Chapter 4 Proposal
4.2 The syntactic analysis of the X shaowei A Y yi-dianer/yi-xie comparative
4.2.2 Late merger of shaowei ‘slightly’
4.2.2.5 Evidence related to intervention effects on A-not-A operators
Yang (2008) points out that in Mandarin Chinese wh-adverbs and A-not-A operators are subject to what he terms the weak intervention effect while wh-nominals are not.7
(89) a. Ta weishemo cizhi?
he why resign ‘Why did he resign?’
b. *{Suoyouderen/Mei-ge-ren} dou weishemo cizhi?
all.person/every-CL-person all whyadv resign ‘Why did all people/everyone resign?’
c. *{Meiyouren/Henshaoren/Zuiduo liang-ge ren} weishemo cizhi?
nobody/few.person/at.most two-CL person whyadv resign
‘Why did nobody/few people/at most two people resign?’
(90) a. Ta zenmo dun niurou?
he howadv stew beef ‘How did he stew beef?’
b. *{Suoyouderen/Mei-ge-ren} dou zenmo dun niurou?
all.person/every-CL-person all howadv stew beef
‘How did all people/everyone stew beef?’
c. *{Meiyouren/Henshaoren/Zuiduo liang-ge ren} zenmo dun niurou?
nobody/few.person/at.most two-CL person howadv stew beef
7 Yang (2008) points out that there are at least two types of intervention effects in Mandarin Chinese.
One is called the “weak intervention effect” where only the wh-adverbs and A-not-A operators are ruled out whereas the wh-nominals are fine. The other type is called the “strong intervention effect”
where all the wh-elements are ruled out.
‘How did nobody/few people/at most two people stew beef?’
(91) a. Zhangsan qu-bu-qu Taibei?
Zhangsan go-not-go Taipei
‘Will Zhangsan go to Taipei or not?’
b. *{Suoyouderen/Mei-ge-ren} dou qu-bu-qu Taibei?
all.person/every-CL-person all go-not-go Taipei ‘Will all people/everyone go to Taipei or not?’
c. *{Meiyouren/Henshaoren} qu-bu-qu Taibei?
nobody/few.person go-not-go Taipei ‘Will nobody/few people go to Taipei or not?’
(92) a. {Suoyouderen/Mei-ge-ren} dou chi shemo?
all.person/every-CL-person all eat what ‘What did all people/everyone eat ___?’
b. {Meiyouren/Henshaoren/Zuiduo liang-ge ren} gan gen shei dajia? (Soh 2005) nobody/few.person/at.most two-CL person dare with who fight
‘Who is the person x such that nobody/few people/at most two people dare(s) to fight with x?’
The weak intervention effect is reminiscent of the LF-movement property of wh-adverbs and A-not-A operators. Huang (1982) shows that Chinese wh-adverbs are subject to island effects.
(93) a. *Ta xihuan [CP [DP Zhangsan weishemo xie] de shu]?
he like Zhangsan whyadv writeDE book ‘Why does he like the book(s) that Zhangsan writes t?’
b. *Ta xihuan [CP [DP Zhangsan zenmo dun] de niurou]?
he like Zhangsan howadv stew DE beef ‘How does he like the beef that Zhangsan stew t?’
On the other hand, Chinese wh-nominals are not subject to island effects because they do not undergo any LF-movement at all (see Tsai 1994).8
(94) Ta xihuan [CP [DP shei xie] de shu]?
he like who write DE book ‘Who does he like the book(s) that wrote?’
Huang (1982) also shows that the A-not-A question exhibits island effects. He suggests that the A-not-A question involves an LF-moving operator which is subject to island constraints.
(95) * Ta xihuan [CP [DP Zhangsan xie-bu-xie] de shu]?
he like Zhangsan write-not-write DE book ‘Does he like the book(s) that Zhangsan writes or not?’
Since only the LF-moving wh-adverbs and A-not-A operators exhibit intervention effects, it is reasonable to conclude that the weak intervention effect is an outcome of LF-movement.
Yang (2008) argues that the factor which triggers the weak intervention effect is a constraint on locality or minimality with respect to feature movement as illustrated in (96), termed as Minimality Effect. Under the minimalist term, the intervention
8 Tsai (1994) suggests the nominal in-situ wh-construal in Chinese is substantiated through the operator-variable pair in an unselective binding fashion; therefore, no movement should be involved.
effect occurs when feature movement of an interpretable [iF1] of Y is blocked by an intervening Z bearing the same feature [iF1].
(96) Minimality Effect
*[[iF1] X[uF1] …[Z[iF1]… [… Y[__, iF2, iF3] … ]]]
The LF-moving wh-adverbs are subject to feature movement where their Q-feature (or wh-feature) is probed by C and thus undergoes feature movement to C in order to check the uninterpretable feature [uQ].
(97) Chinese wh-adverbs
[CP C[uQ] [IP…wh[iQ,…]…]] → [cp [iQ] C[uQ] [IP…wh[ __,…]…]]
It is then during the feature movement that an intervening scope-bearing element bearing a feature of the same type, i.e., operator feature [Op], blocks the feature movement, hence the intervention effect.
(98) *[CP [iQ] C[uQ] [IP Scope-bearing Element[Op] ... ‘why/how/A-not-A’[ __, …] ...]]
We will now demonstrate that the intervention effect on the A-not-A operator occurs in the X shaowei A Y yi-dianer/yi-xie comparative too. The intervention effect on the A-not-A operator existing in the X shaowei A Y yi-dianer/yi-xie comparative receives a natural explanation if the covert quantificational operator binding
yi-dianer/yi-xie ‘a little’, which is treated as a variable, is allowed to adjoin in a scope position, followed by late merger of shaowei ‘slightly’ with the covert quantificational operator. Therefore, the intervention effect on the A-not-A operator existing in the X shaowei A Y yi-dianer/yi-xie comparative can be taken as evidence for the proposal that shaowei ‘slightly’ can be merged late.
Consider the example in (99a), in which the A-not-A constituent gao-bu-gao
‘tall-not-tall’ is involved. The LF-moving A-not-A constituent gao-bu-gao
‘tall-not-tall’ is subject to feature movement where its Q-feature is probed by C and thus undergoes feature movement to C in order to check the uninterpretable feature [uQ]. It is then during the feature movement that the covert quantificational operator binding yi-xie ‘a little’, which is treated as a variable, blocks the feature movement, as the simplified skeletal structure of (99a) shown in (99b) illustrates. Therefore, the ungrammaticality of (99a) is directly predicted by our proposal.
(99) a. *Wangwu shaowei gao-bu-gao Zhaoying yi-xie?
Wangwu slightly tall-not-tall Zhaoying a-little
b. *[CP [iQ] C[uQ] [TP Wangwu OPi shaowei gao-bu-gao[ __, …] Zhaoying yi-xiei]]?
In a word, the intervention effect on the A-not-A operator exisitng in the X shaowei A Y yi-dianer/yi-xie comparative provides strong support for the proposal that shaowei ‘slightly’ is merged countercyclically as the complement of the covert quantificational operator binding yi-dianer/yi-xie ‘a little’, which is treated as a variable, after the covert quantificational operator adjoins in a scope position.