Lin (2009)

Starting with an overview of superiority comparatives in English as well as some basic

data of Chinese bi-comparatives presented in Tsao (1989), Lin (2009) reviews and argues

against several previous studies on comparatives first. Then, he discusses the syntax of

bi-comparatives and the semantics of bi. Since this thesis mainly focuses on the syntactic structure

of bi-comparatives, I will skip the semantic part and summarize Lin’s syntactic analysis in

detail in the following paragraphs.

To begin with, in order to explain the difference in grammaticality between (30) on the

one hand and (31) & (32) on the other hand, Lin (2009) claims that bi-comparatives in

Mandarin Chinese is an argument-dependent comparison; in other words, both the target of

comparison and the standard of comparison have to be arguments of the gradable predicate.

This contrast is shown below:

(30) *Zhe-zhang zhuozi bi na-zhang zhuozi chang kuan

this-CL table BI that-CL desk long wide

Intended: ‘This table is wider than that desk is long.’

(31) [CP Ni qu] bi [CP wo qu] hao

you go BI I go good

‘It’s better for you to go than for me to go.’

(32) [CP Ta zai tushuguan nian shu] bi [CP ta zai jiali nian shu] renzhen

he at library read book BI he at home read book serious

‘He studies more seriously in the library than he does at home.’

According to Lin (2009), sentence (30) is ungrammatical because the two compared items are

not arguments of the predicate. It is quite difficult to treat na-zhang zhuozi chang ‘that desk is

long’ as the argument of kuan ‘wide’. As for (31), both CPs are arguments of the predicate hao

‘good’, which can take sentential subjects. At first glance, the example in (32) seems not to fit

Lin’s analysis. However, following Davidson (1967), Lin (2009) suggests if we assume that

verbs can have event arguments, the two CPs in (32) can be regarded as the event arguments

of the predicate renzhen ‘serious’. In this way, that (32) is grammatical can be explained with

Lin’s proposal.

Based on the contrast above, Lin proposes the Argument requirement of Chinese

comparative, as presented in (33).

(33) Argument requirement of Chinese comparatives

In Mandarin Chinese, compared constituents must be arguments of a gradable predicate of comparison.

With this requirement, Lin (2009) successfully accounts for the fact that adjuncts such as

manner adverbs and reason clauses are not comparable, as shown in (34).

(34) a. *Wo-de shengri hui manmandi bi kuaikuaidi dao

my birthday will slowly BI quickly arrive

Intended: ‘My birthday will come more slowly than quickly.’

b. *Mama yinwei Xiaoming shuohuang bi yinwei ta tou

mother because Xiaoming say-lie BI because he steal

qian geng shengqi

money more angry

Intended: ‘Mother was angry more because Xiaoming told a lie than because he stole

money.’

Furthermore, another point worth mentioning in Lin (2009) is that he treats time and location

expressions as arguments of the main predicate. Therefore, times and locations are comparable,

as illustrated in (35). According to Lin (2009), this proposal gains support from the phenomena

of Chinese wh-extraction discussed in Tsai (1994). Besides, with respect to the issue of tense

and aspect, time is often considered an argument of the predicate, which is demonstrated in Lin

(2003, 2006).

(35) a. Ta jintian bi zuotian shufu

he today BI yesterday feel-good

‘He feels better today than he did yesterday.’

b. Ta zai xuexiao bi zai jiali kuaile

he at school BI at home happy

‘He is happier at school than at home.’

After proposing the Argument requirement of Chinese comparatives, Lin (2009) presents a

syntactic structure for bi-comparatives which can be seen as a variant of the structures proposed

in Xiang (2005) and Erlewine (2007). In this proposal, bi is analyzed as a functional Degree

head, and it can move to other Degree heads. Those Degree heads together with their

complements and specifiers form a DegP-shell, which serves as an adjunct adjoined to the

predicate of comparison. With this structure, the triple-topic comparison¹² in Tsao (1989) can

be well-explained, as shown in (36).

(36)

Here, we can see that bi is base-generated under the lowest Degree head, and then undergoes

12 Tsao (1989) adopts a topic-comment approach to analyze bi-comparatives. It is claimed that both the target of comparison and the standard of comparison should be topics of an equal rank. According to Tsao (1978, 1979,1982), we can make a distinction between primary and non-primary topic based on some qualities shared by primary topics in Chinese. The term “triple-topic comparison” refers to comparative sentences that involve three pairs of topics, one primary one and two non-primary ones. An example of triple-topic comparison is given in (i) below:

(i) Ta zuotian zai xuexiao bi wo jintian zai jiali kaixin he yesterday at school BI I tody at home happy

‘He was happier yesterday at school than I am today at home.’

two head-to-head movements to the highest Degree head. Lin (2009) analyzes bi as a dyadic

degree operator. This means that bi can quantify over more than one indefinite, just like an

adverb of quantification. Thus, Lin’s proposal can be dubbed a dyadic DegP-shell analysis.

To summarize, Lin (2009) is an advocate of the Direct Analysis since there is only one

token of the gradable predicate and that no comparative deletion occurs in the derivation.

However, although Lin’s (2009) proposal can adopt a phrasal analysis to deal with sentences

of triple comparison that are often regarded as instances of clausal comparatives, it is not

without problems.

As pointed out in Liu (2011), Lin’s analysis will encounter the following three problems.

First of all, the comparison between two reason clauses in bi-comparatives (example (37)) are

grammatical and acceptable for the 23 native speakers Liu consulted.

(37) Mama yinwei Xiaoming shuohuang bi baba yinwei ta tou mother because Xiaoming say-lie BI father because he steal

dongxi haiyao shengqi things even angry

‘His mother gets angry more because Xiaoming lies than his father gets angry because he steals things.’

I also checked with some native speakers about the sentence in (37). Most of them consider

(37) grammatical and acceptable. Only few of them think (37) is grammatical but unnatural. In

this way, Lin’s (2009) Argument requirement of Chinese comparatives is challenged since

non-arguments like reason clauses seem to be comparable.

The second problem of Lin (2009) presented in Liu (2011) concerns the following

example, which is an object-preposing construction.

(38) *Zhangsan shuxue bi Lisi wuli xihuan

Zhangsan math BI Lisi physics like

‘Zhangsan likes mathematics more than Lisi likes physics.’

Since both Lisi and wuli ‘physics’ are arguments of the predicate xihuan ‘like’, the sentence in

(38) will turn out to be grammatical under Lin’s analysis. However, according to Liu (2011),

this sentence is ungrammatical. Again, the Argument requirement of Chinese comparatives

fails to account for this example.

Thirdly, Liu (2011) also claims that pivotal constructions in Chinese pose challenges to

Lin’s (2009) proposal. The example given in Liu (2011) is shown in (39).

(39) Wo rang ni bi ni rang ta duo zhu-le san-tian

I let you BI you let he many live-ASP three-day

‘I let you stay for three more days than you let him.’

As suggested in Tang (2010), pivotal verbs select a VP as their complement. Following this

logic, it is impossible for the predicate of comparison duo zhu-le san-tian ‘stay for three more

days’ to take the whole post-bi constituent ni rang ta ‘you let him’ as its external argument.

Therefore, Lin’s (2009) analysis will encounter problems in explaining pivotal constructions

in bi-comparatives.