University of Southern California Dissertations and Theses

Locatives in Chinese and Japanese: distribution, case assignment and c-command

(USC Thesis Other)

Locatives in Chinese and Japanese: distribution, case assignment and c-command

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content LOCATIVES IN CHINESE AND JAPANESE:
Distribution, Case Assignment and C-command
by
Ebony Leigh Bostic
A Thesis Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF ARTS
(East Asian Languages and Cultures)
December 1994
Copyright 1904
Ebony Leigh Bostic
DEDICATION
I dedicate this work to my parents and brother, Mr. Raphael T. Bostic, Mrs.
Viola W. Bostic and Mr. Raphael W. Bostic, who encouraged my natural curiosity and
desire to learn throughout my childhood. I also dedicate this to my ancestors who
were never granted such an educational opportunity.
ACKNOWLEDGMENTS
I would like to express my deepest thanks and appreciation to my thesis
committee members Dr. Hajime Hoji, Dr. Audrey Li, and Dr. Mieko Han without
whose help none of this would be possible. I would also like to thank Dr. Peter Nosco,
Chairman of the East Asian Languages and Cultures Department, for giving me the
opportunity to prove myself as a scholar and to study in his department. His help and
constant encouragement enabled me to overcome some personal doubts and continue
my education. As a role model he was, and continues to be, paramount.
Further thanks are due to Mrs. Hazel Bates for her endless warmheartedness
and kindness; to my fellow graduate students in East Asian Languages and Cultures for
their patience; to my native speaker informants who graciously endured my constant
badgering.
Finally I would like to thank my family and friends for providing me with a
strong support system. To know that they unequivocally support me is a comfort
wherever I happen to be in the world.
CONTENTS
Dedication ii
Acknowledgments iii
Abstract v
Introduction 1
2. Distribution — Chinese 6
2.1. Noun Phrase (NP) Distribution 7
2.2. Prepositional Phrase (PP) Distribution 9
2.3. Locative Phrase (LP) Distribution 13
2.4. Conclusion 15
3. Case and Theta Theory 15
3.1. Case and Theta Theory in Chinese 17
3.2. Conclusion 20
4. Distribution — Japanese 20
4.1. Noun Phrase Distribution 21
4.2. Locative Phrase Distribution 25
4.3. Conclusion 27
5. Case Assignment in Japanese 27
6. Binding Observations 29
6.1. Binding and C-Command in English 30
6.2. Binding and C-Command in Chinese 34
6.3. Binding and C-Command in Japanese 38
7. Final Conclusions 43
Appendices
A Location Words 46
B Prepositions/Co-Verbs 46
C Brief Summary of Constituent Relations Through Use of a
Labeled Tree Diagram 47
D Acceptability-native speaker judgments 48
References 5 1
IV
ABSTRACT
This thesis will discuss two concepts in Chinese and Japanese linguistics,
classification and binding. First, it will discuss location words in Chinese and Japanese
in terms of their syntactic patterning in order to classify them as one part of speech,
preposition, or another, noun. The patterning is first observed in terms of distribution
within a sentence and second examined in terms of case assignment. This structure is
drawn from the work of Li (1990) in which a similar discussion was undertaken.
Within this discussion, previous work done on this topic will be included. The key
scholars within this discussion are Li, Li and Thompson, and Chao in Chinese and
Martin and Isami in Japanese. Within the field of Chinese, Li (1990) and Chao (1968)
argue that location words are nominals while Li and Thompson (1981) classify' them as
particles without going into detail why or how they arrived at this classification. In
reference to Japanese location words, both Isami (1964-5) and Martin (1987) refer to
them as nominals. This paper will support the work of those scholars who classify
these words as noun-like or nominals.
Second, based on the conclusions drawn in the first section, i.e. that location
words are nominals in Chinese and Japanese, the paper will embark upon a discussion
of binding and specifically constituent-command theory (hereafter, c-command). C-
command theory strives to predict coreferential possibilities within a sentence based on
syntactic relationships. Generally it claims that if the second noun phrase in a sentence
is c-commanded by the first noun phrase then coreference between these two noun
phrases is impossible. Conversely, if the second noun phrase is not c-commanded by
first noun phrase, then coreference should be possible.
The expectation is that since location words are indeed nominals, they affect
certain structural relationships, namely the c-command relationship between the
locative-preceding noun (the first nominal) and its possible referent within the predicate
of the sentence (the second nominal). That is to say, this paper will investigate the
referential relationship between two items within a sentence containing a nominal
locative.
VI
1. INTRODUCTION
This analysis will be a study of a group of words in Chinese and Japanese that
will be referred to as location words (see appendix A). I have observed that phrases in
Chinese such as [NP + location word] and phrases in Japanese such as [NP + genitive
no + location word], hereafter called locative phrases, appear on the surface to be
adpositional phrases. However, upon the following more detailed examination of the
facts, these phrases pattern and behave similar to noun phrases. The general basis of
evaluation will be an analysis of the distribution and case assignment of these phrases.
The analysis will involve summarizing in the first few sections the research
previously done in Chinese on the classification of locative phrases. This research
involves comparing the distribution of noun phrases, prepositional phrases and locative
phrases. (For a more detailed discussion, see work by Li (1990), Li and Thompson
(1981) and Hagege (1975).) In the subsequent sections, the distribution and case
assignment of those phrases in Japanese will be tested following the structure of the
Chinese instance and referring to previous classifications in Japanese linguistic
scholarship. These sections will provide support for the theories that classify these
phrases as noun phrases.
Next, I will discuss binding and specifically c-command theory based on the
conclusion that locatives are nominals. If locatives are nominals then it should follow
that these words affect the coreferential relationship between two words within the
1
sentence. In other words, I expect that these locatives block the c-command
relationship and thereby allow coreference.
In a broad perspective, this discussion is part of a larger issue dedicated to
revealing properties o f individual languages and how these properties work within
Universal Grammar. The purported purpose of Universal Grammar is to explain the
behavior of all languages in the most genera] terms. Universal Grammar is “the system
of principles, conditions, and rules that are elements or properties of all human
languages.” 1 It is a theory concerned with knowledge or competence, not of behavior
or performance.
Two people may share exactly the same knowledge of a language but
differ markedly in their ability to put this knowledge to use. Ability to
use a language may improve or decline without any change in
knowledge. This ability may also be impaired...with no loss of
knowledge, a fact that would become clear if injury leading to
impairment recedes and lost ability is recovered.2
Universal Grammar assumes that all speakers, regardless of language, are familiar with
the same system of rules and principles.
Universal Grammar asserts that all of this is innate, not learned. This assertion is
. the result of studies done with children in the process of mastering their native
language. It was found that children do not make certain mistakes despite a “poverty of
1 Chomsky, Noam, Reflections on Language. London: Temple Smith, 1976,
p.29.
= Chomsky, Knowledge of Language: Its Nature. Origin, and Use. New
York: Praeger, 1986, p.9.
stimulus.’0 The children consistently employed complex computations to create
sentences instead of opting for a computationally simpler linear theory. The innateness
of the Universal Grammar mechanism, called a “language acquisition device,” explains
why and how children are able to accomplish such an incredible fete without being
taught. Every human is bom with an internal language device pre-coded with rules
about language. Differences in languages only arise as a result of limits placed on
Universal Grammar by that particular language, otherwise all languages employ the
same principles.
An example of a Universal Grammar principle that applies to all languages is
structure-dependency. This principle claims that sentences rely on structural
relationships instead of linear ordering. An example in English is passivisation. If
evaluated in terms of linear ordering, changing an active sentence into a passive
sentence might be viewed as follows.
1 2 3 4 3 4 (was) (changed 2)
I ate the fish. The fish was eaten.
However, this does not apply to all sentences as is witnessed below.
1 2 3 4 5 3 4 (was) (changed 2)
I ate with my brother. *With my was eaten.
In order to transform an active sentence into a passive sentence it is essential to know
the syntactic categories and structural relationships existing within the sentence and not
5 Chomsky, 1986, p.7. A situation in which there are limitations on the data
available.
simply the linear order. Below are examples o f structure-dependency in Chinese and
Japanese dealing with the Projection Principle.4
Chinese
Zhangsan fanjian-1i shui le.
(name) room-in sleep-ed
Zhangsan slept in the room.
Zhangsan changhu guan le.
(name) window close-d
Zhangsan closed the window.
Japanese
John ga ringo o tabeta.
(name) GA apple O ate
John ate an apple.
John ga ringo o kitta.
(name) GA apple O cut
John cut an apple.
Zhangsan shui le.
(name) sleep-ed
Zhangsan slept.
*Zhangsan guan le.
(name) close-d
* Zhangsan closed.
John ga tabeta.
(name) GA ate
John ate.
*John ga kitta.
(name) GA cut
*John cut.
As stated above, structure-dependency, a Universal Grammar principle, applies to all
languages. However, “if knowledge of language consisted simply of unvarying
principles, all human languages would be identical.”5 The ordering of constituents in a
phrase is an example of variations between languages.
Universal Grammar states that every sentence can be broken down into
constituent phrases— Noun Phrase (NP), Verb Phrase (VP), Prepositional Phrase (PP)
and so on. It also provides that each phrase has a “head” after which the phrase is
4 The Projection Principle is a principle asserting that there are restrictions
on what words in the lexicon can occur in what constructions.
5 Cook, Vivian C., Chomsky’s Universal Grammar. Oxford, UK: Basil
Blackwell Ltd., p.7.
4
named--noun, verb, preposition and so on. The head of the phrase is the most important
constituent in the phrase in that all other constituents revolve around it. In each phrase,
there are two possible positions for the head of a phrase. In a head-left phrase, the head
