Tải bản đầy đủ - 0 (trang)
3 The cases of cross-categorial disharmony in Chinese – what you see is what you get

3 The cases of cross-categorial disharmony in Chinese – what you see is what you get

Tải bản đầy đủ - 0trang

Free ebooks ==> www.ebook777.com

The cases of cross-categorial disharmony in Chinese: What you see is what you get | 305

(2) Selection of correlation pairs from Dryer (1992: 108, table 39; 2009:186,

table 1)3

(order changed and subdivision added for ease of exposition)

Verb patterner

Group 1:

a. verb

b. copula verb

c. negative auxiliary

Group 2:

d. adposition

e. adjective

f. verb

g. verb

Group 3:

h. noun

i. noun

j. complementizer

k. question particle

l. adverbial subordinator

Object patterner





standard of comparison


manner adverb

relative clause





Chinese is “well-behaved” with respect to the first group. This “harmony” is

not surprising, though, because in fact it does not go beyond the format of the

VP. The ordered pair ‘verb object’ in (2a) is not a correlation pair, but instead

serves as the standard of comparison for the other categories. The pair (2b)

‘copula – predicate’ (cf. [4]) can in turn be subsumed under (2a), the copula just

being a particular type of verb. The pair (2c) ‘negative auxiliary – VP’ (cf. [3], [4])

still refers to the order within the verbal projection and therefore does not illustrate cross-categorial harmony in the strict sense, either.


3 The correlation pairs have remained stable over nearly thirty years, modulo the absence in

Dryer (2009) of the pair ‘verb subject’, exemplified by (There) entered a tall man in Dryer (1992:

108). Since in the corresponding construction in Chinese the unique (internal) argument of the

verb is also to its right (cf. [i]) and on a par with ‘verb object’ order, the (non-)inclusion of this

correlation pair does not change the picture we obtain for Chinese.


Lái -le


come-PERF guest

‘Guests have arrived.’


Free ebooks ==> www.ebook777.com

306 | Chinese from a typological point of view: Long live disharmony!


Tā (méi) dǎsǎo fángzi

3SG NEG sweep room

‘He has (not) cleaned the room.’


Tā (bù) shì fǎgrén

3SG NEG be French

‘She is (not) French.’

8.3.2 Where Chinese is harmonic and disharmonic at the same time

Chinese is partly well-behaved with respect to the correlation pairs in Group 2

([2d] – [2g]). The restriction “partly” is necessary, because in all cases, the opposite order is likewise observed. While prepositions pattern with verbs in taking

their complement to the right (cf. [5a]), postpositions do not (cf. [5b]). Furthermore, in the so-called transitive comparative (cf. Erlewine 2007) the standard of

comparison (here Lǐsì) indeed follows the adjective (cf. [6a]) and thus qualifies

as object patterner, but in the comparative construction with bǐ ‘compared to’,

the standard of comparison precedes the adjective (cf. [6b]). Finally, when arguments, PPs follow the verb, on a par with object NPs (cf. [7a]), but are confined to preverbal position when having adjunct status (cf. [7b]).


a. Tā [PP wàng nán] zǒu -le ]


toward south go -PERF

‘She went towards the south.’

b. Wǒ [PostP chúxī

yǐqián] yào huí



New.Year’s eve before need return home

‘I need to go home before New Year’s eve.’


a. Tā gāo Lǐsì shí gōngfēn

3SG tall Lisi 10 cm

‘He is ten centimeters taller than Lisi.’

b. Tā bǐ

Lǐsì gāo (*bǐ

Lǐsì) shí gōngfēn

3SG compared.to Lisi tall

compared.to Lisi 10 cm

‘He is ten centimeters taller than Lisi.’


a. Tā jì

-le yī -ge

3SG send-PERF 1 -CL



[PP gěi Měilì]

to Mary

Free ebooks ==> www.ebook777.com

The cases of cross-categorial disharmony in Chinese: What you see is what you get | 307

‘He sent a parcel to Mary.’

b. Tā [PP gěi péngyou] chàng ge gē (*[PP gěi péngyou])


to friend

sing CL song

to friend

‘He sings a song for his friends.’

