towards a morphosyntactic analysis of mandarin mood/aspect

Microsoft Word - LL13.4-03-Shu 20120704.doc2012-0-013-004-000335-1
Chih-hsiang Shu
Academia Sinica
This paper provides a novel descriptive and theoretical account for the ‘down-play’ mood marker ge in Mandarin Chinese. One essential property that is not covered by previous studies is its clitic status. Incorporating and enhancing Yang’s (2001) clitic treatment of numeral-less classifiers, I propose a morphosyntactic analysis that identifies the morphological-template for the mood marker ge − [VP V-(Xσ)-ge-YP[+N]]. It is shown that this template is quite similar to the morphological template for internal affixes like -de- in Chinese. These findings suggest an updated treatment of clitics as ‘phrasal affixes’ (Klavans 1982, 1985, Anderson 1992, 2005). Furthermore, it supports Aronoff’s (1994) treatment of morphology as an independent component in grammar, and Chomsky’s (2000 et seq.) Agree framework, which allows syntactic operations to apply before morphological operations. Key words: mood, aspect, classifiers, clitics, morphology, Agree theory
1. Introduction
Typically, Mandarin Chinese ge () serves as the most common classifier (cf. Chao 1968, Tang 1990, Yang 2001, etc.). As a classifier, it is obligatory, attached to numerals, and precedes count nouns, as illustrated below.
(1) a. lisi jian-le yi-ge ren L. meet-Pfv one-Cl person1 ‘Lisi met a person.’
* I am deeply grateful to Mark Aronoff, Alice Harris, Robert Hoberman, and Richard Larson, who
helped the early drafts of this paper to take shape. I have also benefitted greatly from two reviewers, whose comments and suggestions improved the content and presentation of the paper.
1 The abbreviations for the glosses used in this paper are as follows: CAI/JIU = connective adverbs, Cl = nominal classifier, DE = nominal/verbal modification marker, TA = non-referential use of third-person singular pronoun, VC = verbal classifier, Pfv = perfective aspect, Q = question particle.
Chih-hsiang Shu
664
b. ta xie-le liang-ge zi he write-Pfv two-Cl character ‘He wrote two characters.’
c. zhe shi-ge fangfa hen qiguai this ten-Cl method very strange ‘These ten methods are very strange.’
In a different set of syntactic and semantic environments, ge can serve as a ‘down-play’ mood marker (Chao 1968, Lü 1980, and Lin 2001) which imposes the telic aspect (Wu 2002). In this article I shall alternatively call it a ‘verbal classifier’ for convenience. This ‘verbal’ ge is attached to verbs or verbal complex and precedes various indefinite nominal constituents, including those already containing a classifier, as shown below:
(2) a. lisi xiang he ge shui (mass noun) L. want drink VC water ‘Lisi wants to drink water.’
b. zhangsan zhe-ci gai fu ge ze (object part of a VO compound) Z. this-time should bear VC responsibility ‘Zhangsan should take responsibility this time.’
c. zhangsan pao-le ge liang-fenzhong (indefinite temporal NP) Z. run-Pfv VC two-minute ‘Zhangsan ran for two minutes.’
d. zhangsan zhi kan-le ge ji-bu dianying2 (indefinite NP) Z. only see-Pfv VC several-Cl movie ‘Zhangsan only saw several movies.’
In two further sets of distinct syntactic and semantic environments, ge can serve as a ‘maximum degree’ mood marker (Biq 2004) and ‘excessiveness’ mood marker (Yeh 2006), respectively, as illustrated below:
(3) a. lisi ke wan-de ge tongkuai L. indeed play-DE VC2 thorough.satisfaction ‘Lisi indeed played to his heart’s content.’
2 The sentence is not well-formed without the marker of exhaustive identification marker zhi.
This is presumably due to the ‘down-play mood’ semantic requirement of ge. I shall leave the semantic details for future research.
Towards a Morphosyntactic Analysis of Mandarin Mood/Aspect Marker ge
665
b. zhangsan xiao ge bu ting Z. laugh VC2 no stop
‘Zhangsan keeps laughing (and there is no sign of his stopping.)’
(4) a. mai ta ge yi bai-ben shu (Examples from Yeh 2006) buy TA VC3 one hundred-Cl book ‘To buy one hundred books.’
b. wan ta ge guoyin play TA VC3 to.one’s.satisfaction ‘To play to a thorough satisfaction of it.’ The issue we are interested in is the morphological and syntactic status of the mood/aspect marker ge. What is its syntactic category? What are the syntactic and morphological conditions that constrain its morphosyntactic distributions? What is its syntactic position?
Previous studies have either treated the mood/aspect marker as a vP-level element (Lin 2001), based on the assumption that there is V-to-v movement but no V-to-T movement in Chinese and ge occurs post-verbally, or a DP-level element (Wu 2002), based on the assumption that ge is derived from the nominal classifier ge and that ge has selectional restrictions with its nominal complement, namely the latter has to be indefinite. These analyses are problematic since an expression cannot be both a D element and a v element at the same time, unless it is composed of two elements, and they also face an important theoretical problem: why does a mood expression occupy a vP-internal position instead of a vP-external position?
In this paper, I shall focus on describing and accounting for the (near) affixal nature of the mood/aspect marker ge illustrated in (2), which will provide clues to the morphosyntactic status of ge in (3) and (4), and offer an up-to-date theoretical account. It will be shown that ge is neither an auxiliary verb nor an adverb, and does not have the morphosyntactic status of a phrase nor a free-standing word. More precisely, it should be treated as a clitic attached to a verb or a verbal complex, just as a nominal classifier is attached to a numeral or other types of hosts in Chinese (Yang 2001). Furthermore, it will be shown that the verbal classifier ge has the additional morphosyntatic requirement that it has to be overtly followed by a nominal constituent, due to its morphological templatic coding. None of these morphosyntactic properties can be fully accounted for in the verb movement and ge-as-a-D-head analysis. I shall, alternatively, adopt an updated Chomsky’s (2000 et seq.) Agree framework and Aronoff’s (1994) idea that morphology is an independent component in grammar, which can jointly provide a natural account for these morphosyntactic properties.
The rest of this paper is organized as follows. In §2, I shall discuss the general properties of the verbal classifier ge. In §3, I shall address the pros and cons of previous
Chih-hsiang Shu
666
studies. In §4, I shall show that the problems can be resolved once we identify the affixal nature of ge. In §5, I shall show that the properties of ge support the thesis that there is an independent Morphology component in grammar, and provide an Agree analysis. Section 6 concludes the paper.
2. Major semantic and syntactic properties of the verbal classifier ge
2.1 Semantic properties As mentioned in the previous section, the major semantic functions of the verbal
classifier ge in (2) is to express a ‘down-play’ mood and the telic aspect. The presence of the ‘down-play’ mood can be shown in the following examples:
(5) a. jie wo ge qian, hao ma? lend me VC money ok Q ‘(Just) lend me some money, will you?’
b. xi ge yifu bu yong hua duoshao shijian wash VC clothes no need cost much time ‘(Just) doing laundry doesn’t take much time.’
c. chuan ge lifu zheme tongku ma? wear VC tuxedo so painful Q
‘Is it so painful to (just) wear a tuxedo?’
(6) a. jie wo (*ge) qian, buran wo yihou ye bu jie ni qian lend me VC money otherwise I afterwards no lend you money ‘Lend me some money, otherwise I won’t lend you money in the future.’
b. xi (*ge) yifu hui hua hen duo shijian wash VC clothes will cost very much time ‘Doing laundry takes a lot of time.’
c. chuan (*ge) lifu zheme suibian ma? wear VC tuxedo so casual Q ‘Is it right to wear a tuxedo so carelessly?’