of the phrase occurs to the left of the other phrasal elements. Conversely, in a head-
right phrase, the head appears to the right of the other constituents. This parameter
illustrates a variation between languages. Some languages, such as English, are head-
left while other languages, such as Japanese, are head-right.
English: Japanese
Once again, the aim of Universal Grammar is to explain language behavior in the most
general tenns, therefore it is not necessary to make an exhaustive list of every possible
phrase in both English and Japanese. It suffices to state the generalization that all
phrases in English are left-headed while those in Japanese are right-headed.
Realizing that individual languages must utilize what is available in Universal
Grammar to create a particular grammar gives rise to some interesting questions
somewhat related, if tangentially, to the present discussion of location words in Chinese
and Japanese. These questions will be noted despite the fact that they will not be
addressed within the scope of this discussion. First, if all languages operate within
Universal Grammar, why does English require two words with different classifications
VP = V(PP)
go to the store
VP = (PP) V
mise e iku
PP = P (NP)
to the store
PP = (NP) P
mise e
and functions to describe one location (i.e. top, noun; above, preposition) while Chinese
and Japanese only need one?6 Second, do these location words in Chinese and
Japanese actually serve two functions or is it the case that they only serve one function
and these languages have another, innate mechanism to deal with the other function?
My goal here is not to answer these very intriguing questions but to describe an
observed phenomena, namely, the behavior of location words, apply it to c-command
theory and draw conclusions, gleaned therefrom. Finally, I will briefly discuss
implications of these conclusions.
2. DISTRIBUTION - CHINESE
Distribution refers to the acceptable syntactic positions that words or phrases
can occupy in a sentence. In this section I will examine and summarize findings in Li
(1990) and Li and Thompson (1981) among others, with respect to the distribution of
noun phrases, prepositional phrases and locative phrases. Each section will include a
brief explanation of the phrasal category under study as well as demonstrations of its
distribution in Chinese. Finally, the results will be synthesized in a concluding section.
‘ I use the words “require” and “need” based on the assumption that
Universal Grammar favors parsimony and therefore would not foster superfluous
syntactic/lexical constituents.
6
2.1. NOUN PHRASE (NP) DISTRIBUTION
Stated simply, a noun phrase (NP) is a phrase constructed around a noun. It can
consist of one word or more than one word and have different acceptable syntactic
positions. For example, in English, both "the big, red apple” and "it" are noun phrases
and they can occupy subject positions and object positions with equal ease. In Chinese,
there are similar rules governing noun phrase composition and distribution. In addition
to being able to occupy the subject and object positions in a sentence, noun phrases can
also be the head of three types of phrases: classifier/measure phrases, associative
phrases and other modifying phrases involving adjectives and relative clauses.
Classifier/measure phrases, as the terms themselves denote, seek to qualify the
noun into a measurable unit. This type of phrase consists of a classifier/measure word
preceded by either a number, a demonstrative or certain quantifiers. Examples are san-
ge, na-ge and mei-ge. Classifiers must be used only if the noun itself does not already
denote a measurable unit. For example, the correct way to say three days in Chinese is
san tian. It is incorrect to say san ge tian since tian (day) already denotes a measure.
Chao (1968) corroborates the creation of noun phrases through the usage of
classifier/measure words which he categorizes in a more general grouping called "D-M
compounds."7 According to Chao, who provides a fairly complete list of measures,
A D-M compound is simply a compound of determiners and measures. Chao
makes no group distinction between the terms measure and classifier. Instead, he uses
the term classifier within the group of measures.
7
however, the proper use of classifiers before nouns is "a matter of words and not
things." In several cases, despite the fact that two classifiers have the same general
definition, they are not interchangeable. For example, all three measure words zhang.
feng and ge can refer to flat, paper objects, but their usage is very restricted. In order to
say “one letter” a speaker must choose i feng shin and not i ge shin or i zhang shin.
Similarly, to say “one envelope,” a speaker must choose i ge shin feng and not i shin
shin feng or i feng shin feng. Also, i zhang zhao pian is the only acceptable way to say
“one picture."
The associative phrase is the second type of phrase in which a noun phrase can
serve as the head. Associative phrase modifications are formed by connecting an
associative phrase with a noun phrase. The associative phrase itself consists of a noun
phrase and the particle de. This type of phrase is so named because the two nouns
within this modification are somehow associated.8 One veiy common type of
association is possession as in Gubo de bi (Gubo's pen) and wo de shu (my book). The
unique aspect of noun phrase usage within associative phrases, though, is that a noun
phrase can be placed both preceding and following the particle de such that the resulting
structure is [NP + de + NP].
8 The work of Li and Thompson does not specify each particular association.
Instead they defer and report that the nature of each association depends entirely on the
two noun phrases involved.
8
"A relative clause in any language is a [nominalized] clause that restricts the
reference of the head noun."9 A third acceptable position for noun phrases is the head
of a modifying phrase involving relative clauses or adjectives. Some examples of noun
phrases heading relative clauses are as follows.
Relative clauses not only consist of nominalized clauses but also sometimes occur with
adjectives.
These attributive adjectives can either modify a noun phrase with the particle de,
in which case it is called a relative clause, or without the particle de in which case it is
known as a compound noun. Despite the terminology, however, both phrases serve to
restrict the reference of the head noun, flower.
hong de hua a flower that is red
hong hua a red flower1 0
This section on noun phrase distribution has shown that noun phrases in Chinese
can function as the head of classifier/measure phrases, associative phrases and other
modifying phrases involving relative clauses and adjectives.
Charles Li and Sandra Thompson, Mandarin Chinese: A Functional Reference
Grammar. Berkeley: University of California Press, 1981, p.579.
1 0 This example was taken from Li and Thompson p. 118 their #90.
ni chuan de yifu
wo chi de yu
km dianying de ren
the clothes you are wearing
the fish that I ate
the people who watch movies
9
2.2. PREPOSITIONAL PHRASE DISTRIBUTION
Historically, discussions concerning prepositional phrases in Chinese have been
controversial. As far back as the 1800s, sinologists and linguists have debated the
existence of prepositions. For example, scholars such as Maspero from France contest
the existence of the preposition as a separate category in Chinese, proposing instead that
it is a subcategory of verbs. These scholars are guided by two ideals: that the
differences in meaning are not important and that it is wrong to analyze and categorize a
non-western language in western terms.
In further attempts to categorize prepositions and verbs scholars have explored
recurrent trends based on phonology. The 1975 work of Claude Hagege discusses
those results as well as problems related to those methods of classification. He writes,
for example, that zai is most often pronounced with an intense accent when it has the
meaning of "to exist, to be alive," and less strong in the usage as a preposition. He
invalidates this type of result by saying that "...this is only a general tendency, and we
have encountered counterexamples from the work of some informants."1 1 Instead of
clarifying the debate, Hagege only seems to further complicate the issue.
In 1981 Li and Thompson presented a convincing historical explanation for this
debate. According to this work so-called prepositions in Chinese are referred to as "co
verbs" because they share properties of verbs and prepositions. Originally, these words
1 1 Claude Hagege, Le Probleme Linguistique des Prepositions et la Solution
Chinoise, Paris: Editions Peeters, 1975, p.40.
10
were verbs but through the evolution of Mandarin over the years, they have come to be
used as prepositions in varying degrees. As a result, some are more verb-like while
others are more preposition-like. That is to say, they overlap syntactic categories. Co
verbs do primarily function as prepositions, though, and to better serve the needs of this
discussion, phrases involving co-verbs will hereafter be referred to as prepositional
phrases. The [co-verb + NP] pattern results in a phrase that modifies the verb, which is
the function of a prepositional phrase.
Concerning distribution, in addition to having the capacity to serve as the
subject in a sentence, prepositional phrases can occupy the V| position in a Vr V2
sequence creating a PP-V2 structure. Chao (1968) defines a Vj-V2 series as a
subordinative construction where V2 has the same meaning as the whole and therefore is
the center of the expression while V) is simply a modifier modifying V2 . The term
"subordinative construction" is used in opposition to the term "coordinative
construction." The two differ in that the actions in a coordinative construction are
reversible without changing the semantic value of the phrase while, in a subordinative
construction, the actions follow a certain order and changing the order in turn changes
the semantic value of the phrase. That is to say, one action in a subordinative
construction is subordinate to the other. Although he uses the term V), Chao notes that
it is "often translatable by a prepositional phrase."1 2
1 2 Yuen-Ren Chao, A Grammar of Spoken Chinese. Berkeley: University of
California Press, 1968, p.326.
1 1
mao congfangjian poo dm lai the cat comes running from the room
wo zai tushuguan du slut I read books at school
In the above sentences pao dm lai (come running) and du shu (read books) are the
semantic centers of the sentences while the prepositional phrases in bold modify those
verbs.
Referring back to section 2.1, it will be recalled that the structure of associative
phrases was introduced as [NP + de + NP]. Prepositional phrases, specifically those
which can serves as predicates, can successfully be inserted into this structure.
However, there is one restriction; prepositional phrases can only be inserted in the
position preceding de such that the resulting phrase structure is [PP + de + NP].1 3