8.3.3 Necessary digression on manner adverbs in Chinese

The correlation pair ‘verb – manner adverb’ stated in (2g) is not easy to evaluate

for Chinese, either. First, as already observed above for the other phenomena in

group 2, which involve opposite orderings, manner adverbs can occur in both

preverbal and postverbal position:


Tā mànyōuyōude zǒu yī quān

3SG leisurely

walk 1 round

‘She walks around at a leisurely pace.’


Tā hěn dàfāngde z -le zìwǒjièshào

3SG very natural

make-PERF self.introduction

‘She introduced herself very naturally.’


Tā zǒngshì chī de tài kuài

3SG always eat DE too fast

‘He always eats too fast.’

Second, as observed by Ernst (1994: 48), adverbs in preverbal position can be

ambiguous between a subject-oriented reading (i) and a strict manner reading



Tāmen hěn bùlǐmàode dui

lǎoshī shuō huà


very impolite towards teacher speak word

(i) ‘Impolitely/rudely, they spoke to the teacher

(ii) ‘They spoke to the teacher impolitely/rudely.’

Under the first reading (11i), it was rude of the students to speak to the teacher

at all, irrespective of the manner used, whereas under the second reading (11ii),

the manner itself used when addressing the teacher was rude. By contrast, an

adverb in postverbal position is not ambiguous and only allows for the strict

manner reading, as again pointed out by Ernst (1994: 48):


Free ebooks ==> www.ebook777.com

308 | Chinese from a typological point of view: Long live disharmony!


Tāmen duì

lǎoshī shuō de hěn bùlǐmào


towards teacher speak DE very impolite

‘They spoke to the teacher impolitely/rudely.’

It is evident that the choice made here will directly influence the picture obtained for Chinese. When only counting the postverbal non-ambiguous manner

instances, manner adverbs pattern with objects and nicely fit in with the “expected” harmonic picture; if, however, both pre- and postverbal manner adverbs are included, the picture obtained will be much less neat.

The third problem related to manner adverbs in Chinese which makes it difficult to obtain a clear result for their role in a word order typology is the lack of

a precise analysis for ‘de XP’ in postverbal position, including the exact status

of de.4 While the best translational equivalent is indeed a manner adverb, there

exist quite a few properties challenging the standard analysis of ‘de XP’ as

manner adverb (cf. Ernst 2002 and references therein for an adverb analysis in

terms of right adjunction).

First, only predicative adjectives (e.g. duì ‘right’, zìrán ‘natural’) are

acceptable following de, to the exclusion of non-predicative adjectives (e.g. cuò

‘wrong’, tiānrán ‘natural’) and verbs ((cf. [13] and [14]). The opposition between

predicative and non-predicative adjectives is illustrated in (15) (also cf. the

discussion in chapter 5.1.1 above.).


Tā cāi

de d /*c

3SG guess DE right/wrong

‘She guessed right/wrong.’


Tā hdá de hěn zìrán / *tiānrán

3SG answer DE very natural/ natural

‘He answered very naturally.’


4 The lack of a precise analysis of de is in general covered up by hyphenating it with the preceding verb, as is the convention for aspect suffixes, and thus presenting it as part of the verb,

as in e.g. Tā cāi-de duì (cf. [13]). My glossing it as DE and assigning it the status of a functional

head (cf. below) is only a first preliminary step and illustrates the necessity for further research. Note that in the following I limit myself to the so-called descriptive complement, to the

exclusion of the result/extent complement, which has the same surface form ‘de XP’ and is

analysed as a head-intial CP by Huang (1982: 96, footnote 15). For further discussion of the

different postverbal ‘de XP’ types, cf. among others Y.-H. Audrey Li (1990, ch. 3), Sybesma

(1991a; 1999a, ch. 2), C.-C. Jane Tang (2001).

