• 沒有找到結果。

漢語動詞及形容詞的質與量

N/A
N/A
Protected

Academic year: 2021

Share "漢語動詞及形容詞的質與量"

Copied!
79
0
0

加載中.... (立即查看全文)

全文

(1)

國 立 交 通 大 學

外國文學系外國文學與語言學碩士班

士 論 文

漢語動詞及形容詞的質與量

The Quality and Quantity Readings of Verbs and

Adjectives in Mandarin Chinese

研 究 生:劉晉廷

Chin Ting Jimbo Liu

指導教授:劉辰生 教授

Prof. Chen-Sheng Luther Liu

(2)

漢語動詞及形容詞的質與量

The Quality and Quantity Readings of Verbs and

Adjectives in Mandarin Chinese

研 究 生:劉晉廷

Chin

Ting

Jimbo

Liu

指導教授:劉辰生 教授

Prof. Chen-Sheng Luther Liu

國立交通大學

外國語文學系外國文學與語言學碩士班

碩士論文

A Thesis

Submitted to Department of Foreign Languages and Literatures

Graduate Institute of Foreign Literatures and Linguistics

National Chiao Tung University

In partial Fulfillment of the Requirements

For the Degree of Master in Arts

Graduate Institute of Foreign Literatures and Linguistics

July 2012

Hsinchu, Taiwan, Republic of China

(3)

漢語動詞及形容詞的質與量

研 究 生:劉晉廷 指導教授:劉辰生 教授

國立交通大學外國文學與語言學研究所

中文摘要

根據Bolinger(1972)對程度詞及Kennedy and McNally (2010)對顏色詞的的研究, 本文主要探討漢語動詞及形容詞的「質」與「量」的語意。藉由兩個漢語的動詞 「透」與「遍」,本文主張「透」可以表達漢語動詞及形容詞中「質」的語意, 而「遍」可以表達漢語動詞中「量」的語意。在分析前人的研究後(Liu, 2011), 本文進一步指出漢語的形容詞也帶有「量」語意。本文的觀察為Bolinger (1972) 的主張提供佐證;也就是說,程度性(gradability)-包念「質」與「量」-為動詞 及形容詞皆具有的特性。

(4)

The Quality and Quantity Readings of Verbs and Adjectives

in Mandarin Chinese

Student: Chin Ting Jimbo Liu Advisor: Prof. Chen-Sheng Luther Liu

Graduate Institute of Foreign Literatures and Linguistics National Chiao Tung University

ABSTRACT

Following Bolinger’s (1972) study on degree words as well as Kennedy and McNally’s (2010) research on color quality and quantity, this study aims to investigate the quality and quantity interpretations of verbs and adjectives in Mandarin Chinese by examining two Chinese lexical verbs-tou, a verb meaning penetrate or through in English, and bian, a verb meaning distribute or all over in English. It is argued that

tou induces the quality readings of verbs and adjectives in Mandarin Chinese, while bian quantizes the subsequent noun phrase (NP) and expresses the quantity reading of

verbs. From a cross-linguistic point of view, this study further claims that quality and quantity are possibly primitive elements in our Universal Grammar (Chomsky, 1981).

(5)

ACKNOWLEDGEMENTS 在過去三年碩士班的求學過程中,要感謝的人太多了。我會盡力把最後這個部份 做好。我要謝謝帶領我探索語言學這個美好世界的師長們、一直不停和我討論並 交換意見的同學和學長姊、學弟妹們、總是給我最大包容及體諒的工作伙伴們、 常常鼓勵我並聽我分享的朋友們以及永遠做為我後盾的家人們。 首先,我要謝謝我的指導教授劉辰生老師。上老師的課以及和老師討論時,總能 感受到老師對學術研究及教學的熱情。除了謝謝老師在過去三年的指導外,也要 謝謝老師在做人處事上的教導,讓我獲益良多。我期許自己未來也能和老師一 樣,總是享受學術研究的過程並成為一位受學生尊重的老師。我非常感謝我的口 試委員林若望老師。老師課堂中的細心講解,讓形式語意學變得非常有趣,也謝 謝老師在資格考及口試時給我的寶貴意見,雖然有很多現象我無法立即找出解釋 方式,但我會把這些問題當成我未來研究的動力。我要謝謝另一位口試委員廖秀 真老師。謝謝老師給我很多練習發表的機會。每當遇到問題時,老師總是願意給 我方向,幫助我解決眼前的難題,謝謝老師的耐心指導。謝謝清華大學語言所的 蘇宜青老師。上了老師開的課後,讓我對心理語言學和語言習得的研究更有興 趣。謝謝雲林科技大學應用外語系的楊孝慈老師。老師從二技到研究所這一路上 給我很多指導以及提點,並時時關心我的生活。沒有老師的幫忙,我不會有機會 考上交大外文所。最後我要感謝在交大這三年教過我的老師,劉美君老師、許慧 娟老師、賴郁雯老師、呂菁菁老師以及林惠芬老師。妳們的課讓我看到語言學不 同的面貌,也讓我對這門學問更著迷。 過去這三年,因為有很多同學、學長姊以及學弟妹的幫忙,讓我的求學過程更順 利、更有趣。謝謝蕭宇吟同學。每次看到妳面對事情的方式,總提醒我以後遇到 問題要以更開闊的視野去面對。謝謝妳在論文撰寫過程中的討論、幫忙以及鼓 勵。每當過程中遇到困難時,我第一個想到的就是和妳討論。妳總是會以妳的經 驗告訴我可能的解決方法,並時時提醒我不用想太多。真的很謝謝妳,也祝妳未 來的校園生活一切順利。我由衷感謝蔡慧瑾博士及陳奕勳學長在碩士論文初期和 我的討論。謝謝你們不吝嗇的提供你們的研究經驗,並指引我論文可能的發展方 向。我的同班同學徐睿良及吳聲弘在過去三年中也給我很大的幫助。在修課過 程、做報告、討論及生活上,有你們的陪伴及體諒,使得這三年的研究生生活過 得很精彩。這本碩士論文能順利的產出也要感謝學長黃皓志、學姊林郡儀、蘇琬淇、 學妹朱育瑩、陳俞汶、蔡幸珊、常鈺、黃慧婷、余素幸、鄭詅月、李靜汶以及學 弟倪孝勇。每當我需要幫忙或有疑問時,你們的熱心協助與解答大大地減輕了我 的焦慮及不安。在此,我要表達我對系辦助理陳雅玲以及陳旅櫻的謝意。每當我 要出國發表時,妳們永遠盡全力幫我找各個不同的補助來源,讓我能順利出國參 加研討會。每當我有任何問題時,妳們總是很有耐心的一一為我解答。最後,我

(6)

必修課只要遇到問題,你也願意花時間和我討論並引導我找出可能的解決辦法。沒有你 的幫忙,我很難想像我要怎麼熬過碩一的課程。直到今天,只要我有任何學業上或生活 上的問題,你都願意聽我分享並提供你的意見,真的很感謝你。我也祝你博士學業順利, 永遠享受你正在進行的研究。

這三年中,我要謝謝工作上老闆及同事的體諒,讓我能在工作及課業上兼顧。萬

分感謝我的老闆及老闆娘David and Sophia。你們是我遇過最好相處、最關心員

工的老闆。謝謝你們總是包容我並給我很大的發揮空間。謝謝你們總是相信我並 給我最大的方便。你們時時替我及我的家人禱告、願我喜樂平安、祝我課業順利、 給予我精神上最大的支持及無數的關心及問候。我一定會很想念這一切的。謝謝 所有工作的伙伴,包括數學班主任、Heidi、Grace、Holly、Rita、Belinda、Elizabeth M. Lefavour、Sephen J. Clark、Katie、Ansel、Kerina、Judy、Stacy、May、Wendy、 Reeve、Sheryl、Dennis、Oliver、Allie、Danny、Bruce、Matt、Nicky、Oliver、 Sammi and Jonathan.