zai jia de rizi the days at home
dui pengyou de taidu attitude toward friends
The phrases in bold above are the prepositional phrases, which, combined with the
particle det modify the ensuing nominal expressions.
This section has demonstrated that prepositional phrases can occupy the
positions of a) subject, b) the V| position in a VpV; sequence and c) preceding the
particle de in associative phrase structures.
1 3 Chao, p.326. Chao notes that there are instances when prepositional phrases
follow the particle de in Associative phrases. However, they are rare.
2.3. LOCATIVE PHRASE (LP) DISTRIBUTION
Locative phrases in Chinese are composed of a noun phrase and a location
word. In certain situations locative phrases may contain another constituent such as,
zai.1 * In Li (1990) it has been determined through an analysis of distribution that these
locative phrases have the same distribution as an ordinary noun phrase. They both can
occur in subject, object and prenominal modifier environments.
1) zhuozi-shang you wode shu. (subject)
table-on to have I DE book
There are some of my books on the table.
qing ni kan zhuozi-shang, hao bu hao? (object)
please you look table-on ok not ok
Will you please look at the area above the table?
zhouzi-shang de bi (prenominal modifier)
table-on DE pen
The pen on the table
2) ta you wo de shu (subject)
He to have I DE book
He has my book
qing ni kan zhaopian, hao bu hao? (object)
please you look picture ok not ok
Will you please look at the picture?
Gubo de bi (prenominal modifier)
(name) DE pen
Gubo's pen
1 4 Infra. Discussion of the Chinese coverb zai occurs in section 3.
13
Sentences in 1) and 2) show that both locative phrases and noun phrases can occur in
the same syntactic position. In contrast, locative phrases do not have the same
distribution as prepositional phrases. When comparing the distributions of locative
phrases and prepositional phrases the following examples illustrate the disparity in their
distribution. This disparity is related to the fact that prepositional phrases are not
argument positions, a discussion of which is forthcoming. Locative phrases cannot
occur in the same positions as prepositional phrases as is evidenced by the following
examples.
mao cong fang-jian pao-chu-lai. (PP in adjunct position)
cat from room run-leave-come
The cat comes running from under the chair.
*mao yizi-xia pao-chu-lai. (LP in adjunct position)
cat chair-under run-leave-come
The cat came from under the chair.
wo zai tushuguan du shu. (PP in adjunct position)
I at library read book
I read books at the library.
*wo xuexiao-li du shu. (LP in adjunct position)
I school-in read book
I read books at school.
Similarly, prepositional phrases cannot occur in the same positions as locative
phrases. Specifically, they cannot occur in prenominal modifier environments.
shu-li de hua (LP in prenominal modifier phrase)
book-in DE words
the words in the book
14
*dui ta de hua (PPinprenominalmodifierphra.se)
to him DE words
the words to him1 5
In order to use a prepositional phrase in prenominal modifier environments, a verb must
be added changing the type of modification into a relative clause.
dui ta shuo de hua
to him to say DE words
the words spoken to him1 6
2.4. CONCLUSION
The results of comparing the distribution of locative phrases with that of noun
phrases and again comparing the distribution of locative phrases with prepositional
phrases strongly suggest that locative phrases should be classified as noun phrases and
not prepositional phrases. However, the analysis does not end there. In conjunction
with distribution within a sentence, case assignment needs also to be examined to
provide more evidence concerning the nature of locative phrases.
3. CASE AND THETA THEORY
The concept of case refers to the encoding of information about a noun phrase's
syntactic role in a sentence. It is governed by a distributional limitation on constituents
1 5 Yen-hui Li, Order and Constituency in Mandarin Chinese. Boston: Kluwer
Academic Publishers, 1990, p.5. Her example 7a.
1 6 This example was taken from Yen-hui Li, p.5 her example 7b.
15
in a sentence. In other words, certain elements in a sentence can only occur in certain
positions. In English, case is overt. The value of the words "he, him and his" is
assigned to the same referent, however, the form of the word must change depending
upon its syntactic role in the sentence. For example the underlined portions in the
following sentences have a direct correlation to the syntactic role/position in
parentheses.
He hit the ball, (nominative: subject)
I hit him, (accusative: direct object)
This is hjs book, (genitive: possessor)
I gave it to him, (dative: indirect object)
This is an example of general distributional limitation because the underlined forms can
only be used in the specified syntactic positions of subject, direct object and so on. The
following sentences, which break this limitation, are unacceptable.
*Him hit the ball.
*1 hit hjs.
*This is he book.
*1 gave it to he.
Some other pronominal case paradigms in English include "I, me, my, mine" and "You,
you, your, yours." Otherwise, the contrast of nouns used as subjects and direct objects
in English is not distinguished by case. Instead, it is marked by word order and the use
of prepositions.
For some time linguists have been trying to explain the link between case
assignment and theta theory.1 7 Although there are several theories on the exact nature
1 7 Richard K. Larson, "Bare-NP Adverbs," Linguistic Inquiry 16:595-621.
16
of the link, it is commonly agreed that case is automatically assigned to those positions
assigned a theta role in theta theory. These positions are called argument positions, and,
once again, correspond to essential constituents in a sentence such as subject and object.
On the other hand, non-essential constituents are referred to as non-argument positions
or adjunct positions. They are called such precisely because they are adjuncts; they
convey superfluous information regarding where, when and/or in what manner the main
verb takes place.
An example sentence is given below.
(At 3pm,) John ran (to the store)(quickly).
The nuclear sentence is underlined. The verb "to run" assigns only one theta role, that
of agent, and the subject of the sentence, John, receives that theta role. "John" is
licensed by the verb and fills the only argument position in this sentence. The other
three phrases contribute to the sentence semantically by adding the time, destination and
manner of John's running. Ignoring the adverb "quickly" because adverbs of manner are
beyond the scope of this discussion, these positions are adjunct positions and are in need
of case assignment.
3.1. CASE and THETA THEORY IN CHINESE
Case in Chinese is referred to as abstract case because unlike English, the form
of the words do not change. With one look at a sentence in Chinese, case assignment is
not as easily distinguishable as it is in English. However, case in Chinese still involves a
17
distributional limitation on noun phrase positioning. According to the Case Filter, case
must be assigned to lexical noun phrases.1 3 Therefore, noun phrases in typical noun
phrase positions - subject, object and so on - all have a case assignment. These
positions automatically are assigned case. However, when noun phrases occur in a non-
typical noun phrase position, they do not receive this default case assignment given to
typical noun phrase positions. Therefore, noun phrases in these non-typical noun phrase
positions need to have case assigned from some other source.
In Chinese, the source for obligatory case assignment in order to fulfill the Case
Filter is any coverb such as zai. The positions of the lexical noun phrases in the below
three examples are all argument positions. They are all licensed in the sentence and are
therefore obligatory.
to you wo de shu. (subject)
he to have I DE book
He has some of my books.
qing ni kan ta, haobu hao? (object)
please you look he ok not ok
Will you please look at him?
to de bi (prenominal modifier)
he DE pen
His pen
Similarly, locative phrases can replace the lexical noun phrase in these argument
positions and receive default case assignment.
1 8 Yen-hui Li, p.25.
18
zhouzi-shang you wo de shu. (subject)
table-on to have I DE book
There are some of my books on the table.
qing ni kan zftuozi-shang, hao hu hao? (object)
please you look table-on ok not ok
Will you please look at the area above the table?
zhouzi-shang de bi (prenominal modifier)
table-on DE pen
The pen on the table
In contrast to the above default case assignment, compare the following sentences.
mao cong fang-jian pao-chu-lai. (PP in adjunct position)
cat from room run-leave-come
The cat comes running from under the chair.
* mao yizi-xia pao-chu-lai. (LP in adjunct position)
cat chair-under run-leave-come
The cat came from under the chair.
wo zai tushuguan du shu. (PP in adjunct position)
I at library read book
I read books at the library.
*wo xuexiao-li du shu. (LP in adjunct position)
I school-in read book
I read books at school.
When a locative phrase occurs in a non-argument position it requires any case-
assigning coverb for acceptability. This is because without a coverb, the locative phrase
is not assigned case and therefore violates the Case Filter. The occurrence of a coverb,
a case assigner, satisfies the Case Filter and the sentences are acceptable.
mao cong yizi-xia pao-chu-lai. (LP in adjunct position)
cat from chair-under run-leave-come
The cat came from under the chair.
19
wo m i xuexiao-li du shu. (LP in adjunct position)
I at school-in read book
I read books at school.
3.2. CONCLUSION
With results similar to the distribution analysis, an investigation of abstract case
assignment in Chinese contributes to the positive identification of locative phrases as
noun phrases. The Case Filter theory demands that all noun phrases in a sentence be
assigned case. When a locative phrase occurs in a position that does not receive default
case marking, it absolutely must occur with a coverb in order to receive case marking.
This identifies locative phrases as having noun phrase-like characteristics.
4. DISTRIBUTION - JAPANESE
The Japanese language exhibits very similar characteristics. There exists a
group of words denoting location (see Appendix A) which when paired with a noun
create a locative phrase. One difference from the Chinese, however, is that the Japanese
grammar necessitates the use of a genitive particle, no, between the location word and
the noun. Examine the following examples: hon no shita (under the book), rajio no
ushiro (behind the radio), and eki no mae (in front o f the station). The Japanese
language is also very similar to Chinese in terms of the intermittent use of other
20
constituents in conjunction with these locative phrases, the phrase particles ni, o and
de,1 9
Just as the locative phrases in Chinese pattern similar to noun phrases, the
following section will support the claim that these locative phrases in Japanese should
also be classified as noun phrases instead of postpositional phrases. The term adposition