The only difference between (5) and (6) is the perceived importance of the event, which determines whether ge is present or not. In (5a), because of the use of ge, the request is not perceived as urgent or imperative, so the speaker can be polite by hedging the question with hao ma. On the other hand, when the request is given with a threat (6a), which suggest it is important that the addressee fulfill the request, ge cannot be used. If
667
ge is used, (6a) would both convey that the request is not important and that the request is important, which is contradictory. Similarly, in (5b) the use of ge suggests the laundry-doing is perceived as an insignificant task, and it is natural to say that such a task does not take much time. On the other hand, ge is not allowed in (6b) since it does not make sense to say a task perceived as insignificant generally takes a lot of time. Lastly, in (5c) the use of ge suggests wearing a tuxedo is a significant task, and it’s natural to assume an insignificant task not to be painful, therefore the question makes sense. However, in (6c), it is anomalous to assume that a task perceived as insignificant should be done seriously, so ge should not be allowed.
Further contrast can be seen in sentences modified by different evaluative adverbs jiu and cai, which mark consequent clauses in two types of conditional clauses: jiu marks the consequent clause of if-clauses, while cai marks the consequent clauses of only if-clauses.
(7) a. ni chi ge yi-pian pisa, jiu neng qu kan dianshi you eat VC one-slice pizza JIU can go watch TV ‘(Just) eat one slice of pizza, then you can go watch TV.’
b. ni chi (*ge) yi-pian pisa, cai neng qu kan dianshi you eat VC one-slice pizza CAI can go watch TV ‘Only if you eat one slice of pizza can you go watch TV.’
The difference between (7a) and (7b) is that the antecedent clause in (7a) is a sufficient condition, and the antecedent clause in (7b) is a necessary condition. By the law of logic, the positive truth value of a sufficient condition is not necessarily crucial (and therefore may not be important) for the positive truth value of its consequent clause. A sufficient condition should be compatible with the mood marker ge. On the other hand, since the positive truth value of a necessary condition is crucial (and therefore important) for the positive truth value of its consequent clause, a necessary condition should not be compatible with the ‘play-down’ mood marker ge. The facts in (7) thus conform to the semantics of ge.
The semantic property of telicity can be shown in the following sentences: (8) a. [A: What did Lisi do yesterday?]
B: ta da-le ge quan he hit-Pfv VC fist ‘He practiced some martial arts.’
b. [A: What does Lisi like to do as a hobby?] B: da (*ge) quan
Chih-hsiang Shu
668
(9) da (*ge) quan shi hao yundong hit VC fist be good sport ‘Practicing martial arts is a good sport.’
(10) lisi zai da (*ge) quan L. Prog hit VC fist ‘Lisi is practicing martial arts.’ In (8a), the sentence depicts an episodic event, and there is no problem for the occurrence of ge. In (8b), however, the sentence depicts a generic event, and ge is not possible here. The same is true in (9). (10) shows that the verbal classifier ge is incompatible with progressive aspect, which is also observed by Wu (2002). In general, telic events are incompatible with generic aspect and progressive aspect, so these facts show ge marks telic aspect.
Further evidence of the telicity-marking function comes from the fact that the ge can only co-occur with indefinite noun phrases.3 Consider the following examples:
(11) a. zhangsan zhi kan-le (*ge) zhe ji-bu dianying Z. only see-Pfv VC this several-Cl movie ‘Zhangsan just saw these movies.’
b. lisi chi-le (*ge) na liang-wan fan L. eat-Pfv VC that two-bowl rice ‘Lisi ate those two bowls of rice.’
c. lisi hua-le (*ge) na wu-kuai qian L. spend-Pfv VC that five-Cl dollar ‘Lisi spent those 5 dollars.’
According to Liu (2006), there is a relationship between the telicity/atelicity of an
event and the definiteness/indefiniteness of the object noun phrase in the sentence. More precisely, enhancing Tenny’s (1994) Aspectual Interface Hypothesis, Liu argues that a telic sentence requires an event measure, and that while an indefinite noun phrase is an event measure, a definite noun phrase is not. This means that a definite noun phrase does not contribute to a telic event. She listed the following examples as evidence for her proposal:
3 I am grateful for a reviewer who pointed out to me that Liu (2006) has insightful accounts of
the relations between telicity and definiteness/indefiniteness of object noun phrases.
669
(12) ni neng zai shi fenzhong nei gei women shuo (a) san-ge gushi ma? you can at 10 minute in to us tell three-Cl story Q (b) * naxie gushi those story ‘Can you tell us (a) three stories in ten minutes?’ (b) *those stories According to Liu, the reason behind the contrast in (12) is that only indefinite noun phrase objects are event measures, and are therefore compatible with telic events (marked by zai shi fenzhong nei ‘in ten minutes’). If this line of reasoning is correct, then we can argue that the VC ge has the similar function as zai shi fenzhong nei, i.e. the telicity- marking function. Sentences marked by the VC ge can take indefinite but not definite object noun phrases (11), because the latter are not event measures and are not compatible with telic events.4 2.2 Syntactic properties
Basically, the syntactic distributions of the VC ge is somewhat similar to those of the typical nominal classifiers and verbal measure words, in that, among other things, the VC ge must occur after the verb and before a noun phrase. However, it is also clearly distinct from typical nominal classifiers and verbal measure words when we examine it more carefully. 2.2.1 Similarities and differences between the VC ge and nominal classifiers
The VC ge has a number of syntactic properties that are shared by nominal classifiers. The most obvious one is that it occurs between a verb and a nominal expression, as we
4 In fact, the notion of telicity is not very precise. As noted by Lin (2001), the VC ge can co-occur
with stative predicates: (i) a. dong ge fawen you sheme liao-bu-qi?
understand VC French Foc what big.deal ‘Is it a big deal to understand French?’
b. lao ge ji-sui bu shi hen zhongyao old VC several-year not be very important ‘It is not very important to be a few years older.’
In general a stative predicate is not compatible with telicity. It seems, however, the states ‘to understand French’ and ‘to be a few years older’ are regarded as bounded. Pursuing this boundedness notion will take us too far afield from the main concern of this paper, so it will stick to the telicity notion in this paper.
Chih-hsiang Shu
670
have seen above. Another important, albeit less discussed property they share is that they both need a host as if they are affixes or clitics. The most extensive discussions of nominal classifiers as suffixes and enclitics are provided by Yang (2001).5 According to Yang, nominal classifiers are either suffixes or enclitics. When numerals are present, they are suffixes, since they always attach to the numerals. When the numerals are absent (the semantic quantity can only be ‘one’), however, they are enclitics, due to two sets of facts: they always need another host word to precede it, and they appear in a wider range of contexts.
Typical cases of nominal classifiers with clitic status (the numerals are absent) are illustrated as follows:
(13) a. nei ben shu hen gui that Cl book very expensive ‘That book is expensive.’
b. mei ben shu dou hen gui every Cl book all very expensive ‘Every book is expensive.’
c. lisi mai-le ben shu L. buy-Pfv Cl book ‘Lisi bought a book.’
In (13a), the classifier is attached to a determiner. In (13b), it is attached to a quantifier. In (13c), it is attached to a verb-aspectual suffix complex. We see that the hosts can belong to a variety of syntactic categories. As Yang noted, this property indicates that a numeral-less classifier is less like a suffix but more like a clitic.
The need for an adjacent host is illustrated in the following examples:
(14) a. * (yi) ben shu bu gou one Cl book not enough ‘One book is not enough.’
b. lisi song-le yi-ge pengyou *(yi) ben shu L. give-Pfv one-Cl friend one Cl book ‘Lisi gave a friend a book.’
c. lisi mai-le yi-ben shu he *(yi) zhi bi L. buy-Pfv Cl book and one Cl pen ‘Lisi bought a book and a pen.’