這一路以來,我要感謝我的同學及朋友,包括陳立靜、史宛霖、李宛珊、李宛真、 郭世昕、黃彥華、楊昆峰、尤健源、黃煜婷、許峰銘、陳柏祥、江芷羚、丁敦仁、 Iris Cheng、林子涵、張閔芳、吳曄彤、蘇于珊、莊惠涵、張孟玉、楊菁菁、郭 千慈、王雅嬅、方家琪、陳姿吟、林聖心、Ray Chuang、Austin Chen 以及周信 甫。你們總是給我信心,你們相信我能做得到,也要我相信自己做得到。 我要特別感謝家人一路上的支持,謝謝爸爸劉金財、大姑姑劉鳳霞、二姑姑劉春 美、三姑姑劉金花、小姑姑劉娃花、叔叔劉漢生、外公蔡鳳白、大姨蔡旻蓉、舅 舅蔡富川、舅媽周秀滿、阿姨段玥慈、表哥方先哲、三嫂盧秋雅、表姊王夢潔、 表哥蔡忠佑、表姊蔡佳紋以及表弟吳凌宇。謝謝妳們從小到大不間斷的關心、鼓 勵與支持,讓我的求學之路無後顧之憂。 這一段是特別留給辛苦的姊姊劉怡孜。謝謝妳一直以來的幫助與體諒。媽媽過世 後的這十五年中,妳對家裡的付出最多。從小妳除了要讀書以外還要照顧我。在 我讀大學時,妳開始工作,同時也開始分擔家裡的開銷和我的生活費。我唸研究 所後,妳也決定再次回到校園攻讀碩士學位。這時的妳除了之前的負擔外還增加 了課業壓力。不過無論壓力再大,妳總是無怨無悔的為家裡付出。如果沒有妳, 我一定無法完成五專、二技及現在的碩士學位。除了再次感謝妳的付出,也恭喜 妳完成碩士學位,邁向另一個新的里程碑。 我有很多話想對來不及和我分享這一切的媽媽說。小時候我告訴她我想成為數學 博士,她很高興。現在看起來要我拿到數學博士的機會不大。不過至少很喜歡我 現在所做的研究,希望有一天我可以真的拿到博士學位並在博士論文中的致謝詞

(7)
(8)

TABLE OF CONTENTS

中文摘要... I ABSTRACT... II ACKNOWLEDGEMENTS...III TABLE OF CONTENTS...VI CHAPTER 1 INTRODUCTION ...1

CHAPTER 2 THE DATA ...2

CHAPTER 3 THE MEANING...9

3.1 Color Quality and Quantity (K&M, 2010) ...9

3.2 The Quality Reading of Tou and the Quantity Reading of Bian ...11

3.3 Selectional Restriction of Tou and Bian...26

3.3.1 Selectional Restriction of Adj-tou...28

3.3.2 Selectional Restriction of V-tou/bian...36

CHAPTER 4 THE SYNTACTIC POSITION ...40

4.1 The Basic Syntactic Positions of V/Adj-tou and V-bian...40

4.2 The Fronted Object ...50

4.3 Locative NP Inversion ...54

CHAPTER 5 DISCUSSION...57

CHAPTER 6 CONCLUSION ...61

(9)

CHAPTER 1

_____________________________________________________________________

CHAPTER 1

INTRODUCTION

Bolinger’s (1972) study on degree words has clearly asserted that gradability is a property for four major categories: adjective, adverb, noun, and verb. The gradability in his work refers to the quality reading of the four major categories (Bolinger, 1972:17). However, nowadays, when discussing gradability, one may refer to not only

quality but also quantity interpretations.1 For instance, Kennedy and McNally (2010)

indicate that gradable adjectives may have two sub-readings between quality and quantity. One question raised by this is as follows: If, as shown by Bolinger, quality is a universal property of four major categories, is this true of quantity? The current study approaches this issue by investigating two Chinese words-tou, a verb meaning

penetrate or through in English, and bian, a verb meaning distribute or all over in

English. More specifically, I aim to examine if Chinese verbs and adjectives (adj.) encompass both quality and quantity interpretations. I will first present some relevant data pertaining to tou and bian below.

(10)

CHAPTER 2

_____________________________________________________________________

CHAPTER 2

THE DATA

Tou, a verb meaning penetrate or through in English, has four major usages. This

study will focus on the two of them which are listed in (1) and (2).2,3 For the other

two usages, please see Appendix A.

(1) Adjective-tou

ta hong- / shi- / zaogao- / wuliao-tou le.

he red- wet- terrible- boring-through SFP

‘He is extremely popular / wet / terrible / boring.’

(2) V-tou with a location ta wan-tou taiwan le.

he play-through Taiwan SFP

‘He has traveled to every part of Taiwan and thus he knows a lot about Taiwan.’

2 Abbreviations used in this study include: ASP: aspect marker, CL: classifier, DE

1: the verbal suffix or the marker for modifying phrases like genitive phrases, relative clauses, and noun complement clauses, DE2: resultative marker, NEG: negation and SFP: sentence final particle. symbols used in this study include *: ungrammatical sentences, ?: sentences which are marginally acceptable and #: sentences that are semantically anomalous.

(11)

First, adjectives can combine with tou, as shown in (1). Before I provide a more detailed analysis of the meaning of tou below, the meaning of tou can be roughly construed as extremely or very. That is, if the state denoted by an adjective “penetrates” the subject of a sentence, such as ta ‘he’ in (1), the subject must be fully equipped with some obvious features of the state, giving rise to the extremely reading.

The second type of tou introduces a location as a complement. The location is obligatory because the sentence is out when the location is omitted:

(3) * ta wan-tou le.

he play-through SFP

Intended meaning: ‘He has traveled and knows a lot.’

This phenomenon brings my attention to another verb bian, meaning distribute or all

over in English, which has a similar requirement indicated in (4):

(4) a. ta wan-bian taiwan le.

he play-all-over Taiwan SFP

(12)

b. *ta wan-bian le. he play-all-over SFP

Intended meaning: ‘He has traveled to every place.’

The V-bian construction also selects a location that can not be absent; otherwise the sentence is out, as in (4b). In spite of the similarities between tou and bian, there exist some differences:

(5) a. ta shi-tou le.

he wet-through SFP ‘He is extremely wet’

b. * ta shi-bian le.

he wet-all-over SFP

Intended meaning: ‘His body is all wet. ’

(6) a. ta yazhou wan-tou le.

he Asia play-through SFP

(13)

b. * ta yazhou wan-bian le.

he Asia play-all-over SFP

Intended meaning: ‘He has traveled to every part of Asia.’

First, as I noted earlier, an adjective can combine with tou, forming the Adj-tou construction in (5a), while bian seems to be incompatible with an adjective, as in (5b). Secondary, a location is allowed to precede V-tou (6a) but not V-bian (6b). I will argue in sections 3.2 and 4 that the locations in (6) are in fact preposed objects and with a contrastive sentence, (6b) could be saved.

Another difference between tou and bian is the rigidity of their selectional restriction with the preceding verbs:

(7) a. * yinghua kai-tou le shangu.

cherry-blossoms bloom-through Asp valley

(14)

b. yinghua kai-bian le shangu.

cherry-blossoms bloom-all-over Asp valley

‘The cherry blossoms bloom in every part of the valley.’

c. *ta zou-tou taiwan le.

he walk-through Taiwan SFP

Intended meaning: ‘He has walked to every part of Taiwan and thus he knows a lot about Taiwan.’

d. ta zou-bian taiwan le.

he walk-all-over Taiwan SFP

‘He has walked to every part of Taiwan.’

