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
漢語平比句式的句法語意研究
The Syntax and Semantics of Chinese Equatives
研 究 生: 陳奕勳 Yi-Hsun Eason Chen
指導教授: 劉辰生 博士 Dr. Cheng-Sheng Luther Liu
漢語平比句式的句法語意研究
The Syntax and Semantics of Chinese Equatives
研 究 生: 陳奕勳 Yi-Hsun Eason Chen
指導教授: 劉辰生 博士 Dr. Cheng-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
Submitted to Department of Foreign Literatures and Linguistics
Graduate Institute of Foreign Literatures and Linguistics
National Chiao Tung University
In partial Fulfillment of the Requirements
For the Degree of
Master
In
Graduate Institute of Foreign Literatures and Linguistics
June 2010
Hsinchu, Taiwan, Republic of China
漢語平比句式的句法語意研究
學生: 陳奕勳 指導教授: 劉辰生博士
國立交通大學外國語文學系外國文學與語言學碩士班
摘要
本篇論文主要探討中文的”一樣”和”不一樣”的句法語意特質，並進一步探討 三種比較句(由這兩個詞彙所核心構成(headed))的句法語意特性。這三種比較句 分別為程度平比句(scalar (un-)equatives)、像似比較句(similarity comparatives)及相 同比較句(identity comparatives)。首先，我們提供了許多證據，例如句法刪略 (ellipsis)、問句助詞的語意範域(scope of question particle)及結構歧義性(structural ambiguity)等，證明中文的”(不)一樣”在句法分布上的確可以擔任程度副詞和形容 詞(謂語)。再來，我們提出中文的”跟”以及”和”在比較句中做為介紹比較標準的 比較標記(comparative marker)時，其詞性為介係詞。關於程度比較句，我們提出 此類比較句是由程度副詞的”(不)一樣”所核心構成。程度副詞的”(不)一樣”在語意 上會引介出程度序列關係(ordering relation)，從這個角度來說，程度副詞的”(不) 一樣”在中文程度比較句中所擔任的句法及語意工作跟 more 在英文的程度比較 句中所擔任的非常相似，除了詞彙本身所引介的程度序列關係不同。另外，我們 提出中文的”(不)一樣”和英文中的 same 及 different 相同，均呈現了詞彙歧義性 (lexical ambiguity)，英文的部份請參考 Alrenga (2007)。進一步說，”(不)一樣”在 中文同時具有”像似語意(similarity reading)”以及”相同”語意(identity reading)。我 們針對”(不)一樣”的歧義性，進行句法和語意上的研究。一方面，我們提出像似 形容詞”一樣”可以帶一個控制結構的補語子句，然而像似形容詞”不一樣”卻不 行。我們進一步提出，如果把”一樣”所帶的補語子句看作是像似比較句中的衡量 詞組(measure phrase)，這個句法不對稱則可以歸因於”一樣”和”不一樣”引介的區間 性 質(interval nature) 不 同 ， 而 非 單 純 詞 彙 表 現 的 任 意 性 (idiosyncrasy of
lexicons)。另一方面，我們提出中文的像似比較句及相同比較句分別由”(不)一樣” 的兩種詞項所核心構成。進一步說，我們認為”(不)一樣”不僅是這兩種比較句在 語法語意上的核心成分，同時表現也像程度謂語一樣，決定了一個在相關量級 (scale)上正向或負向的區間。這樣的分析也幫助我們見到了—―—那些具有比較 語意的程度形容詞(comparative-like gradable adjectives)更深一層的本質。
The Syntax and Semantics of Chinese Equatives
Student: Yi-Hsun Eason Chen Advisor: Dr. Cheng-Sheng Liu
Graduate Institute of Foreign Literatures and Linguistics
National Chiao Tung University
Abstract
This thesis explores both syntax and semantics of yiyang ’same’, buyiyang ‘different’ and three types of comparative constructions (i.e. scalar (un-)equatives, similarity comparatives and identity comparatives) headed by them in Mandarin. In the first place, we present several pieces of evidence (e.g. ellipsis, the scope of question particle and structural ambiguity) for a necessary distinction between two different uses of yiyang and buyiyang, namely, degree adverbs and adjectival predicates. Secondly, we argue that the comparative marker gen/ he, which introduces the comparative standard, is prepositional in these three types of comparatives.This in turn suggests an adjunction analysis for the structural configuration of comparatives (e.g. Liu 1996, Kennedy 1999, Lin 2009). Regarding scalar (un-)equatives, we propose that they are both syntactically and semantically headed by degree adverbs
yiyang and buyiyang. Seen in this way, yiyang and buyiyang resemble the English
degree morpheme more in two respects. First, all of them are the head of comparatives. Second, all of them are degree morphemes introducing an ordering relation between individuals with respect to possessing some gradable property.
Turning to similarity comparatives, we first propose that yiyang and buyiyang, resembling their counterparts same and different in English, are lexically ambiguous between similarity and identity readings in Mandarin. Regarding the syntax of similarity predicates, we propose that yiyang (but not buyiyang) syntactically combines with a clause which is complement in nature, since the extraction of elements from it does not render island effects (i.e. CED effects in the sense of Huang 1982). Further, this complement clause functionally serves as a further specification of the dimension of similarity.
For another, by relating dimensions of similarity to degrees of similarity, we propose that the complement clause can be considered as measure phrases in similarity comparatives. Seen in this light, dubbed with Alrenga’s (2007) insight that comparative adjectives such as different and like determine positive and negative
intervals of a scale (i.e. similarity), we suggest that the syntactic asymmetry between
yiyang and buyiyang can not be attributed to the idiosyncrasy of lexicons; rather, it is
better considered as a reflex of the deeper syntax-semantics of measure phrases and the interval nature of buyiyang.
Concerning the semantics of similarity predicates, we argue against Alrenga ’s (2007) treatment of similarity same and different as a pair of total/partial adjectives; rather, we suggest that the two pairs of adjectives same/ different and yiyang/
buyiyang be better considered as the adjectives with totally closed scale (e.g. full/ empty, open/ closed), rather than the adjectives with partially closed scale (e.g. dry/ wet, straight/ bent) (see Rotstein & Winter 2004, Kennedy & McNally 2005). Finally,
we present a syntax-semantics analysis of similarity yiyang/ buyiyang and similarity comparatives in Mandarin. In particular, we propose that similarity comparatives are both syntactically and semantically headed by similarity predicates yiyang and
buyiyang.
As for identity comparatives, we point out two potential problems for Alrenga’s (2007) semantic analysis of identity same and different. The first problem is an empirical one, concerning the combination of same and proportion modifiers such as
almost and completely. The second one is theoretical in nature, concerning the
postulation of an abstract measure phrase. Given these considerations, I propose a syntax-semantics analysis of identity yiyang/ buyiyang and identity comparatives in Mandarin. Specifically, following Alrenga’s conception that individual identity itself constitutes as the dimension of comparison in identity comparatives, I propose that identity predicates yiyang and buyiyang not only syntactically and semantically head identity comparatives, but also determine positive and negative intervals of a scale (i.e. cardinality). Importantly, our analysis requires yiyang to return a positive interval on the relevant scale, and this move leaves room for how proportion adverbs semantically contribute to identity comparatives. Obviously, our analysis thus fares better than Alrenga’s with respect to the empirical problem. However, with respect to the theoretical problem, our analysis suffers the same pain as Alrenga’s does, since both analyses have to postulate an abstract measure phrase and the truth value of a comparative sentence relies on the combination of such degree morphemes.
