found in the four children’s spontaneous speech in our study. Under careful observation, three sets of condition found in the spontaneous production data can be the
effective evidence to argue that our hypothesis that ba is firstly acquired as a lexical verb may be valid. The three sets of condition are reorganized and presented in (1)-(3), and examples are also shown to illustrate these conditions:
(1) NP1 [[BA NP2] (VP)]: ‘omission of VP in the ba construction’
Investigator: 這塊綠色的跟它放在一起就會逼逼。
zhe kuai lusede gen ta fang zai yiqi jiu hui bibi this CL green with it put to together then will ‘sound’
‘Put this green one beside it, and then it will make the ‘beebee’
sound.’
Child: 這樣把它就沒有了 (HY: 3;0)
zheyang ba ta jiu meiyou le this BA it then without LE
‘Put it in this way, and it will not (make that sound).’
(2) canonical ba sentence: ‘only canonical ba sentence appears before 3;9’
Child: 我們先把車車收起來 (HY: 2;11)
women xian ba che che shou qilai we first BA car car gather up ‘Firstly, we gather the car.’
(3) verbs in a bare form: ‘the bare verb form ceases to appear at the age of 4’
Child: 我把它切 (JC: 3;1)
wo ba ta qie I BA it cut
‘I cut it.’
In the following section, we are going to discuss the roles that the three sets of condition played in the acquisition of ba sentence in child language.
4.1 Omission of VP
Adopting Huang & Li & Li (2009), the construction of ba sentence is that a subject is followed by ba and a NP, and a VP follows the NP. The structure has been presented in (18) in Chapter 2 as repeated here:
(4) [baP Subject [ba’ ba [vP NP [v’ v [VP V XP]]]]].
It is perfectly accepted that in the ba construction there must be a VP following the post-ba NP. The ba construction expresses that an object (that is the NP in (4)) is affected, dealt with, or disposed of by a VP and the complement of the VP. Take (5) for example.
(5) 我把張三打斷了手
wo ba Zhangsan da duan le shou I BA Zhangsan hit break LE hand
‘I hit Zhangsan and break his hand.’
In (5), 張三 Zhangsan is affected by the verb phrase 打斷 da duan ‘hit and break’ and its complement 手 shou ‘hand’. Furthermore, the post-ba NP is assigned an “affected”
theta-role by the complex verb phrase following the post-ba NP. However, in the data we collected we can find that before a certain age children might sometimes produce ba sentences without a VP following the post-ba NP. Examples are shown below:
(6) a. Child: 然後，然後把 taxi (HY: 2;11)
‘Put this green one beside it, and then it will make the ‘beebee’
sound.’
From the data above, we can deduce that, at the beginning stage of language acquisition of ba sentence, children take ba as a more verb-like morpheme. Ba can serve as a verb, which can directly take an NP, or an NP followed by a complement. The semantic properties of ba might retain the original verb meaning ‘take, grasp, deal with.’
The assumption is not too odd, since there is no doubt that ba is a lexical verb
historically. Also, it is undeniable that ba underwent semantic bleaching. And then ba becomes “grammaticalized” and does not behave like a verb (Wang, 1958). For this reason, we can argue that a grammaticalized category is, firstly, too complicated for children to acquire. Therefore, children might treat ba as a verb at the beginning.
Another example presented in (7) can support this assumption.
(7) Investigator: 修理，啊修理不好是不是？
xiuli, a xiuli bu hao shi bu shi ? fix PAR fix not well yes not yes
‘You did not fit it well, did you?’
Child: 修理很好
xiuli hen hao.
fix very well ‘I fixed it very well.’
Investigator: 修理很好是不是?
xiuli hen hao shi bu shi?
fix very well yes not yes ‘You fix it very well, didn’t you?’
Child: 嗯，就是他把這邊把好了 (HY: 3;4)
en, jiushi ta ba zhebian ba hao le um then he BA here BA well LE
‘Yes, then he fixed this place well.’
In the example (7), we can find that there is no VP in the ba construction. The structure is [subject [ba NP [ba complement]]]. It seems that this sentence is uninterpretable.
However, a replacement of ba by a verb xiu ‘fix’ gives rise to a grammatical sentence:
(8) 他修這邊修好了
ta xiu zhebian xiu hao le he fix here fix well LE
‘He fixed this place well.’
From the examples above, we can, reasonably, have evidence to support our hypothesis that ba is a more verb-like morpheme in child’s language but not a functional category as in adult’s speech.
4.2 Canonical BA Sentence
It has been mentioned in chapter 2 that there are two kinds of ba construction. One is the canonical ba construction, and the other is causative ba construction. In causative ba construction, the subject of the ba sentence is generally inanimate causers, and ba can not form a constituent with the post-ba NP. In our spontaneous speech data of the four children, we can easily observe that we can not find any causative ba sentence until a certain age. The first causative ba sentence found in the spontaneous speech data of JC appears at the age of 3;9. The sentence is presented below:
(9) 他的槍可以把怪獸的劍變成劍 (JC: 3;9)
ta de qiang keyi ba guaishou de jian biancheng jian
he POSS gun can BA monster POSS sword become sword ‘His gun can turn the monster’s sword to another sword.’
Furthermore, to verify whether (9) is a causative ba sentence, we can have a test to examine whether the ba can form a constituent with the post-ba NP. This is illustrated in (10):
(10) *把怪獸的劍，他的槍可以變成劍
*ba guaishou de jian, ta de qiang keyi biancheng jian BA monster POSS sword, he POSS gun can become sword
It seems that this sentence in which we preposed the ba and post-ba NP is odd.
In the following, we present the first causative ba sentence from the other three children, and the age at which the causative ba sentence appeared.
(11) 這很容易把我東西弄下來 (HY: 3;8)
zhe hen rongyi ba wo dongxi nong xialai this very easy BA my thing make fall-down
‘This might make my thing fall down easily.’
(12) 這個圓的剛好可以把她拿掉 (AN: 4:00)
zhege yuan-de ganghao keyi ba ta nadiao this round-POSS happen-to can BA her take-off
‘This round one happens to take her off.’
‘This round one happens to take her off.’