It seems that bian embraces a wider range of verbs while the “V-tou location” construction is more restricted.

More interestingly, when the subject of the “V-bian location” construction is a theme, the location can be inversed, as shown in (8).

(15)

(8) a. yinghua kai-bian le shangu.

cherry-blossoms bloom-all-over Asp valley

‘The cherry blossoms bloom in every part of the valley.’

b. shangu kai-bian le yinghua.

valley bloom-all-over Asp cherry-blossoms

‘In every part of the valley bloom the cherry blossoms.’

The similarities and differences between tou and bian are summarized in (9):

(9) The similarities and differences between tou and bian

tou bian

1. compatibility with a preceding verb and a

location complement

V V

2. compatibility with a preceding adj. V *

3. object fronting V *

4. selectional restriction restricted free

(16)

With some basic observations of tou and bian in this section, I will argue in depth in section 3 that the tou alone in the V-tou construction expresses the quality of the verb while the bian in the V-bian construction quantizes the NP follows bian and expresses the quantity reading. Additionally I will pinpoint the selectional restriction of Adj-tou and V-tou/bian. Section 4 deals with the syntactic aspect of tou and bian. A discussion of the quality and quantity readings of verbs and adjectives will be formulated in section 5. Finally, section 6 concludes the paper.

(17)

CHAPTER 3

_____________________________________________________________________

CHAPTER 3

THE MEANING

In this section, I intend to argue that the V/Adj-tou construction denotes a quality reading whereas the V-bian construction expresses a quantity meaning. In order for readers to access the discussion, in 3.1, I first summarize Kennedy and McNally’s (K&M) (2010) research pertaining to color terms. With this preliminary idea, I will show how the same concept can be applied to tou and bian in 3.2. Finally, 3.3 studies the selectional restriction between V/Adj and tou/bian.

3.1 Color Quality and Quantity (K&M, 2010)

K&M provide empirical arguments asserting that color adjectives are ambiguous

between gradable and nongradable interpretations.4 The gradable readings contain

two subtypes, which they term color quantity and color quality (including degree of hue, color saturation, and brightness). The color quantity, as they define, “expresses a measure of how much of the object is the relevant color; this kind of reading is brought out by proportional modifiers (p. 90).” (10) contains the examples they provide:

(18)

(10) a. Pia painted the leaves part/half/completely green.

b. Instead of jerseys with blue stripes this year, the team is wearing shirts that are completely/entirely/100% blue.

c. His pants are half/part green, half/part blue: one leg of each color.

The compatibility of proportional modifers such as part, completely, 100% etc. and color adjectives such as green and blue reveals that what is concerned is how much of the object is the relevant color instead of the degree of hue, color saturation or brightness.

The quality interpretation, on the other hand, “involves a measurement of how closely an object’s color approximates or diverges from a “center” or prototype (p. 91).” There are many ways to measure the similarities or divergence between a color and its prototypical characteristics, including not only hue, saturation, and brightness but also other factors based on the physical and perceptual characteristics of color. (11), which is from K&M (2010: 91), shows that it is the color quality, but not the color quantity, that is being evaluated:

(19)

(11) a. I see that your leaves are all completely painted, but some are greener than others. Try to ensure that the colors are identical.

b. Your painting is coming along, though it still needs some work: all of the sky is blue, but it isn’t blue enough, and the clouds are too white. Try modifying your pigment mixtures.

After introducing the color quality and color quantity, I will turn back to tou and bian and illustrate how these concepts can be applied to our data.

3.2 The Quality Reading of Tou and the Quantity Reading of Bian

I would like to use (12) and (13) to argue that tou and bian convey different semantic contents:

(12) a. taiwan san-bai-yi-shi-jiu-ge xiangzhen ta dou wan-guo

taiwan three-hundred-one-ten-nine-CL country-and-town he all play-ASP

le, zhishi ta dou mei wan- *bian / tou.

SFP but he all not play- all-over through

‘He has traveled and had fun in all the 319 towns and countries in Taiwan. However, he hasn’t *visited all of them / learned too much about them.’

(20)

b. jiashang suoyou de shu ta dou kan-guo le, zhishi haimei du *bian /

shelf-top all DE1 book he all read-ASP SFPonly not-yet read all-over

tou. through

‘He has read all the books in the shelf, but he hasn’t *read all of them / understood all of them.’

(13) ta dui Chomsky zuixin de na-pian wenzhang hen shouxi, suiran ta

she to Chomsky latest DE that-CL article very familiar although he

haimei du-bian ta suoyou de wenzhang / *du-tou na-pian

not-yet read-all-over he all DE article study-through that-CL

wenzhang. article

‘She is extremely familiar with Chomsky’s latest article, although she hasn’t read all of his articles / *fully understood that article.’

There are 319 and only 319 countries/towns in Taiwan. If a person has traveled and had fun in all the countries/towns, the person must visit all of these 319 countries/towns. So, when the later part of the sentence negates, by using

(21)

“not-V-bian”, that the person hasn’t visited and had fun in all these 319 countries/towns, the meaning of the sentence is contradictory itself and is as expected out. The problem of (12b) is identical. If the first part of the sentence asserts that a person has read all the books in the shelf, it is contradictory to negate the sentence with V-bian in the later part of the sentence. To be more specific, what (12) suggests is that when the earlier mentioned quantity (san-bai-yi-shi-jiu-ge xiangzhen ‘319 countries/towns’ in (12a) and suoyou de shu ‘all the books’ in (12b)) is later negated by using “not V-bian,” the sentences are unacceptable, showing that bian denotes a quantity reading. V-tou, on the other hand, does not merely focus on the quantity reading. That is, the use of ‘not V-tou’ does not necessarily entail that the person hasn’t visited and had fun in all these 319 countries/towns in (12a) or the person hasn’t read all the books in the shelf in (12b); rather, V-tou seems to show that to what degree the action is performed. More specifically, V-bian is a preliminary condition for V-tou. For instance, when a person wan-tou taiwan ‘Someone has traveled to every part of Taiwan and thus he knows a lot about Taiwan’, it implies that the person has wan-bian taiwan ‘Someone has traveled to every part of Taiwan’. The difference between V-tou and V-bian is that the later focuses on the quantity reading while the quality reading is the center of the former. This initial observation can be further supported by (13), where the first part of the sentence reveals that the person is

(22)

familiar with Chomsky’s latest article. When the later part of the sentence asserts that the person hasn’t fully understood the article by negating V-tou, the sentence is unacceptable. When V-bian, which denotes a quantity reading, is used to replace V-tou, the sentence is perfect. This implies that it is V-tou, but not V-bian, expresses a quality reading.

One question, then, arises: Are the quality and quantity readings triggered solely by

tou and bian respectively? As I am going to argue below, tou itself can trigger the

quality reading whereas bian and the subsequent NP are responsible for the quantity interpretation. Recall that in (5), repeated as (14) below, an adjective can combine with tou, forming the Adj-tou construction, as in (14a), while bian seems to be incompatible with adjectives, as in (14b):

(14) a. ta shi-tou le.

he wet-through SFP ‘He is extremely wet.’

b. * ta shi-bian le.

he wet-all-over SFP

(23)

(14a) is grammatical when the subject ta ‘he’ is interpreted as a location for the predicate Adj-tou. In terms of (14b), even the subject ta ‘he’ is interpreted as a location, the sentence is still out. It appears that bian needs an obligatorily present complement. However, in Mandarin Chinese, when adjectives are the main predicates of a sentence, they do not take a complement. It is, therefore, impossible to add a location, giving rise to the ungrammaticality of (14b). I claim that bian quantizes the NP follows bian. As Krifka (1989, 1992) states, a quantized NP denotes an object with precise limits. In the case of bian, the precise limit is delimited by the range denoted by the NP. This view predicts that what bian selects should not be merely restricted as a locative NP. As (15) shows, the prediction is born out:

(15) a. ta shang-bian suoyou xishang de ke.

he attend-all-over all department-on DE1 class

‘He has taken all the courses offered by the department.’ b. ta du-bian shishu.

he read-all-over history-book

(24)

The NPs that follow bian (i.e. ke ‘class’ in (15a) and shishu ‘history book’ in (15b)) are themes instead of locations. After being quantized by bian, the NPs provide a range within which the action takes place. As a consequence, the “V-bian NP” construction can be paraphrased as “the processes are repeatedly performed to/within the complement NP.” Now, the new question is: Why is the complement NP necessary? Recall that the color quantity is defined as to express a measure of how much of the object is the relevant color (K&M, 2010). In other words, an affected object is required to show the quantity of the color. By analogy, an affected NP (be it a location or a theme) must be present to show the quantity of the action. This line of argument naturally illustrates why the NP can not be omitted.