Last but not the least, our analysis of similarity and identity comparatives sheds light on the nature of those comparative-like gradable adjectives such as yiyang and
buyiyang. More specifically, yiyang and buyiyang not only serve as the head of
comparatives (i.e. similarity and identity), but also resemble gradable adjectives in determining an interval on the relevant scale (i.e. similarity and cardinality).
致謝辭
論文終於完成了！此刻心中除了喜悅，還有滿滿的感激。首先，我要感謝我 的指導教授劉辰生老師。第一次遇見劉老師，是我在輔大語言所就讀時邀請他到 學校演講，演講前給予我的一席話，讓我受益良多。後來有幸進入交大，立即決 定請他指導論文。他這三年來給予我的指導與關懷，開拓了我學業上及人生上的 新視野，他對我的鼓勵和磨練讓我這三年在各方面成長了很多。他細心又耐心的 教學樹立了老師的典範，他努力不懈的精神樹立了學者風範，而這也時時刻刻提 醒著我不能偷懶懈怠。另外，我要感謝林若望老師。他是我的形式語意學啟蒙老 師，他深入淺出的教學讓我體驗到語意學真的是非常有趣也非常奧妙。林老師在 上課時候常常跟我們分享他讀書做研究的小秘訣，讓我覺得很幸運，得以窺探傑 出學者成功的方式。再來，我要感謝林宗宏老師。他是我去清大語言所上課認識 的一位好老師。他提綱挈領、簡明扼要的教學方式，讓我在不知不覺中快速吸收 了很多重要的知識。他犀利精闢的見解常常點出了很多我在思考上的盲點，這也 帶給我多元思考的可能性。非常感謝以上三位老師抽空參加我的論文口試，給予 我許多重要且寶貴的批評和意見，您們三年來對我的教導讓我成長很多也學習到 很多，謝謝您們！ 接著，我要感謝攻讀碩士期間曾教導過我的交大與清大的老師們 (許慧娟、 潘荷仙、劉美君、蔡維天、連金發)、輔大的老師們 (李子瑄、許洪坤)、師大的 老師們 (丁仁、謝妙玲、吳瑾瑋) 及政大的老師們 (詹惠珍、徐嘉慧、何萬順)， 謝謝你們讓我看到豐富多樣的語言現象及研究方法，謝謝！ 在此，我要感謝一些非常重要的朋友。首先，我要感謝輔大尤煒中學長，陪 我散步聊天喝多多，抒發心情的煩悶。輔大廖佩瑜及梁育群同學，很懷念和你們 一起吃粉圓、念瘋話的日子。台大林國喬同學，很高興在人生與學術的旅途中交 到你這個朋友，感謝你在許多方面給予我的建議，每次和你辯論或是聊彼此心中 理想的時候，時間總是過得特別快，希望我們都能早日完成心中那一幅名為夢想 的拼圖。清大學長姊李釗麟、蕭佩宜、蔡慧瑾，謝謝你們三年來對我的照顧、鼓 勵和關心。同梯戰友清大林仕岳同學與交大楊謦瑜同學，很開心在研究所時期認 識你們，祝福我們往後的研究生涯一切順利。交大的同學與學弟妹們盧柏宏、簡 蔓婷、黃皓志、蘇琬淇，感謝你們這些日子的陪伴，如果沒有你們，研究室的生 活真的是非常孤獨也非常苦悶。最後還要謝謝系辦助理陳雅玲、陳旅櫻一直以來 提供了各種學術資源的訊息。Last but not least, 我要感謝好友吳佳霖，謝謝妳一直以來默默的付出，體諒 我照三餐的有口無心，包容我所有的抱怨與碎碎念。謝謝妳給我的愛情，讓我成 長了很多。最後，我要感謝一直支持我的家人。我要對摯愛的爸爸陳炳堯、摯愛 的媽媽連秀卿與親愛的弟弟陳奕瑋獻上最深的敬意。因為有你們辛苦工作養家、 無私的奉獻付出，我才能專心完成這一趟碩士旅程。我要再次感謝我的家人，你 們無怨無悔的愛讓我覺得自己是個很幸運也很幸福的人，我愛你們！
Table of Contents
Chinese Abstract……….i English Abstract……….ii Acknowledgements………...iv Table of Contents………...v Chapter 1 Introduction………...1Chapter 2 Literature reviews………. 7
2.1 Previous Analyses of Chinese Equatives ………7
2.2 The Semantics of the Positive Form of Adjectives and Implicit Comparison………9
2.3 The Positive Morpheme in Mandarin and the Adjectival Structure………...14
2.4 Basic Facts about Chinese Equatives……….17
Chapter 3 Scalar (Un-)Equatives in Mandarin ………22
3.1 Truth Condition……….. 22
3.2 The Deictic Reading………...23
3.3 The Structural Ambiguity………...25
3.4 The Clausal Complement of Yiyang………...26
3.4.1 Ellipsis Behavior of Yiyang and the Scope of Question Particle Ma… 27
3.4.2 CED Effect………32
3.4.3 The Syntactic Status of Comparative Marker Gen/He in Chinese Equatives………36
3.4.4 Control Structure and the Antecedent of pro……….41
3.4.5 The Function of the Complement Clause………..43
3.5 Un-equatives in Mandarin………..45
3.6 The Syntax-Semantics of Scalar (Un-)Equatives in Mandarin……….. 48
Chapter 4 Similarity Comparatives in Mandarin……….52
4.1 The Lexical Ambiguity of Same and Different in English……….53
4.2 The Lexical Ambiguity of Yiyang and Buyiyangt in Mandarin………. 56
4.2.1 Scale Structure of Yiyang /Buyiyangt and Degree Modifications…….60
4.2.2 A Solution to the Syntactic Asymmetry between Yiyang and Buyiyang………..67
4.3 The Syntax-Semantics of Similarity Comparatives in Mandarin………….. 72
Chapter 6 Conclusions……….90 References………95
Chapter 1
INTRODUCTION
All languages have syntactic categories that express grade concepts, and also have designed comparative constructions, which are used to express ordering between two objects with respect to the degree or amount to which they possess some property (Sapir 1944). Although comparative constructions syntactically vary quite a bit from language to language, comparatives may universally have the following semantic constituents (the labels are meant to be descriptive), illustrated with an example from English.
(1) a. John is taller than Bill.
b. TARGET OF COMPARISON: John; GRADABLE PREDICATE: tall; COMPARATIVE MORPHEME: -er; STANDARD MARKER: than STANDARD OF COMPARISON: Bill
Concerning the ordering relation conveyed by comparatives, (un-)equatives are the comparative constructions used to express an (in-)equality relation between individuals with respect to the degrees of possessing some gradable property (i.e., the one introduced by the gradable predicate). On the other hand, aside from the comparatives involving the ordering relation between individuals along some quantitative dimension (e.g., height, width, length), there exist some comparative constructions involving a comparison relation between individuals along both quantitative dimensions and qualitative dimensions (e.g., color, shape).