is used in this circumstance because Japanese is a postpositional language and therefore
has no prepositions. In fact, unlike Chinese, within the Japanese grammar, there is a
consensus regarding the classification of location words as nominals. Therefore this
section will discuss noun phrase distribution and illustrate how locative phrase
distribution is similar to it.
4.1. NOUN PHRASE DISTRIBUTION
The first step in determining the classification of locative phrases in Japanese is
to examine the distribution of ordinary noun phrases. In addition to taking the subject
and object roles in a sentence, ordinary noun phrases can do all of the following.
A noun phrase can be the head of a relative clause. In Japanese as in English
and Chinese, relative clauses restrict the reference of the head noun. The following are
examples of noun phrase-headed relative clauses.
watashi ga katta kuruma the car that I bought
watashi ga tabeta sakana the fish that I ate
eiga o mint hito people who watch movies
1 9 Infra. A discussion of the Japanese phrase particle ni will occur in section 5.
21
kare ga itta tokoro the place where he went
kanojo ga sooji shita heva the room that she cleaned
Noun phrases are determinable by the pre-nominal "ko/so/aM series. Martin
(1987) quotes Yasuo Isami (1964-5) stating that the exceptions to noun phrases being
determinable by the ko/so/a series are action and abstract nouns.2 0 Isami claims that the
words he classifies as abstract nouns can be quantified (“a little ...") but not directly
counted or measured, nor are they actions that can be engaged in. His classification of
action nouns relates to words which can connect with ... o suru (to do), but with the
nominalizer no replacing suru. This subcategorization of nouns corresponds to the
renyookei stem described by Iwabuchi, in the Japanese school grammar endorsed by the
Japanese Ministry of Education. Examples are oyogi, tsuri, mane and so on.
Evidence is available, however, to call into question the absolute validity of this
exception. Although Isami painstakingly divides the open category of noun into eleven
different sub-categories, there is still some discrepancy. His categories consist of
countable nouns, measurable nouns, human nouns, non-human animate nouns, place
nouns, time nouns, mass nouns, relational nouns, action nouns, abstract nouns and
quantity nouns.2 1 Certain words, such as shiawase (happiness), qualify as abstract
nouns yet are indeed determinable by the pre-nominal "ko/so/a" series. It is completely
2 0 Isami, qtd. in Samuel Martin, A Reference Grammar of Japanese. Vermont:
Charles Tuttle Company, p. 177.
2 1 Martin, p. 177.
acceptable to say kono shiawase. Despite this incongruity however, Isami's assertion is
largely valid; most nouns can be determined or qualified by the pre-nominal "ko/so/a"
series as is demonstrated as follows: kono nikagetsu (these two months), sono san-doru
(those three dollars), ano hito (that person), kono tokoro (this temporal place: recently),
sono renshuu (that practice).
Noun phrases can be substituted by sore (that), Martin notes, with the
exception of relative place and time nouns.2 2 This exception is valid for two reasons.
First, sore is classified as a nominal, however its use is restricted to things and, in a
derogative manner, people. In order to refer to places, a speaker must use the ko/so/a
nominal specifically designated for places - koko, soko, osoko. Second, within the
ko/so/a series, there is no specific designator of time. Therefore, a speaker must use the
more general, default pre-nominal ko/so/a series - kono toki, sono toki, ano toki.
Examples of sore replacing noun phrases follow.
Anata no shoorai wa akarui ne. Your future is bright.
Sore wa akarui ne. It's bright.
Watashi ga hoshii no wa aisu kuriimu desu. What I want is ice cream.
Watashi ga hoshii no wa sore desu. What I want is that.
It is interesting to note that sore is the only member of the nominal ko/so/a
paradigm that has the ability to replace any noun phrase. If all noun phrases were
charted according to the tangibility of their possible referents on an "abstractness
2 2 Martin, p. 177-83.
23
continuum” with the low end of the continuum representing physical nouns and the high
end representing completely abstract nouns, the different possible usages of each
member of the ko/so/a paradigm becomes more clear. Kore "this" must refer to a
physical object within close proximity to the speaker and listener; kore cannot refer to an
abstract noun. That is to say that it cannot be used non-deictically. As a result, it is
placed on the low end of the continuum. Are on the other hand, can not only refer to a
physical object, but also to an abstract idea of which the speaker and listener share a
common knowledge. An example is the utterance "Are, are, are,,." when a speaker is
trying to remember something which both parties can recall. Since it can function
deictically and somewhat non-deictically, on the "abstractness continuum," (ire is placed
near the center. Of the ko/so/a paradigm, sore can refer to physical objects as well as
abstract nouns and therefore is placed at the high end of the continuum It can refer, for
example, to a lecture or an attitude and there are also cases where it has no direct
referent and functions similar to the empty pronoun "it" in English. Viewing this
"abstractness continuum" as a modified hierarchy, the all-encompassing nature of sore's
ability to replace any noun phrase - to function deictically and non-deictically, to the
exclusion of kore and are, is revealed.
A noun phrase can attach to another noun through the obligatory use of the
particle no. This structure, similar to the associative phrase in Chinese, can also apply to
locative phrases and this relationship will be addressed in section 4.2. In Japanese, this
associative structure can be created in two distinct manners. The first version allows the
24
noun phrase to precede the particle no and the attaching noun such that [NP + no +
(noun)] results. The second version is the exact reverse; the noun phrase follows the
attaching noun and the particle no. What results is f(noun) + no + NP]. Both
renderings are completely acceptable.
4.2. LOCATIVE PHRASE DISTRIBUTION
The next step is to illustrate the compatibility of locative phrases with noun
phrases in what has been previously determined to be noun phrase distribution. A
locative phrase can be the head of a relative clause which restricts its reference.
However, when dealing with locative phrases, other items need be considered.
Consider the following example.
Watashi ga sooji shita tana no shita...
I GA clean-ed bureau GEN under
This sentence fragment can be interpreted in two ways. It can mean "under the bureau
that I have cleaned" or it can mean "under the bureau where I have cleaned." The
difference between the two sentences is that in the first sentence, the bureau is what was
cleaned and in the second sentence, the area under the bureau is what was cleaned. This
distinction is simply transmitted by the speaker through accent. For the purposes of
discussing locative phrases serving as the head of a relative clause, however, the latter
interpretation is relevant.
25
Excluded from the exceptions of action and abstract nouns,2 3 locative phrases
are determinable by the pre-nominal ko/so/a series. Two examples are as follows.
kono ie no shita sono teburu no ue
this house GEN under that table GEN above
Once again, these phrases have two semantic interpretations, however this discussion
necessitates the interpretation where in the first case, for example, kono (this) modifies
the entire locative phrase ie no shita such that the phrase would mean "this area under
the house."
In terms o f the third quality listed for noun phrase distribution, namely, noun
phrases can be substituted by sore (that), locative phrases do not qualify because they
are one o f the exceptions - relative place nouns. As stated earlier, Japanese has a
specific mechanism to express locative pronominals. It takes the form of the koko, soko,
asoko pronominal paradigm and can only be used for locations.
Locative phrases can be used in an associative structure. They can attach to a
noun through the obligatory use of the particle no. In the discussion in section 4.1 on
noun phrase distribution, it was noted that this associative relationship is reversible. The
same is true of locative phrase usage within an associative phrase.
tsukue no ue no hon (LP + no + noun)
desk GEN top GEN book
the book on top of the desk
2 3 Martin, p. 177.
26
Jon no liana no shita (noun + no LP)
John GEN nose GEN under
John's top lip
(LP + no + noun) and (noun + no + LP) are equally acceptable, however, in order for
the listener to arrive at the latter bracketing of the phrase, more mental machinations are
necessary.
4.3. CONCLUSION
As demonstrated in sections 4.1 and 4.2, noun phrases and locative phrases have
similar distributions. This corroborates the conclusions of scholars in the field of
Japanese linguistics, among whom Martin and Isami figure largely, who classify location
words as nominals.
5. CASE ASSIGNMENT IN JAPANESE
Case assignment is inherently linked to particles in Japanese. There are over
seventy words that qualify as "particles" in Japanese which behave similar to English
prepositions and conjunctions. A number of them are also placed at the end of
sentences to contribute the speaker's attitude toward what has been said. However, of
these seventy-plus particles, less than ten are case assigners. Among these are ga, o, ni
and de, some of which serve more than one purpose.
27
According to the Case Filter, "every phonetically realized noun phrase must be
assigned (abstract) Case."2 4 Japanese grammar is such that all nouns are followed by
particles, called postpositions, which express functional relations and assign case.2 5 For
example, consider the following sentence.
Hanako ga sakana o tabeta.
(name) GA fish O ate
Hanako ate fish.
Hanako and sakana (fish) are both nouns and recalling the Theta Theory from section 3,
their existence is licensed by the verb. Both nouns are followed by particles which fulfill
the Case Filter by assigning case to these nouns. The particle ga assigns nominative
case while the particle o assigns accusative case.
Similarly, when locative phrases are included within a sentence, they obligatorily
adhere to the Case Filter and general Japanese grammar rules. Within Japanese,
location is marked differently for different verbs,2 6 however, it is marked. To mark
location in sentences containing stative verbs such as iru (to be), and another group
called "attachment verbs" which contains words such as tsukeru (to attach),2 7 the
2 4 Chomsky, Knowledge, p.74.
2 5 Susumu Kuno, The Structure of the Japanese Language. Cambridge: The
MIT Press, 1973, p.4-5.
2 6 Susan Shinkawa, "English 'at,' 'in,' 'on,' and 'by,' Compared with Japanese
de, ni, and o," English and Japanese in Contrast. New York: Regents, 1979, p. 160.
2 7 Harvey Taylor, Case in Japanese. South Orange: Seton Hall University Press,
1971, p.34.
28