The fact that bian has to quantize an NP to show the quantity of an action also explains why its object fails to be fronted to a preverbal position, a phenomenon which is allowed in the V-tou construction. However, when a contrastive sentence is added, the acceptability of the sentence is boosted. The contrast is presented below:

(16) a. ta yazhou wan-tou le.

he Asia play-through SFP

(25)

b. * ta yazhou wan-bian le.

he Asia play-all-over SFP

Intended meaning: ‘He has traveled to every part of Asia.’

c. ? ta yazhou wang-bian le, danshi ouzhou haimei.

he Asia play-all-over SFP but Europe not-yet

‘He has traveled to every part of Asia, but not Europe.’

The reason that (16b) is unacceptable is because, after the NP is moved to a preverbal position, bian does not succeed in quantizing an NP and hence fails to express the quantity of an action. That is, once the NP is missing, the scope/range that the action takes place in can not be defined, leading to the failure of quantizing. In (16c), the additional sentence helps readers delimit the scope of the NP. More specifically, although the theme object yazhou ‘Asia’ is moved to a preverbal position, the presence of the contrastive sentence shows that what is focused is the theme object

yazhou ‘Asia’ and this emphasis also helps readers understand that yazhou ‘Asia’ can

be quantized by bian, giving rise to a higher degree of acceptability. I will turn back to

this issue and show that (16c) indeed involves focus in 4.2.5

5 Liao (p.c.) points out a problem of this analysis. Let’s look at (i). (i) a. ta chi shi-ke pingguo.

(26)

Tou, on the other hand, only requires one of its arguments to be understood either as a

location or a theme. As I mentioned earlier, (14a) is only acceptable when the subject

ta ‘he’ is understood as a location of the predicate shi-tou le ‘extremely wet’.

Additionally, (17) below shows that Adj-tou can be legitimate when the subjects are understood as themes:

(17) ta huai- / zaogao- / wuliao-tou le.

he bad- terrible- boring-through SFP

‘He is extremely bad / terrible / boring.’

In fact, the view that tou only requires one of its arguments to be understood either as a location or a theme explains why (18) is ambiguous between two possible readings.

b. ta pingguo chi shi-ke. he apples eat ten-CL ‘He eats ten apples.’

The shi-ke pingguo ‘ten apples’ in (ia) is a quantized NP. However, as shown in (ib), pingguo ‘apple’ can be fronted, which is contrary to my proposal.

There exist two possibilities to explain the differences between (i) and (16). The first argument is that the pingguo ‘apple’ in (ib) is in fact a sentence-internal topic expressing new information or new events rather than a pre-posed object. The second possibility is that tou has a strong feature to license / agree with the covert form in (16a) while bian contains a weak feature which fails to license / agree with the covert form in (16b). As noted earlier in 3.2, V-bian is a preliminary condition for V-tou. That is, the

(27)

(18) ta hong-tou le.

he red-through SFP

Reading a: ‘His whole body is extremely red.’ Reading b: ‘He is extremely popular.’

The meaning of (18a) can be established when ta ‘he’ is understood as a location, meaning that the person’s body is extremely red. Conversely, when ta ‘he’ is construed as a theme, the only meaning is ‘he is extremely popular,’ leading to the interpretation in (18b). Still, another question emerges. It is well-known that hen ‘very’ in Mandarin Chinese also describes the quality of adjectives. Thus, one may naturally wonder if there is any semantic divergence between hen-Adj and Adj-tou. (19) offers some hints for the differences between hen-Adj and Adj-tou:

(19) a. zhe-tiao yu kanqilai yijing jian de hen shou le, dan qishi

this

-

CL fish look-like already fry DE2 very well-cooked SFP but in-fact

limian de rou haimeiyou shou-tou.

inside DE1 meat not-yet well-cooked-through

‘Although the fish looks well-cooked on the outside, actually it is not well-cooked on the inside.’

(28)

b. zhe-ke putaoyou kanqilai hen hong, dan qishi limian de guorou

this

-

CL grapefruit look-like very red but in-fact inside DE1 pulp

haimeiyou hong-tou.

not-yet red-through

‘Although the grapefruit looks red on the outside, actually the pulp isn’t very red.’

For (19a), although the fish appears to be well-cooked on the outside, it does not guarantee that it is well-cooked on the inside. The similar phenomenon is observed in (19b). Even though the grapefruit looks extremely red, it does not warrant that the fruit inside the peel is also extremely red. Bearing this observation in mind, I argue that the hen-Adj in this case describes the visually observable quality, while Adj-tou describes the quality that distributes over not only the surface but also the inner domain of the object. If a kind of quality prevails over the whole object both on the outside and inside, it is spontaneous to yield an extreme reading of the quality. The view that Adj-tou describes the quality that distributes over not only the surface but also the inner domain is further supported by (20):

(29)

(20) a. # zhe-tiao yu kanqilai yijing jian de shou-tou le.

this-CL fish look-like alreadyfry DE2 well-cooked-through SFP

Intended meaning: ‘The fish looks well-cooked on the inside.’

b. # zhe-ke putaoyou kanqilai hong-tou le.

this

-

CL grapefruit look-like red-though SFP

Intended meaning: ‘The grapefruit looks red on the inside.’

The kanqilai ‘looks like’ in (20) signifies that the speaker’s subsequent statement is based upon his/her visual observation. It is, therefore, expected that the use of Adj-tou is unacceptable in (20) because merely providing the surface quality of a given object fails to confirm the semantic requirement of Adj-tou.

One more thing to add is that, in the V-tou construction, if the subject can not be understood as a location or a theme, a location or theme complement must be present, as indicated in (21) and (22):

(30)

(21) Adjective-tou

a. ta hong-tou quan shijie le.6

he red-through whole world SFP

‘He has become extremely famous around the world.’

b. ta hong-tou le.

he red-through SFP

Reading a: ‘His whole body is extremely red.’ Reading b: ‘He is extremely popular.’

(22) V-tou with a location

a. ta wan-tou taiwan le.

he play-through Taiwan SFP

‘He has traveled to every part of Taiwan and thus he knows a lot about Taiwan.’

b. * ta wan-tou le.

he play-through SFP

Intended meaning: ‘He has traveled and knows a lot.’

(31)

The subject in (21a) can be understood as a theme; therefore, the omission of the internal argument is allowed in (21b). On the contrary, the subject of (22a) is an actor and as it is neither a location nor a theme, the sentence requires an internal argument. Otherwise, these sentences are not interpretable, as the second sentence in (22) suggests. So far, I have argued that the tou along in the V/Adj-tou construction can induce a quality reading. In what follows, I am going to claim that the derivation of such meaning can be accounted in terms of K&M’s (2010) color quality.