What are these comparative constructions? As pointed out by Alrenga (2007), they are similarity comparatives (see also Huddleston & Pullum 2002). Similarity comparatives are the comparatives used to express a similarity or dissimilarity relation between individuals. As suggested by Alrenga (2007: 3), a key insight into the interpretation of similarity comparatives is that these are concerned with the distances that separate individuals’ locations along various dimensions of comparison (i.e., both quantitative and qualitative). In addition to the two types of comparatives (i.e., ordinary (scalar) comparatives and similarity comparatives) mentioned above, a third type of comparatives concerns the identity or non-identity relation between individuals. They are identity comparatives (see Heim 1985: 21, Beck 2000, Alrenga 2007: 5). Below, scalar (un-)equatives, similarity comparatives, and identity comparatives are illustrated with examples from English, in (2a), (2b) and (2c) respectively.
(2) a. John is (not) as tall as Bill/ John is (not) equally tall as Bill. b. John is the same/different as he was ten years ago.
c. The presenters at this year are the same as/ different from the presenters at last year.
According to Alrenga (2007), example (2c) reveals a variation of truth conditions. Under its identity reading, (2c) asserts that the presenters at this year are similar to the presenters at last year in all relevant repects. Under its similarity reading, (2c) asserts that the set consisting of the presenters at this year is identical in its membership to the set consisting of the presenters at last year.
Additionally, as is well observed by Alrenga, in scalar comparatives, the dimension of comparison is introduced grammatically by a gradable adjective
occurring in construction with the comparative head (e.g., tall-er, as tall). In contrast, similarity comparative heads (e.g., same, different) do not combine with any dimension-introducing expressions. This suggests that the dimensions relevant to their interpretation must be provided to them in some other fashion. On the other hand, concerning the dimension of comparison in identity comparatives they seem to take individual identity to itself constitute an attribute with respect to which individuals may differ, so that the dimension of comparison relevant to the identity readings in (2c) simply is the dimension of individual identity.
Turning to Mandarin, interestingly, three types of comparatives in (2) all involved the morphological forms yiyang ‘same’ and buyiyang ‘different’. See the examples below.
(3)Scalar (Un-)Equatives
a. Zhangsan gen Lisi yiyang gao. Zhangsan with Lisi same tall ‘Zhangsan is equally tall as Lisi.’
b. Zhangsan gen Lisi buyiyang gao. Zhangsan with Lisi different tall ‘Zhangsan is not equally tall as Lisi.’
cf. Lit. Zhangsan is unequally tall than Lisi.
(4) Identity comparatives and Similarity comparatives
a. Zhangsan de zhidaojiaoshou gen Lisi de zhidaojiaoshou yiyang ma? Zhangsan POSS adviser with Lisi POSS adviser same Q ‘Are Zhangsan’s adviser and Lisi’s adviser the same one?’
contextually salient properties?’
b. Zhangsan de zhidaojiaoshou gen Lisi de zhidaojiaoshou buyiyang Zhangsan POSS adviser with Lisi POSS adviser different ‘Zhangsan’s adviser and Lisi’s adviser are different ones.’
‘Zhangsan’s adviser is different from Lisi’s adviser with respect to some contextually salient properties.’
Note that both examples (4a-b) are ambiguous between similarity and identity readings. A more interesting fact is the example (5) below reveals a variation in truth-conditions as well.
(5) Zhangsan gen Lisi yiyang xihuan Mali. Zhangsan with Lisi same like Mary
Reading A: ‘Zhangsan is the same as Lisi in that they both like Mary.’
Reading B: ‘The degree to which Zhangsan likes Mary is the same as the degree to which Lisi likes Mary.’
Given these intriguing facts, I intend that (un-)equatives, similarity comparatives, and identity comparatives in Mandarin call for both syntactic and semantic analysis. In this thesis, following the terminology of Alrenga (2007), I use the term “identity comparatives” to refer to comparative constructions headed by yiyang and buyiyang when these adjectives receive identity interpretations. On the other hand, I use the the term “similarity comparatives” to refer to comparative constructions headed by yiyang and buyiyang when these adjectives receive similarity interpretations. Finally, I will use the term “Chinese Equatives” (henceforth CE) to refer to the three types of comparatives, namely, scalar (un-)equatives, similarity comparatives, and identity
comparatives in Mandarin.
This thesis is organized as follows. Chapter 2 gives a brief introduction about previous analyses on CEs, and some common views on gradable adjectives and comparatives in the formal literature. Chapter 3 presents a detailed discussion on three issues. One concerns the argument that both yiyang and buyiyang involve two uses, namely degree adverbials and adjectival predicates. Specifically, I provide several pieces of evidence (e.g., ellipsis, the scope of question particle and structural ambiguity) for a necessary distinction between these two different uses. Another concerns the syntax of similarity predicate yiyang and its clausal complement. In particular, I suggest that similarity predicate yiyang syntactically combines with a clause which is complemental in nature, since the extraction of elements from it does not render island effects (i.e., CED effects in the sense of Huang 1982). Further, this complement clause functionally serves as a further specification of the dimension of similarity. On the other hand, it is observed that the adjectival predicate buyiyang can not combine with such a complement clause. Thus, a syntactic asymmetry exists between yiyang and buyiyang. The final issue concerns the syntax-semantics of scalar (un-)equatives. Precisely, I show that they are both syntactically and semantically headed by degree adverbs yiyang and buyiyang. Chapter 4 consists of three themes. First, I show that yiyang and buyiyang, resembling their counterparts same and
different in English, are lexically ambiguous (between similarity and identity readings)
in Mandarin. Second, the syntactic asymmetry between yiyang and buyiyang can not be attributed to the idiosyncrasy of lexicons; rather, it is better considered as a reflex of the deeper syntax-semantics of measure phrases and the interval nature of buyiyang. Third, I provide a syntax-semantics analysis for similarity yiyang/buyiyang and similarity comparatives in Mandarin. Chapter 5 is dedicated to identity readings of
yiyang/ buyiyang and identity comparatives in Mandarin. Chapter 6 is the conclusion
Chapter 2
LITERATURE REVIEWS
2.1 Previous Analyses of Chinese Equatives
Chao (1968: 342) points out equatives X gen Y yiyang Adj in Mandarin have two different structural interpretations, depending on the categorial status of gen/he. If
gen/he is a preposition, then the interpretation would be ‘X is equally Adj with Y’; in
contrast, if gen/he is a coordinator, the interpretation would be ‘X and Y are equally Adj’. Inaddition, he also points out that the negation counterpart of the first structure is X bu gen Y yiyang Adj ‘X is not equally Adj with Y’, and that of the second one X
gen Y buyiyang Adj ‘X and Y are not equally Adj’.1 Seen in this way, there are two important issues here, one concerns the categorial status of comparative marker (i.e., standard marker), since it would result in different interpretations. The other concerns the syntactic category of yiyang ‘same’. It seems to Chao that yiyang is an adverbial element in equatives. Li and Thompson (1981: chapter 19) make this point more explicitly. They suggest a generalized schema for all comparative constructions in
1 Chao (1968) does not touch the issue whether bu ‘not’ is a lexical negation or a sentential negation in
buyiyang ‘different’. Morphologically speaking, bu may be a prefix or a free morpheme. However, in
this paper, it is shown that buyiyang demonstrates some idiosyncratic lexical properties distinct from
yiyang ‘same’. For example, buyiyang receives different patterns of degree modification, and buyiyang
(crucially but not yiyang) can occur in bi-comparatives.