5 I am grateful to a reviewer for referring me to Yang’s (2001) work.
671
d. na [houhou-de] *(yi) ben shu hen gui that thick-DE one Cl book very expensive ‘That very thick book is expensive.’
(14a) shows that in the sentence-initial position, a full NP containing a classifier
can never occur without a numeral. (14b-c) shows that a numeral-less classifier cannot be separated from the verb by other words. Similarly, (14d) shows that a numeral-less classifier cannot be separated from the determiner by an adjective. As noted by Yang, these facts show that a numeral-less classifier needs a host and has to be adjacent to its host.
How about the VC ge? There is, in fact, some evidence showing that it also has some properties that suggest its clitichood. First, the VC ge either follows a verb or a verb-aspectual suffix complex, as we have seen in (2). This shows it allows a wider selection than suffixes. Second, it has to be adjacent to its host, as illustrated in the following examples:
(15) a. wo xiang song ta ge ji-pen hua I want give him VC several.pot flower ‘I want to give him several pots of flowers.’
b. * wo xiang song linju ge ji-pen hua I want give neighbor VC several.pot flower
(16) a. wo zhi song-le ge ji-pen hua I only give-Pfv VC several.pot flower ‘I only gave somebody several pots of flowers.’
b. * wo zhi song-le ta ge ji-pen hua I only give-Pfv him VC several.pot flower
The contrast between (15a-b) shows that while the VC ge can be separated from
the verb by a pronoun, it cannot be separated from the verb by a non-pronominal noun phrase. (16) shows us that while ge can be separated from the verb by either a pronoun or an aspectual suffix, it cannot be separated by both a pronoun and a suffix. The general principle is thus the VC ge cannot be too far away from the verb, a property that is shared by numeral-less nominal classifiers.
Beyond the similarities, however, there are also obvious differences. There are four syntactic properties that distinguish the VC ge from nominal classifiers: the VC ge selects a wider range of nominal expressions, it can co-occur with nominal classifiers, it
Chih-hsiang Shu
672
does not generally co-occur with numerals, it generally follows only verbs or verbal complexes.
We have seen in (2) that the VC ge can co-occur with various types of noun phrases. Compare these cases to (17), where the same types of noun phrases are attached to nominal classifiers.
(17) a. * lisi xiang he san-ge shui (mass noun) L. want drink three-Cl water
b. * zhangsan zhe-ci gai fu san-ge ze (object part of a VO compound)
Z. this-time should bear three-Cl responsibility
c. * zhangsan pao-le san-ge liang-fenzhong (indefinite temporal NP) Z. run-Pfv three-Cl two-minute
d. * zhangsan zhi kan-le san-ge ji-bu dianying (indefinite NP) Z. only see-Pfv three-Cl several-Cl movie
In (17) I attach ge to a random numeral san ‘three’ to indicate in these examples it is unmistakably a nominal classifier. It is clear that none of the sentences are acceptable, showing that nominal classifiers can only co-occur with unquantified count nouns, while the VC ge can co-occur with a wider range of nominal expressions.6 It is also clear that the VC ge always occurs outside a quantified noun phrase that already contain a nominal classifier, as we have seen in (2d). (17) also shows us that the VC ge does not co-occur with numerals. Finally, the following examples show that the VC ge does not co-occur with determiners and quantifiers:7
6 A reviewer noted that the VC ge can sometimes co-occur with the numeral yi ‘one’, making it
difficult to distinguish the nominal ge from the VC ge. However, it is clear that the nominal ge can co-occur with any numeral, not just yi. I assume that when ge is attached to yi and serves as a verbal classifier, yi does not serve as a real numeral identifying the number of entities or events, but is actually a dummy morpheme that does not have semantic functions.
7 Note, however, that ge can sometimes occur with a determiner and appears to serve as an event measuring expression, such as the following (see also Huang 2009):
(i) zhe ge niu, ni chui de tai guohuo le this VC cow you blow DE too outrageous Pfv ‘This act of bluffing, you did it too outrageously.’
It is unclear whether this ge serves the same function as the one we are concerned with in this paper, but it appears that there is no ‘down-play’ mood here. I shall not discuss this case further. In any case, whether the VC ge can or cannot attach to determiners or even quantifiers does not affect the main analyses of this paper.
673
(18) a. * lisi xiang he zhe-ge shui L. want drink this-Cl water
b. * zhangsan zhe-ci gai fu mei-ge ze Z. this-time should bear every-Cl responsibility In sum, we find that the VC ge and nominal classifiers differ in their syntactic distribution, but are similar in that they can be regarded as clitics. 2.2.2 Similarities and differences between the VC ge and verbal measure words
Although I tentatively call ge discussed throughout this paper a verbal classifier, there are no other verbal classifiers in Chinese. Instead, the type of vP-level expressions that resembles the VC ge the most are perhaps the verbal measure words. Like the VC ge, verbal measure words can occur between a verb and the direct object, and also conveys aspectual, and sometimes even mood information. In addition, verbal measure words resemble the VC ge in that they do not share some of the properties of nominal classifiers: they do not have strict selection restrictions with the following nominal constituents, as they can co-occur with nominal classifiers. These properties are illustrated as follows:
(19) a. lisi he-le san-ci shui L. drink-Pfv three-time water ‘Lisi drank water three times.’
b. zhangsan du-le san-bian na-ben shu Z. read-Pfv three-time that-Cl book ‘Zhangsan read that book three times.’ (19a) shows that verbal measure words can occur before a mass noun. (19b) shows that it can occur before a definite noun phrase, which already contains a classifier.
However, it is clear that the VC ge has a number of properties that distinguish itself from verbal measure words: the former behaves like a clitic, rather than a suffix, it generally does not attach to numerals, it cannot co-occur with a definite noun phrase, nor can it occur without a complement. We have seen some properties that indicate the clitichood of the VC ge above. Numeral-less verbal measure differs from the VC ge and numeral-less nominal classifiers in that the former generally do not attach to verbs or verbal complexes:
Chih-hsiang Shu
(20) a. * lisi he-le-ci shui L. drink-Pfv-time water
b. * zhangsan du-le-bian na-ben shu Z. read-Pfv-time that-Cl book Unlike the VC ge, verbal measure words cannot occur with indefinite noun phrases:
(21) * lisi du-le san-bian wu-ben shu L. read-Pfv three-time five-Cl book The ungrammaticality of (21) can be explained by Tenny (1994), who argues that an event can be measured in one way. In (21), the event is measured twice, since both the direct object and the frequency adjunct provides event measure, so it is ill-formed. This also shows that while the VC ge marks telicity, it does not of itself have the event measuring function.
Verbal measure words can occur without overt ‘complements’. That is, they can occur in the sentence-final position. This property is not shared by the VC ge.
(22) a. zhe-ben shu, lisi du-le san-ci this-book book L. read-Pfv three-time
‘This book, Lisi read three times.’
b. zhangsan jiao-le liang-ci Z. shout-Pfv two-time
‘Zhangsan shouted two times.’
(23) a. * lisi zhi pao-le ge L. only run-Pfv VC
b. * zhangsan xiang chi ge Z. want eat VC
Generally, the VC ge is both semantically and syntactically distinct from verbal measure words, except that they both have looser selection restrictions of their complements then the nominal classifiers.
Let us sum up the properties of the VC ge discussed so far. It conveys mood and aspect information, and has the properties of a clitic. Generally, it follows a verb or a verbal complex, and precedes a nominal constituent. Its morphosyntactic distribution can be informally represented as follows (Xσ represents a monosyllabic functional morpheme):
675
(24) [VP V-(Xσ)-gevc-DP] 2.3 Two other cases
How about the instances of ge in (3) and (4) (repeated below)?
