Mandarin Grammar Cheng Kai Ko
Object Shift in Mandarin Chinese
1. Introduction—The main goal of this research is to account for the syntactic
motivation and properties of Chinese object shift.Ex: 1(a)我看了這本書 VS 1(b)
我這本書看了。
--Adopting the Minimalist approach (Chomsky 1995), I will propose that all types of
object shift in modern Chinese check a strong nominal feature of v triggered by the
focus marker of the object. Thus there is no arbitrary optionality in object shift.
--The Trigger Hypothesis proposed in this paper claims that languages fidder in the
default strength of a formal feature, rather than the absolute strong/weak parameters,
as assumed in the present linguistic theories. The default strength can be changed
under certain conditions. Exploring such conditions enables us to explain the
variations between covert checking and overt checking of the same formal feature
in a single language.
--Three types of focusing: contrastive, additive, and restrictive types.
--Contrastive focus(2)他是星期三看了那場電影
--the additive or inclusive focusing highlights that some alternative is included as a
variable of the sentence meaning. The restrictive or exclusive focusing, on the
contrary, highlights that none of the alternatives under consideration satisfies the
relevant meaning of the sentence. Ex: (3a) Even John has come.(additive focused)
(3b) Only John has come.(exclusively focused)
2.The Issues
--2.1 Three Types of Preverbal Objects
--Chinese is an SVO language. ???
--Three types of preverbal objects in Mandarin Chinese. They share many properties.
One is that they all occur to the left of a sentence negation word, if there is one.
Another property: all of them are focused. When an object occurs to the left of a
transitive verb, it is semantically focused.
--Ex: Contrastive: (4)a. 他是看了那本書 b.他那本書看了
Additive: (5)a. 他甚至背過這篇散文 b.他(連)這篇散文都背過
Restrictive: (6)a. 他只喝紅茶 b.他只有紅茶可喝
--2.2 The Absence of Focus Markers with O in VO Order: One important observation
with respect to focus markers and a focused element is that if a focused element
which is not an object occurs in its canonical position, it can be preceded by a focus
marker. However, a postverbal object cannot be preceded by a focus marker. See
P205 (7).
--The additive focus marker in Chinese is 甚至 or 也. If 甚至 is used, the XP
following 甚至 can be any categories except CP; and if 連 is used, the XP
following 連 must be a nominal or a CP. Thus, if 連 is used before a nominal, it is
1
Mandarin Grammar Cheng Kai Ko
changeable with 甚至.
--Two kinds of focus markers are used in restrictive focus constructions: 只有 can
only be used before nominal elements, while 只 is usually used before non-nominal
elements. As with 是 and 連 cannot occur to the left of an object when it occurs in
its canonical position. Ex: See P207 (9)a~f.
3.structural Properties of Various Focus Markers
--3.1 Archi-forms of Focus Markers: In the additive type of focalization, there are four
focus markers:連、甚至、都、也。連 occurs only to the left of a nominal or clause,
甚至 occurs to the left of various categories except a clause, and 都 and 也 occur
only to the left of non-nominal/clausal categories, such as a verb or a preposition.
-- The focus marker 都 can exchange with another focus marker 甚至 in some
cases, if the focalized element is to its right, and can exchange with another focus
marker 也 if the focalized element is to its left.
-- Secondly, the additive focusing is that if the focalized element is to the left of this
arch-form of the focus marker, another focus marker 連, or 甚至, or 甚至連, can
precede the focalized element.
--In the restrictive type of focalization, there are three focus markers:只有、只、and
才. 只 and 才 are interchangeable, if the focalized element is a post-verbal
quantified element. See P210 (12)a.b.
--Unlike the additive focus markers, the occurrence of the restrictive focus marker is
constrained by aspectual factors. See P211 (13)a~e.
--3.2 The Syntactic Status of Focus Markers
--3.2.1 Against Modal Status of Focus Markers: Shi(1994) claims that focus marker
shi can be analyzed as a modal verb, since it shares two properties with the latter:
first, both can have A-not-A form; and second, both can occur in sentence initial
position, i.e. pre-subject position.
--What is relevant here is that if an element can have A-not-A form it is not
necessarily a verb or a modal verb.
--Thus, syntactically, focus markers do not behave like a modal verb. Semantically, a
modal verb expresses modality meaning, either epistemic or deontic, however, a
focus marker does not have modality meaning. Therefore, focus markers are not
modal verbs.
--3.2.2 Against Adjective Status of Prenominal Focus Markers
The categorical status of focus markers has been taken to be adverb generally.
Adverbs have been assumed to be licensed by the head of a verbal projection (the
term verbal projection is from Grimshaw 1991 and 1993)
--Shyu(1995) assumes that a pronominal focus marker is an adjective and a focus
marker elsewhere is an adverb. Sybesma(1996:13) comments on Shyu’s treatment of a
2
Mandarin Grammar Cheng Kai Ko
focus marker, stating that “calling it an adjective is unfortunate, but the idea that it is
adjoined to the phrase it forms a constituent with is correct.”
Thus, although both AP and a focus marker can adjoin to nominal phrase, they have
different structural positions. The former is lower than the latter. In this way, calling
focus markers adjectives is indeed “unfortunate”.