Regarding the latter contrast, as suggested in Liu (2010a:17), the gradable adjective with lexical negation (but not the gradable adjective with sentential negation) are permitted to occur in
bi-comparatives, as shown in (i).
(i) a. *Zhangsan bi Lisi (geng) bu gao Zhangsan BI Lisi even more not tall b. Zhangsan bi Lisi (geng) bu-shufu. Zhangsan BI Lisi even more uncomfortable
In brief, there is a good amount of supporting evidence for us to make the assumption that bu ‘not’ is a lexical negation in buyiyang ‘different’, though I do not formally argue for this point in this paper.
Mandarin:
(6) X comparison word Y (adverb) dimension
In this schema, it is evident that yiyang and buyiyang are considered as adverbial elements in CEs, under the view of Li and Thompson. On the other hand, Lü (1980: 609) and Liu et al. (2001: 833) observe that the elements following yiyang could be verbs or adjectives (i.e., either semantically gradable or non-gradable). In addition, Zhu (1982: 177) and Liu et al. (2001) also observe that yiyang can serve as a predicate, when no elements follow it in the sentence.
However, a common problem for all previous analyses on CEs is that they are descriptive. They do not pay much attention to the syntax-semantics of gradable predicates and comparatives. Furthermore, regarding the contrasts among (3)-(5), none of the previous analyses can accommodate them. Despite these problems, these previous analyses have offered some important observations. First, yiyang can be employed either as an adverbial element or a predicate in Mandarin. Secondly, the categorial status of gen/he seems to determine the different interpretations of comparatives.
Based on these two basic observations, I explore the syntax and semantics of Chinese Equatives in this thesis. The main themes of this thesis are demonstrated as follows: (a) I argue that yiyang ‘same’ and buyiyang ‘different’ both can be used either as degree adverbs or adjectival predicates. Being degree adverbs, yiyang and buyiyang semantically establish an ordering relation (i.e., equality and inequality) between individuals/ objects with respect to some gradable property (the one introduced by the gradable predicate). In that case, yiyang/ buyiyang headed the (un-)equatives in Mandarin. (b) I argue that, resembling same and different in English, yiyang and
buyiyang are lexically ambiguous in Mandarin as well. More precisely, they are
ambiguous between similarity and identity readings. In these cases, yiyang and
buyiyang both headed the similarity and identity comparatives in Mandarin. (c) I
argue that yiyang (but not buyiyang) is syntactically permitted to take a complement clause, which functions as a specification of the dimension of similarity. Further, this syntactic asymmetry between yiyang and buyiyang can be regarded as a reflex of the deeper syntax-semantics of measure phrases and the interval nature of buyiyang. (d) Regarding the categorial status of gen/he, it is argued that gen/he is prepositional in nature, when it functions as a comparative marker introducing the comparative standard in the sentence. Otherwise, it may be a coordinator. In that case, the whole coordination complex semantically serves as the target of comparison, while the standard of comparison may be recovered from the context. (e) I will propose a syntax-semantics analysis for Chinese Eqauatives in detail.
In the following two sections (section 2.2 and 2.3), I briefly review some standard assumptions on the semantics of gradable predicates and comparatives in the formal literature. Finally, in section 2.4 I demonstrate some basic facts about CEs.
2.2 The Semantics of the Positive Form of Adjectives and Implicit
Comparison
In the formal semantics literature, it is widely assumed that gradable predicates do not themselves denote properties of individuals; rather, they map objects onto abstract representations of measure (i.e., scales) formalized as sets of values (i.e., degrees) ordered along some dimension (e.g., height, length, width) (see e.g., Cresswell 1977, von Stechow 1984, Heim 1985, Kennedy 1999, Graff 2000, Barker 2002, Kennedy & McNally 2005, Kennedy 2007a and Kennedy 2007b). In such a degree analysis of gradable predicates (in contrast to “the vague predicate analysis”),
a gradable adjective expensive is given a denotation like (7), where tall represents a measure function that takes an individual and returns its value, a degree on the scale associated with the adjective, so that tall(x) represents x’ height2.
(7) [[ tall ]] =λdλx.tall(x) ≥d
Pursuant to Graff (2000), Barker (2002), Kennedy & McNally (2005) and Kennedy (2007a), most gradable predicates have contextually dependent interpretation in the positive form (with a few exceptions). In addition, the positive form of a gradable adjective lacks overt morphology, in contrast to its comparative form (i.e., more
expensive and wider).
(8) a. This elephant is small. b. This ant is big.
(8a) could be judged true if asserted as part of a discussion about the size of elephants, but false in a discussion about the size of an ant versus an elephant. Likewise, (8b) could be judged true if asserted as part of a discussion about the size of ants, but false in a discussion about the size of an ant versus an elephant. One possible explanation
2 In fact, the denotation given in (7) is the relational analysis; under such a view gradable predicates
are analyzed as relations between individuals and degrees. On the other hand, some authors noted above treat gradable adjectives as functions from individuals to degrees (e.g., Kennedy 1999), as shown in (i).
(i) [[ tall ]] = λx.tall(x)
As pointed out in Kennedy (2005b:10), the crucial differences between the relational analysis and measure function analysis boil down to the following: “In the former, gradable adjectives introduce degree arguments which must be saturated to generate a property of individuals; while in the latter, gradable adjectives must combine with some other expression (possibly something that introduces a relation and a degree) in order to generate a property of individuals.”
In this thesis, I basically take the relational analysis, though shifting to the measure function analysis for the convenience of demonstrations on some occasion.
for this variability, as Kennedy (2005a, 2007a) and Kennedy & McNally (2005) argues, is to assume a degree morpheme pos (i.e., a covert positive morpheme) with a denotation in (9), where s is a context-sensitive function from measure function to degrees: it returns a contextually significant degree (i.e., the standard of comparison) of the gradable property measured by the adjective g.
(9) [[Deg pos]] = λgλx.g(x) ≥ s(g)
In other words, the positive form of adjectives is evaluated with respect to the context-sensitive function denoted by the covert positive morpheme: a DELINEATION FUNCTION (in the terminology of Kennedy) which maps a measure function to a degree that represents the standard of comparison based on the context of utterance. Furthermore, as pointed out in Graff (2000) and Kennedy (2005a), one fundamental semantic property of the positive form of a gradable adjective is that it is vague, and this vagueness leads to borderline cases: the cases in which it is not clear whether the predicate holds for the object or not (i.e., crisp judgment).
Most importantly, Kennedy (2005a) uses this semantic characteristic of the positive form to divide comparison in natural languages into two different modes, namely, explicit comparison and implicit comparison. Crucially, it is the latter that involves borderline cases (i.e., the cases leading to crisp judgment) but not the former. The definitions of explicit and implicit comparison is illustrated in (10), and the relevant examples are demonstrated in (11) and (12) respectively (see also Kennedy, 2007a and 2007b).
(10) a. Implicit comparison
gradable property g using the positive form by manipulating the context in such a way that the positive form true of x and false of y.
b. Explicit comparison
Establish an ordering relation between objects x and y with respect to gradable property g using special morphology (e.g.,, more/-er, less, or as) whose conventional meaning has the consequence that the degree to which x is g exceeds the degree to which y is g.