particle ni can be used but the particle de cannot. Conversely, sentences with active
verbs use the particle de and not the particle ni to mark location. In addition, location is
marked in sentences containing movement verbs by the particle o. Thus three different
particles--de, ni and o, are used to mark location and thereby assign case in Japanese
sentences. Below are examples containing locative phrases from each type of class
assignment.
Hon ga teburu no ue ni aru.
book GA table GEN top LOC to be
The book is on the table.
Yumi ga butai no mae de odotte iru.
(name) GA stage GEN front LOC dancing
Yumi is dancing in front of the stage.
Fune ga hashi no shita o tooru.
boat GA bridge GEN bottom LOC to pass
The boat passes under the bridge.
As demonstrated above, locative phrases behave similarly to noun phrases in
terms o f case assignment. This exercise supports the general consensus among Japanese
language scholars regarding the classification of locative phrases as nominals.
6. BINDING OBSERVATIONS
Using the conclusions from the previous sections of this thesis, this section will
discuss the expected behavior of locative phrases in terms of a theory called c-command
theory established by Reinhart (1983). This theory is based on structural relationships
and attempts to predict the behavior of certain sentence constituents in relation to
(stative verb + ni)
(active verb + de)
(movement + o)
29
others. C-command theory claims that if the second noun phrase in a sentence is c-
commanded by the first noun phrase then coreference between these two noun phrases
is impossible. Conversely, if the second noun phrase is not c-commanded by the first
noun phrase, then coreference should be possible. This section will investigate whether
or not locative phrases in Chinese and Japanese adhere to this theory.
6.1. BINDING AND C-COMMAND IN ENGLISH
The central idea in binding theory is that constituents in a sentence have certain
relationships with each other. Before embarking upon a discussion using specific terms,
this section will explain them in detail.
A domain is a set of nodes within a tree diagram that are all structurally related
to one element. The domain of an element is determined by the least maximal projection
that contains it.2 8 In the tree diagram below, in order to determine the domain of A, it is
necessary to find the least maximal projection, in other words, the most immediately
dominating branching node, which in this case is a. The domain of A contains all
elements excluding A that fall below a. Here, the domain of A is B.
a
B A
2 8 Noam Chomsky, Knowledge, p. 162.
30
C-command theory is stated as "Node A c(onstituent)-commands node B iff [if
and only if] the branching node alphai most immediately dominating A either
dominates B or is immediately dominated by a node alphas which dominates B, and
alpha2 is of the same category type as alpha\."2 9 In the tree diagram above, A not only
contains B in its domain, but it also c-commands B. Reinhart collapses the relationship
between domain and c-command as "the domain of node A consists of all and only the
nodes c-commanded by A."3 0 Thus, the domain is a crucial structural relationship for
c-command theory.
Binding theory uses the notion of c-command to investigate whether
constituents can designate the same referent. It "is concerned with connections among
noun phrases that have to do with such semantic properties as dependence of reference,
including the connection between a pronoun and its antecedent."3 1 As Chomsky notes,
binding does not only apply to anaphors and pronominals and their antecedents,
however it is the most common application and this section will be based on that
example. The three relevant word-classes are anaphors, pronominals and referring
2 9 T. Reinhart, Anaphora and Semantic Interpretation. London: Croom Helm, 1983,
p.23. Reinhart's c-command theory is an evolution of Langacker's 1966 command
theory which restricted linear ordering of constituents by using the notion "precede" and
stipulated that in order to determine a domain, the most immediately dominating S
(sentence) node must be located.
3 0 Reinhart, p. 19.
3 1 Chomsky, 1988, p.52 qtd. in Vivian Cook, p.43.
31
expressions (hereafter, r-expressions). Anaphors consist of reflexives, such as himself or
herself, and reciprocals, such as each other. Pronominals consist of pronouns such as
he, she and them. Those words or phrases that quality as r-expressions are nouns that
refer to an entity in the real world such as John or the dog.
The definition of binding is stated as "alpha binds beta if alpha c-commands and
is co-indexed with beta."1 2 ’ Coindexation refers to the linguistic convention assigning
two constituents the same "index" to show that they refer to the same entity. For
example, take the sentence “I| love myselfi.” “I” and “myself’ are coindexed, meaning
they refer to the same person. In terms of binding theory stated above, “I,” the alpha,
binds “myself,” the beta, because it both c-commands and is coindexed with beta.
The final component of binding theory as is relevant to the current discussion is
a group of conditions or principles which concisely state the required situations under
which binding can occur.
A: An anaphor is bound in a local domain.
B: A pronominal is free in a local domain.
C: A referring expression is free.3 3
Conditions A and C are demonstrated in the sentence “Jane wanted (the girl] to help
herselfi).” 3 4 The anaphor “herself’ is bound by “the girl” within the local domain inside
3 2 Chomsky, Knowledge, p. 164.
” Chomsky, Knowledge, p. 166.
1 J This example was taken from Cook, p.46.
32
the parentheses while “Jane,” the r-expression, is free— not bound. Conditions A and B
are demonstrated in the sentence “Jane said (shei shot herselfi).” Once again the
anaphor “herself’ is bound by “she.” However, in this sentence, “she” is not bound
within a local domain and therefore it can refer to Jane or someone else not specifically
named in the sentence.
Having briefly described binding theory and c-command, it is time to turn to a
discussion of coreference of definite noun phrases. The core sentences that will be used
as illustrations are as follows.
Shei loves herselfi.
*Shef loves Maryi.
Hisi father loves Johnj.
Johni loves hisi father.
*/? Het loves John^s father.
The first sentence is an example of obligatory coreference. “Herself’ and “she”
must refer to the same person according to Condition A. The second sentence,
however, is slightly different. First, this sentence does not contain an anaphor and
therefore Condition A is not applicable. Second, it is impossible to coindex “she” and
“Mary” because “she,” a pronoun, c-commands “Mary,” a non-pronoun. It was to
explain sentences such as this that Langacker (1966) developed his theory in terms of
precede and command. Reinhart restructured it as follows. “A pronoun cannot be
interpreted as coreferential with a non-pronoun in its domain.”3 5
1 5 Reinhart, p. 16.
The next three sentences will be evaluated as a group. Coreference within the
sentence “Johni loves hisi father.” is perfectly acceptable because the pronoun is c-
commanded by the non-pronoun and also is in the domain of the non-pronoun. The
sentence “Hisi father loves Johni.” is acceptable for two reasons. First, since “his
father” constitutes a noun phrase, the word “his” does not c-command "John" thereby
allowing coreference. The second reason is related in that since the pronoun “his” does
not c-command the non-pronoun “John,” the non-pronoun is not in the domain of the
pronoun. This adheres to Reinhart’s rule of coreference given above.
The third sentence, “Hei loves Johnt’s father.” is deemed questionable in terms
of acceptability. Some native speakers judge it to be completely unacceptable while
others judge it to be marginally acceptable. (Hence the notation “*/?.”) According to c-
command theory and pronoun/non-pronoun coreference theory, this sentence should be
unacceptable. The pronoun “he” c-commands and therefore has within its domain the
non-pronoun “John.” These conditions should theoretically block coreference, and the
unacceptable evaluation of this sentence is corroborated.
6.2 BINDING AND C-COMMAND IN CHINESE
Binding and c-command in Chinese adhere to all of the same rules and
conditions as English with the exception of the definition of the term domain. By
34
defining the syntactic domain differently, that is, in terms of cyclic c-command, scholars
are able to account for the fact that sentences with structures similar to “Hisj father
loves Johni.” are not acceptable.
This section will discuss cyclic c-command in Chinese. It will also investigate
whether'the classification of locatives as nouns or prepositions affects the outcome of
the application of the cyclic c-command theory.
Cyclic c-command is defined as the following. “A cyclic c-commands B iff the
cyclic node that immediately dominates A c-commands B.”1 6 Cyclic nodes are noun
phrase and sentence nodes. The following tree diagram illustrates how the syntactic
domain differs when defined in terms of c-command and when defined in terms of cyclic
c-command.
S
NP VP
A ... B
When defined in terms of c-command, the domain of A includes everything that falls
below the NP node; B is not in the domain of A and therefore coreference between A
and B should be acceptable. However, when defined in terms of cyclic c-command, the
domain of A is found by identifying the cyclic node immediately dominating A, which in
J * Kitagawa, Yoshihisa, Subjects in Japanese and English, doctoral
dissertation, University of Massachusetts, 1986.
35
this case is the NP node. The domain of A includes everything that the noun phrase
node c-commands; B is in the domain of A and therefore coreference between A and B
should be unacceptable.
The theory of cyclic c-command helps explain why sentences in Chinese with
the same syntactic stiucture as “Hist students love Johni.” are judged unacceptable or
questionable by native speakers. The generalization describing this phenomena can be
simplified by stating that a noun phrase in the possessor position can still c-command the
object within its syntactic domain, defined in terms of cyclic c-command. As illustration,
take the following sentence.
*/? Tcii de mama xihuan Zhangsan/.
he DE mother likes (name)
*/? Hisi mother likes Zhangsam.
S
NP VP
ta/ de mama xihuan Zhangsan/
In this sentence, ta does not c-command Zhangsan, but it does cyclic c-command
Zhangsan, explaining the unacceptable/questionable judgment.
Additionally, prepositions do not block the c-command relationship in Chinese
as is illustrated in the unacceptable results shown below.
36
* Wo gen tai zhao daole Zhangsan/ de shu.
I with him found (name) GEN book
* With himi I found Johni’s book.
PP
... gen ta{ zhao daole Zhangsan/ de shu
In the example, the coverb gen does not block the c-command relationship between the
two coindexed nouns. If it did, then coreference in this sentence would be acceptable to
native speakers. However, it is not, demonstrating that Zhangsan must be in the
domain of ta (he). This phenomena is particularly interesting in relation to the current