(25) a. lisi ke wan-de ge tongkuai L. indeed play-DE VC2 thorough.satisfaction ‘Lisi indeed played to his heart’s content.’
b. zhangsan xiao ge bu ting Z. laugh VC2 no stop ‘Zhangsan keeps laughing (and there is no sign of his stopping.)’
(26) a. mai ta ge yi bai-ben shu buy TA VC3 one hundred-Cl book ‘To buy one hundred books.’
b. wan ta ge guoyin play TA VC3 to.one’s.satisfaction ‘To play to a thorough satisfaction of it.’ I assume that these are distinct mood markers and do not directly bear on the VC ge with which we are concerned in this paper. They do, however, exhibit similar morphosyntactic properties, which merit some discussion. I shall return to this issue in §5.
3. Pro and cons of previous studies
3.1 The VP-external analysis
Based on the mood property of ge, Lin (2001) placed ge in a position higher than VP to account for the semantic effect of ge on the whole event, and adopts a V-to-v analysis to account for its post-verbal position. This analysis can be represented by the following structure:
Chih-hsiang Shu
676
(27) vP 3 v XP 3 X VP 3 V DP paoi ge ti liang fenzhong run VC two minute This analysis has the advantage of accounting for ge’s semantic scope over an event, the loose selection restriction between ge and its NP complement, the fact that ge can co- occur with nominal classifiers, and its post-verbal position. However, it cannot account for the fact that various other mood expressions (mood adverbs, particles, modal auxiliaries) do not occur post-verbally, the fact that the VC ge has to be adjacent to the verb or the verbal complex, and the fact that it has to precede its overtly realized complement. 3.2 The DP-internal analysis
Based on the fact that the VC ge is phonetically and orthographically identical to the default nominal classifier ge, and its requirement of an indefinite noun phrase complement, Wu (2002) placed the VC ge as a D head inside DP, and adopts an analysis that treats overt determiners as overt realizations of a D head. This analysis is represented by the following structure:
(28) DP 3 D NumP 3 Num ClP ge liang fenzhong VC two minute This analysis apparently answered the question why the VC ge requires a complement and has to co-occur with an indefinite noun phrase. It is, however, unable to account for
677
the fact that the VC ge cannot occur in the sentence-initial position, the fact that it has to be adjacent to the verb or verbal complex, and its semantic scope over the whole event. Furthermore, it is unclear why telicity is encoded in the D head in (28), but encoded in the aspectual/temporal adjunct in (12).
4. An analysis of the VC ge as a clitic
We have seen that the VC ge has properties of a clitic that overlap with those of numeral-less nominal classifiers, yet it also possesses some unique properties. Further- more, it is unclear how a clitic in Chinese can be accounted for in modern syntactic and morphological theories. To provide an up-to-date analysis of the VC ge, therefore, we need to incorporate the insights of Lin (2001), Yang (2001), and Wu (2002), as well as search for a coherent theoretical solution.
In this section I shall provide some background of the descriptive literature about clitics and incorporate the insight of previous three studies to provide a descriptive account for the VC ge. In §5 I shall provide a theoretical treatment. 4.1 Clitics and morphosyntactic clitics
Before we provide a more precise descriptive account of the morphosyntactic VC ge, we need to review what literature has to say about clitics in general. There are four important descriptive generalizations observed in the literature:
(29) a. Clitics are distinct from words and phrases b. Clitics are distinct from affixes c. Special/morphosyntactic clitics are distinct from simple clitics d. Clitics can be described as phrasal affixes
(29a) refers to perhaps the most prominent property of a clitic: it is similar to an
affix in that it is bound in a specific way. More specifically, a clitic needs a host, they have to be adjacent, and the order of the host and the clitic is fixed. Zwicky (1985) employs numerous other tests to distinguish clitics from words, some of which are based on unsettled theoretical assumptions, but the major ones are pretty straightforward and can be used to detect clitichood of a given expression. As we have seen in §2, Yang (2001) uses this criterion to help determine the clitichood of numeral-less classifiers.
(29b) refers to a set of properties that are not shared by affixes. Among these properties, the most prominent one is as follows (Zwicky & Pullum 1983):
Chih-hsiang Shu
678
(30) Clitics can exhibit a low degree of selection with respect to their hosts, while affixes exhibit a high degree of selection with respect to their stems.
As noted by Zwicky & Pullum, English contracted auxiliaries like -s/-ve can attach to words of various syntactic categories:
(31) a. The person I was talking to’s going to be angry with me. b. The ball you hit’s just broken my dining room window. c. Any answer not entirely right’s going to be marked as error.
As we have seen in §2, Yang (2001) uses this criterion to test whether a numeral-less nominal classifier is a clitic or not.
(29c) and (29d) are closely interconnected properties of clitics. Among elements that fall into the above criteria of clitics, some have the same distribution as that of associated free forms, while the others either do not have free counterparts, or have a different distribution from their free counterparts. Zwicky & Pullum (1983:510) terms the former ‘simple clitics’ and the latter ‘special clitics’. The behaviors of special clitics are not well understood and there have been conflicting analyses of the same phenomenon, but it seems clear that they are subject to principles that do not belong to the syntactic or phonological component. Anderson (2005:31) terms them ‘morphosyntactic clitics’ and provides the following definition:
(32) Morphosyntactic clitic: a linguistic element whose position with respect to the other elements of the phrase or clause follows a distinct set of principles, separate from those of the independently motivated syntax of free elements of the language.
To provide a more precise characterization of morphosyntactic clitics, Klavans (1982, 1985) and Anderson (1992, 2005) developed the concept of special clitics as phrasal affixes. A characterization of special clitics as phrasal affixes is as follows (Anderson 2005:82):
(33) A clitic is located a. within the domain of some syntactic constituent (X0 or Xmax for some
value of X); b. by reference to the first versus last daughter constituent of that domain
(interpreted either syntactically or prosodically); and c. preceding or following this anchor point.
679
This characterization of special clitics resembles descriptions of an affix, hence the term ‘phrasal affixes’.
According to Anderson, a case in point is second-position clitics. Such a clitic has CP as its domain, the anchoring point is the first daughter of the domain, and it follows the anchoring point. The following Czech examples discussed in Anderson (2005), cited from Richardson (1997), can serve as an illustration:
(34) a. Vera =jsem =se =mu konene omluvil yesterday PAST-1SG REFL 3SG-DAT finally apologized
‘Yesterday I finally apologized to him.’
b. ervené tulipány =se objednal red tulips REFL ordered ‘He did order red tulips.’
From the perspective of phrasal affixation, the second position clitic in (34) is
comparable to an infix, which is basically a prefix but is realigned to the stem-internal position due to some other independent principles or constraints. The main differences are that the ‘stem’ of the clitic is a phrase instead of the word, and that the anchoring point, can be a syntactic element, such as the noun phrase in (34b).
Properties (29c) and (29d) are important descriptive tools for studying clitics, but have not received much attention in the studies of the VC ge and clitics in Chinese in general.8 In what follows I shall show that they are essential for describing the morphosyntactic distributions of the VC ge.
4.2 A clitic analysis of ge
Let us now review the essential morphosyntactic distributions of ge established in §2 in light of linguists’ current understanding of clitics. 4.2.1 The VC ge is neither a free word/phrase nor an affix
We have seen in §2 that like numeral-less nominal classifiers, the VC ge behaves like a clitic instead of a free word/phrase, because it needs a host and needs to be adjacent to its host. Furthermore, we see that it has a wider range of selection restrictions with its preceding host, and it also requires a nominal constituent to follow it. These two
8 The only studies that have employed the notion ‘phrasal affixes’ to account for clitics in Chinese
that I know of are C. Huang (1989) and Liu (1995, 1998).