(11) Context 1: A 600-word essay and a 200-word essay (Kennedy 2005a:11) a. This essay is longer than that one.
long(e1) > long(e2)
b. Compared to that essay, this one is long. long(e1) > s[e2](long)
(12) Context 2: A 600-word essay and a 590-word essay a. This essay is longer than that one.
long(e1) > long(e2)
b. ??Compared to that essay, this one is long.
long(e1) > s[e2](long)
Explicit comparison in (12a) simply requires an asymmetric ordering relation between the degrees to which two objects possess the relevant property (i.e., the length of essay), the crisp judgments thus are not problematic.
However, implicit comparison in (12b) requires the first novel to have a degree of length that is significant relative to the region of the length scale whose lower bound is the length of the second essay. In other words, the differences between the
two degree values of length (i.e., the differences between the length of 600 words and the length of 200 words), as shown in Context 1, must be significantly greater than some contextually determined threshold specifying the degrees of length of that essay. Before leaving this section, I want to mention another common view on the meaning of comparative constructions in many recent analyses: the definite description of degrees (e.g., von Stechow 1984, Heim 1985, Kennedy 1999, Kennedy 2005a and Kennedy 2007b). The basic idea behind the view of definite description of degrees is that it presupposes an exactly-reading for the degree variable. That is, “Mary is d-tall” abbreviates that Mary has exactly the degree d on the tallness scale. In an at
least-reading for the degree variable the uniqueness presuppositions would not be
satisfied.
(13) a. Mary is taller than Bill (is).
b. [the d: Mary is d-tall] > [the d’: Bill is d’-tall]
In order to derive definite descriptions of degrees, a maximality operator is introduced as an essential component of meaning of comparatives in many recent analyses.
(14) a. Mary is taller than Bill (is).
b. max [d: Mary is d-tall] > max [d’: Bill is d’-tall]
To be brief, in this paper, following some common views on gradable predicates in the formal semantics literature, I adopt a degree analysis of gradable predicates (i.e., specifically, a relational analysis). Furthermore, I assume a maximality operator as a basic component in an analysis of meaning of comparatives. In the next section, I briefly discuss the positive morpheme and the adjectival structure in Mandarin, by
reviewing the work of Liu (2010a).
2.3 The Positive Morpheme in Mandarin and the Adjectival
Structure
As shown in the previous discussion, the positive morpheme is covert (i.e., without overt morphology) in a language such as English. Interestingly, different languages may vary in this regard. For example, as demonstrated in Sybesma (1999), the positive form of gradable adjectives in Mandarin is morphologically marked by the morpheme hen. Hen is sometimes glossed as very, but it also has a neutral interpretation that marks the positive form (see Sybesma 1999: 27 for discussion).
(15) a. Zhangsan hen gao. Zhangsan HEN tall ‘Zhangsan is tall.’ b. Zhangsan gao. Zhangsan tall
‘Zhangsan is taller (than X).’
Additional support for hen as an overt positive morpheme in Mandarin comes from the extensive study of Liu (2010a) on the occurring environment of the positive form of gradable predicates in Mandarin. Two important points are concluded in Liu (2010a).The first one is that the positive morpheme in Mandarin has two allomorphs: a covert one and an overt one (i.e., the degree word hen). Pursuant to Liu (2010a), the former, behaving like a polarity item, only occurs in a predicate-accessible domain with a structure where the head X0 carries the predicate-accessible operator[-wh] feature.
also licenses the occurrence of a degree phrase headed by the covert positive morpheme (i.e., Deg0). In contrast, the latter (i.e., the degree word hen) occurs in contexts elsewhere. Without running into the theoretical complexities of covert positive morpheme in Mandarin, I simply demonstrate the empirical contexts where the covert positive morpheme occurs in (16) ~ (21).
(16) The bu negation sentence a. Zhangsan bu gao. Zhangsan not tall
‘Zhangsan is not tall, and the possibility of Zhangsan’s being short is not excluded.’
(17) The contrastive focus construction a. Zhangsan gao, Lisi bu gao. Zhangsan tall Lisi not tall
‘Zhangsan is tall, but Lisi is not tall.’ b. Zhangsan gao, Lisi ai.
Zhangsan tall Lisi short
‘Zhangsan is tall, but Lisi is short.’
(18) The ma particle question a. Zhe duo hua hong ma? This CL flower red SFP ‘Is this flower red?’
(19) The conditional
a. Zhangsan yaoshi gao dehua, Lisi jiu bu ai. Zhangsan if tall PAR Lisi then not short ‘If zhangsan is tall, then Lisi is not short.’
(20) The epistemic adjectival small clause a. Zhangsan xiao [EA-SC ni sha].
Zhangsan deride you silly
‘Zhangsan derided you as being silly.’
(21) The construction ending with sentence final particle le a. Tian hei/ liang le.
Sky black/bright SFP ‘It got dark/It dawned.’
Another conclusion in Liu (2010a) is that Mandarin has a simpler adjectival structure than English. More specifically, English has a QP between the lower adjectival phrase and its functional degree projection (see Bresnan 1973, Corver 1997 and Neeleman et al. 2004 for discussions). In contrast, Mandarin simply has an adjectival structure introduced by a functional degree projection headed by the positive morpheme without having a QP in-between (see Liu 2010a: 44). Example (22) illustrates this point.
(22) a. Adjectival phrase in English: [DegP [Deg [QP[Q [AP [A ]]]]]]
In short, in this paper, following Sybesma (1999) and Liu (2010a), I assume that the degree word hen is the overt positive morpheme in Mandarin. I further assume that Mandarin has a simpler adjectival structure than English; in particular, an adjectival phrase does not contain a QP in Mandarin.
2.4 Basic Facts about Chinese Equatives
In the previous sections, I have introduced some common views concerning gradable adjectives and comparatives in the literature, which I assumed in this paper. In the following I will first review two semantic parameters concerning comparative construction (proposed in Huddleston & Pullum 2002), and I will indicate some interesting facts when we consider the Mandarin data with the two semantic parameters.
Huddleston & Pullum (2002: 1099) consider that there are two semantic parameters concerning comparative constructions, namely, (a) whether comparisons are concerned with relative position on some scale, such as that denoted by gradable adjective old or not; and (b) whether comparisons are concerned with equality relation or not. As shown in (23), two semantic dimensions of contrast yield the four types of comparative construction.
(23) Four types of comparative construction
EQUALITY INEQUALITY
SCALAR Kim is as old as Pat. Kim is older than pat.
NON-SCALAR I took the same bus as last time. I took a different bus from last time.
Note that the non-scalar comparison, in Huddleston & Pullum (2002), includes both identity vs. non-identity and likeness vs. unlikeness (i.e., similarity). For another,
when we consider the Mandarin data with two semantic dimensions indicated above, an interesting fact emerges: In Mandarin, comparisons concerned with equality relation (i.e., including both scalar and non-scalar) may all involve yiyang ‘same’; and comparisons concerned with inequality relation (i.e., including both scalar and non-scalar) may all involve buyiyang ‘different’. Examples (24)-(26) illustrate this point.