Free ebooks ==> www.ebook777.com

The cases of cross-categorial disharmony in Chinese: What you see is what you get | 309


Tā de kànfǎ { bù d / hěn zìrán }/{*c /*tiānrán}

3SG SUB opinion NEG right/very natural/ wrong/ natural

‘His point of view is not correct/ is natural / is wrong.’

Under an analysis of ‘de XP’ as manner adverb, the restriction to exclusively

predicative adjectives for XP is completely unexpected. This constraint can,

however, be captured by analysing the adjectival phrase (AP) as a complement

selected by de as a head, where the resulting de-phrase is in turn selected by the


Second, this new analysis can also account for the obligatory adjacency

between the de-phrase ‘de XP’ and the preceding verb as well as for that

between de and the following AP; hence both positions for wèntí ‘question’ are

bad in (16). This property remains mysterious under an analysis as adverb.


a. Tā hdá (*wèntí) de (*wèntí) hěn zìrán

3SG answer question DE question very natural

‘He answered the question very naturally.’

b. Tā hdá (*-le) de hěn zìrán

3SG answer -PERF DE very natural

‘He answered very naturally.’

Third, unlike adverbs (cf. [19]), the AP following de can be negated ([17a]),

questioned in the ‘A-bù-A’ form ([18]) (cf. C.-T. James Huang 1988b, Y.-H.

Audrey Li 1990:45, among others) and modified by adverbs ([17b]), thus

providing further evidence in favour of its predicate status and against its

adverbial status:


a. Tā shuō [de [AP bù qīngchǔ]]

3SG speak DE

NEG clear

‘He doesn’t speak clearly.’

b. Tā shuō [ de [AP[PP bǐ

nǐ] [AP gèng qīngchǔ]]]

3SG speak DE

compared.to 2SG

more clear

‘He speaks even more clearly than you.’


Tā chàng de [ dàshēng bù dàshēng]?

3SG sing DE loud

NEG loud

‘Does she sing loudly?’


Free ebooks ==> www.ebook777.com

310 | Chinese from a typological point of view: Long live disharmony!


* Tā [mànyōuyōude] bù [mànyōuyōude] zǒu yī quān?

3SG leisurely

NEG leisurely

walk 1 round

(cf. [8] above)

I would therefore like to propose that the AP is a predicative projection which

denotes a subevent that enters into the composition of a complex predicate with

the matrix verb: ‘V de AP’. This not only accounts for the syntactic properties

just described, but also for the strict manner interpretation observed for

postverbal ‘de AP’ (cf. [12] above), which contrasts with the availability of both a

strict manner interpretation (ii) and a subject-oriented reading (i) for preverbal

adverbs (cf. [11] above).5

This short digression on manner adverbs in Chinese reveals two major

sources of problems, apparently neglected by word order typology as it is

currently practiced. One is the possibly insufficient state of knowledge of the

language at hand, which makes it impossible to establish a correlation pair, the

phenomena involved simply not having been studied enough (as e.g. ‘de XP’ in

Chinese). The other problem is directly linked to the general format imposed by

correlation pairs aiming at testing cross-categorial harmony (X either precedes

or follows X). This format leaves no room for semantic ambiguities displayed by

particular items in a given position, as observed for Chinese manner adverbs in

the preverbal vs the postverbal position (assuming for the sake of the argument

adverbial status for ‘de XP’ here). To my knowledge, these not infrequent cases

where no 1:1 relationship between form and meaning exists have not been discussed explicitly in word order typology; nor has any heuristic device been

proposed of how to deal with them, i.e. whether and how to count them. The

same critique applies to the much more straightforward cases where no subtle

semantic differences are involved, but where simply two opposite orders are

possible within the same language; once again, one is at a loss which phenomenon to count. Needless to say, the temptation to exclusively count the

harmonic one and to discard the disharmonic one is great.