It is reasonable for adjectives to express a quality reading. As I argue in section 2, if the state denoted by an adjective “penetrates” the subject of a sentence, the subject must be fully equipped with some obvious features of the state, giving rise to the

extreme reading. However, the verb wan ‘play’ in type 4 is an activity verb. Therefore,

the next question I am about to answer is how the quality interpretation of an activity verb such as wan ‘play’ is derived. Recall that K&M (2010) advocate that color quality includes the degree of hue, saturation, and brightness of the color. It is more than obvious that an action does not have hue or brightness. Thus, I argue that the concept of saturation is the key to the question. Please pay attention to (2), which is repeated as (23) below:

(32)

(23) ta wan-tou taiwan le.

he play-through Taiwan SFP

‘He has traveled to every part of Taiwan and thus he knows a lot about Taiwan.’

I argue that in the “V-tou location” construction, the meaning can be paraphrased as “the processes denoted by the verb repeatedly happen and increased and thus saturated at a given location.” In this case, tou signifies the concept of saturation. The location symbolizes a referent range/volume where the action is increased and therefore must be present. One relevant comment in order here is that a range is always required to show the concept of saturation. That is, when we say ‘The job market is saturated’, the range (i.e. ‘the job market’) must be present. Analogously, in (23), when wan ‘play’ combines with tou, it means that the action of playing has saturated. However, the subject of the sentence is an actor, instead of a theme or a location which can show a range. In this case, an obligatory locative NP must present to show a range that enables the concept of saturation to survive. In this sense, (23) indicates that the event of wan ‘play’ is increased and saturated in Taiwan, generating the quality reading of ‘He has traveled to every part of Taiwan and thus he knows a lot about Taiwan.’ The derived meaning (i.e. ‘knows a lot about Taiwan’) is a natural consequence. When something or some action is repeatedly performed at/within a

(33)

specific location, the actor must be familiar with the thing or the action.

If I argue that tou expresses a quality reading and follow K&M (2010) that the quality reading involves a measurement of how closely an object’s color approximates or diverges from a “center” or prototype, it should be expected to show that some different acceptability may result from different speakers’ divergence of a “center” or prototype. This prediction is born out in (24):

(24) a. A: haimian baobao hong-tou quan shijie.

sponge baby red-through whole world

‘Sponge Bob is extremely famous in the world.’

b. B: dui a! dajia dou renshi ta.

yes SFP everyone all know it

‘Yes! Everyone knows it.’

c. C: nayou! suiran dajia dou renshi ta, danshi zhiyou xiaohai xihuan ta

no although everyone all know it but only kid like it

ba. SFP

(34)

(24b) and (24c) are two possible responses for (24a), although they contrast dramatically in their meanings. Speaker B’s acceptance of (24a) indicates that as long as everyone knows Sponge Bob, it can be counted as extremely famous in the world. On the contrary, to speaker C, Sponge Bob is popular among kids but not among adults; therefore, it is not extremely famous in the world. The divergence of speakers B and C’s acceptability on (24a) thus supports the view that K&M’s (2010) definition on color quality is applicable to tou.

3.3 Selectional Restriction of Tou and Bian

Last section mainly focuses on how the quality and quantity readings are derived from

tou and bain respectively. However, one significant fact pertaining to V/Adj-tou and

V-bian is that not all given verbs or adjectives are compatible with tou or bian. (25) and (26) list some examples:

(25) Adj-tou

a. xin liang-tou le. heart cool-through SFP ‘I feel dejected.’

(35)

b. * beizi man-tou le

cup full-through SFP

Intended meaning: ‘The cup is totally full.’

(26) V-tou/bian

a. ta wan-tou taiwan le.

he play-through Taiwan SFP

‘He has traveled to every part of Taiwan and thus he knows a lot about Taiwan.’

b. ta wan-bian taiwan le. he play-all-over Taiwan SFP

‘He has traveled to every part of Taiwan.’

c. *ta zou-tou taiwan le.

he walk-through Taiwan SFP

Intended meaning: ‘He has walked to every part of Taiwan and thus he knows a lot about Taiwan.’

d. ta zou-bian taiwan le.

he walk-all-over Taiwan SFP

(36)

In (25), in spite of the fact that both liang ‘cool’ and man ‘full’ are adjectives, only the former adjective, but not the later one, fits in with tou. Similarly, in (26), both wan ‘play’ and zou ‘walk’ are activity verbs. Interestingly, zou ‘walk’ is only compatible with bian, as in (26d), while wan ‘play’ can combine with bian and tou, as in (26a,b). It appears that bian embraces a wider range of verbs than tou does. In what follows, I aim to provide a more fine-grained selectional restriction on tou and bian.

3.3.1 Selectional Restriction of Adj-tou

According to Kennedy and McNally (2005), adjectives can be categorized by their context-dependency. Context-dependent adjectives, or relative adjectives in Unger’s (1975) term, require a standard of comparison, which is itself determined relative to a comparison class of objects that are similar in some way to whatever is being discussed (Klein, 1980). For instance, the truth condition of the adjective tall may vary. A man who is 175 cm tall is tall in Indonesia where men’s average height is 158 cm. (Frankenberg & Jones, 2004). However, the same person will not be considered

as tall in Netherlands where men’s average height is 183.7 cm..7 Context-independent

adjectives, or absolute adjectives in Unger’s (1975) term, are demonstrably gradable

7 The information is retrieved from Statline

(37)

but whose standards of comparison are not context-dependent (Kennedy and McNally, 2005). These adjectives simply require their arguments to possess a minimal or a maximal degree of the property in question. (27) and (28) present some examples from Kennedy and McNally (2005):

(27) Context-independent adjectives with minimal standards a. The door is open.

b. The rod is bent.

(28) Context-independent adjectives with maximal standards a. The glass is full.

b. The door is closed.

In (27a), as the door is not fully closed, it is open. The property of “open” does not need to exceed any standard of comparison. Similarly, in (27b), once the rod is minimally bent, the truth condition is confirmed. On the other hand, (28a) requires the glass to be completely full and (28b) requires the door to be fully closed. These two sentences are established once the property in question reach a maximal degree and do not involve any context-dependent standards.

(38)

Now, let’s turn back to the Adj-tou construction. It is apparent that only context-dependent adjectives are compatible with tou:

(29) Context-independent adjectives with minimal standards

a. * men kai-tou le.

door open-through SFP

Intended meaning: ‘The door is fully open.’ b. * zhe-gen shuzhi wan-tou le.

this-CL branch bent-through SFP

Intended meaning: ?‘The branch is extremely bent.’

(30) Context-independent adjectives with maximal standards

a. * shui man-tou le.

water full-through SFP

Intended meaning: ‘The glass of water is full.’

b. * men guan-tou le.

door close-through SFP

(39)

(31) Context-dependent adjectives a. xin liang-tou le. heart cool-through SFP ‘I feel dejected.’

b. zhe-ben shu wuliao-tou le.

this-CL book boring-through SFP

‘This book is extremely boring.’

The reason that tou is only compatible with context-dependent adjectives is understandable. Since tou expresses the ‘extreme’ quality of adjectives, adjectives with minimum standards, which show the minimal degree of the quality denoted by the adjectives, are semantically confused with tou. Additionally, an adjective with a maximal standard has itself indicated the ‘extreme’ reading of the quality; it is therefore redundant to add tou, which conveys a similar meaning. Context-dependent adjectives, nevertheless, encode neither maximal nor minimal standards and their standards of comparison are subject to contexts. The use of tou serves to indicate the ‘extreme’ quality denoted by the adjectives without any semantic contradiction.

(40)

However, requiring an adjective to be context-dependent is only a sufficient condition instead of a necessity condition. (32) lists some counterexamples with context-dependent adjectives which fail to combine with tou.