(24) Comparison of Degrees
a. Zhangsan gen Lisi yiyang gao. Zhangsan with Lisi same tall ‘Zhangsan is exactly as tall as Lisi.’
cf. ‘Zhangsan is equally tall as Lisi.’
b. Zhangsan gen Lisi buyiyang gao. Zhangsan with Lisi different tall ‘Zhangsan is not exactly as tall as Lisi.’ cf. Lit. Zhangsan is unequally tall than Lisi.
(25) Comparison of Identity and comparison of Similarity (properties)
a. Zhangsan de zhidaojiaoshou gen Lisi de zhidaojiaoshou yiyang ma? Zhangsan POSS adviser with Lisi POSS adviser same Q ‘Are Zhangsan’s adviser and Lisi’s adviser the same one?’
‘Is Zhangsan’s adviser the same as Lisi’s adviser with respect to some contextually salient properties?’
b. Zhangsan de zhidaojiaoshou gen Lisi de zhidaojiaoshou buyiyang Zhangsan POSS adviser with Lisi POSS adviser different ‘Zhangsan’s adviser and Lisi’s adviser are different ones.’
‘Zhangsan’s adviser is different from Lisi’s adviser with respect to some contextually salient properties.’
(26) Comparison of Similarity (properties)
a. Ta de daan gen wo de daan youdian/ jihu/chabuduo/ His answer with my answer slightly almost/ nearly/
wanquan yiyang. completely same
‘His answer is slightly/almost/nearly/completely the same as mine.’
b. Ta de daan gen wo de daan youdian/ xiangdang/ hen/ jihu/ His answer with my answer slightly/ quite/ very/almost/
wanquan buyiyang. completely different
‘His answer is slightly/quite/very/almost/completely different from mine.’
For one thing, in (24a-b), it seems that yiyang and buyiyang in Mandarin serve as degree adverbs (i.e., which saturates the degree argument of a gradable adjective). For another, (25a-b) is truth-conditionally ambiguous between the so-called
token-identity reading and type-identity reading in the literature. Last, yiyang and
buyiyang in (26a-b) act like gradable predicates in the sense that they receive degree
modifications. In addition to the interesting properties indicated above, consider the semantic contrast in (27a-b) below:
(27) a. Zhangsan gen Lisi yiyang gao. Zhangsan with Lisi same tall ‘Zhangsan is exactly as tall as Lisi.’
cf. ‘Zhangsan is equally tall as Lisi.’
b. Zhangsan gen Lisi yiyang hen gao. Zhangsan with Lisi same very tall
‘Zhangsan is the same as Lisi in that they both are (very) tall.’
The truth condition of (27b) requires that the height of Zhangsan and Lisi must exceed some contextually determined standard of tallness (i.e., they both must be tall and they are not necessarily of the same height), while that of (27a) does not. In other words, Zhangsan and Lisi in (27a) could both be short or tall as long as they are of the same height. On the other hand, it appears that yiyang is followed by an embedded clause in (28a-b). Note that temporal adverbials and evaluative adverbials are normally assumed to be IP-level and CP-level, respectively.
(28) a. Zhangsan gen Lisi yiyang mingtian hui qu taibei. Zhangsan with Lisi same tomorrow will go Taipei
‘Zhangsan is the same as Lisi in that they both will go to Taipei tomorrow.’ b. Zhangsan gen Lisi yiyang hen xingyundi zhong le letou. Zhangsan with Lisi same very luckily win ASP lottery
‘Zhangsan is the same as Lisi in that they both luckily win the lottery.’
Given all contrasts demonstrated above, several questions are immediately raised: (a) is it possible to propose a unified account for all contrasts illustrated above? (b) If not, how many yiyang and buyiyang are necessary to be semantically and syntactically distinguished in Mandarin? (c) What are the semantic contributions of
yiyang and buyiyang to the truth condition of a sentence in Mandarin? (d) What are
(dis-)similarity comparatives in Mandarin? In this thesis I will answer these questions in order.
Chapter 3
SCALAR (UN-)EQUATIVES IN MANDARIN
In this chapter, from section 3.1 to 3.3 I first present several pieces of supporting evidence for the argument that yiyang in Mandarin potentially can occupy two different syntactic positions, specifically, a degree adverb and a predicate position. Then, in section 3.4, I explore the syntax of predicate yiyang, and examine the syntactic nature of comparative marker gen/he. Finally, in section 3.5 and 3.6, I return to the data about scalar (un-)equatives, and propose a syntax-semantics analysis for them.
3.1 Truth Condition
The first piece of evidence comes from the variation of truth conditions with respect to the two different syntactic positions of yiyang. As pointed out in section 2.4, sentences (27a-b) are truth-conditionally different, here repeated as (29a) and (30a). The logical representations and truth conditions of (29a) and (30a) are shown in (29b-c) and (30b-c), respectively (Note that c stands for some contextually determined standard of tallness).3
(29) a. Zhangsan gen Lisi yiyang gao. Zhangsan with Lisi same tall ‘Zhangsan is exactly as tall as Lisi.’
3 Notice that the descriptive version (30b) here does not involve the semantics of yiyang “same”. As it
will become clear, (30a) actually involves a comparison of similarity. The formal version will be provided in chapter 4.
cf. ‘Zhangsan is equally tall as Lisi.’
b. max [d: Zhangsan is d-tall] = max [d’: Lisi is d’-tall]
c. The degree d such that Zhangsan is d-tall equals the degree d’ such that Lisi is d’-tall.
(30) a. Zhangsan gen Lisi yiyang hen gao. Zhangsan with Lisi same very tall
‘Zhangsan is the same as Lisi in that they both are (very) tall.’ b. [gao(Zhangsan)≥ c ∧∧∧∧ gao(Lisi)≥ c]
c. The degree d such that Zhangsan is d-tall exceeds the contextual standard of tallness and the degree d’ such that Lisi is d’-tall exceeds the contextual standard of tallness.
Descriptively speaking, yiyang in (29a) introduces an equality relation between individuals in terms of degrees (i.e., the sameness of degrees of tallness), while in (30a) in terms of properties (i.e., the sameness of property of being (very) tall). To put it differently, it would be not unreasonable for us to propose that yiyang could potentially occupy two different syntactic positions since (29a) and (30a) are truth-conditionally different. This in turn suggests that (29a) and (30a) should be considered as two different comparative constructions in Mandarin.
3.2 The Deictic Reading
The second piece of evidence stems from the deictic reading of same in English. Carlson (1987: 532) points out that there are two readings concerning the use of same in English. One is deictic reading. Under such reading, the comparative standard is identified in the previous context. The other is the interpretation of an internal reading,
which requires the comparative standard be identified within the sentence (i.e., without referring to the previous context). The two readings are illustrated in (31c-d) respectively (see also Dowty 1985, Moltmann 1992, Lasersohn 2000, Beck 2000 and Barker 2007 for discussions).
(31) a. Mary read The Old Man and The Sea.
b. John and Bill read the same book. (Ambiguous between two readings) c. John and Bill both read The Old Man and The Sea. (The deictic reading) d. John read the book that Bill read. (The sentence internal reading)
Interestingly, the deictic reading of yiyang in Mandarin yields two different syntactic patterns, which is completely unexpected under the view that yiyang is located in exactly one syntactic position.