5 In fact, C.-T. James Huang (1992) already proposed a complex predicate analysis for ‘V de AP’

(although for reasons different from those presented here); apparently, this was not taken up

by subsequent studies of adverbs. Also cf. Y.-H. Audrey Li (1990, ch. 3) and C.-C. Jane Tang

(1990, ch. 4) for some of the observations integrated into the analysis presented here.

Free ebooks ==> www.ebook777.com

The cases of cross-categorial disharmony in Chinese: What you see is what you get | 311

8.3.4 Where Chinese is disharmonic throughout

Let us now turn to the correlation pairs in group 3. Here Chinese shows exactly

the opposite order of the one predicted for a SVO language and displays instead

the cross-categorial disharmony which is so often cited in the literature. The nominal projection

In contrast to the orders ‘noun – relative clause’ and ‘noun – genitive’ expected

for a SVO language (cf. the correlation pairs [2h] and [2i] above), in the Chinese

nominal projection all modifying elements as well as relative clauses and complement clauses precede the NP.


yī jiàn zāng/ gānjìng yīfu

1 CL dirty/ clean dress

‘a dirty/pretty/clean dress’


yī ge [NP[NP hēi

] yīguì]

1 CL

black lacquer wardrobe

‘a black-lacquered wardrobe’


[DP Měilì/ tāmen] de péngyou

Mary/ 3PL

SUB friend

‘Mary’s friend/their friend’


[PP d

wèntí ] de kànfǎ

towards problem SUB opinion

‘an opinion about the problem’


[DP zhèxiē [TP Øi mǎi xiǎo qìchē] de réni]


buy small car

SUB person

‘the persons who bought a small car’


[DP[TP Bālí xià xuě ] de xiāoxi].

Paris fall snow SUB news

‘the news that it is snowing in Paris’

(Fan Jiyan 1958: 215)

(Lü 2000 [1980]: 157)

As discussed in chapter 5.1.3 above, under certain circumstances the subordinator de can be absent and the adjectival or nominal modifier can be simply juxtaposed with the head noun (cf. [20] and [21]).


Free ebooks ==> www.ebook777.com

312 | Chinese from a typological point of view: Long live disharmony! The head-final CP

Of the remaining three “exceptions” to the word order predicted for a SVO language, two cases, i.e. (2j) and (2k), reduce to the unexpected, hence disharmonic head-final character of the CP in Chinese; the order ‘clausal complement

– complementiser’ is “unexpected” insofar as here the complementiser visibly

does not pattern with the verb.

As argued for in chapter 7, in the light of Rizzi’s (1997) split CP it makes

sense to extend the notion of complementisers from exclusively subordinating

items such as that and whether in English to the so-called sentence-final particles (SFP) in matrix sentences in Chinese, among them the yes/no-question

particle ma.


[ForceP[TP Tā huì chàng gē] ma]?

3SG can sing song FORCE

‘Can he sing?’

As a consequence, Dryer’s (1992) “question particle” involves a C element as

well and the relevant correlation pair (2k) can therefore be subsumed under (2j)

predicting the order ‘complementiser – sentential complement’ for SVO languages. Recall that chapter 7 also provided evidence for de in the propositional

assertion and dehuà in conditional clauses as exclusively non-root complementisers, thus consolidating the head-final character of the Chinese CP in both

matrix and embedded contexts.

The disharmony between SVO order and head-final CP displayed by Chinese is all the more significant as Dryer (1992: 102), referring to his own work

(Dryer 1980) as well as Hawkins (1990: 225), concludes that “[…] in fact it may be

an exceptionless universal that final complementizers are found only in OV

languages. […] complementizers are therefore verb patterners, while the Ss they

combine with are object patterners.”6 This is confirmed in Dryer (2009, table

[24]) where no case of sentence-final C for the 140 VO languages examined is

attested.7 Unfortunately, Dryer (2009) only indicates language genera;

accordingly, there is no way to know whether Mandarin Chinese or any other


6 This goes back to Greenberg’s (1963: 81) universal 9: “With well more than chance frequency, when question particles or affixes are specified in position by reference to the sentence

as a whole, if initial, such elements are found in prepositional languages, and, if final, in postpositional.” Recall that in general VSO languages and SVO languages are associated with

prepositions, and SOV languages with postpositions.