discussion.
The first part of this thesis sought to determine the classification of location
words and concluded that they pattern similar to nouns. This second part uses the
notion of c-command to test whether these location words, as nominals, behave as
expected. However, in light of the fact that Chinese grammar employs cyclic c-
command and that prepositions do not block the c-command relationship, the
classification of location words as nouns or prepositions is found not to be relevant in
this experiment.
Below are two tree diagrams with similar syntactic structures.
37
(A) S (B) S
NP VP PP VP
/ \ A . / \ / \
N,(P) N V N2 (P) N,(P) P V N2 (P)
The difference between tree diagram A and tree diagram B is the classification of one
node (denoted in bold face). This node is where a location word would be inserted. In
diagram A, the location word is assumed to be a noun. According to cyclic c-
command, Nj cyclic c-commands N2 thereby blocking coreference. Ni and N2 cannot
refer to the same entity. Diagram B assumes the location word to be a preposition.
However, since prepositions do not block c-command relationships in Chinese, Ni c-
commands N2 once again blocking coreference. Ni and N2 cannot refer to the same
entity; the result is the same.
Therefore, as demonstrated above a nominal classification or a prepositional
classification does not affect the syntactic relationship between nodes in terms of c-
command behavior. Both classifications create the situation where N2 is within the
domain o f N]. As expected, there is no contrast as far as this test is concerned.
6.3 BINDING AND C-COMMAND IN JAPANESE
The rules for binding and c-command in Japanese are slightly different from
those in Chinese or English. Japanese employs something similar to the cyclic c-
38
command phenomena of Chinese, however, it only applies to postpositions. Certain
postpositions in Japanese do not block the c-command relationship and therefore allow
coreference between coindexed nominals. The key difference between Japanese and
Chinese is that in Japanese there is no theory providing for noun phrases to behave
similar to the adpositional phrases which foster c-command relationships. Noun phrases
block c-command relationships. It is for precisely this reason that I expect the syntactic
classification of location words as nouns or prepositions to be of the utmost relevance in
terms of c-command theory.
For convenience, the two similar diagrams from the previous section are
repeated.
(A) S (B) S
Ni(P) N V N:(P) N|(P) P V N2 (P)
Once again, the node denoted in boldface print is where the location word would be
inserted. If the location word is assumed to be a noun, as in diagram A, N i cannot c-
command N2. Coreference in diagram A should be permitted. If the location word is
assumed to be a postposition, as in diagram B, Ni does c-command N2 and coreference
in diagram B should be prohibited.
39
In order to test this hypothesis, I gathered acceptability judgments from native
speaker informants based upon a group of sentences in which two words were
coreferenced. The sentences (see Appendix D) were created based on two basic
structures.
2)
John
1)
Johni
These two structures were chosen because within the field of Japanese linguistics,
scholars studying coreference have found this distinction to be enlightening.3 7 In order
to apply it to the current discussion, additional sentences with similar structures were
created substituting kare with sore or soko. In creating these sentences it was
important to avoid sentences in which interpretation of the sentences could be affected
by other concerns, pragmatics being one. Pragmatics deals with the inherent meaning of
a particular word within a sentence or of the situation or context that the sentence
creates.
Based upon the general linguistic assumption that sentence type 1 is
unacceptable and sentence type 2 is acceptable, I created the following chart.
TYPE KARE SORE(KO)
1
*
2 ok
1 1 See Hoji (1985) and Saito (1985) among others.
40
Statistically there could be four results: i) ok, ok, ii) ok, *, iii) *, ok, and iv):|i, *
Scenarios i) and ii) could be excluded, because based on the general linguistic
assumption for kare, it is illogical to expect the outcome of sore/soko to be diametrically
opposed. Condition D has been proven to function consistently across languages, a
conclusion that discourages the idea that Japanese would behave any differently.
Therefore, as far as theory was concerned, there could be two possible outcomes.
A)
TYPE KARE SORE(KO)
1
* *
2 ok
*
B)
TYPE KARE SORE(KO)
1
* *
2 ok ok
Of the two, neither was particularly ideal, however, outcome B had more potential
than outcome A in terms of corroborating the first part of this thesis.
The problem in scenario A would arise in explaining why type 2 was acceptable
for kare but not acceptable for soko/sore. This would suggest that the workings of
Condition D must be more closely examined. Condition D states “a less referential
expression may not bind a more referential one.”3 8 An example hierarchy of
referentiality is names>epithets>pronouns, with names being the most referential
) S Lasnik, Howard, “On the Necessity of Binding Conditions,” Essays on
Anaphora. Dordrecht: Kluwer Academic Publishers, 1989, p. 12-13.
41
member of the group. The definition of Condition D might need revision in order to
encompass just such an outcome.
Despite my expectations based on standard judgments in scholarly literature, the
results of my experiment differed greatly. According to my twenty-five informants,
none of the sentence types incorporated within my chart were completely acceptable.
Instead, the results were as follows.3 9
TYPE KARE SORE(KO)
1
* *
2
*? / * *? / *
Based upon these surface observations, very interesting conclusions can be
drawn. Since a random pool of native speaker informants rejected coreference
possibilities in both sentence types for both sets of words, it is necessary to compare the
theory-driven results with the actual results. This difference suggests that the theory
controlling native speaker judgments is different from the prevailing view espoused by
prominent scholars of how noun-embedding affects syntactic domains. In fact, it
suggests that nouns do not seem to block c-command relationships. If nouns did block
c-command relationships, coreference would be permitted and the two bottom
quadrants of the chart would be filled with lok’ judgments. This was not the case.
” Of the informants, some had difficulty identifying to which end of the
spectrum questionable judgments belong-closer to the acceptable side or closer to
the unacceptable side. This turned out not to have a great bearing on the evaluation
of the results however, as the percentage within the sample set was small— less than
0.05.
42
If this conclusion is combined with the established fact that in Japanese
postpositions do not block c-command relationships either, then this experiment does
not show conclusively whether location words behave as postpositions or as nouns in
terms of c-command theory. The syntactic classification, postposition or noun, has no
relevance in this case because both classifications behave similarly— they both allow c-
command and thereby prohibit coreference.
7. FINAL CONCLUSIONS
The similarity between the Japanese case and the Chinese case is striking.
Neither conclusively indicate nominal or adpositional behavior in c-command. The
results show that Chinese native speaker intuition and Japanese native speaker intuition
are on a par with each other; the judgments of acceptability are similar. These results
are very different from standard results within the field of linguistics. The possible
implications of these results are fourfold.
One possibility for the discrepancy between scholarly judgments and native
speaker judgments is bias on the part of theoreticians. It is quite possible that linguists,
after working with a theory for an extended period of time, become biased by that
theory and lose the ability to be objective when judging sentences. Instead of beginning
with intuition, they begin with theory, and proceed to note what “should” be acceptable.
Another possibility closely related to the first is a methodological problem of
recording acceptability. An ‘ok’ acceptability judgment from one informant may not
43
equal the value of an ‘ok ’ acceptability judgment from another informant. Linguists, for
example, only rule out a sentence, with the notation **,’ when there is absolutely no
conceivable way that the sentence can be construed as acceptable. A non-linguist, on
the other hand, may call a sentence unacceptable and accordingly mark it **’ if they
would not use it. They may also indicate that a sentence is unacceptable if coreference,
for example, is not the. first, most accessible reading. It is difficult a) to force informants
to adhere to the same value scale and b) to know what they are thinking as they mark
their acceptability judgments.
The third possible implication favors the theory-driven judgments of theorists
over native speaker judgments. It is possible, although unlikely, that the judgments of
one group of native speakers-either the Chinese or the Japanese are not accurate. This
possibility assumes that other elements might be at work, forcing incorrect judgments
from native speakers. I believe this scenario is unlikely because theoretically Universal
Grammar creates a theory which is proven or disproved by empirical evidence,
specifically native speaker judgments. In this scenario, the theory refutes the judgment
and that is inconsistent with the purpose of Universal Grammar. According to
Chomsky, there are two possible paths to take in this type o f situation.
We may disregard the information as fallacious (or as an intuition about
something other than grammatical form [pragmatics, for example]), or we may
reconstruct the theory. Between these two poles of reliance on the results of a
given theory and reliance on intuition, there are many possible positions and
44
attitudes...In the absence of clear criteria of adequacy and relevance, behavioral
or otherwise, for theories, it is difficult to determine a correct position.4 0
One might speculate that these results favor native speaker judgment. Assuming
that the native speaker judgments are accurate, there is a possibility that, contrary to the
generally held belief, while there is a difference between Chinese and Japanese, it is not
rigid. In fact, these results might suggest that the distance between the two languages is
decreasing and the two languages are approaching alignment. Conversely, there is also
the possibility that the distance is increasing and these results betray lasting vestiges of a
former alignment. The point here is that these two languages may be in the midst of a
synchronic evolution and this movement is evidenced by the results of this experiment.
Unfortunately, the study undertaken here is not able to choose one possible
implication over another. The only way to come closer to achieving this is to conduct