(32) a. Zhangsan gao yi bai wushi gongfen, Lisi ye yiyang gao.4
4 Under the unified account, one may argue that the underlying structure of (32a) is indeed (i), which
undergoes PF-deletion of the measure phrase yi bai wushi gongfen “one hundred and fifty centimeters”, so that yiyang has the same syntactic status (i.e., a predicate) as in (32b). .
(i) Zhangsan gao yi bai wushi gongfen, Lisi ye yiyang gao [yi bai wushi gongfen].
However, there are both theoretical and empirical problems for this analysis. Theoretically speaking, such a PF-deletion analysis would predict that the degree argument of an adjective such as gao can be saturated simply at LF (i.e., without PF-realization of the degree argument), contrary to the fact, as shown in the empirical data (ii). Alternatively, one may still argue that (i) involves a covert positive morpheme, which saturates the degree argument. However, if that is the case, it would lead to semantic
anomaly, since the sentence Zhangsan gao yi bai wushi gongfen ‘Zhangsan is one hundred and fifty
centimeters tall’ does not entail that Zhangsan is tall, and one hundred and fifty centimeters does not constitute as the contextual standard of tallness (i.e., the standard for being tall) in normal situations. This semantic anomaly is demonstrated in (iii). In short, (i) can not be justified as the underlying structure for (32a).
Zhangsan tall one hundred fifty centimeter, Lisi also same tall ‘Zhangsan is one hundred and fifty centimeters tall, Lisi is equally tall.’ b. Zhangsan gao yi bai wushi gongfen, Lisi ye yiyang
Zhangsan tall one hundred fifty centimeter, Lisi also same ( gao yi bai wushi gongfen ).
tall one hundred fifty centimeter
‘Zhangsan is one hundred and fifty centimeters tall; Lisi is the same as Zhangsan, too.’
Crucially, the deictic reading of yiyang in (32a) refers to the contextually salient degree to which Zhangsan is tall, namely, the measure phrase yi bai wushi gongfen “one hundred and fifty centimeters”. On the other hand, however, the deictic reading of yiyang in (32b) presupposes an individual holding the contextually salient property, namely, being one hundred and fifty centimeters tall. The basic idea here is that the deictic reading of yiyang revealing two different patterns is actually predicted under a non-unified account. Imaginably, a unified account has to make some stipulations in order to explain why this is so.
3.3 The Structural Ambiguity of Yiyang
The final piece of evidence comes from the case of structural ambiguity. Recall that it is argued that yiyang potentially can occupy two different syntactic positions (either degree adverbs or adjectival predicates) in Mandarin. Since a degree adverb is normally incompatible with a non-gradable predicate, it is expected that the
(iii) ??Zhangsan gao yi bai wushi gongfen, Lisi ye yiyang gao [ yi bai wushi gongfen]. .
Intended meaning: Zhangsan is one hundred and fifty centimeters tall; Lisi is the same as Zhangsan in that he is (very) tall, too.
modification of the degree adverb yiyang to a non-gradable predicate would lead to semantic anomaly. Example (33) verifies this expectation.
(33) Zhangsan gen Lisi yiyang you yi ge erzi. Zhangsan with Lisi same have one CL son ‘Zhangsan is the same as Lisi in that they both have a son.’
*’The degree to which Zhangsan has a son is the same as the degree to which Lisi has a son.’
Crucially, if the predicate is gradable, the sentence becomes ambiguous. Compare (33) and (34):
(34) Zhangsan gen Lisi yiyang xihuan Mali. Zhangsan with Lisi same like Mary
‘Zhangsan is the same as Lisi in that they both like Mary.’
‘The degree to which Zhangsan likes Mary is the same as the degree to which Lisi likes Mary.’
Again, the contrast between (33) and (34) is unexpected under a unified account. The important point here is that (34) can be regarded as a case of structural ambiguity. More specifically, the two different syntactic positions of yiyang structurally contribute to the variation of truth conditions of a sentence.
3.4 The Clausal Complement of Yiyang
In the following several subsections, I present a detailed discussion about the syntax of yiyang. In particular, I will first show two pieces of evidence for the
predicative status of yiyang, and propose that yiyang is a similarity predicate in the sense that it semantically introduces a similarity relation between individuals/objects. Secondly, I argue that yiyang takes a clause involving control structure, and further that the embedded clause is complement in its syntactic nature since extraction of syntactic elements from the clause does not render island effects (i.e., CED effects, in the sense of Huang, 1982). Next, I argue that the categorial status of gen/he is preposition in nature, when they functions as comparative markers introducing the
comparative standard in Chinese Equatives. Importantly, this does not exclude the
possibility for gen/he to be a coordinator in Chinese Equatives. Finally, I propose that the clausal complement functions as a specification of the dimension of similarity, by reviewing McCawley’s (1970) discussion about in that clause in similarity construction in English.
3.4.1 Ellipsis Behavior of Yiyang and the Scope of Question Particle
Ma
The first supporting evidence comes from the ellipsis behavior of yiyang. As well observed in the literature, in Mandarin Chinese, sentences involving ellipsis need an operation like the do-support in English to insert an auxiliary after the application of PF-deletion; otherwise, a sentence will be ungrammatical, as the contrast in (35a-b) indicate. Alternatively, some predicative elements but not adverbial elements seem to license the elided elements as well, as the contrast in (36a-b) and (37a-e).5
5
Luther Liu (p.c.) points out that the function of elements (such as shi ‘is’, hui ‘will’ and xihuan ‘like’) may be to support the predicate position in the second conjunct involving ye ‘also’, instead of licensing the elided elements. He provides the following contrast:
(i) a. Zhangsan mingtian qu Taibei, Lisi houtian b. *Zhangsan mingtian qu Taibei, Lisi ye houtian
‘Tomorrow Zhangsan will go to Taipei; the day after tomorrow, Lisi will, too.’
As shown above, (ia) is well-formed despite the fact that there are no supporting elements (i.e., predicative elements) to license the elided elements in the second conjunct. For this moment, I leave
(35) a. Zhangsan xihuan chi pinguo, Lisi ye shi [ e ]. Zhangsan like eat apple Lisi also is
‘Zhangsan likes eating apples, so does Lisi.’
b. *Zhangsan xihuan chi pinguo, Lisi ye [ e ]. Zhangsan like eat apple Lisi also
(36) a. Zhangsan xihuan chi pinguo, Lisi ye xihuan [ e ]. Zhangsan like eat apple Lisi also like
‘Zhangsan likes eating apples, Lisi likes, too.’
b. Zhangsan mingtian hui qu Taibei, Lisi ye hui [ e ]. Zhangsan tomorrow will go Taipei Lisi also will ‘Zhangsan will go to Taipei tomorrow, Lisi will, too.’
(37) a. *Zhangsan changchang qu Taibei, Lisi ye changchang [ e ]. Zhangsan often go Taipei Lisi also often
Lit. Zhangsan often goes to Taipei, Lisi often, too.
b. *Zhangsan dashengdi chang zhe ge, Lisi ye dashengdi [ e ]. Zhangsan loudly sing ASP song, Lisi also loudly
Lit. ‘Zhangsan is singing songs loudly, Lisi loudly, too.’
c *Zhangsan zai gongyuan chi pinguo, Lisi ye zai gongyuan [ e ]. Zhangsan at park eat apple Lisi also at park
Lit. ‘Zhangsan eats apples at park, Lisi at park, too.’
d. *Zhangsan zuotian chi le pinguo, Lisi ye zuotian [ e ].
open the issue whether the function of the elements (such as shi ‘is’, hui ‘will’ and xihuan ‘like’) is to rescue the second conjunct involving ye ‘also’ by supporting the predicate position, or to license the elided elements. The crucial point here is that yiyang resembles shi ‘is’, hui ‘will’ and xihuan ‘like’ in that they all are predicates.