7 Explicit reference is made to English that as illustrating a clause-initial C and to Japanese to

as illustrating a clause-final C, respectively.

Free ebooks ==> www.ebook777.com

The cases of cross-categorial disharmony in Chinese: What you see is what you get | 313

Sinitic language was included under the very vast genus ‘South-East-Asian and

Oceanic languages’ in this survey. (Note that Chinese is not included in the

database used in Dryer 1992). Dryer’s (1992, 2009) unwieldy adverbial subordinator

Finally, the last correlation pair (2l) ‘adverbial subordinator – sentence’ (as in

Dryer’s example because Bob left) cannot be directly transposed to Chinese,

because the term adverbial subordinator is very vague and turns out to involve

several different categories. This holds not only for Chinese, but for other languages as well. In English, for example, items with lexical content such as

before, after are in general analysed as prepositions (selecting an NP or a

clausal complement), in contrast to that and if analysed as complementisers.8

However, both groups of items would probably be considered as falling under

the pre-theoretical labels adverbial subordinators or subordinating conjunctions.

The question as to what items can count as possible equivalents of “adverbial subordinators” in Chinese leads us to another poorly understood domain in

Chinese syntax. While the following section can evidently not accomplish an indepth analysis, the discussion should suffice to show that whatever categories

turn out to be included under the cover term “adverbial subordinator”, they are

all clearly different from the various types of complementisers realized by sentence-final particles (SFP) in Chinese (cf. chapter 7).

In fact, just as in English, the Chinese candidates for subordinating conjunctions such as yàoshi ‘if’, rúguǒ ‘if’, suīrán ‘although’, jìrán ‘since’, yīnwèi

‘because’, zìcóng ‘since (temporal)’ do not represent a homogeneous group, but

include (sentence-level) adverbs on the one hand and prepositions on the other.

As Lu Peng (2003, 2008) has argued in great detail, rúguǒ/yàoshi ‘if’, suīrán

‘although’, and jìrán ‘since’ are sentence-level adverbs on a par with e.g. xiǎnrán ‘obviously, naturally’ and xìnghǎo ‘fortunately’; like adjunct NPs and PPs

they can occupy either the TP-external or the TP-internal topic position (Spec,

TopP) (cf. chapter 6). For reasons of space, this will be shown only for the pair

xìnghǎo ‘fortunately’ and rúguǒ ‘if’. (For further discussion, cf. Lu Peng 2003,

2008: §3.2.)9


8 Prepositions in English behave differently from C such as that, if in that they may allow

sluicing. (Thanks to John Whitman for pointing this out to me.)


I left before Bill left, but Jane left after [e]


* I know that Bill left, but Jane doesn’t know that/whether [e]

9 C.-T. James Huang (1982: 85) left open the P vs C status of items such as yīnwèi ‘because’,

concentrating on the head-initial character of their projection. Note that he analysed rúguǒ ‘if’


Free ebooks ==> www.ebook777.com

314 | Chinese from a typological point of view: Long live disharmony!


a. [ext.TopP Xìnghǎo [TP wǒ [int.TopP [ nà fù huà]

[AspP mài-le

fortunately 1SG

that CL painting


ge gāo jià]]]]

CL high price

‘Fortunately, I sold that painting at a high price.’ (Lu Peng 2008: 164)

b. [ext.TopP [ Nà fù huà ] [TP wǒ [int.TopP xìnghǎo [AspP mài-le

that CL painting




ge gāo jià]]]]

CL high price

‘That painting, I fortunately sold at a high price.’