(32) a. * zhe-ben shu gui-tou le.

this-CL book expensive-through SFP

Intended meaning: ‘This book is extremely boring.’

b. * zhangsan gao-tou le.

Zhangsan tall-through SFP

Intended meaning: ‘Zhangsan is extremely tall.’

Recall that in 3.2, I argue that tou requires the quality denoted by the adjective to “penetrate” the modified entity. That is, all the subparts of the modifiee must be equipped with the quality. This view explains why the exemplar sentences in (32) are not acceptable. For (32a), the reason that a book is expensive is not because each leaf in the book is expensive. Specifically, the subparts (each leaf) of the entity (the book) are not equipped with the quality (expensive). In the same vein, in (32b), it is awkward to say that Zhansang is tall because his legs, hands, waist, chest, neck and

(41)

etc. are all tall. More supports of this criterion are in (33).

(33) a. zhe-ben shu wuliao-tou le.

this-CL book boring-through SFP

Intended meaning: ‘This book is extremely expensive.’

b. # suiran zhangsan wuliao-tou le, danshi ta shuo de xiaohua

although Zhangsan boring-through SFP but he speak DE2 joke

hen youqu.

very interesting

#Although Zhangsan is extremely boring, his jokes are interesting.

(33a) is modified from (32a). The difference between these two sentences is their adjectives. However (33a) is perfect while (32a) is unacceptable. (33a) is perfect because the reason that the book is extremely boring is caused by the fact that the contents in each leaf are extremely boring. In other words, the subparts of the modifiee are equipped with the quality (i.e. boring). Similarly, when a person articulates the first part of (33b), he or she thinks that all of Zhangsan’s behavior and performances are boring. For example, his jokes are bad and he is not humorous at all. Thus, it is semantically confusing when the last part of the sentence asserts that

(42)

Zhangsan is good at talking jokes.

My analysis predicts that when the subject NP is plural, each individual of the plural subject can be understood as the subparts of the whole noun. That is, if the shu ‘book’ in (32a) is replaced by a plural form, the sentence should be saved. This prediction is borne out:

(33) zhe-xie shu gui-tou le.

this-CL book expensive-through SFP

‘These books are extremely boring.’

Some counterexamples, however, appear. (34) lists some.

(34) a. *zhe-xie ren gao-tou le.

this-CL person tall-through SFP

Intended meaning: ‘These people are extremely tall.’ b. *zhe-xie xuesheng congming-tou le.

this-CL student smart-through SFP

(43)

These adjectives are individual-level predicates. The qualities denoted by those adjectives are true throughout the existence of an individual; therefore, it is odd to say that some individual is equipped with certain qualities by using Adj-tou when the qualities doom to associate with the individual permanently. When adjectives are

stage-level predicates, the sentences are acceptable, as shown in (35).8

8 Liao (p.c.) points out the following contrasts:

(i) a. *zhe-xie xuesheng congming-tou le. this-CL student smart-through SFP

Intended meaning: ‘These students are extremely smart.’ b. zhe-xie xuesheng ben-tou le.

this-CL student stupid-through SFP ‘These students are extremely stupid.’

It is still unclear to me why ben ‘stupid’ seems to be an exception. However, as indicated in (ii), individual-level predicates with tou are always out.

(ii) *zhe-xie xuesheng ai / pang / shou / mei / chou-tou le this-CL student short / fat / thin / beautiful / ugly-through SFP

Intended meaning: ‘These students are extremely short / fat / thin / beautiful / ugly.

When temporal adverbs such as yizhi ‘always’ or yixiangyilai ‘always’ which express a long period of time are added, (ib) becomes bad.

(ii) *zhe-xie xuesheng yizhi / yixiangyilai dou ben-tou le. this-CL student always / always all stupid-through SFP Intended meaning: ‘These students are extremely stupid for a long time.’

This shows that tou is incompatible with elements expressing a long-term property. I thus suspect that

ben ‘stupid’ may be ambiguous between individual-level and non-individual-level interpretations. The

following example from Liu (p.c.) supports this idea.

(iii) ni shou ben jiao ben meiyou yi-yang xing. ni zhenshi you hand stupid foot stupid no one-CL can you really ben-toule.

stupid-through SFP

(44)

(35) a. zhe-xie yundongyuan dou e-tou le.

this-CL athlete all hungry-through SFP

‘All of these athletes are extremely hungry.’

b. zhe-ji-tian de tianqi leng-tou le.

this-many-day DE weather cold-through SFP

‘The weather in these days is extremely cold.’

In short, the adjectives in the Adj-tou construction must be the context-dependent ones. Additionally, all the subparts of the modified entity must be equipped with the quality denoted by those adjectives.

3.3.2 Selectional Restriction of V-tou/bian

Thus far, it is still unclear what kinds of verbs can be used in the “V-tou/bian Location/NP” construction. The picture is not as clear as the one in the Adj-tou construction. However, some insights can be seen in the definitions of V-tou and V-bian. As I stated earlier that, in the “V-tou location” construction, the meaning can be paraphrased as “the event denoted by the verb is repeatedly taken place and increased and thus saturated at a given location.” Tou in this case signifies the concept

(45)

of saturation and the location symbolizes a referent range/volume where the action is increased and therefore must be present. Therefore, for V-tou, the event denoted by the verb must be one that can be increased:

(36) a. ta wan-tou yazhou le.

he play-through Asia SFP

‘He has traveled to every part of Asia and thus knows a lot about Asia.’

b. ? ta guang-tou le zheng-dong baihuogongsi.

he shop-through Asp whole-CL department-store

‘He has shopped every part of the department store and thus knows a lot about the department store.’

c. ? ta chi-tou le zheng-tiao yeshi.

he eat-through Asp whole-CL night-market

‘He has tasted almost all the food in the night market and thus knows a lot about the night market.’

According to the informants I consulted, (36a) is natural and (36b) and (36c) are not perfect, but are marginally acceptable. These verbs have one thing in common: The

(46)

events denoted by these verbs can be incrementally increased. For instance, the meaning of (36a) can be established only if the event of wan ‘play’ takes place at every corner of yazhou ‘Asia’. With the accumulation of these events, the range/volume encoded by the location can be saturated, giving rise to the extreme reading. Likewise, (36b) can be established when the event denoted by guang ‘shop’ takes place at every corner of baihuogongsi ‘department store’ and (36c) can be established when the event denoted by chi ‘eat’ takes place at every corner of yeshi ‘night market’. Again, with the accumulation of these events, the scale/volume encoded by the location (zheng dong baihuogongsi ‘the whole department store’ in (36b) and zheng tiao yeshi ‘the whole night market’ in (36c)) can be saturated, giving rise to the quality reading.

In terms of the “V-bian NP” construction, it is paraphrased as “the event is repeatedly performed to/within the complement NP.” This implies that the event doesn’t need to be incrementally increased. Therefore, it is expected that the “V-bian” construction can embrace a wider range of verbs. As shown in (37):

(47)

(37) a. ta xing- / ta-bian quan tai. he walk walk-all-over whole Taiwan

‘He has been to every part of Taiwan (by walking).’

b. * ta xing- / ta-tou quan tai.

he walk walk-through whole Taiwan

Intended meaning: ‘He has been to every part of Taiwan (by walking) and

(48)

CHAPTER 4

_____________________________________________________________________

CHAPTER 4

THE SYNTACTIC POSITION

In this section, the focus is shifted to the syntactic aspects of tou and bian. I will first argue that V/Adj-tou and V-bian are compound words. The internal structure of the compounds will be introduced as well. I will, then, provide the syntactic structures for them. Finally, some space will be devoted to the fronted objects and the inversion construction observed in these constructions.