Chih-hsiang Shu
680
properties indicate that the VC ge is a clitic instead of an affix.9 More specifically, this clitic is subject to the morphosyntactic templatic (24) ([VP V-(Xσ)-gevc-DP]). This certainly cannot be a template for affixes. 4.2.2 The VC ge is a special/morphosyntactic clitic
In §2 I have shown clearly that ge conveys the speaker’s under-evaluation and marks the telic aspect. In theories of syntax, mood is encoded either in I(nflection) or C(omplementizer). We have also seen in §2 that ge is phonetically realized between a verb and a nominal complement. This position certainly does not correspondent to either I or C, nor does it correspond to the syntactic positions of other realizations of mood expressions. In addition, we have also seen that its distribution is very different from typical measure words, since measure words do not have a fixed position and can be free. These properties of ge all suggest ge is a special/morphosyntactic clitic, since its distribution is different from what we would expect of a free syntactic element. 4.2.3 The VC ge as a phrasal affix
In light of the useful set of parameters given in (33) and template (24), we can provide the following (partial) characterizations for the VC ge:10
(35) a. domain: VP b. anchor: first c. orientation: follow
The domain is VP since a proper descriptive account has to refer to the whole VP, not just the verb, as we have seen in (24). The anchor is the first daughter constituent of the domain, since, as we have seen in (15) and (16), ge has to be adjacent to the verb or verbal complex. The VC ge also has to follow the anchor.
It thus seems the VC ge can be regarded as a phrasal affix. However, despite the fact that the parameters in (33) can somewhat accommodate
the VC ge, they are certainly insufficient. First of all, they fail to account for the fact that the VC ge can and must follow a monosyllabic pronoun, as shown in (15a). In addition, the fact that it has to precede a nominal constituent is also not captured. Furthermore, it has not yet been made clear how those parameters are encoded in grammar and how and when the clitic enters syntactic derivations. I shall address these issues in the next section.
9 See Liu (1995) for similar accounts of the verbal de.
10 This is also Liu’s (1995) characterization of the verbal de.
681
5. The Morphology component and the Agree theory
To augment the descriptive power of our linguistic theory while staying coherent, I propose that we need a theory that allows Morphology as a component independent from narrow syntax, and allows morphological information to be encoded in the lexicon. Aronoff’s (1994) framework provides just such a theory. In addition, I argue that the existence of clitics such as the VC ge provides evidence for Chomsky’s (2000) Agree theory. 5.1 Morphology as a component of grammar
It is not surprising that the parameters in (33) are not sufficient to account for all the distributional properties of the VC ge, if it can indeed be regarded as a phrasal affix. This is because affixal morphology is far more complicated than what parameters like (33) can accommodate. If ‘phrasal affixal morphology’ parallels affixal morphology, then (33) is certainly not adequate. In this section, I should like to show there is indeed a word-level morphological process in Chinese that resembles the process that derives the VC ge, which cannot be adequately accounted for by parameters in (33). I argue that we need to employ a morphological template to derive the facts, and the template for the VC ge is (24), repeated below:
(36) [VP V-(Xσ)-gevc-DP]
Due to the fact that ge occurs in the middle of the template, I call the VC ge an internal clitic.11 The existence of this template can be accommodated by Aronoff’s (1994) framework in which Morphology is an independent component of grammar. Let us see how.
In word-level morphology in Chinese, a morphological process typically labeled as the ‘potential’ construction (cf. Chao 1968, Li & Thompson 1981) involves internal affixation of a modal element in a verb-resultative compound. Examples are as follows:
(37) a. ren - de - chu
identify able out ‘able to recognize’
11 This term should be distinguished from endoclitic, because an endoclitic splits apart a root and
does not refer to templates.
Chih-hsiang Shu
682
b. sou-xun - de - dao find-seek able reach ‘able to find out’
c. na - de - qilai take able rise ‘able to pick (something) up’
In these examples, ren-chu, souxun-dao, and na-qilai are word-level verbal expressions, since chu, dao, and qilai are bound morphemes. The process that derives realization of de should therefore be a word-level affixal morphological process. Here the placement of de is sensitive to the morphological and semantic properties of the preceding and the following elements. The preceding element, no matter how many syllables it has, must be an activity verb that does not contain a resultative morpheme. The following element, no matter the number of its syllables, has to be a resultative morpheme that does not contain the activity verb. So the template for the internal affix de looks like this:
(38) The morphological template of de ‘able’
[V VActivity - de - VResult]
Any violation of the template yields ungrammaticality. This can be shown by comparing the following examples with (37):
(39) a. * na - de take able
b. * na-qilai-de take rise able
c. * de - na - qilai able take rise
d. * sou - de - xun find able seek
The existence of the internal affix de shows that (33) is inadequate for word-level
affixal morphology and we need a word-level morphology template in our repertoire of descriptive tools12. Therefore, if phrase-level morphology parallels word-level mor-
12 Contra Liu (1995:141), who employs parameters in (33) to account for the internal affix de:
683
phology, we should abandon (35) and favor (36) to describe the morphosyntactic distributions of the VC ge.
Further evidence for a morphological template analysis comes from the fact that ge exhibits properties that indicates direct correspondence between morphological form/class and morphology template, regardless of its meaning. First of all, when nominal classifiers are merged with the numeral yi and acts as a clitic, it is subject to the following template:13
(40) [DP/VP Dem/Quant/V-(Xσ)-Cl-NP]
Template (40) can be derived from facts in (13) and (14), and the fact that the clitic is also dependent on the following element, as shown in the following examples:
(41) a. lizi, wo chi-le yi-ge pear I eat-Pfv one-Cl
b. * lizi, wo chi-le-ge pear I eat-Pfv-Cl
The template shows that a numeral-less nominal classifier is an internal clitic, and can be separated from its preceding host only by a monosyllabic functional morpheme. Interestingly, this template resembles the template for the VC ge (24), in terms of the shape of the template and the syntactic categories of the participants involved: the preceding host can be a verb, the following host is a nominal constituent.
In addition to numeral-less nominal classifiers, two different instances of the VC ge are also subject to the same morphological template. These instances of ge have been briefly discussed in §1 and §2.3, repeated below:
(42) a. lisi ke wan-de ge tongkuai
L. indeed play-DE VC2 thorough.satisfaction ‘Lisi indeed played to his heart’s content.’
(i) a. scope - V
b. anchor - head c. orientation - follow
This account cannot rule out (39a) and (39d). 13 This template is modified from Yang’s (2001) analysis of numeral-less nominal classifiers.
Yang acknowledges the fact that they are dependent on both the preceding and the following element, but did not spell-out the whole template.
Chih-hsiang Shu
684
b. zhangsan xiao ge bu ting Z. laugh VC2 no stop ‘Zhangsan keeps laughing (and there is no sign of his stopping.)’
(43) a. mai ta ge yi bai-ben shu buy TA VC3 one hundred-Cl book ‘To buy one hundred books.’
b. wan ta ge guoyin play TA VC3 to.one’s.satisfaction ‘To play to a thorough satisfaction of it.’
We have seen in §2.3 that semantically the instances of ge in these examples are distinct from the ‘down-play’ mood marker VC ge. What is interesting for us is that they are still subject to the morphological template (24). The template is the same in that there is evidence that the adverbials in (42) are in fact nominalized adverbials, and that ge has properties of an internal affix. As the following examples show, adverbials modified by the typical adverbial hen or bu are not allowed:14
(44) a. lisi ke wan-de *(ge) hen tongkuai L. indeed play-DE VC2 very thorough.satisfaction ‘Lisi indeed totally enjoyed his time playing.’
b. zhangsan wan-de *(ge) bu tongkuai Z. play-DE VC2 Neg thorough.satisfaction ‘Lisi didn’t enjoy his time playing.’