[ext.TopP [ Nà fù huà] [ext.TopP xìnghǎo [TP wǒ [AspP mài-le

that CL painting

fortunately 1SG


ge gāo jià]]]]

CL high price

‘That painting, fortunately, I sold it at a high price.’

As illustrated in (27a) and (27b), xìnghǎo ‘fortunately’ as a sentential adverb can

occur either in the external or the internal topic position to the left or the right

of the subject, respectively. Furthermore, both the DP nà-fù huà ‘that painting’

and the adverb xìnghǎo ‘fortunately’ can co-occur in the external topic positions

(cf. [27c]), in either order: nà fù huà, xìnghǎo,…or xìnghǎo, nà fù huà, …

The same holds for both items in the TP-internal topic positions, where they

are likewise interchangeable:


a. [TP Wǒ [int.TopP xìnghǎo [int.TopP [ nà fù huà ] [AspP mài-le



that CL painting


ge gāo jià]]]]

CL high price

b. [TP Wǒ [int.TopP [nà fù huà ] [int.TopP xìnghǎo [AspP mài-le


that CL painting



ge gāo jià]]]]

CL high price

‘I fortunately sold that painting at a high price.’


and suīrán ‘although’ as P/C-heads on a par with yīnwèi ‘because’, an analysis which remained

unchallenged up to Lu Peng’s (2003) dissertation.

Free ebooks ==> www.ebook777.com

The cases of cross-categorial disharmony in Chinese: What you see is what you get | 315

(29) below shows rúguǒ ‘if’ to have the same distribution as xìnghǎo ‘fortunately’. It can precede or follow the subject in the conditional clause; when to

the right of the subject, it is interchangeable with an internal topic DP, both

occupying the internal TopP. (Recall from chapter 6.1.1 that the external topic

position is the default position for a conditional clause.)


a. [ext.TopP[cond.clause Rúguǒ [TP nǐ [int.TopP [DP yīngyǔ kǎoshì] [AuxP néng



English exam


kǎo ge dìyī ]]]] [TP wǒ jiù jiǎnglì nǐ yī liàng xīn zìxíngchē]]

pass CL first

1SG then award 2SG 1 CL new bicycle

‘If in the English exam you can pass as first, I’ll reward you with

a new bicycle.’

b. [ext.TopP[cond.clause Nǐ [int.TopP rúguǒ [int.TopP [DP yīngyǔ kǎoshì] néng



English exam can

kǎo ge dìyī]]]…

pass CL first

‘If in the English exam you can pass as first,…’


[ext.TopP [cond.clause Nǐ [int.TopP [DP yīngyǔ kǎoshì] [int.TopP rúguǒ néng


English exam



kǎo ge dìyī]]]…

pass CL first

‘If in the English exam you can pass as first,…’

Accordingly, rúguǒ ‘if’ is not a head and the following clause is not its complement. Instead, rúguǒ is a sentence-level adverb which shows the same distribution as adjunct NPs and PPs, viz. it occupies the specifier of the TP-external or

TP-internal TopP. 10 (Note, though, that adjunct NPs and PPs can also occur to

the right of auxiliaries, a position excluded for sentence-level adverbs.)

By contrast, yinwèi ‘because’, zìcóng ‘since (temporal)’ etc. are prepositions,

i.e. heads and must therefore always precede their complement clause. Note


10 While semantically the sentence-level adverb rúguŏ ‘if’ may fulfill a function similar to that

of the non-root C dehuà, it clearly belongs to a different syntactic category, as witnessed by the

co-occurrence of the two:


[ClowP[TopP[CP(-root) Rúguǒ tā lái

dehuà] [TP wǒ jiù bù cānjiā huìyì ]


1SG then NEG attend meeting CLOW


3SG come C(-root)

‘If he comes, then I won’t attend the meeting.’


Tài liệu bạn tìm kiếm đã sẵn sàng tải về

3 The cases of cross-categorial disharmony in Chinese – what you see is what you get

Tải bản đầy đủ ngay(0 tr)