--3.3 The Interactions of Focus Markers with Question Elements in a Simple Sentence
A well-known property of the focus marker shi is that it cannot occur in direct WH
questions unless the WH element is the focused element. Ex: (21)a. *誰是買了那本
書? (21)b.是誰買了那本書?
--If a focus marker occurs to the left of a WH element, the sentence is always
unacceptable; however, if a focus marker occurs to the right of a WH element, the
sentence is acceptable in some case. See (22)(23)a~c
--shi is that it cannot co-occur with an A-not-A word, although shi itself in
A-not-A.See (24)a.b
-- When we observe other focus marker, we can see that if a focus marker occurs to
the left of an A-not-A word, the sentence is always unacceptable, however, if a focus
marker occurs to the right of an A-not-A word, which can be a focus marker in
A-not-A form, the sentence is acceptable in some cases.
--The generalization: The licensing relation between the [Q] pf C and a question word,
which is either a WH word or an A-not-A word, can be blocked by a non-interrogative
focus marker which has a shorter distance to C than the question word, regardless
whether the licensing relation is covert checking or unselective binding.
--3.4 M-command of Dou/Cai on Focused Elements: when an focused element occurs
to the left of the focus marker dou or cai, another focus marker such as lian or zhiyou
may occur in a sentence and adjoin to the focused element. Dou and cai are adverbs
and occurs to the left of a VP. See P219 (27)a.b.
-- The adverbs dou and cai must M-command the focused elements. Specifically,
when lian, shenzhi or shenzhi lian adjoins to an element, dou must M-commend the
element; while when zhi, or zhiyou adjoins to an element, cai, if it shows up, must
M-commend the element.
--If the focused nominal is a subject, which is at Spec of IP, dou or cai be either to the
left or the right of an IP adjunct since in either case dou/cai M-commends the focused
nominal.
--3.5 Principle of Lexical Association
--Tancredi’s (1990) Principle of Lexical Association: It states that an operator like
only must be associated with a lexical constituent in its c-command domain. See P223
(36)a~c.
4. Optionality and Triggered Strong Features
3
Mandarin Grammar Cheng Kai Ko
--4.1 Triggering Hypothesis: Modern Chinese is generally taken to an SVO language.
Contrary to Travis (1984) among others, many present generative suntacticians adopt
a hypothesis argued by Kayne (1994) that all languages are base-generayed as SVO.
--An assumption of the Minimalist Program is that all syntactic movement is driven
by strong feature checking. The preverbal object is moved from a postverbal position
to check a strong feature. In other words, the syntactic motivation of object shift in
Chinese is string feature checking.
--Object shift is neither obligatory nor absent in Chinese.
--Trigger Hypothesis: (37)
--Since there are both SVO and SOV orders in Chinese, it has been that object shift in
Chinese is optional.
--4.2 Locality Constraint on Triggering: One Spec of v is filled by the trace of the
subject. A shifted object must move to a different Spec of v. The relative order of
these two Specs of v, according to Chomsky, is that the shifted object is at the inner
Spec and the subject is at the outer Spec. This order is also compatible with Travis’s
(1993) and Koizumi’s (1995) Split VP hypothesis, which claims that the base position
of a subject is higher than both of the base position and the derived position of an
object (AgrOP).
--4.3 Accusative Case and Clause Bound Conditions of Object Shift: Chinese object
shift is Accusative Case related movement. Chinese CPs can have abstract Case
according to Tsai (1995). When a CP functions as the direct object of a transitive verb,
it has the Accusative Case and is able to be raised.
--Accusative Case feature is checked along with the triggered strong feature. Thus, we
do not need to take object shift to be a movement of an object to Spec of Agro solely
for the sake of checking Accusative Case feature, as assumed by Qu (1994).
--Restructuring can unify the domains of the embedded nonfinite verb and the matrix
control verb into one complex, thus licensing long distance object shift, shown in the
(48).
--4.4 Triggered [D] and the Trigger’s [Specificity]
--One might think that the triggered strong feature is [Focus], an uninterpretable
feature, and that the focused object moves to the checking domain of v to check the
strong [Focus] feature of v, as in the case that a WH phrase n English moves and
checks the strong [Q] of CQ. Alternatively, one can assume that the triggered
difference between a [Focus] and a categorical feature hypotheses.
--The specificity requirement on the contrastive type of object shift has been pointed
out by Tsao (1990) and Qu (1994). P233, (49)~(51)
--Classifier-Noun phrase P234
--Diesing’s assumption (1997): movement can be either syntactically driven by
4
Mandarin Grammar Cheng Kai Ko
checking of uninterpretable features such as Case, or semantically driven in the sense
that movement is necessary as a result of the interpretive requirements of certain types
of noun phrases.
--Definite objects must be raised out of VP, while indefinite objects need not, since
they can be bound by the existential closure by default.
--Tow kinds of economy requirements: The syntactic economy requires that a
movement should be delayed of possible, while the semantic economy requires that
indefinite object need not be raised.
5. Conclusion: In this paper, based on the investigation of Chinese object shift, I
have made one major theoretical claim, namely, that the strength of a feature can
be triggered to be strong. This claim implies that languages differ in default
strength of a formal feature, rather than in terms of absolute strong/weak
parameters.
5