4.1 The Syntactic Analysiss of V/Adj-tou and V-bian

The most straightforward evidence in support of the view that V/Adj-tou and V-bian are compound words is from (38) to (40):

(38) Adjective-tou

a. * ta shi-le-tou.

he wet-ASP-through

(49)

b. ta shi-tou-le.

he wet-through-Asp9

‘He is extremely wet.’

(39) V-tou with a location

a. * ta wan-le-tou taiwan.

he play-Asp-through Taiwan

Intended meaning: ‘He has traveled to every part of Taiwan and thus he knows a lot about Taiwan.’

b. ta wan-tou-le taiwan.

he play-through-Asp Taiwan

‘He has traveled to every part of Taiwan and thus he knows a lot about Taiwan.’

9 One may notice that earlier in (5), the same –le was classified as a sentential –le (SFP) rather than a verbal –le (Asp). In fact, the classification of this kind of –le is still debatable in the literature. See Li and Thompson (1981), Ross (1995), Smith (1997) and Soh (2009) for a sentential –le analysis, but see Klein, Li & Hendriks (2000), Liu (1988), Sybesma (1997, 1999) and Wu (2005) for a verbal –le

(50)

(40) V-bian NP

a. * tachi-le-bian taiwan.

he eat-Asp-all-over Taiwan

Intended meaning: ‘He has tasted a lot of food at every part of Taiwan.’

b. ta chi-bian-le taiwan.

he eat-all-over-Asp Taiwan

‘He has tasted a lot of food at every part of Taiwan.’

The aspectual marker –le can not be inserted between V/Adj and tou/bian in every (a) sentence of (38) to (40). When the aspectual marker –le is suffixed to the end of V/Adj-tou and V-bian in the (b) sentences of (38) to (40), the sentences are interpretable. These empirical facts suggest that V/Adj-tou and V-bian are compounds instead of separate elements.

Although the –le insertion test shows that V/Adj-tou and V-bian are compound words, a more fine-grained categorization is required. More specifically, there are five kinds of internal structures for compound words, including Subject-Predicate (SP), Verb-Object (VO), Verb-Complement (VC), Modifier-Head (MH) and Coordinative-Construction (CC). One may wonder what kinds of structures V/Adj-tou

(51)

and V-bian are. I will eliminate them one by one. First, two out of the five can be eliminated. They are SP and VO. The reason is that there are neither subjects nor objects in V/Adj-tou and V-bian. At first glance CC may be a potential candidate because tou and bian are verbs, with their preceding verbs, the internal structure can be analyzed as CC. However, in CCs, the two elements with an identical syntactic category should be either semantically similar or semantically opposite. This is not the case for V-tou and V-bian because no obvious similarities or contraries are found between tou/bian and their preceding verbs. The remaining possibilities are VC and MH. The complement in a VC compound shows the result of the action denoted by

the verb. Li and Thompson (1981:56) assert that, the insertion of debetween the two

elements can take place in resultative compounds. This test is implemented in (41)-(43):

(41) Adjective-tou

* ta shi-de-tou le.

he wet-DE2-ASP SFP

(52)

(42) V-tou with a location

* ta wan-de-tou taiwan.

he play-DE2-through Taiwan

Intended meaning: ‘He travels to every part of Taiwan and thus he knows a lot about Taiwan.’

(43) V-bian NP

* tachi-de-bian taiwan.

he eat-DE2-all-over Taiwan

Intended meaning: ‘He tastes a lot of food at every part of Taiwan.’

Although Adj-tou and V-tou/bian fail this test, it is still too early to claim that they are not VCs because some well known VCs fail this test as well. For instance, in the literature, da-pao ‘hit-run’ and qi-lei ‘ride-tired are regarded as VCs (Huang, 2006; Huang, Li & Li, 2009; Li, 1990), but they fail the de insertion test:

(44) a. Zhangsan da-pao le Lisi.

Zhangsan hit-run Asp Lisi.

(53)

b. *Zhangsan da-de-pao Lisi le.

Zhangsan hit-DE2-run Lisi SFP

Intended meaning: ‘Zhangsan hit Lisi and Lisi ran away.’

(45) a. Zhangsan qi-lei le ma le.

Zhangsan ride-tired Asp horse SFP

(i) ‘Zhangsan rode a horse and got tired from it.’ (ii) ‘Zhangsan rode a horse and got the horse tired.’

b. * Zhangsan qi-de-lei ma le.

Zhangsan ride-DE2-tired horse SFP.

Intended meaning: (i) ‘Zhangsan rode a horse and got tired from it.’

(ii) ‘Zhangsan rode a horse and got the horse tired.’

These examples indicate that not all the VCs are compatible with de insertion. This is to say that compounds that are compatible with de insertion are VCs, but it does not amount to saying that those which fail this test are not. Therefore, I employ an additional test-the “V-V resultative compound” and “V-de VP resultative phrase” alternation.

(54)

As Huang (2006:21) contends, the derivation of V-V resultative compounds from the

de phrasal resultatives involves, if any, verb movement over the object. This is

tantamount to saying that VCs should have phrasal resultative counterparts. (46) and (47) show the alternations:

(46) a. Zhangsan da-pao le Lisi.

Zhangsan hit-run Asp Lisi.

‘Zhangsan hit Lisi and Lisi ran away.’

b. Zhangsan da-de Lisi dou pao le.

Zhangsan hit-DE2 Lisi all run SFP.

‘Zhangsan hit Lisi and Lisi ran away.’

(47) a. Zhangsan qi-lei le ma le.

Zhangsan ride-tired Asp horse SFP

(i) ‘Zhangsan rode a horse and got tired from it.’ (ii) ‘Zhangsan rode a horse and got the horse tired.’

b. Zhangsan qi-ma qi-de hen lei.

Zhangsan ride-horse ride-DE2 very tired.

(55)

c. Zhangsan qi-de ma hen lei.

Zhangsan ride-DE2 horse very tired.

‘Zhangsan rode a horse and got the horse tired.’

Now, let’s turn to Adj-tou and V-tou/bian:

(48) Adjective-tou

* ta shi-de quanshen dou tou le.

he wet-DE2 whole-body all through SFP

Intended meaning: ‘His body is extremely wet.’

(49) V-tou with a location

* ta wan-de taiwan dou tou le.

he play-DE2 Taiwan all through SFP

Intended meaning: ‘He travels to every part of Taiwan and thus he knows a lot about Taiwan.’

(56)

(50) V-bian NP

* tachi-de taiwan dou bian le.

he eat-DE2 Taiwan all all-over SFP

Intended meaning: ‘He tastes a lot of food at every part of Taiwan.’

The incompatibility of the “V-V resultative compound” and “V-de VP resultative phrase” alternation shown in (48) to (50) indicates that the only possible internal structure for V/Adj-tou and V-bian is MH. This analysis predicts that tou and bian are heads while the preceding verbs and adjectives are modifiers. Semantically, this view is plausible because the meanings of such constructions always convey that some quality has “penetrated” something or some events have been widely performed, whose meanings are derived from the original lexical words tou and bian respectively. Conversely, the verbs and adjectives in such compounds always encode the manner and can be replaced depending on the way some quality “penetrates” an entity or some events have been widely/repeatedly performed to/within an entity or a location. For instance, in ta chi-bian taiwan ‘He has tasted a lot of food at every part of Taiwan’, bian signifies that the subject has been to every part of Taiwan and chi ‘eat’ reveals the manner (i.e. by tasting food). If the person visited every part of Taiwan by having fun instead of tasting food, the first word of the compound, chi ‘taste’, has to

(57)

be replaced by wan ‘play’. However, the change of the first word, which encodes the manner, does not change the core meaning of the compound. That is, the person still has visited every part of Taiwan, irrespective of the way he traveled. In short, after eliminating the other four possibilities, I claim that the only possible internal structure of such compounds is MH. Moreover, semantically, the meanings of such constructions always convey that some quality has “penetrated” something or some events have been widely performed, whose meanings are derived from the original lexical words tou and bian respectively. Syntactically, the first words of the compounds can be replaced depending on the way some quality “penetrates” an entity or some events have been widely/repeatedly performed at an entity or a location.