The examples show that there is nothing wrong with the sentences if ge is not present. When ge is present, adverbial modifiers are not allowed. These facts receive a natural explanation if the adverbials in (42) as nominalized adverbials and ge is subject to template (24).15, 16
14 Some adverbials cannot be modified by hen or bu directly in any context, such as bu ting ‘no
stop’ in (42b), so this test is not applicable. I assume the adverbial is still a nominal adverbial in these cases.
15 A reviewer questioned why ge is obligatory in certain cases when adverbials are involved: (i) wangwu ba xiaotou da-le *(ge) ban si
W. BA thief hit-Pfv VC2 half.dead I assume that verbal adverbials in Chinese in general have to be licensed by specific morphemes. The licensing morphemes are either de or ge. As for why the licensing is required and how it works, I leave these issues for future research.
685
It is also clear that when ge occurs in the ‘excessive’ construction, it has to follow the same template, as its following adverbial cannot be modified by hen:
(45) wan ta ge (*hen) guoyin
play TA VC3 very to.one’s.satisfaction ‘To play to a thorough satisfaction of it.’ Under no circumstances can those instances of ge occur without a preceding and a following host:
(46) Q: [How much fun did Lisi have?] A: a. wan-le-ge tongkuai Play-Pfv-VC2 thorough.satisfaction ‘He played to his heart’s content.’ b. * ge tongkuai c. tongkuai
(47) [Q: How many books do you want to buy?] a. mai ta ge yibai-ben shu buy TA VC3 one.hundred-Cl book ‘I want to buy 100 books.’ b. * ge yibai-ben shu
(48) * mai ta ge buy TA VC3 So far, we have seen that a single general morphological template can accommodate
four types of clitics with distinct semantic and syntactic properties. It appears that individual affixes or clitics in Chinese are encoded with ‘template class’ in the lexicon. That is, certain morphological forms are associated with specific morphological templates, no matter their semantic and syntactic function. This suggests that Morphology is a grammatical component separate from narrow syntax. This process is similar to the situations of English past participle discussed in Aronoff (1994), in which exactly the same inflected morphological form (stem-ed, stem-en, etc.) of any given verb can express either passive voice or perfect aspect. The two different functions also cannot
16 Another question raised by the reviewer concerns the reason why ge can follow a verbal
modification marker de if the adverbal is nominalized (cf. (42a)). I assume that the template of the internal clitic de allows the following constituent to be either nominal or adverbial, as long as the semantic requirements are met.
Chih-hsiang Shu
686
be semantically or syntactically derived. According to Aronoff, the two functions of passive and perfect are paired by the abstract morphological entity Fen, regardless of their syntactic or semantic differences. The function Fen is purely morphological (or, in Aronoff’s term, ‘morphomic’). In our analysis of the VC ge, we can also postulate a function Fge to represent the mapping between morphological form/class and template class for the class of clitics represented by the VC ge.
In sum, we argue that clitics and affixes form a natural class, since they can both be regulated by morphological templates, which support the thesis that Morphology is a grammatical component. However, it is undeniable that postulating a Morphology component does not solve all the analytical problems, since inflectional affixes have syntactic properties as well. Morphology has to interact with narrow syntax somehow. I shall address this issue next. 5.2 Support for the Agree theory
If Morphology is an independent grammatical component, it still needs to interact with narrow syntax. This view is also shared by Aronoff (1994:71), who provides the following schema for ‘the mapping from syntax to phonology through morphology’:
(49) Level Entity (morpho)syntax (morpho)syntactic properties (gender) (morpho)syntax to morphology morphosyntactic rules (gender to inflectional class) morphology by itself inflectional classes realization pairs morphology to phonology realization rules phonology phonological forms
According to (49), syntax provides information to morphology, which in turn pro-
vides information to phonology. However, not all syntactic theories are compatible with such a mapping. In Chomsky’s (1995:Ch.3) framework, all morphological operations apply before syntactic operations, and no words can remain uninflected at any point in the syntactic derivations due to the Extension Condition. For example, in the sentence John saw Mary yesterday, the verb saw enters the numeration/lexicon fully inflected. When the tense head (T0) is merged with vP, covert feature movement applies and the tense feature on T is checked. It is not possible for syntax to feed information to the Morphology component. Instead, morphological information encoded in the lexicon has to be ‘checked off’ by syntactic operations. It is clear that this framework is not compatible with Aronoff’s framework.
687
Under Chomsky’s (2000 et seq.) Agree framework, however, the situation is quite different. According to Chomsky (2001:5, 2004:16), the simplest assumption about an uninterpretable feature F is that it enters the derivation without value. This is so because the value is determined only in the syntactic context by Agree. For example, in the sentence John saw Mary yesterday the inflected verb saw is uninflected in the numeration, and is only inflected after Agree applies. This framework appears to be compatible with Aronoff’s framework, since syntactic operations can apply before morphological operations, and words can remain uninflected until Agree applies. I shall now show that this framework indeed accommodates quite well the descriptive facts discussed in the paper thus far.
In this framework, the mismatch between the wide semantic scope and its post- verbal position noted by Lin (2001) indicates that Agree has to apply. More specifically, I argue that mood information is encoded at the C0, and correspondent uninterpretable unvalued features are encoded at the V head. Agree between C0 and V0 induces feature matching and valuation. These operations are defined as follows:
(50) Agree a. Match: A feature F (a probe) on a head H at syntactic location α searches
its c-command domain for another F (a goal) at location β with which to agree.
b. Valuation: Replace any unvalued feature with valued feature at α and β. Feature valuation induces merge of ge to VP, where ge has to follow the morphological template TCge, because of the mapping rule Fge in the Morphology component. The function of ge is to realize feature valuation. These processes are laid out below:
(51) Agree a. C V C V valued [iM] unvalued [uM] valued [iM] valued [uM] b. Select ge and merge it to the edge of V.17 ge realizes the feature valuation
of [uM]. c. The morphosyntactic context of ge has to conform to TCge, specified in
the Morphology component as a mapping rule Fge. d. TCge = [VP V-(Xσ)-ge-YP[+N]]
17 This operation may seem counter-cyclic, since the merging site is not the outermost syntactic
constituent. However, Chomsky’s assumptions about uninterpretable features readily accommodate ‘delayed-Merge.’ See Shu (2011:Ch.4) for discussion.
Chih-hsiang Shu
688
At the final stage of derivation, a sentence like (2d) can be represented by the following tree diagram:18
(52) CP 3 C TP 3 DP VP 3 Adv VP 3 V DP ZS zhi kan-le-ge ji-bu dianying Z. only see-Pfv-VC several-Cl movie This analysis of ge captures the essential morphosyntactic properties of the VC ge laid out in §2. The semantic properties of ‘play-down’ mood and telicity aspect are encoded at the C0 head. It has properties of a clitic because it is associated with TCge according to its morphological class, so the merge operation in which it is involved has to conform to the template. Its distributions are distinct from those of nominal classifiers and verbal measure words because nominal classifiers and verbal measure words have different semantic and syntactic functions: the former are reflexes of Agree between a noun and a numeral/verb/determiner, and the latter are verbal event measure words. None of these are encoded at C0.