Zhangsan yesterday eat ASP apple Lisi also yesterday Lit. ‘Zhangsan ate apples yesterday, Lisi yesterday, too’
e. *Zhangsan hen xingyundi zhong le letou, Lisi ye hen Zhangsan very luckily win ASP lottery Lisi also very
xingyundi [ e ].
luckily
Lit. ‘Zhangsan luckily wins the lottery, Lisi luckily, too.’
Importantly here, as indicated in (37a-e), adverbial elements cannot license the elided elements regardless of the syntactic levels of the adverbs. More specifically, under normal circumstances, frequency adverbs changchang ‘often’, manner adverbs
dashengdi ‘loudly’ and locative adverbial phrases zai gongyuan ‘at park’ are assumed
to be VP-level; temporal adverbs zuotian ‘yesterday’ associate with IP-level; and evaluative adverbs hen xingyundi ‘very luckily’ are assumed to be CP-level. In contrast, predicate elements such as shi ‘is’, xihuan ‘like’ and hui ‘will’ can license the elided elements. Given the contrasts above, consider the following examples (38) and (39), where yiyang surprisingly licenses the elided elements.
(38) a. Zhangsan you yi ge erzi, Lisi ye yiyang you yi ge erzi. Zhangsan have one CL son Lisi also same have one CL son ‘Zhangsan has a son; Lisi is the same as Zhangsan in that he has a son,
too.’
b. Zhangsan you yi ge erzi, Lisi ye yiyang [ e ]. Zhangsan have one CL son Lisi also same
(39) a. Zhangsan hen gao, Lisi ye yiyang hen gao.. Zhangsan very tall Lisi also same very tall
‘Zhangsan is (very) tall; Lisi is the same as Zhangsan in that he is (very) tall, too.’
b. Zhangsan hen gao, Lisi ye yiyang [ e ]. Zhangsan very tall Lisi also same
‘Zhangsan is (very) tall; Lisi is the same as Zhangsan, too.’
In contrast, without the occurrence of yiyang, the sentences became ungrammatical since ye is an adverbial element, which can not license the elided elements.
(40) a. *Zhangsan you yi ge erzi, Lisi ye [ e ]. Zhangsan have one CL son Lisi also Lit. Zhangsan has a son; Lisi, too.
Intended meaning: ‘Zhangsan has a son; Lisi is the same as Zhangsan, too.’
b. *Zhangsan hen gao, Lisi ye [ e ]. Zhangsan very tall Lisi also Lit. Zhangsan is (very) tall; Lisi, too.
Intended meaning: ‘Zhangsan is (very) tall; Lisi is the same as Zhangsan, too.’
The ability of licensing the elided elements suggests that yiyang in Mandarin be predicative in its syntactic nature, rather than an adverbial element, in these cases.
Additional support comes from the scope of question particle ma. As widely assumed in the literature, the scope of question particles such as ma can not be
embedded, it must take matrix scope. The prediction here is that the question particle
ma would scope over the whole comparative construction, since yiyang is employed
as the predicate (as previous discussions suggested above) in the matrix clause and question particle ma can not be embedded. Fortunately, example (41) witnessed the prediction.
(41) a. Zhangsan gen Lisi yiyang you yi ge erzi ma? Zhangsan with Lisi same have one CL son Q ‘Is Zhangsan the same as Lisi in that they both have a son?’
b. Zhangsan gen Lisi yiyang hen gao ma ? Zhangsan with Lisi same very tall Q
’Is Zhangsan the same as Lisi in that they both are (very) tall?’
Last, the temporal adverbial such as mingtian ‘tomorrow’ and the evaluative adverb such as hen xingyundi ‘very luckily’ are normally assumed to associate IP-level and CP-level respectively. This in turn suggests that the syntactic nature of the element following yiyang ‘same’ be exactly a clause. Example (28) is repeated here as (42).
(42) a. Zhangsan gen Lisi yiyang mingtian hui qu taibei. Zhangsan with Lisi same tomorrow will go Taipei
‘Zhangsan is the same as Lisi in that they both will go to Taipei tomorrow.’ b. Zhangsan gen Lisi yiyang hen xingyundi zhong le letou. Zhangsan with Lisi same very luckily win ASP lottery
In short, I have presented two pieces of evidence for the predicative status of
yiyang in Mandarin. Also, I show that yiyang in Mandarin indeed takes an embedded
clause67. In the next section 3.4.2, I argue that this embedded clause is complement in its syntactic nature by demonstrating the fact that extraction of syntactic elements from the clause does not render island effects.
3.4.2 CED Effect
Huang et al. (2009, Chapter 6) suggest that relatives in Mandarin could be syntactically formed in two ways; one involves movement, whereas the other does not. In particular, they make the following generalizations (Huang et al. 2009: 225):
(43) a. Relatives with a gap in argument position:
A relative can be derived by directly raising the nominal to be relativized to the Head position. The Head is related to the trace in an argument inside the relative.
b. Relatives with the Head related to an adjunct or a pronoun in an argument position:
The Head of the relative is base-generated. The Head-relative clause relation is via a relative operator at the peripheral position of the relative clause.
6 In fact, it is not surprising that adjectives can take an embedded clause. In English, an evaluative
adjective can take an infinitival clause. Further, this infinitival clause is adjunct in its syntactic nature (see Stowell 1991 and Landou 2009 for discussions). Example (i) is borrowed from Landou (2009). (i) a. %? To whom was it stupid of John to talk?
b. *John went home, as it was smart of john___/ as he was smart__. c. *How stupid to leave town was it of John?
d. *How stupid of John to leave town was it?
Regarding the infinitive, it is shown that it creates a weak island (ia), cannot be gapped in a clause (ib), and cannot be pied-piped (ic, id).
7
I am grateful to Jonah Lin for bringing my attention to the argument structures of adjectives in English.
Regarding the relatives involving movement, island conditions are undoubtedly relevant. Consider the example below:
(43) a.*[[Wo renshi henduo[[ ei xihuan] de] ren de] na ge nuhaii]
I know many like DE person DE that CL girl Intended meaning: the girl that I know many people who ei likes
b. *[[Wo hen xihuan [[ei chang ge] de] shengyin de] na ge nuhaii]
I very like sing song DE voice DE that CL girl Intended meaning: the girl that I like the voice with which ei sings
c. *Wo xiang kan [[ni [yinwei ei bu hui lai] hen shengqi
I want see you because not will come very angry de] [na ge xueshengi].
DE that CL student
Intended meaning: I want to see the student with whom you are angry because he would not come.
Now, to make it more complex, consider the cases where the relativized nominal is originally within the embedded clause following the predicate yiyang ‘same’ in Mandarin.
(44) a. *Zhangsan gen Lisi yiyang [xiangxin [[Wangwu xihuan ei] de]]
Zhangsan and/with Lisi same believe Wangwu like DE
yaoyan de] [na ge nuhaii]
rumor DE that CL girl