another, more extensive study in which location words in relation to referential
dependency is again the subject. In the future, a study under more controlled conditions
in which the noun phrase/locative phrase is farther embedded, could be useful. In
Chinese, for example, the deeper the phrase is embedded, the more acceptable the
coreference becomes. A study in which similar deep embedding is created in Japanese
would most likely be more illuminating in determining the behavior of location words as
nominals or adpositions within c-command theory.
Chomsky, Noam, The Logical Structure of Linguistics Theory. Chicago:
University of Chicago Press, 1975a, p. 103.
45
Appendix A : Location words
Japanese Chinese English noun
gloss
English preposition
gloss
shita xia (bian) bottom below
ue shang (bian) top above
ushiro hou (bian) rear, back behind, in back of
mae qian (bian) front before, in front of
Appendix B: Prepositions/Co-verbs
Chinese English gloss Chinese English gloss
gen with chule except, beside
cong from li apart from
he with zai at
zhiyu concerning wei for
xiang like xiang facing
tong with wang facing
guanyu concerning dao to (a place)
chao facing bi compare
gei for, by dui
4 1
to
il This is only a small sampling of the most common co-verbs. An exhaustive list
can be found in Li and Thompson, p.368. Chao also enumerates the co-verbs and
includes their alternative syntactic categorizations along with their alternative definitions.
46
Appendix C : Brief summary of constituent relations through use of a labeled tree
diagram
Every point on the diagram is a node.
All capital letters are terminal nodes.
All small letters are branching nodes.
B, C, and c are sister nodes.
A and b are sister nodes.
B is a domain head. It heads the domain [C, c, D, E],
E is a domain head. It heads the domain [D].
A c-commands b, B, C, c, D, and E.
D c-commands E.
C c-commands B, c, D, and E.
A dominates A and b.
b dominates B, C and c.
Example sentence: I gave it to Mary.
a S
NP VP
N
(I)
B C c V NP PP
(gave)
D E N P NP
(it) (to) N
(Mary)
47
Appendix D: Acceptability-native speaker judgments
John wa sono CDi no ue ni sorej o tsukutta hito no shashin o kasaneta.
(name) WA that CD GEN to LOC that O made person GEN picture O displayed
John displayed on top of that CDj a picture of the person who made itj.
John wa sorei no ue ni sono CD; o tsukutta hito no shashin o kasaneta.
(name) WA that GEN top LOC that CD O made person GEN picture O displayed
John displayed on top of itj a picture of the person who made that CDi.
John wa sono CDj ni sorej o tsukutta hito no shashin o kasaneta.
(name) WA that CD LOC that O made person GEN picture O displayed
John displayed on that CDi a picture of the person who made it,.
John wa sore* ni sono CDi o tsukutta hito no shashin o kasaneta.
(name) WA that LOC that CD O made person GEN picture O displayed
John displayed on itj a picture of the person who made that CDj.
John wa sono dansu hoorUj no naka e soko; ni ikitagatte iru hito o tsurete itta.
(name) WA that dance hall GEN inside E there NI wanting to go person O took
John took inside that dance halls the person who wanted to go there;.
John wa sokoi no naka e sono dansu hoorui ni ikitagatte iru hito o tsurete itta.
(name) WA there GEN inside E that dance hall NI wanting to go person O took
John took inside there; the person who wanted to go to that dance hall,.
John wa sono dansu hoorUj e sokos ni ikitagatte iru hito o tsurete itta.
(name) WA that dance hall E there Ni wanting to go person O took
John took to that dance halls the person who wanted to go therej.
John wa soko; e sono dansu hoorui ni ikitagatte iru hito o tsurete itta.
(name) WA there E that dance hall NI wanting to go person O took
John took therej the person who wanted to go to that dance hallj.
SokOj no mae ga John ga tatte iru tokoroj kara sasu kage de kakusarete ita.
there GEN front GA (name) standing place from fall shadow DE covered
In front of therei is covered by the shadow falling from the place; where John’s
standing.
48
John ga tatte iru tokoroj no mae ga soko! kara sasu kage de kakusarete ita.
(name) GA standing place GEN front GA there from fall shadow DE covered
The place; in front of where John is standing is covered by the shadow falling from
there;.
Soko; no mae ga sono biru; kara dete kita hito de konde iru.
there GEN front GA that building from emerged people DE crowded
The front of there; is crowded with people who emerged from that building;.
Sono biru; no mae ga soko; kara dete kita hito de konde iru.
that building GEN front GA there from emerged people DE crowded
The front of that building; is crowded with people who emerged from there;.
Soko; no shacho wa A to iu kaisha; de tsutomete iru hito ni shokai shimashita.
There GEN chief WA “A” named company DE working people NI introduced
The chief from there; introduced (someone) to the people who work at “A”
company;.
A to iu kaisha; no shacho wa soko; de tsutomete iru hito ni shokai shimashita.
“A" named company GEN chief WA there DE working people NI introduced
“A” companyj’s chief introduced (someone) to the people working there;.
John; ga kare; no atarashii kuruma no shashin o misete kuremashita.
(name) GA him GEN new car GEN picture O showed (us)
John; showed us a picture of his; new car.
Kare; no sensei ga John; o shikatta.
Him GEN teacher GA (name) o scolded
His; teacher scolded John;.
John; no sensei ga kare; o shikatta.
(name) GEN teacher GA him O scolded
John’s teacher scolded him.
Kare; ga John; no atarashii kuruma no shashin o misete kuremashita.
him GA (name) GEN new car GEN picture O showed (us)
He; showed us a picture of John;’s new car.
Kare; no tomodachi ga John; no kuruma o aratta koto.
Him GEN friend GA (name) GEN car O washed
His; friend washed Johnj’s car.
49
Karej ga Johnj no tomodachi o suisen shita koto.
Him GA (name) GEN friend O recommended
Hei recommended Johnj's friend.
Karej no sensei ga Johnj o suisen shita koto.
He GEN teacher GA (name) O recommended
His; teacher recommended Johnj.
Karej ga watashi ni Yamadaj-san no kinkyoo ni tsuite oshiete kureta.
He GA I NI Mr. Yamada GEN recent conditions concerning told (me)
He* told me what has been going on with Mr. Yamadaj.
Kare; ga Johnj no kuruma o aratta koto.
He GA (name) GEN car O washed
He; washed Johnj’s car.
Karej no tomodachi ga Johnj o suisen shita koto.
He GEN friend GA (name) O recommended
HiSj friend recommended Johnj.
SOURCES
Beijing Language Institute, Practical Chinese Reader. Elementary Course Books 1 and
2, Beijing: The Commercial Press, 1981.
Chao, Yuen-Ren, A Grammar of Spoken Chinese. Berkeley: University of California
Press, 1968.
Chino, Naoko, All About Particles. Tokyo: Kodansha International, 1991.
Chomsky, Noam, The Logical Structure of Linguistics Theory. Chicago: University of
Chicago Press, 1975a.
Chomsky, Noam, Reflections on Language. London: Temple Smith, 1976.
Chomsky, Noam, Knowledge of Language: Its Nature. Origin, and Use. New York:
Praeger, 1986.
Chomsky, Noam, Lectures on Government and Binding: The Pisa Lectures. New York:
Mouton de Gruyter, 1993.
Cook, Vivian C., Chomsky's Universal Grammar. Oxford, UK: Basil Blackwell Ltd.,
1988.
Emonds, Joseph E., "The Invisible Category Principle," Linguistic Inquiry 18: 613-632.
Hagege, Claude, Le Probleme Linguistique des Prepositions et la Solution Chinoise.
Paris: Editions Peeters, 1975.
Hoji, Hajime, Logical Form Constraints and Configurational Structures in Japanese.
Doctoral dissertation, University of Washington, 1985.
Jorden, Eleanor Harz with Mari Noda, Japanese : The Spoken Language. New Haven:
Yale University Press, 1990. Vols. 1-3.
Kitagawa, Yoshihisa, Subjects in Japanese and English, doctoral dissertation, University
of Massachusetts, 1986.
Kuno, Susumu, The Structure of the Japanese Language, Cambridge: The MIT Press,
1973.
Larson, Richard K., "Bare-NP Adverbs," Linguistic Inquiry 16:595-621.
51
Lasnik, Howard, “Remarks on Coreference,: Linguistic Analysis 2:1-22.
Levinson, S.C. "Pragmatics and the Grammar of Anaphora," Journal of Linguistics.
23.2:400-410.
Li, Charles N., and Sandra A. Thompson, Mandarin Chinese: A Functional Reference
Grammar, Berkeley: University of California Press, 1981.
Li, Yen-hui, Order and Constituency in Mandarin Chinese. Boston: Kluwer Academic
Publishers, 1990.
Martin, Samuel, A Reference Grammar of Japanese. Vermont: Charles Tuttle Co.,
1987.
Miyagawa, Shigeru, "Structure and Case Marking in Japanese," Structure and
Semantics vol.22., San Diego: Academic Press, Inc., 1989.
McCawley, James D., "Notes on Li and Thompson, Mandarin Chinese: A Functional
Reference Grammar," Journal of the Chinese Language Teachers Association. 14: 19-
42.
O'Grady, William, Contemporary Linguistics. New York: St. Martin's Press, 1989.
Otsu, Yukio and Ann Fanner, MIT Working Papers in Linguistics Volume 2
Theoretical Issues in Japanese Linguistics. 1980. '
Reinhart, T., Anaphora and Semantic Interpretation. London: Croom Helm 1983.
Saito, Mamoru, Some Asymmetries in Japanese and Their Theoretical Consequences,
Doctoral dissertation, MIT, 1985.
Shinkawa, Susan H., "English 'at,' 'in,' 'on,' and 'by,' Compared with Japanese de, ni,
and o " English and Japanese in Contrast. New York: Regents, 1979.
Taylor, Harvey M., Case in Japanese. South Orange: Seton Hall University Press, 1971.
Whitman, John, “Configurationality Parameters,” Issues in Japanese Linguistics, The
Netherlands: Foris Publications, 1987.
52
INFORMATION TO USERS
This manuscript has been reproduced from the microfilm master. UM I
films the text directly from the original or copy submitted. Thus, some
thesis and dissertation copies are in typewriter face, while others may
be from any type of computer printer.
The quality of this reproduction is dependent upon the quality o f the
copy submitted. Broken or indistinct print, colored or poor quality
illustrations and photographs, print bleed through, substandard margins,
and improper alignment can adversely affect reproduction.
In the unlikely event that the author did not send UMI a complete
manuscript and there are missing pages, these will be noted. Also, if
unauthorized copyright material had to be removed, a note will indicate
the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by
sectioning the original, beginning at the upper left-hand corner and
continuing from left to right in equal sections with small overlaps. Each
original is also photographed in one exposure and is included in
reduced form at the back of the book.
Photographs included in the original manuscript have been reproduced
xerographicaliy in this copy. Higher quality 6" x 9" black and white
photographic prints are available for any photographs or illustrations
appearing in this copy for an additional charge. Contact UMI directly
to order.
A Befl & Howell Information Company
300 North Zeeb Road. Ann Arbor. M l 48106-1346 USA
313/761-4700 800/521-0600
□M X Number: 1376447
UMI Microform 1376447
Copyright 1995, by UMI Company. All rights reserved.
This microform edition is protected against unauthorized
copying under Title 17, United States Code.
UMI
300 North Zeeb Road
Ann Arbor, MI 48103