6. Conclusion
In this paper I argue that the verbal classifier ge has the properties of an internal clitic, since it shows properties of being locally bound to its preceding and following elements, and those elements can have phrasal status. Furthermore, it is shown that not only the VC ge, but also affixes, numeral-less nominal classifiers, and two other instances of the VC ge are also subject to ‘internal bound morpheme’ analyses. We thus have two sets of general patterns. On the one hand, affixes and clitics can form a natural class; on the other, clitics with different syntactic properties can be subject to the same type of morphological template. The natural conclusion is that clitics and affixes are
18 Here I abstract away from light verb structures.
689
both subject to operations in the Morphology component, which is independent from narrow syntax. Furthermore, the state of affairs suggest direct mapping between morphological form/class and template class. Finally, it is shown that these morphological operations are compatible with the Chomsky’s Agree framework, but incompatible with his older Checking framework.
Our treatment of the VC ge has the advantage of wider empirical coverage and adopting theoretical frameworks that have more descriptive power since they can capture the most crucial facts in our current understanding of these clitics. Nevertheless, there is still much unfinished business. The Morphology component, if it indeed exists, is an under-explored territory. The bulk of research is focused on Indo-European languages. Very little modern research is done with Chinese, which is sometimes regarded as morphology-free. There are certainly numerous other clitics and affixes in Chinese, including ta, de, briefly discussed in this paper, and perhaps sentence-final particles, which belong to distinct template classes. Theoretically, it is also worth pursuing whether templates can be derived from more primitive operations or rules, and refining the workings of word- and phrase-level morphology with relevant theoretical consequences. Eventually, we also need to address the challenging yet important question of how all the differences between different morphemes and among languages can be characterized in a theory of language parametrization.
Chih-hsiang Shu
References Anderson, Stephen R. 1992. A-Morphous Morphology. Cambridge & New York:
Cambridge University Press. Anderson, Stephen R. 2005. Aspects of the Theory of Clitics. Oxford & New York:
Oxford University Press. Aronoff, Mark. 1994. Morphology by Itself: Stems and Inflectional Classes. Cambridge:
MIT Press. Biq, Yung-O. 2004. Construction, reanalysis, and stance: ‘V yi ge N’ and variations in
Mandarin Chinese. Journal of Pragmatics 36.9:1655-1672. Chao, Yuen Ren. 1968. A Grammar of Spoken Chinese. Berkeley: University of Cali-
fornia Press. Chomsky, Noam. 1995. The Minimalist Program. Cambridge: MIT Press. Chomsky, Noam. 2000. Minimalist inquiries: the framework. Step by Step: Essays on
Minimalist Syntax in Honor of Howard Lasnik, ed. by Roger Martin, David Michaels & Juan Uriagereka, 89-155. Cambridge: MIT Press.
Chomsky, Noam. 2001. Derivation by phase. Ken Hale: A Life in Language, ed. by Michael Kenstowicz, 1-52. Cambridge: MIT Press.
Chomsky, Noam. 2004. Beyond explanatory adequacy. Structure and Beyond, ed. by Adriana Belletti, 104-131. Oxford & New York: Oxford University Press.
Huang, Chu-Ren. 1989. Mandarin Chinese NP de: A Comparative Study of Current Grammatical Theories. Taipei: Institute of History and Philology, Academia Sinica.
Huang, C.-T. James. 2009. Lexical decomposition, silent categories, and the localizer phrase. Yuyanxue Luncong [Essays on Linguistics] 39:86-122. Beijing: The Com- mercial Press.
Klavans, Judith L. 1982. Some Problems in a Theory of Clitics. Bloomington: Indiana University Linguistics Club.
Klavans, Judith L. 1985. The independence of syntax and phonology in cliticization. Language 61.1:95-120.
Li, Charles N., and Sandra A Thompson. 1981. Mandarin Chinese: A Functional Reference Grammar. Berkeley: University of California Press.
Lin, T.-H. Jonah. 2001. Light Verb Syntax and the Theory of Phrase Structure. Irvine: University of California dissertation.
Liu, Feng-hsi. 1995. A note on clitics and affixes in Chinese. Proceedings of the 6th North American Conference on Chinese Linguistics (NACCL 6), Vol. 2: Language Acquisition, Pragmatics, Phonology and Morphology, ed. by Jose Camacho & Lina Choueiri, 137-143. Los Angeles: GSIL.
Liu, Feng-hsi. 1998. A clitic analysis of locative particles. Journal of Chinese Linguistics 26.1:48-70.
691
Liu, Feng-hsi. 2006. Event measures in Chinese. Snippets 12:12-13. Lü, Shuxiang. 1980. Xiandai Hanyu Babai Ci [800 Words in Modern Chinese]. Beijing:
The Commercial Press. Richardson, Matthew. 1997. Czech clitics: 2P or not 2P that is the question. Yale A-
Morphous Linguistics Essays: Studies in the Morphosyntax of Clitics, ed. by Lizanne Kaiser, 131-150. New Haven: Department of Linguistics, Yale University.
Shu, Chih-hsiang. 2011. Sentence Adverbs in the Kingdom of Agree. Stony Brook: Stony Brook University dissertation.
Tang, Chih-Chen Jane. 1990. Chinese Phrase Structure and the Extended X'-theory. Ithaca: Cornell University dissertation.
Tenny, Carol L. 1994. Aspectual Roles and the Syntax-Semantics Interface. Dordrecht & Boston: Kluwer.
Wu, Zoe. 2002. Wo pao-le ge feikuai and reanalysis of the classifier ge. On the Formal Way to Chinese Languages, ed. by Sze-Wing Tang & Chen-Sheng Luther Liu, 163-188. Stanford: CSLI.
Yang, Rong. 2001. Common Nouns, Classifiers, and Quantification in Chinese. New Brunswick: Rutgers University dissertation.
Yeh, Jui-chuan. 2006. On the excessive construction in Mandarin Chinese. Concentric: Studies in Linguistics 32.2:93-118.
Zwicky, Arnold M. 1985. Clitics and particles. Language 61.2:283-305. Zwicky, Arnold M., and Geoffrey K. Pullum. 1983. Cliticization vs. inflection: English
n’t. Language 59.3:502-513.
[Received 16 August 2010; revised 19 October 2011; accepted 28 October 2011] Institute of Linguistics Academia Sinica 130, Sec. 2, Academia Road Nankang, Taipei 115, Taiwan [email protected]
Chih-hsiang Shu

Klavans (1982, 1985) Anderson (1992, 2005)
Aronoff (1994)
Chomsky2000

<< /ASCII85EncodePages false /AllowTransparency false /AutoPositionEPSFiles true /AutoRotatePages /None /Binding /Left /CalGrayProfile (Dot Gain 20%) /CalRGBProfile (sRGB IEC61966-2.1) /CalCMYKProfile (U.S. Web Coated \050SWOP\051 v2) /sRGBProfile (sRGB IEC61966-2.1) /CannotEmbedFontPolicy /Error /CompatibilityLevel 1.4 /CompressObjects /Tags /CompressPages true /ConvertImagesToIndexed true /PassThroughJPEGImages true /CreateJobTicket false /DefaultRenderingIntent /Default /DetectBlends true /DetectCurves 0.0000 /ColorConversionStrategy /CMYK /DoThumbnails false /EmbedAllFonts true /EmbedOpenType false /ParseICCProfilesInComments true /EmbedJobOptions true /DSCReportingLevel 0 /EmitDSCWarnings false /EndPage -1 /ImageMemory 1048576 /LockDistillerParams false /MaxSubsetPct 100 /Optimize true /OPM 1 /ParseDSCComments true /ParseDSCCommentsForDocInfo true /PreserveCopyPage true /PreserveDICMYKValues true /PreserveEPSInfo true /PreserveFlatness true /PreserveHalftoneInfo false /PreserveOPIComments true /PreserveOverprintSettings true /StartPage 1 /SubsetFonts true /TransferFunctionInfo /Apply /UCRandBGInfo /Preserve /UsePrologue false /ColorSettingsFile () /AlwaysEmbed [ true ] /NeverEmbed [ true ] /AntiAliasColorImages false /CropColorImages true /ColorImageMinResolution 300 /ColorImageMinResolutionPolicy /OK /DownsampleColorImages true /ColorImageDownsampleType /Bicubic /ColorImageResolution 300 /ColorImageDepth -1 /ColorImageMinDownsampleDepth 1 /ColorImageDownsampleThreshold 1.50000 /EncodeColorImages true /ColorImageFilter /DCTEncode /AutoFilterColorImages true /ColorImageAutoFilterStrategy /JPEG /ColorACSImageDict << /QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /ColorImageDict << /QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /JPEG2000ColorACSImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /JPEG2000ColorImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /AntiAliasGrayImages false /CropGrayImages true /GrayImageMinResolution 300 /GrayImageMinResolutionPolicy /OK /DownsampleGrayImages true /GrayImageDownsampleType /Bicubic /GrayImageResolution 300 /GrayImageDepth -1 /GrayImageMinDownsampleDepth 2 /GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true /GrayImageFilter /DCTEncode /AutoFilterGrayImages true /GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict << /QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /GrayImageDict << /QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /JPEG2000GrayACSImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /JPEG2000GrayImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /AntiAliasMonoImages false /CropMonoImages true /MonoImageMinResolution 1200 /MonoImageMinResolutionPolicy /OK /DownsampleMonoImages true /MonoImageDownsampleType /Bicubic /MonoImageResolution 1200 /MonoImageDepth -1 /MonoImageDownsampleThreshold 1.50000 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict << /K -1 >> /AllowPSXObjects false /CheckCompliance [ /None ] /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile () /PDFXOutputConditionIdentifier () /PDFXOutputCondition () /PDFXRegistryName () /PDFXTrapped /False /CreateJDFFile false /Description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> /CHS <FEFF4f7f75288fd94e9b8bbe5b9a521b5efa7684002000410064006f006200650020005000440046002065876863900275284e8e9ad88d2891cf76845370524d53705237300260a853ef4ee54f7f75280020004100630072006f0062006100740020548c002000410064006f00620065002000520065006100640065007200200035002e003000204ee553ca66f49ad87248672c676562535f00521b5efa768400200050004400460020658768633002> /CHT <FEFF4f7f752890194e9b8a2d7f6e5efa7acb7684002000410064006f006200650020005000440046002065874ef69069752865bc9ad854c18cea76845370524d5370523786557406300260a853ef4ee54f7f75280020004100630072006f0062006100740020548c002000410064006f00620065002000520065006100640065007200200035002e003000204ee553ca66f49ad87248672c4f86958b555f5df25efa7acb76840020005000440046002065874ef63002> /CZE <FEFF005400610074006f0020006e006100730074006100760065006e00ed00200070006f0075017e0069006a007400650020006b0020007600790074007600e101590065006e00ed00200064006f006b0075006d0065006e0074016f002000410064006f006200650020005000440046002c0020006b00740065007200e90020007300650020006e0065006a006c00e90070006500200068006f006400ed002000700072006f0020006b00760061006c00690074006e00ed0020007400690073006b00200061002000700072006500700072006500730073002e002000200056007900740076006f01590065006e00e900200064006f006b0075006d0065006e007400790020005000440046002000620075006400650020006d006f017e006e00e90020006f007400650076015900ed007400200076002000700072006f006700720061006d0065006300680020004100630072006f00620061007400200061002000410064006f00620065002000520065006100640065007200200035002e0030002000610020006e006f0076011b006a016100ed00630068002e> /DAN <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> /DEU <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> /ESP <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> /ETI <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> /FRA <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> /GRE <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a stvaranje Adobe PDF dokumenata najpogodnijih za visokokvalitetni ispis prije tiskanja koristite ove postavke. Stvoreni PDF dokumenti mogu se otvoriti Acrobat i Adobe Reader 5.0 i kasnijim verzijama.) /HUN <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> /ITA <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> /JPN <FEFF9ad854c18cea306a30d730ea30d730ec30b951fa529b7528002000410064006f0062006500200050004400460020658766f8306e4f5c6210306b4f7f75283057307e305930023053306e8a2d5b9a30674f5c62103055308c305f0020005000440046002030d530a130a430eb306f3001004100630072006f0062006100740020304a30883073002000410064006f00620065002000520065006100640065007200200035002e003000204ee5964d3067958b304f30533068304c3067304d307e305930023053306e8a2d5b9a306b306f30d530a930f330c8306e57cb30818fbc307f304c5fc59808306730593002> /KOR <FEFFc7740020c124c815c7440020c0acc6a9d558c5ec0020ace0d488c9c80020c2dcd5d80020c778c1c4c5d00020ac00c7a50020c801d569d55c002000410064006f0062006500200050004400460020bb38c11cb97c0020c791c131d569b2c8b2e4002e0020c774b807ac8c0020c791c131b41c00200050004400460020bb38c11cb2940020004100630072006f0062006100740020bc0f002000410064006f00620065002000520065006100640065007200200035002e00300020c774c0c1c5d0c11c0020c5f40020c2180020c788c2b5b2c8b2e4002e> /LTH <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> /LVI <FEFF0049007a006d0061006e0074006f006a00690065007400200161006f00730020006900650073007400610074012b006a0075006d00750073002c0020006c0061006900200076006500690064006f00740075002000410064006f00620065002000500044004600200064006f006b0075006d0065006e007400750073002c0020006b006100730020006900720020012b00700061016100690020007000690065006d01130072006f00740069002000610075006700730074006100730020006b00760061006c0069007401010074006500730020007000690072006d007300690065007300700069006501610061006e006100730020006400720075006b00610069002e00200049007a0076006500690064006f006a006900650074002000500044004600200064006f006b0075006d0065006e007400750073002c0020006b006f002000760061007200200061007400760113007200740020006100720020004100630072006f00620061007400200075006e002000410064006f00620065002000520065006100640065007200200035002e0030002c0020006b0101002000610072012b00200074006f0020006a00610075006e0101006b0101006d002000760065007200730069006a0101006d002e> /NLD (Gebruik deze instellingen om Adobe PDF-documenten te maken die zijn geoptimaliseerd voor prepress-afdrukken van hoge kwaliteit. De gemaakte PDF-documenten kunnen worden geopend met Acrobat en Adobe Reader 5.0 en hoger.) /NOR <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> /POL <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> /PTB <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> /RUM <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> /RUS <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> /SKY <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> /SLV <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> /SUO <FEFF004b00e40079007400e40020006e00e40069007400e4002000610073006500740075006b007300690061002c0020006b0075006e0020006c0075006f00740020006c00e400680069006e006e00e4002000760061006100740069007600610061006e0020007000610069006e006100740075006b00730065006e002000760061006c006d0069007300740065006c00750074007900f6006800f6006e00200073006f00700069007600690061002000410064006f0062006500200050004400460020002d0064006f006b0075006d0065006e007400740065006a0061002e0020004c0075006f0064007500740020005000440046002d0064006f006b0075006d0065006e00740069007400200076006f0069006400610061006e0020006100760061007400610020004100630072006f0062006100740069006c006c00610020006a0061002000410064006f00620065002000520065006100640065007200200035002e0030003a006c006c00610020006a006100200075007500640065006d006d0069006c006c0061002e> /SVE <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> /TUR <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> /UKR <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> /ENU (Use these settings to create Adobe PDF documents best suited for high-qua

towards a morphosyntactic analysis of mandarin mood/aspect

Documents