After the above discussion, the syntactic structures of V/Adj-tou and V-bian can be generated. The tree drawn in (51) is based on V-bian:

(51) a. ta wan-bian quan tai.

he play-all-over whole Taiwan

(58)

b. VP qp DP V’ qp V DP 6

ta wan-bian quan tai

4.2 The Fronted Object

Although I have provided the syntactic structures for V/Adj-tou and V-bian, the proposed structures can not account for the problem stated at the end of section 2. The problematic sentences in (6) are repeated as (53):

(53) a. ta yazhou wan-tou le.

he Asia play-through SFP

‘He has traveled to every part of Asia and thus knows a lot about Asia.’ b. * ta yazhou wan-bian le.

he Asia play-all-over SFP

(59)

In 3.2, I also provided the reason that the unacceptability of (53b) can be dramatically reduced when a contrastive sentence is added:

(54) ? ta yazhou wang-bian le, danshi ouzhou haimei.

he Asia play-all-over SFP but Europe not-yet

‘He has traveled to every part of Asia, but not Europe.’

The fact that the addition of a contrastive sentence can rescue (53b) is a piece of evidence for the object fronting analysis. As Tsai (2008) points it out that object fronting occurs in Chinese only when contrastive focusing is involved.

Another piece of evidence comes from (55):

(55) (Tsai, 2008, (1))

a. * wo liang-ben shu nian-guo, san-ben shu mei nian-guo.

I two-CL book read-Asp, three-CL book not read-Asp

(60)

b. wo you liang-ben shu nian-guo, you san-ben shu mei nian-guo.

I have two-CL book read-Asp, have three-CL book not read-Asp

‘I read two of the books, but not the other three.’

c. wo zhe liang-ben shu nian-guo, na san-ben shu mei nian-guo.

I this two-CL book read-Asp, that three-CL book not read-Asp

‘I read these two books, but not those three.’

According to Tsai’s (2008) observation, when numeral object NP is in a preverbal position, it must be bound either by the existential modal you ‘have’ (55b) or by a demonstrative like zhe ‘this’, as in (55c). Otherwise, the sentence is out, as in (55a). This implies that, if the preverbal elements are fronted NPs in (54), once the fronted bare NP is replaced by a numeral NP, the syntactic behaviors should be no difference from (55). The prediction is born out:

(61)

(56) V-tou with fronted objects

a. * ta liang-ge chengshi wan-tou le, san-ge chengshi haimei.

he two-CL city play-through SFP three-CL city not-yet

Intended meaning: ‘He has visited and had fun in two cities and thus he

knows a lot about the cities. However, he hasn’t had fun in those three cities and thus he doesn’t know much about the cities.’

b. ta you liang-ge chengshi wan-tou le, you san-ge chengshi haimei.

he have two-CL city play-through SFP have three-CL city not-yet

‘He has visited and had fun in two cities and thus he knows a lot about the cities. However, he hasn’t had fun in those three cities and thus he doesn’t know much about the cities.’

c. ta zhe liang-ge chengshi wan-tou le, na san-ge chengshi haimei.

he this two-CL city play-through SFP that three-CL city not-yet

‘He has visited and had fun in these two cities and thus he knows a lot about the cities. However, he hasn’t had fun in those three cities and thus he doesn’t know much about the cities.’

(62)

(57) V-bian with fronted objects

a. * ta liang-ge chengshi wan-bian le, san-ge chengshi haimei.

he two-CL city play-all-over SFP three-CL city not-yet

Intended meaning: ‘He has visited and had fun in two cities, but not in those

three cities.’

b. ta you liang-ge chengshi wan-bian le, you san-ge chengshi haimei.

he have two-CL city play-all-over SFP have three-CL city not-yet

‘He has had fun in two cities, but not in those three citie.’

c. ta zhe liang-ge chengshi wan-bian le, na san-ge chengshi haimei.

he this two-CL city play-all-over SFP that three-CL city not-yet

‘He has had fun in these two cities, but not in those three cities.’

It is quite robust to assert that (53) involves object fronting and the ungrammaticality

of (53b) can be saved when a proper contrastive sentence is supplied.10

4.3 Locative NP Inversion

One remaining question that I haven’t dealt with is that, when the subject of V-bian is a theme, the locative NP following bian can be inversed with the theme. The instance

(63)

shown in (8) is given below.

(8) a. yinghua kai-bian le shangu.

cherry-blossoms bloom-all-over Asp valley

‘The cherry blossoms bloom in every part of the valley.’

b. shangu kai-bian le yinghua.

valley bloom-all-over Asp cherry-blossoms

‘In every part of the valley bloom the cherry blossoms.’

To account for this inversion construction, I follow Wu’s (2008) proposal. The verb (kai-bian ‘bloom-all-over’) is overtly moved out from its based-generated position (V head) to T head in order to avoid the problem of locality and hence allow the lower DP (LP) to move upward. The locality restriction is relativezed to minimal domains proposed by Chomsky (2001), as shown in (58). The LP is subsequently moved to [Spec, TP] to fulfill EPP, giving rise to the outcome in (8b). A tree diagram of these steps is shown in (59).

(64)

(58) a. Terms of the same minimal domain are equidistant to the probe.

b. The minimal domain of a head H is the set of terms immediately containedin

the projection of H.11 (59) TP qp shangu j T’ qp kai-bian le i vP qp yinghua VP qp t i t j 11

(65)

CHAPTER 5

_____________________________________________________________________

CHAPTER 5

DISCUSSION

Thus far, I have argued that tou induces quality interpretations of adjectives and verbs. On the other hand, bian quantizes the subsequent NP and expresses the quantity reading of verbs. If the proposed ideas are correct, the study bears following implications.

First, the results of the study echo with Bolinger’s (1972) view on gradability. Recall that, in his view, quality is a universal property for adjectives, adverbs, nouns, and verbs. Although this study only focuses on the verbs and adjectives in Mandarin Chinese, it provides some insights. It is not surprising that adjectives have a quality interpretation, but the fact that tou induces quality readings of verbs reveal that, in addition to adjectives, verbs in Mandarin Chinese also contain a quality reading. Thus, it can be inferred that, universally, if not all of the four major word classes, at least verbs and adjectives contain a quality reading. This conclusion reminds us the question brought up in the introduction: If quality is a universal property of four major categories, would it be true for quantity? Although, at first glance, the findings of the study may provide a negative answer, I will argue below that quantity may as well be

參考文獻

相關文件

Methodologies used in this study comprise two parts, the first part is questionnaires, which investigate customers’ assessment of the quality and satisfaction of service

The study derives five parameters as background factors, personality trait, community trait, and aggression behavior for the research based on the literature studies to

This study aims to explore whether the service quality and customer satisfaction have a positive impact on the organizational performance of the services and whether the

In order to serve the fore-mentioned purpose, this research is based on a related questionnaire that extracts 525 high school students as the object for the study, and carries out

The purposes of this research was to investigate relations among learning motivation, learning strategies and satisfaction for junior high school students, as well as to identify

The purpose of this study is to investigate the researcher’s 19 years learning process and understanding of martial arts as a form of Serious Leisure and then to

The purpose of this research is to study the cross-strait visitor’s tourist experience.With the research background and motives stated as above, the objectives of this research

This dissertation utilizes the Kano Two-dimension Quality Model and the Importance-Performance Analysis to investigate the following studies: (1) Travel Motivation and