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

The Japanese demonstratives ko, so and a

PDF

Grammaticalization and the development of functional categories in Chinese

PDF

Japanese pronunciation by English speaking students and English-Korean speaking students: contrastive and error analysis

PDF

Ellipsis constructions in Chinese

PDF

Form and meaning: Negation and question in Chinese

PDF

The syntax of cleft constructions in Japanese: A base-generation analysis

PDF

Sluicing and stripping in Japanese and some implications

PDF

A comparative study of focus constructions

PDF

Dominant language influence in acquisition and attrition of the Chinese reflexive ziji By Chinese-English bilinguals

PDF

Direct objects in Persian

PDF

Perceptions and realities: women, equality and discriminatory barriers in Japanese society and workplace

PDF

Intonation Of Single-Word Utterances In Japanese: An Acoustic Phonetic Study

PDF

Japanese Accent: An Analysis Based On Acoustic - Phonetic Data

PDF

An investigation of the pseudohomophone effect: where and when it occurs

PDF

"It was in the air": Adolph Gottlieb, the pictographs, New York, and the Zeitgeist of the 1940s

PDF

Binding and scope dependencies with 'floating quantifiers' in Japanese

PDF

Reduplication and distributivity in Kannada

PDF

Hand function in older adults: the relationship between performance on the Jebsen Test and ADL status

PDF

Functional categories: the syntax of DP and DegP

PDF

Dye laser characterization of two films of baceriorhodopsin

Asset Metadata

Creator Bostic, Ebony Leigh (author)

Core Title Locatives in Chinese and Japanese: distribution, case assignment and c-command

School Graduate School

Degree Master of Arts

Degree Program East Asian Languages and Cultures

Degree Conferral Date 1994-12

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag language, linguistics,OAI-PMH Harvest

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Nosco, Peter (committee chair), Han, Meiko (committee member), Hoji, Hajime (committee member), Li, Audrey (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c18-910

Unique identifier UC11356854

Identifier 1376447.pdf (filename),usctheses-c18-910 (legacy record id)

Legacy Identifier 1376447-0.pdf

Dmrecord 910

Document Type Thesis

Rights Bostic, Ebony Leigh

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA