An Introduction To English Syntax
An Introduction To English Syntax
An Introduction To English Syntax
Compiled by
Nguyen Van Huy
Than Trong Lien Nhan
HCFL
1
CHAPTER 1
INTRODUCTION TO SYNTAX
I. Introduction
SYNTAX is the central component of human language. Language has often been
characterized as the systematic correlation between certain types of oral/graphic forms for
spoken/written language; and, for signed language, they are manual.
It is not the case that every possible meaning that can be expressed is correlated with a
unique, un-analyzable form. Rather, each language has a stock of meaning-bearing elements
and different ways of combining them to express different meanings, and these ways of
combining them are themselves meaningful. The two English sentences Chris gave the
notebook to Dana and Dana gave the notebook to Chris contain exactly the same meaning-
bearing elements, i.e. words, but they have different meanings because the words are
combined differently in them. These different combinations fall into the realm of syntax; the
two sentences differ not in terms of the words in them but rather in terms of their syntax.
Syntax can thus be given the following characterization, taken from Matthews (1982:1):
The term ‘syntax’ is from the Ancient Greek syntaxis, a verbal noun which literally
means ‘arrangement’ or ‘setting out together’. Traditionally, it refers to the branch of
grammar dealing with the ways in which words, with or without appropriate inflections,
are arranged to show connections of meaning within the sentence.
First and foremost, syntax deals with how sentences are constructed, and users of human
languages employ a striking variety of possible arrangements of the elements in sentences.
One of the most obvious yet important ways in which languages differ is the order of the main
elements in a sentence. In English, for example, the subject comes before the verb and the
direct object follows the verb. In Lakhota (a Siouan language of North America), on the
other hand, the subject and direct object both precede the verb, while in Toba Batak (an
Austronesian language of Indonesia; (Schachter 1984b), they both follow the verb.
In Lakhota, the subject comes first followed by the direct object, whereas in Toba Batak the
subject comes last in the sentence, with the direct object following the verb and preceding the
subject. The basic word order in Toba Batak is thus the opposite of that in Lakhota. There
are also languages in which the order of words is normally irrelevant to the interpretation of
which element is subject and which is object. This is the case, for example, in Russian
sentences.
In Russian the order of the words is not the key to their interpretation, as it is in the sentences
from the other languages. Rather, it is the form of the words that is crucial.
The changes in the form of the words to indicate their function in the sentence are what
Matthews referred to as ‘inflections’, and the study of the formation of words and how they
may change their form is called morphology. The relationship between syntax and
morphology is important: something which may be expressed syntactically in some languages
may be expressed morphologically in others. Which element is subject and which is object is
signaled syntactically in theses languages, while it is expressed morphologically in the others.
2
1. Definition
SYNTAX is the study of how words are combined to form sentences in a language. Thus,
syntax concerns the system of rules and categories that underlies sentence formation.
A central part of the description of what speakers do is characterizing the grammatical (or
well-formed) sentences of a language and distinguishing them from ungrammatical or (ill-
formed) sentences. Grammatical sentences are those that are in accord with the rules and
principles of the syntax of a particular language, while ungrammatical sentences violate one
or more syntactic rules or principles. For example, The teacher is reading a book is a
grammatical sentence of English, while Teacher the book a reading is would not be.
Ungrammatical sentences are marked with an asterisk, hence *Teacher the book a reading is.
This sentence is ungrammatical because it violates some of the word order rules for English,
that is (i) basic word order in English clauses is subject-verb-object, (ii) articles like the and a
precede the noun they modify, and (iii) auxiliary verbs like is precede the main verb, in this
case reading. It is important to note that these are English-specific syntactic rules.
Well-formed sentences are those that are in accord with the syntactic rules of the language;
this does not entail that they always make sense semantically. For example, the sentence The
book is reading the teacher is nonsensical in terms of its meaning, but it violates no syntactic
rules or principles of English; indeed, it has exactly the same syntactic structure as The
teacher is reading a book. Hence it is grammatical (well-formed), despite being semantically
odd.
1.2. Grammaticality
• Grammaticality is not based on what is taught in school but on the rules acquired or
constructed unconsciously as children. Much grammatical knowledge is ‘in place’ before we
learn to read.
The ability to make grammaticality judgments does not depend on having heard the sentence
before. You may never have heard or read Enormous crickets in pink socks were dancing at
the ball but your syntactic knowledge will tell you the sentence is grammatical.
Although the sentences do not make much sense, they are syntactically well formed. They
sound ‘funny’ but they differ in their 'funniness" from the following strings:
The grammaticality of this case is based on the ordering of words and morphemes of a
sentence.
• Grammatical sentences may be uninterpretable if they include nonsense strings, that is,
words with no agreed-on meaning, as shown by the first two lines of ‘Jabberwocky’ by Lewis
Carroll:
Such nonsense poetry is amusing because the sentences ‘obey' syntactic rules and sound like
good English. Ungrammatical strings of nonsense words are not entertaining:
• Grammaticality does not depend on the truth of sentences either - if it did, lying would be
impossible - nor on whether real objects are being discussed, nor on whether something is
possible or not.
iv. the fact that sentences with different structures can have the same meaning;
vi. speaker’s creative ability to produce and understand any of an infinite set of possible
sentences.
In the syntactic structure of sentences, two distinct yet interrelated aspects must be
distinguished. The first one has already been mentioned: the function of elements as subject
and direct object in a sentence. ‘Subject’ and ‘direct object’ have traditionally been referred to
as grammatical relations. Hence this kind of syntax will be referred to as relational
structure. It includes more than just grammatical relations like subject and direct object; it
also encompasses relationships like modifier-modified, e.g. tall building or walk slowly (tall,
slowly=modifier, building, walk=modified) and possessor-possessed, e.g. Pat’s car (Pat’s =
possessor, car = possessed). The second aspect concerns the organization of the units which
constitute sentences. A sentence does not consist simply of a string of words; that is, in a
sentence like The teacher read a book in the library, it is not the case that each word is
equally related to the words adjacent to it in the string. There is no direct relationship
between read and a or between in and the; a is related to book, which it modifies, just as the
is related to library, which it modifies. A is related to read only through a book being the
direct object of read, and similarly, the is related to in only through the library being the
object of the preposition in. The words are organized into units which are then organized
into larger units. These units are called constituents, and the hierarchical organization of the
units in a sentence is called its constituent structure. This term will be used to refer to this
second aspect of syntactic structure. Consider the eight words in the sentence The teacher
read a book in the library. What units are these words organized into? Intuitively, it seems
clear that the article the or a goes with, or forms a unit with, the noun following it. Is there
any kind of evidence beyond a native speaker's intuitions that this is the case?
If the article forms a unit with the noun that follows it, we would expect that in an alternative
form of the same sentence the two would have to be found together and could not be split up.
Thus in the passive version of this sentence, A book was read by the teacher in the library,
the unit a book serves as subject, and the unit the teacher is the object of the preposition by.
The constituent composed of a noun and an article is called a noun phrase [NP]; as will be
shown later, NPs can be very complex. The preposition in and the NP following it also form
a constituent in this sentence (in the library); it is called a prepositional phrase [PP]. The
fact that the PP is a constituent can be seen by looking at another alternative form. In the
library the teacher read a book. Finally, the verb plus the NP following it form a unit as well,
as shown by a sentence like I expected to find someone reading the book, and reading the
book was a teacher. The constituent composed of a verb plus following NP is called a verb
phrase [VP]. As with NPs, VPs can be quite complex. In each of these alternative forms, a
combination of words from the original sentence which one might intuitively put together in a
single unit also occurs together as a unit, and this can be taken as evidence that they are in fact
constituents. Using square brackets to group the words in constituents together, the
constituent structure of The teacher read a book in the library may be represented as follows
(‘S’ stands for ‘sentence’):
5
[S [NP [N The [N teacher]] [VP [V read] [NP a [N book]] [PP [P in] [NP the [N library]] PP] VP] S]
Note the nesting of constituents within constituents in this sentence, e.g. the NP the library is
a constituent of the PP in the library which is a constituent of the VP read a book in the
library.
At the beginning of this section it was noted that the two aspects of syntactic structure,
relational structure and constituent structure, are ‘distinct yet interrelated’, and it is possible
now to see how this is the case. For example, a VP was described as being composed of a
verb and the following NP, but it could alternatively be characterized as involving the verb
and its direct object. Similarly, a PP is composed of a preposition and its object. NPs, on the
other hand, involve modifiers, and accordingly the relation between the and teacher could be
described as one of modifier-modified. Thus, these two aspects of syntactic structure are
always present in a sentence, and when one or the other is emphasized, the sentence is being
described from one of the two perspectives. It will be seen later that different grammatical
phenomena seem to be more easily analyzed from one perspective rather than the other.
In the discussion of the constituents of sentences, reference has been made to nouns and noun
phrases, verbs and verb phrases, and prepositions and prepositional phrases. Nouns, verbs
and prepositions are traditionally referred to as ‘parts of speech’ or ‘word classes’; in
contemporary linguistics they are termed lexical categories. The most important lexical
categories are noun, verb, adjective, adverb and prepositions and postpositions (being
subsumed adposition). In traditional grammar, lexical categories are given notional
definitions, i.e. they are characterized in terms of their semantic content. For example, noun
is defined as ‘the name of a person, place or thing’, verb is defined as an action word’, and
adjective is defined as ‘a word expressing a property or attribute’. In modem linguistics,
however, they are defined morpho-syntactically in terms of their grammatical properties.
Nouns may be classified in a number of ways. There is a fundamental contrast between nouns
that refer uniquely to particular entities or individuals and those that do not; the best example
of the first kind of noun is a proper name, e.g. Sam, Elizabeth, Paris or London, and nouns of
this type are referred to as proper nouns. Nouns which do not refer to unique individuals or
entities are called common nouns, e.g. dog, table, fish, car, pencil, water. One of the
important differences between proper and common nouns in a language like English is that
common nouns normally take an article, while proper nouns do not, e.g. The boy left versus
*The Sam left (cf.*Boy left versus Sam left). Common nouns may be divided into mass
nouns (or non-count nouns) and count nouns. Count nouns, as the name implies, denote
countable entities, e.g. seven chairs, six pencils, three dogs, many cars. Mass nouns, on the
other hand, are not readily countable in their primary senses, e.g. *two waters, *four butters,
*six snows. In order to make them countable, it is necessary to add what is sometimes called
a 'measure word', which delimits a specific amount of the substance, e.g. two
glasses/bottles/drops of water, four pats / sticks of butter, six shovelfuls of snow. Measure
words can be used with count nouns only when they are plural, e.g. *six boxes of pencil
versus six boxes of pencils, *two cups of peanut versus three jars of peanuts. Pronouns are
closely related to nouns, as they both function as NPs. Pronouns are traditionally
characterized as ‘substitutes’ for nouns or as ‘standing for’ nouns, e.g. John went to the store,
and he bought some milk, in which he substitutes or stands for John in the second clause.
This, however, is true only of third-person pronouns like he, she, it, or they; it is not true of
6
first-person pronouns like I or second-person pronouns like you. First- and second-person
pronouns refer to or index the speaker and addressee in a speech event and do not replace or
stand for a noun.
Verbs can likewise be categorized along a number of dimensions. One very important
dimension is whether a verb takes just a subject (an intransitive verb), or a subject and a
direct object (a transitive verb), or a subject, direct object and indirect object (a ditransitive
verb). This will be referred to as the ‘valence’ of the verb. Another dimension concerns the
kind of situation it represents. Some verbs represent static situations which do not involve
anyone actually doing anything, e.g. know as in Chris knows the answer, or see as in Pat sees
Dana over by the bookcase. Some symbolize actions, e.g. run as in Kim ran around the track,
or sing as in Leslie sang a beautiful aria. Others refer to a change of state, e.g. freeze as in
The water froze (the change in the state of the water is from liquid to solid), or dry as in The
clothes dried quickly (the change in the state of the clothes is from wet to dry). Some
represent complex situations involving an action plus a change of state, e.g. break as in Larry
broke the window with a rock (Larry does something with a rock [action] which causes the
window to break [change of state]). This classification of verbs is quite complex and is more
appropriately in the domain of semantics rather than syntax.
Some examples of adjectives in English include red, happy, tall, sick, interesting, beautiful,
and many others. Adjectives typically express properties of entities, e.g. a red apple, a tall
woman, a beautiful sunset. Some properties are inherent attributes of an entity; for example,
some apples are red because they are naturally so, whereas some barns are red because they
have been painted red, not because they are inherently red. Hence color is an inherent
property of apples but not of barns. Some languages signal this distinction overtly. In
Spanish, for example, the adjective feliz means ‘happy’, and whether it is an inherent or
permanent property of the person referred to is signaled by the verb it is used with, i.e. Maria
es feliz ‘Maria is happy (a happy person)’ versus Maria esta feliz ‘Maria is happy (now, at this
moment but not necessarily always)’. Spanish has two verbs meaning ‘be’, ser and estar, and
one of the differences between them is that ser plus adjective (es in this example) is used to
signify inherent or permanent attributes, while estar plus adjective (esta in this example)
serves to indicate non-permanent, transitory attributes.
English adverbs typically, but not always, end in -ly, e.g. quickly, happily, beautifully, rapidly
and carefully. Fast and friendly are exceptions; fast is an adverb without -ly (it can also be an
adjective), and friendly, despite the admonitions of road signs in Texas to ‘drive friendly’, is
an adjective, e.g. a friendly waiter. Adverbs modify verbs, adjectives and even other adverbs,
and they can be classified in terms of the nature of this modification; manner adverbs, for
example, indicate the manner in which something is done, e.g. The detective examined the
crime scene carefully, or The ballerina danced beautifully, while temporal adverbs, as the
name implies, express when something happened, e.g. Kim talked to Chris yesterday, or Dana
will see Pat tomorrow. Yesterday and tomorrow do not end in -ly and have the same form
when functioning as an adverb that they have when functioning as a noun, e.g. Yesterday was
a nice day, Tomorrow will be very special. The most common adverbial modifiers of
adjectives and adverbs are words like very, extremely, rather, e.g. a very tall tree, the
extremely clever student, rather quickly. This class of adverbs is referred to as degree
modifiers.
Prepositions are adpositions that occur before their object, while postpositions occur after
their object. English (and Spanish) have only prepositions, e.g. English in, on, under, to,
7
(Spanish en, a, con,) whereas Japanese and Korean have only postpositions. German has
both: in dem Haus ‘in the house’ (preposition in) versus dem Haus gegenilber ‘over across
from the house’ (postposition gegenilber).
There are a number of minor categories. The category of determiners includes articles like a
and the, and demonstratives like this and that. Determiners modify nouns in relation to their
referential properties. Articles indicate roughly whether the speaker believes her
interlocutor(s) can identify the referent of the NP or not; an indefinite article like a(n) signals
that the speaker does not assume the interlocutor(s) can identify the referent of the NP, while
a definite article like the indicates that the speaker does assume that the interlocutor(s) can
identify it. Demonstratives, on the other hand, refer to entities in terms of their spatial
proximity to the speaker; English this refers to an entity close to the speaker, while that refers
to one farther away. (Which book do you mean? This one here or that one over there? versus
*This one over there or that one here?) Many languages make a three-way distinction: close
to the speaker (English this, Spanish esta [FEM]), away from the speaker but not far (English
that, Spanish esa [FEM]), and farther away from the speaker (archaic English yon, Spanish
aquella [FEM]). These distinctions are also expressed by locative demonstratives, e.g.
English here, German hier, Spanish aqui versus English there, German da, Spanish ahi versus
English yonder, German dort, Spanish alii. Quantifiers, as the label implies, express
quantity-related concepts. English quantifiers include every, each, all, many, and few, as well
as the numerals one, two, three, etc., e.g. every boy, many books, the seven sisters.
Classifiers serve to classify the nouns they modify in terms of shape, material, function,
social status and other properties. They are found in many East and Southeast Asian and
Mayan languages, among others. They are similar in many respect to the measure words that
occur with English mass nouns, but they occur with all nouns regardless of the count-mass
distinctions. Conjunctions, like and, but and or, serve to link the elements in a conjoined
expression. There are conjoined NPs, e.g. a boy and his dog, conjoined verbs, e.g. Leslie
danced and sang, and conjoined adjectives, e.g. Lisa is tall and slender. All major lexical
categories can be linked by conjunctions to form conjoined expressions. Complementizers
mark the dependent clause is a complex sentence, e.g. English that as in Sally knows that Bill
ate the last piece of pizza. The final category is particles, which is a classification often
given to elements which do not fall into any of the other categories. Many particles have
primarily discourse functions, e.g. English indeed, German doch, Spanish entonces.
There is an important opposition that divides lexical categories into two general classes, based
on whether the membership of the class can readily be increased or not. Languages can
usually increase their stock of nouns, for example, by borrowing nouns from other languages
or creating new ones through compounding (e.g. black + board yields blackboard) or other
morphological means (e.g. rapid + -ly = rapidly), but they do not normally create or borrow
new adpositions, conjunctions or determiners. Lexical categories such as noun and verb
whose membership can be enlarged are termed open class categories, whereas categories such
as adposition, determiner or conjunction, which have small, fixed membership, are called
closed class categories.
The definitions of lexical categories given so far are primarily the notional ones from
traditional grammar. These definitions seem intuitively quite reasonable to speakers of Indo-
European languages, and they seem to correlate nicely with the syntactic functions of the
different parts of speech. Let us define three very general syntactic functions: argument,
modifier and predicate. In a sentence like the teacher read an interesting book, the teacher
and an interesting book are the arguments, read is the predicate, and the, an and interesting
8
are modifiers. Similarly, in Kim is tall, Kim is the argument and (is) tall is the predicate. The
term ‘argument’ here includes NPs and PPs functioning as subject, direct object or indirect
object. The notions of predicate and argument will be discussed in more detail in the
following chapters, but for now one can say simply that in a sentence the predicate expresses
the state of affairs that the referents of the arguments are involved in. (The terms ‘predicate’
and ‘argument’ are also used in semantics with a different meaning; they are being used here
and elsewhere to refer to syntactic notions, unless otherwise noted.) It is usual to distinguish
1 -place, 2-place and 3-place predicates, depending on how many participants there are in the
state of affairs depicted by the predicate. Being sick is a state of affairs involving only one
participant, hence be sick is a 1-place predicate which takes one argument, e.g. Kim is sick. In
the teacher destroyed the note, there is an action of destroying involving a teacher and a note.
Destroying involves a destroyer and something destroyed; hence destroy is a 2-place predicate
and takes two arguments. Finally, giving involves a giver, something given and a recipient,
and therefore give is a 3-place predicate and takes three arguments, e.g. The teacher gave an
interesting book to Kim. Given these distinctions, it seems intuitively clear that nouns would
be arguments, verbs would be predicates and adjectives would be modifiers, and this is in fact
the case very often. But not always. Nouns and adjectives can function as part of a predicate,
as in Dana is a phonologist and Chris was sick. Even though they are part of the predicate,
they are still formally distinct from verbs; they do not take tense suffixes like verbs do, i.e.
*Dana phonologists or *Chris sicked. The copula be, a kind of verb, carries these verbal
inflections. […]
Every language has noun and verb as lexical categories. This reflects the fundamental role of
reference and predication in communication. One of the most important functions of
language is to allow speakers to depict states of affairs in the world, and in order for them to
do this, there must be linguistic devices which refer to the participant(s) in a state of affairs
and other devices which denote the action, event or situation in a state of affairs. Lexical
items specialized for the first task are nouns, those specialized for the second are verbs.
What about the other major lexical categories? There are languages which lack adpositions
altogether; they express the semantic content of prepositions and postpositions by means of
the kind of suffixes on nouns such as in the Russian language. The concepts expressed by
these endings are called 'case', and the endings are called ‘case markers’. Russian has both
case suffixes and prepositions, but Dyirbal, an Australian Aboriginal language (Dixon 1972),
has only case suffixes and no adpositions at all. Hence the lexical category ‘adposition’ is not
universal. It also appears that adjective is not universal. In Lakhota, for example, the words
expressing properties like ‘red’, ‘tall’, ‘big’, etc., are formally verbs and have basically the
same morphosyntactic properties as verbs. […]. Finally, there has been much less research
done on adverbs cross-linguistically than the other major categories, and therefore it is
difficult to draw any conclusions about their universality.
Thus, it appears that noun and verb are universal lexical categories, but adposition and
adjective are not. It is crucial to keep in mind that when it is claimed that adjective is not a
universally valid lexical category, it does not mean that there are languages which lack words
expressing properties like ‘red’, ‘big’, ‘happy’, etc. Rather, it means that the words
expressing these notions behave morphosyntactically like members of one of the other classes
(verb in Lakhota, noun in Dyirbal and Quechua).
9
In modem linguistics, the determination of the category of a word is not based on its meaning
but rather on its morphosyntactic behavior, i.e. the elements it co-occurs with and the
morphosyntactic environment(s) it occurs in. Meaning is not irrelevant to the function of a
word, but it does not reliably predict it either. The term which is used to refer to classes based
on their morphosyntactic properties is form class. Consider the similarities and differences
between common and proper nouns in English, which was initially characterized
semantically. They are both a type of noun, because they both occur in the major
morphosyntactic environments which nouns (and NPs) occur in, e.g. as the subject or direct
object of a verb, as the object of a preposition in a PP, and with be as a predicate nominal
(The girl gave a book to the teacher. Pat introduced Kim to Dana; Max is my lawyer. My
lawyer is Max). Other form classes cannot occur in these positions, e.g. *The yellow put a
clumsily on the receive. However, they differ in that common nouns can be modified by
determiners and adjectives, while proper nouns cannot, e.g. a tall girl versus *a tall Dana.
Furthermore, common nouns, if they are count nouns, can take plural inflection, while proper
nouns cannot, e.g. the tall girls versus *Danas. Thus there are both syntactic and
morphological differences between common and proper nouns which can be used to
distinguish them as belonging to two distinct subclasses of the category noun.
English verbs can be differentiated from the other major classes by both morphological and
syntactic criteria. Morphologically, only verbs take the suffixes -ing ‘progressive’, -ed ‘past
tense’, or ‘past participle’, -s ‘third-person singular subject-present tense’ and -en ‘past
participle’. Syntactically, they occupy a unique position in a clause, and they may be
modified by adverbs but not by adjectives or demonstratives. There are no consistent
morphological properties that characterize English adjectives; there are distinctive endings
that some adjectives carry, e.g. -y as in slimy (related to the noun slime) or tricky (related to
the noun trick), and -ic as in toxic (related to the noun toxin) or metric (related to the noun
meter). Many adjectives take -er for their comparative forms, e.g. taller, faster, and -est for
their superlative forms, e.g. tallest, fastest. However, many do not, e.g. *beautifuler,
*beautifulest; these adjectives take more and most to indicate their comparative (more
beautiful) and superlative (most beautiful) forms. English adjectives occupy a specific
position within NPs, i.e. DEM- QNT - ADJ - N, as in the seven tall trees (*tall the seven
trees, *the tall seven trees), and they may function predicatively only in combination with the
copula be, e.g. The tree is tall, *The tree talls). Finally, English adverbs, as noted earlier,
often (but not always) end in -ly; they function only as modifiers (but never of nouns), e.g. the
extremely quick rabbit, the rabbit ran very quickly, *the quickly rabbit, and never as
predicates, e.g. *The rabbit is quickly.
This brief discussion of the morphosyntactic properties of the major English classes has not
been exhaustive, but it does illustrate how morphological and syntactic criteria can be used to
characterize the form classes in a language. Even though the criteria for the classes are
ultimately morphosyntactic, the labels for the classes reflect the traditional notional
distinctions. That is, after having established the existence of a form class based on the
morphosyntactic properties of its members, the semantic properties of the prototypical
members of the class determine the name of the class. Hence if the prototypical members of a
class include elements that function as the name of a person, place or thing, then the class will
be given the label ‘noun’.
Summary
There is very little consistency or uniformity in the use of the term ‘category’ in modern
treatments of grammatical theory. It is frequently employed, like ‘class’ or ‘set’, to refer to
any group of elements recognized in the description of particular languages. Following the
more traditional usage, we restrict the application of the term to such features associated with
the ‘parts of speech’ in the languages such as person, tense, mood, etc. By grammatical
category we understand ‘a class or group of items which fulfill the same or similar functions
in a particular language.’ (J.C. Richards, J. Platt and H. Platt 1993:162)
3. Phrases
A Phrase is a group of words that has no subject and predicate of its own and which is used as
a single part of speech,
The fact that she didn’t come makes him very very sad.
=> single word or group of words that do not have a subject and predicate of its own and
which can be used as a single part of speech is a phrase.
4. Sentence
(partial definition)
A sentence is a single free utterance, minimum (= simple sentence) or expanded (=
compound, complex sentence). It is not included in any larger structure by means of any
grammatical device.
Your mother has borrowed the car. She should be back in about an hour.
He is staying with his aunt because the College food is wretched and the rooms aren't heated.
The College food is wretched and the rooms aren’t heated.
The College food is wretched - I am staying with my aunt
5. Clause
A clause is a group of words with its own subject and predicate (a finite, non-finite or implied
verb phrase) if it is included into a larger sentence. A clause forms a sentence (=independent
clause, simple sentence); or, part of a sentence (dependent clauses) and often functions as
noun, adjective or adverb.
Questions:
1. What is syntax? (What are the three key points in the definition of syntax?)
2. How is the notion of grammaticality understood?
3. What are the two aspects of syntactic structure?
4. What are the lexical and functional categories of the English language?
5. Present/Describe the grammatical categories of each English word class.
12
CHAPTER 2
PHRASES
I. Introduction
Sentences are not formed by simply stringing words together like beads on a necklace.
Rather, sentences have hierarchical structures consisting of groups of words that may
themselves consist of smaller groups of words, and so on. This section will focus on the
internal structure of syntactic units built around Nouns, Verbs, Adjectives, Adverbs and
Prepositions, with an emphasis on the organizational properties that they have in common.
Such units are called phrases. Hence, A phrase includes a single word or group of words that
do not contain ‘Subject-Predicate structure’ and is used (i.e., functions) as a single part of
speech.
Heads: Phrases are built around a ‘skeleton’ consisting of two levels. (The symbol P in the
upper level stands for ‘phrase’.)
Each level of phrase structure can be thought of as a sort of ‘hook’ (like a hook on a pole) to
which elements of different types can be attached.
The lowest level is reserved for the word around which the phrase is built - an N in the case of
NPs, a V in the case of VPs, and so on. This element is called the head of the phrase. As the
following examples show, it is possible to have a phrase in which only the head position is
filled. (The material in parentheses provides a context in which these one-word phrases might
occur.)
NP VP
N V
(he likes) books (all animals) eat
AP PP
A P
(she is) certain (he went) in
Although phrases can consist of just one word, they often contain other elements as well. For
example:
a) [NP the books]
b) [VP will eat]
c) [AP quite certain]
d) [PP almost in]
13
In addition to a head (the underlined element), each of these phrases includes a second word
that has a special semantic and syntactic role.
Specifiers: These words (determiners such as the, auxiliaries such as will, and degree words
such as quite or almost) are said to function as specifiers. Semantically, specifiers help to
make more precise the meaning of the head. Hence, the Det the in (a) indicates that the
speaker has in mind specific books, the Aux will in (b) indicates a future event, and the Deg
words quite and almost in (c, d) indicate the degree to which a particular property or relation
is manifested.
Syntactically, specifiers typically mark a phrase boundary. In English, specifiers occur at the
left boundary (the beginning) of their respective phrases. They are attached to the top level of
phrase structure, to the left of the head. Together, these two elements form the phrase
structures depicted in the following tree diagrams.
a NP b VP
Det N Aux V
c AP d PP
Deg A Deg P
The syntactic category of the specifier differs depending on the category of the head. As the
examples in Figure 3 help show, determiners serve as the specifiers of Ns, auxiliaries as the
specifiers of Vs, and degree words as the specifiers of As and (some) Ps.
Some specifiers
In addition to a specifier and a head, the phrases above also contain a complement. These
elements, which are themselves phrases, provide information about entities and locations
whose existence is implied by the meaning of the head. For example, the meaning of eat
implies an object that is eaten, the meaning of in implies a location, and so on.
Complements are attached to the right of the head in English (but to the left in many other
languages). Figure 4 illustrates the structure of a VP and a PP consisting of a specifier, a
head, and a complement.
VP PP
NP NP
As noted above, complements are themselves phrases. Thus, the Complement of the V eat is
an NP that itself consists of a determiner (the) and a head (hamburger). This phrase then
combines with the verb and its auxiliary specifier to form a still larger structural unit.
The head of an adjective phrase (AP) is an Adjective. An AP often contains only a single
word, the head adjective; but the complete functional possibilities are more extensive:
Complements of adjectives are of three types: prepositional phrase, noun clause, and infinitive
verbal phrase (non-finite clause). In other words, an adjective phrase doesn’t always end with
the head adjective; it may contain further grammatical structure. As you become acquainted
with adjectives, you will realize that only some adjectives take complements - particularly
those that semantically refer to mental or emotional states, e.g., aware, afraid, sorry,
disappointed, astonished, hopeful, sad.
Adverb Phrases contain a head adverb and an (optional) intensifier drawn from the same
limited class (very - quite - rather - too - more - most - only - …)
As we noted for single adverb (i.e., adverb phrases with head alone), adverb phrases are
relatively movable within a sentence.
The NP formula states that a noun phrase must contain a headword but need not contain
anything else. If the NP has more elements than the head, it may contain one or more
16
premodifiers (which precede the head) and/or one or more postmodifiers (which follow the
head). The formula thus abbreviates several possibilities:
Noun Head
Premodifier(s) + Noun Head
Noun Head + Postmodifier(s)
Premodifier(s) + Noun Head + Postmodifier(s)
Single word noun phrases will always consist of a headword which is a noun or pronoun.
Simple NPs can also contain a head preceded by a single-word premodifier. The range of
premodifiers of noun heads is large, including nearly all of the parts of speech, at least in
some form. The below examples present some basic possibilities.
Simple Premodifiers
Most of the simple premodifiers above contain one word. The least complex postmodifier -
and by far the most common - is a prepositional phrase (PP). This simple postmodification
will have the structure N = PP;
Our examples so far have provided only single-word premodifiers. In fact, premodifiers can
be multiple:
Multiple Premodifier
Phrasal Premodifiers
Much more common cases in complex noun phrase are the various sorts of phrases and
clauses that follow head nouns. The prepositional phrase that follows head noun contains
NPs, which can contain PPs that contain other NPs that can contain a PP… The following NP
is an example.
The book in the drawer /of the desk //in the office ///of the leader ////of the rebellion
/////against the oppression //////of readers ///////of tales ////////of adventures /////////on far
planets //////////of the galaxy
Adjective Phrase [Anyone fond of kumquats] should buy this recipe book.
Verbal Phrase [The contestant guessing the title] will win a vacation in Tahiti.
Noun Complement The realization [that his hair was false] amused the
Clause audience.
It is possible to repeat NPs twice, thice, …even an infinite number of times. Coordinated NPs
will be joined by a coordinate conjunction, usually and or or, as in:
The verb phrase has a verb as its head. Let’s start with the functional formula for VPs and
then examine the forms that can satisfy that function:
Head
Auxiliary(ies) + Head
Head + (Object(s)/Complement)
Head + (Modifier*)
Combination of the above
Hector walks.
All of the students agree.
The baby cries.
The major auxiliary verbs in English are be, have and do.
A phrase that obligatorily follows a verb head is called an object or complement. These terms
are roughly convertible, although tradition has attached the word “object” to some
constructions and “complement” to others. The reasons for the variation are obscure. The
label “object” dimly suggests the goal or purpose of the verb head, although neither of these
semantic labels applies to every structure so labeled. The term “complement” suggests the
notion of completing (hence the spelling) the verb in some way. This label also isn’t a reliable
19
clue to structure. The below sentences show the main types of objects and complements. A
quick inspection of the sentences will reveal that noun phrases can serve any object or
complement function and that adjective phrases can also act in complement functions. An
important grammatical notion associated with the direct object is that of transitivity. A
transitive verb takes a direct object; an intransitive verb doesn’t.
To distinguish verb modifiers from modifiers of noun, modifiers of verbs often have the
special names such as adverbial and adjunct. Formally, modifiers are of only four types as
indicated in:
Adverbial clauses begin with the subordinating adverbial conjunctions mentioned in the
preceding chapter. Like single adverbs, the phrasal and clausal modifiers are somewhat
movable in the sentence:
Sometimes a short (one- or two-word) adverbial will appear within the VP:
Noun phrase adverbials may be confused with direct objects. However, they will never
become the subject of a corresponding passive sentence:
Example (a) is ungrammatical because a great deal isn’t the true direct object.
The adverbials that modify verbs can be grouped semantically according to the semantic roles
that they express. The most common appear below:
20
Although we have illustrated separately each of the functions accompanying the verb head,
the options in the formula stated at the beginning of this section allow for more than one
function to appear with the verb.
Scott // offered │Zeida │a ride │since her car was out of gas.
(Head + Indirect Object + Direct Object + Adverbial Clause Modifier)
Verb phrases have one prominent purpose in life: to function as predicates along with subjects
and thus to form clauses. That single role is a powerful one, but it would be a shame if such a
linguistic marvel as a verb phrase would have no other use in the language. In fact, English
has arranged for verb phrases to serve a much wider variety of functions - though at a small
cost.
We identified verbs in the previous chapter by their ability to accept a tense marker.
However, a verbal phrase is a verb phrase without tense and modals. The grammatical term
21
nonfinite encapsulates this restriction. Finite verbs are thought to be “limited” by the
presence of tense. (Finis in Latin means “limit or boundary.”) Those VPs without tense are
“unlimited” or nonfinite. Aside from this minor formal restriction - and a few others - verbal
phrases look like other VPs: They have perfect, progressive, and passive auxiliaries, objects,
complements, and modifiers.
One might also extend the notion of being unlimited to the range of functions into which the
verbal phrases enter. While their functions aren't totally unrestricted, they can act as
modifiers (premodifiers, postmodifiers, adverbial modifiers) or can substitute for noun
phrases.
5.3.1 Participles
V-ing
V-en
A participle is a verbal phrase whose first verb is V-ing or V-en; it functions as a premodifier
or a postmodifier of a noun head.
By calling it a verbal phrase, we indicate that the participle lacks tense and modal but may
include other auxiliaries, objects, complements, and modifiers. We also identify an important
formal property of the participle, the use of V-ing or V-en at the beginning. Finally, we
specify precisely the functions of the participle without confusing it with adjectives.
Forms of Participles
5.3.2 Gerunds
A gerund is a verbal phrase whose first verb is V-ing; it functions in the range of NPs.
Formally, gerunds resemble participles, except that they cannot have a verb head with V-en.
They can, however, express passive voice through the be + V-en. Only four verb groups are
possible for gerunds:
Like participles, gerunds are subject to historical change, turning into regular nouns over time.
Such changes are completed when the noun can be pluralized, as in:
22
5.3.3 Infinitives
The word infinitive is used by grammarians in two ways. First, it refers to the basic form of
verb as it would appear if you looked it up in an English dictionary. A second definition is ‘a
verb, usually preceded by to, that is used as a noun or modifier.’ Rephrasing this traditional
definition to recognize infinitives as phrases and to remove the confusion of form and
function, we adopt a definition of infinitive as follows:
An infinitive is a verbal phrase, usually beginning with to, that functions in the range of noun
phrases, or as a modifier or complement.
Forms of Infinitives
to + V to + have to + Be to + Have + Be
V-en V-ing V-ing
Active to sing to have sung to be singing to have been singing
Passive to be sung to have been sung
Questions:
1. How is phrase defined (in English)?
2. What are the elements/components of an English phrase? Give examples of certain
English phrases and describe the elements of each phrase.
3. Describe the functional and formal structures of the English Prepositional phrases,
Adjective phrases, Adverb phrases, Noun phrases, and Verb phrases.
23
CHAPTER 3
WORD CLASSES
I Introduction
Part of our linguistic knowledge involves knowledge of a large number of words, which
constitute our vocabulary or the lexicon as linguists have it. In general, the elements of the
lexicon are what we might think of as words, although different syntactic theories have
slightly different conceptions of what a ‘lexical item’ is, and so it is not always safe to think
of the lexicon as just a stock of words.
However, grammar is neutral, in principle, with respect to analysis and synthesis. In terms of
synthesis, it will be convenient to have the lexicon organized into word classes, given them
symbols such as N for noun, V for verbs, etc.
In this chapter, we examine the individual word classes. It covers nouns, verbs, adjectives,
and adverbs, which contribute the major ‘content’ to a message, and hence are sometimes
called content words, as opposed to other classes known as function words or structure
words. As we will see function words express important meanings and are so grammatically
crucial that nearly every sentence contains one or more of them. However, the content words
allow language to relate to an infinite number of different topics.
1. Nouns
The traditional definition of noun is a ‘word that names a person, place or thing.’ However,
this simple semantic definition has not been agreed upon by other linguists. Nor has the
functional one for nouns been given. For suitable analyses, we consider the forms of nouns.
This classification of nouns has been approached through a series of tests. The tests will help
learners to determine the word class by using the native speaker intuitions that they already
possess. Thus, …
-age -ance/ -ence -ard -cy -dom -er/-or -ess -hood -ism -ist -ity -ment
-ness -th -tion -ude
Single nouns have one dominant function - that of head of a Noun Phrase
horses
the horses
the swift horses
several swift horses
large swift horses
One important subdivision of nouns is that between mass and count nouns. (...) Nouns also
fall into concrete and abstract subclasses. Nouns can also be subdivided into collective
nouns, denoting entities which are collections of individuals (army, jury, the public, ...) and
common nouns. Some grammarians distinguish proper nouns, referring to particular entities,
from common nouns, which refer to classes.
2. Verbs
Traditional grammars typically define verbs semantically, i.e., as ‘words that designate
actions (kiss, run), processes (grow, change), experiences (know), or states of being (be,
have).’ As with most meaning-based criteria, the semantic definition above is somewhat
misleading. For instance, nouns derived from verbs through zero derivation (e.g., strike, kick,
throw, …) will maintain their verbal sense of action. Likewise, verbs derived from nouns -
e.g., man - may appear to maintain whatever naming sense that they have. A far simpler
approach is employ formal consideration to define what a verb is.
2.1.1. Suffixes
25
2.1.2. Prefixes
dis- (disappoint), un- (untie), mis- (misrepresent), mal- (malfunction), out- (outdistance),
over- (overestimate), under- (underestimate), fore- (foresee), re- (reconsider), en- (enlighten),
be- (belabor)
Subclasses of Verbs
Verbs are subdivided into transitive, intransitive and linking verbs. Quirk and Greenbaum
(1973) subdivided verbs into intensive verbs (= linking/copula verbs), which have subject
complements, and extensive verbs. Extensive verbs are then subdivided into intransitive
verbs if they do not permit any of the objects and complements, and transitive verbs.
Transitive verb that takes a direct object are called monotransitive. If it takes a direct and an
indirect object, it is called ditransitive. If it takes object complement, it is referred to as
complex transitive.
3. Adjectives
While traditional grammars usually define nouns and verbs semantically, they often shift to
functional criteria to characterize adjectives. Their definition of an adjective is ‘a word that
modifies a noun or pronoun.’
The definition holds good in simple cases, such as old shoes, offensive remark, and matters
inconsequential, though in the last case, students will have trouble recognizing the second
word, rather than the first, as an adjective. But in each case, the adjective does modify a noun,
which serves as the head of the phrase. However, other words can modify nouns that are
clearly not adjectives. For instance, stone in stone wall is by formal criteria a noun and not an
adjective (e.g., stones and stone's). Likewise, the in the wall shows none of the formal
characteristics of adjectives, although it clearly modifies its head noun. In other words, the
fact that a word modifies a noun doesn't provide sufficient reason to call it an adjective.
The definition suffers also because it extended to functions that don’t include modification.
Note the words ‘or pronoun’ in the definition. Clearly, an adjective cannot modify a pronoun
in any of the following examples: *old them, *offensive it, *they inconsequential,…
26
-ish, -al, -ar, -ful, -some, -y, -ic, -able/ -ible, -ing, -ed
very careful, quite reasonable, thoroughly insane, unusual for its beauty, …
3.2.1. Attributive
Adjectives that directly modify nouns by preceding or following them are often called
attributive adjectives
3.2.2. Predicative
predicative adjectives occur after verbs in the be-become-seem type.
4. Adverbs
Adverbs and adverb phrases seem almost exclusively to modify. But what do they
modify? Our position here will be to distinguish one subclass of adverbs that clearly modify
the sentence and another that modify, in some general sense, the verb group or verb phrase.
The first function is the sentence modifier, the second is the adjunct. Sentence modifiers have
two major functions. They can indicate a speaker's evaluation of the truth of the sentence, or
of what the sentence refers to, which is also called disjunct, and connect one clause or part of
a clause with another, which is called conjunct.
Sentence modifier
Disjunct
Conjunct
Adjunct
5. Pronouns
Pronoun is a word used in place of one or more nouns. Pronouns bear the grammatical
functions of Person, Case, Gender and Number.
First Nominative I we
Accusative me us
Genitive my our
mine ours
Second Nominative you you
Accusative you you
Genitive your your
yours yours
Gender
Masculine Feminine Neutral
Third Nominative he she It they
Accusative him her It them
Genitive his her Its their
his hers Its theirs
Reading:
him and her if it's the object of a verb or a preposition; and his and either her or hers
if they modify or complement a noun or pronoun. We will use the traditional names
to refer to these cases: he/she are in the nominative case; him/her are in the
accusative (AKA objective) case; and hers/his are in the genitive.
English also differentiates other pronouns according to case. Thus I, you, we, and
they are all nominative; me, you, us, and them are all accusative; and my, mine, your,
yours, our, ours, their, and theirs are all genitive.
You will no doubt have noticed that there are two genitive forms of certain pronouns,
such as my and mine. The forms corresponding to my (your, our, their) are used when
the nouns they modify occurs immediately after them. Otherwise, we use the other
genitive forms: e.g., That is my horse, as opposed to That horse is mine. The former
are sometimes referred to misleadingly as possessive adjectives, as they occur before
the nouns they modify in the positions typical of attributive adjectives. The latter are
often distinguished as possessive pronouns because they appear to replace nouns or
noun phrases, e.g., compare That bike is mine with That is my bike.
English nouns functioning as subjects don't differ in form from nouns functioning as
objects, and so we don't distinguish between nominative and accusative cases for
nouns. Grammarians occasionally refer to the nominative/accusative form of nouns
as the common case. English does, however, distinguish between common case and
genitive nouns. The genitive is indicated in written English as 's: Bill versus Bill’s.
Nouns, of course, don't have two genitive forms parallel to the pronouns.
Earlier forms of English, the classical languages such as Latin and Greek, and
modern languages such as Finnish, have much more elaborate case distinctions than
modern English. Table 6.2 provides a list of traditional case names and some of their
functions.
Many languages require case markings on parts of speech besides nouns and
pronouns. Modem German, for instance, makes case differentiations on both articles
and adjectives.
A pronoun may function as the head of a noun phrase, as our revised definition
suggests. Genitives may function as either the head of a noun phrase, as in [4a], or a
modifier of a noun, as in [4b]:
Name Function
Nominative subject
Accusative object
Genitive possessive, partitive
Dative recipient, beneficiary
30
This These
That Those
Third himself
herself themselves
itself
6. Wh-words
7. Articles
There are two articles in English. The definite article the, and the indefinite article a(n).
Articles always function as modifier of the head noun in a noun phrase.
The a(n)
8. Prepositions
Prepositions are important to English because they form phrases that play a wide range of
grammatical roles. Prepositions also express many of the major semantic relations that unite
members of a sentence in a meaningful whole. It's thus important for teachers and students to
become familiar with the approximately fifty members of this class. The common
prepositions appear in the table Single-Word Prepositions below
Grammatically, prepositions are formally recognizable by the fact that they're usually
followed by a noun phrase
of my toe to my closest friends beneath contempt
Single-Word Prepositions
Multiword Prepositions
by reason of by virtue of
by way of except for
in accord(ance) with in addition to
in case of in compliance with
in consequence of in consideration of
in opposition to in place of
in regard to inside of
in spite of instead of
on account of out of
round about with reference to
with regard to with respect to
9. Conjunctions - Conjuncts
Time: after, as, as long as, as soon as, before, just as, now that,
since, until, till, when, whenever, while
Condition: as long as. if, on (the) condition that, provided, provided that,
unless
Nominal clauses function as noun phrases typically function, i.e., as subjects, objects, and
complements. When they do so, they will be introduced by a certain set of subordinating
conjunctions. That set of conjunctions includes most of the wh-words along with the word
that. To illustrate, note the following sentences:
To assure yourself that the clauses truly have a nominal function, replace them with the
pronouns it or that.
Relative clauses function as modifiers of the nouns that they follow. Typically, they're
introduced by members of the wh-word class (traditionally called relative pronouns), and by
the word that. Examples of relative structures appear below:
Questions:
1. What are the formal and functional characteristics of English Nouns, Verbs, Adjectives,
Adverbs, and Prepositions?
2. How are verbs classified according to Quirk and Greenbaum’s view? Draw a diagram
and provides examples to illustrate.
3. How are adverbs classified functionally?
4. What are the grammatical functions or categories of pronoun? How do English pronouns
inflect within the operation of grammatical functions?
5. Which semantic characteristics can be found in pronoun group?
6. What is the difference between nominal conjunctions and relative conjunctions?
34
CHAPTER 4
SENTENCES
I Introduction
Bloomfield's definition of the sentence will serve as a starting-point for our studying.
According to Bloomfield a sentence is ‘an independent linguistic form, not included by virtue
of any grammatical construction in any larger linguistic form’. (…) The point of Bloomfield's
definition can be stated more concisely as follows: the sentence is the largest unit of
grammatical description. A sentence is a grammatical unit between the constituent parts of
which distributional limitations and dependencies can be established, but which can itself be
put into no distributional class. (Bloomfield, cited in J. Lyons 1972: 172-173)
However, the problems concerning a satisfactory definition of the sentence are still unsolved.
Different linguists have different viewpoints. Recently, Hurford and Heasley (1984) in their
discussion of sentence stated that:
“A sentence is neither a physical event nor a physical object. It is conceived abstractly, a
string of words put together by the grammatical rules of a language. A sentence can be
thought of as the ideal string of words behind various realizations in utterances and
inscriptions. (…). A sentence is a grammatically complete string of words expressing a
complete thought.”
II Characteristics of Sentences
What is the internal organization of sentences? (How are units distributed within a sentence?)
To understand the internal organization of sentences and the distribution of the units forming
them, we must consider three major properties of sentence structure:
1. Linearity:
S V
S V O
S V O A
S V C A
S V A
S V O C
S V O.i O.d
Or
Subject - Predicate
No one can utter simultaneously all the words of a sentence. Nor could such an utterance
be understood. Words are spoken (or written) and heard (or read) in a time sequence
from early to later, a sequence represented in the English writing system by a procession
of written forms from left to right. There is a standard order for subjects and objects. In
the English sentence example used earlier: Cassius sees Brutus. The subject of the
35
sentence, Cassius, precedes the verb, while the object, Brutus, follows the verb.
Numbers of other languages follow the same order, Subject-Verb-Object (abbreviated as
SVO). We could try to switch around the subject and the object, converting the SVO
order into OVS, as in this example:
But if we did, English speakers would identify Brutus as the subject. The order would
still be SVO, but the meaning would be different. Other languages may use different
orderings. The range of possible orderings of these words or phrases is known as the
word order parameter. The verb-object parameter discussed earlier is, in fact, part of
this more general parameter. In many languages, word order is less crucial than it is in
English because, as in Latin, there is greater reliance on suffixes and other ways of
marking sentence constituents. Word order therefore appears to be a setting on a yet
more general parameter of function marking. But in no language is word order totally
insignificant. The examples that follow show languages which are like English in that
word order is quite significant, but differ from it in their settings for this parameter.
Gwelodd y dynion y ci
saw the men the dog
V S 0
“The men saw the dog.”
In the Philippine national language, Tagalog, more variation is allowed in word order,
but a very common order is VOS:
Many languages fall into two major groupings regarding the verb-object parameter,
those in which the verb precedes its object and those in which it follows its object. What
is especially interesting is that this difference is associated with other differences in
linear ordering. Thus the relative position of verbs and their objects is very significant.
(Surprisingly, perhaps, subject position seems less important.) Let us focus now on the
clusterings of properties of linear ordering that correspond to the relative order of verbs
and their objects.
36
Although few languages are fully consistent, the following tendencies have been
observed in languages such as Japanese and Turkish, whose verbs follow their objects:
1. Auxiliary verb forms typically follow main verbs, usually as suffixes. Thus the
Japanese for was kidnapped would be literally translated as “kidnapped-was.”
3. Adjectives (red, quiet, circular), relative clauses (who was obstinate, that I wanted
most), and other modifiers of nouns precede rather than follow their head noun.
What about languages in which verbs precede their objects instead of following them?
These reveal the following tendencies:
1. Auxiliary verbs precede main verbs (will talk, not *talk will).
2. There are prepositions rather than postpositions (in Jakarta, not * Jakarta in).
3. Adjectives, relative clauses, and other modifiers of nouns follow their head nouns.
We must emphasize that, especially when large numbers of languages are examined, we
find quite a few exceptions. For example, since English is a language whose verbs
precede their objects, we would expect adjectives, along with other noun modifiers, to
follow their head nouns. In fact, however, adjectives precede their head nouns (e.g.,
lively music rather than *music lively). But adjectives are special; other modifiers of
nouns follow their head nouns, as we would expect.
Despite the irregularities, the correlations are consistent enough to be of interest for
language acquisition research. The task of learning a second language with clusterings
of properties similar to those in one's own language is presumably quite different from
that of learning a second language with different clusterings of properties. Thus, non-
native speakers of languages with linear orderings close to those of English should find
this aspect of English easier than will speakers whose first language orders its
constituents very differently.
This is not necessarily the case, however. Japanese students of English, for example,
rarely seem to have problems with the different positions of modifiers in relation to their
heads, while speakers of German, which is much closer to English, sometimes do.
Perhaps similarities mislead some learners into assuming greater likenesses than
actually exist.
What about first language acquisition? When children learn their mother tongue, one
task confronting them is to find out, on the basis of a limited language input, the
principles of constituent order to which their language conforms. During the so-called
critical learning period,' the child figures out subconsciously the appropriate setting for
each parameter in the language being acquired. For instance, the child must figure out
the position of modifiers with respect to their heads. There are two major options: the
head is to the left of the modifier(s) or the head is to the right of the modifier(s). There is
no interference from knowledge of the parameter settings for another language.
(R. A. Jacobs, 1995: 35-37)
37
2. Hierarchy
Sentences are hierarchically structured, that is, they are not simply sequences of individual
words but are made up of word groupings, which themselves may consist of lesser groupings.
Words are not necessarily the only constituents of sentences; there are also higher-level
constituents that form sentences. This kind of hierarchical organization, like linearity,
represents a more general strategy the mind uses to organize experience. In sentences,
lesser elements are parts of larger wholes, which are in turn parts of yet larger wholes.
Things are easier to deal with if they can be placed within a larger frame, a part-to-
whole strategy, or if they can be seen as consisting of distinguishable parts, a whole-to-
part strategy. The latter strategy, for example, makes it easier for us to memorize this
sequence of numbers:
36, 724, 215, 105. 142, 52, 0, 77
than this one:
3672421510514252077
Now think about the following sentence and look at the tree-style diagram below:
The government expelled the officers from Thailand.
DIAGRAM 4.1
No two words in the diagram group together to form a higher-level constituent. Is this a
correct reflection of sentence organization in English? Clearly not, since the lack of
grouping fails to capture relationships that any native speaker of English can perceive.
English speakers know that the second the in the sentence is tied more closely to the
noun officers than to the verb expelled that precedes it. The closeness of this tie is
indicated by the fact that these two words, forming the phrase the officers, can be
replaced with a single pronoun, them. In contrast, the words expelled the do not form a
constituent phrase replaceable by any single word. The pair the government forms the
same kind of phrase as the officers. Finally, the prepositions from is more closely tied to
the word following, Thailand, than to officers, which precedes it. A more accurate
representation of the structure of our sentence would show these higher-level
constituents too.
38
DIAGRAM 4.2
Note in Diagram 4.2 that the phrase from Thailand does not form a higher constituent
grouping with the phrase the officers. The predicate expel here has one phrase indicating
who was expelled and another indicating the place from which they were expelled. The
object of expelled is the officers. There is evidence to support this constituent structure.
First, the object can be replaced by a pronoun object, them:
Since them replaces the original object, that object must have been just the officers.
Second, the sentence has the following passive voice counterpart:
In the passive voice sentence the object noun phrase, the officers, has been shifted to the
subject slot. The prepositional phrase has not been shifted.
But note that the active voice sentence. The government expelled the officers from
Thailand, has an alternative interpretation, one in which from Thailand does not indicate
the place from which the officers were expelled but simply functions as further
descriptive detail specifying which officers were expelled. Under this interpretation, the
officers from Thailand is a constituent. It can therefore be replaced by the pronoun them:
This time, the object noun phrase the officers from Thailand has been shifted to the
subject slot. In this interpretation, the determiner the makes definite not just officers but
the whole grouping officers from Thailand. So the sequence is a constituent whose
structure can be shown like this:
39
DIAGRAM 4.3
The constituents of the officers from Thailand together form a single higher-level
constituent, that is, a phrase. On this interpretation, the constituent structure tree for the
whole sentence must therefore show this constituent. The two interpretations of the
sentence correspond to two distinct constituent structure trees. There is yet another
higher-level constituent, one headed by a verb. Note that the verb expelled is a transitive
verb, that is, it takes an object. In our example, what the object is depends on which
interpretation is chosen. For the interpretation in which the prepositional phrase from
Thailand is separate from the officers, only the officers is the object. The higher-level
constituent to which these phrases belong, the verb phrase, is made up of three separate
parts: expelled, the officers, and from Thailand, as this next diagram shows:
DIAGRAM 4.4
For the second interpretation, in which the object is the officers from Thailand, the verb
phrase is made up of two separate parts, expelled and the officers from Thailand, as the
following tree diagram shows:
DIAGRAM 4.5
For both interpretations there is an obvious dependence between expelled and the object
constituent that follows. The verb expel requires an object and also allows a
40
prepositional phrase to indicate the place from which someone is expelled. This special
grouping relation is quite different from a relation between verbs and their subjects.
Verbs are not categorized according to whether or not they take subjects; any verb can
have a subject. So the verb and the constituents following it form a higher-level
constituent in a clause.
Note, however, that not all constituents following a verb are necessarily part of the verb
phrase. Forms like yesterday, which can be shifted to other positions in the sentence,
Yesterday the government …, or, The government yesterday … are outside the verb
group.
DIAGRAM 4.6
DIAGRAM 4.7
This analysis has demonstrated that the linearity property alone does not account for the
relation between form and meaning in a sentence. The differences noted in hierarchical
structure correspond to the differences between the two interpretations of the example
sentence. A grammar of English that did not posit hierarchically organized constituents
for sentence structures would find it hard to account for the ambiguity of sentences such
as The government expelled the officers from Thailand. The differences in the groupings
of the forms match up with the differences in meaning.
To make the tree diagram system more useful for representing hierarchy, we need a few
terms. The points on a tree where the branches come together are called nodes. Three
feminine labels from the kinship system are used for the relations in the diagram
between constituents. Two or more constituents attached to the next higher node on the
tree are referred to as sisters. So, in the following tree for expelled the officers:
DIAGRAM 4.8
the constituents the and officers are sisters because they are connected to the same next
higher node. The larger constituent, the officers, is the sister of the verb expelled, since it
is attached to the same higher node as expelled. Not surprisingly, the higher node to
which sisters are attached is known as the mother node, and the sisters are daughters of
the mother node. Thus, in our diagram, the word officers is a daughter of the higher
node to which the and officers are attached.
(R. A. Jacobs, 1995: 37-41)
3. Categoriality
Sentences are made up of parts which belong to a set of distinct categories, each with its
special characteristics.
The constituent structure trees studied so far represent (1) the linear ordering of the
sentence and (2) native-speaker intuitions as to the hierarchical organization of the parts.
But the trees fail to express crucial generalizations about sameness and difference.
Certain constituents are of the same kind, and they are different from others. Without
conscious effort, native speakers exploit the samenesses and differences by using
constituents of the same kind in the same positions within a sentence; that is, the
constituents share the same distribution. A descriptive grammar must differentiate
between items that are the same and those that are different. Words, and the larger
constituents they make up, belong to a set of distinct categories, each with its special
characteristics. This is the third general property of sentence structures, categoriality.
The words car and tree are similar kinds of words, and their distribution - the range of
positions in which they can occur - is very similar. They can, for instance, occur right
after the words a and the. The two words also have counterparts with the -s suffix
indicating plurality: cars, trees. This last similarity is not a matter of distribution but of
the range of forms allowed for particular categories of words. The study of word forms,
morphology, provides useful criteria for determining the category to which a word
42
belongs. These supplement the distributional criteria for a particular category. What
about other words, for example, words like this and unless? Obviously these words don't
belong in the same category with car and tree. They neither occur after a and the nor
take the -s suffix. Moreover, the differences in their distribution indicate that they
themselves fall into two separate lexical categories, the categories of words in the
lexicon. While the lexical item or word this
can follow prepositions like after, on, before, and from, the word unless cannot. We can
say after this but not *after unless.
To show categorial distinctions on constituent structure trees, the words must be labeled
appropriately. The bottom part of the trees could look like this (DET stands for
determiner, words like the, this, a, while N stands for noun, V for verb, and P for
preposition):
DET N V DET N P N
DIAGRAM 4.9
As we've already seen, the higher-level constituents - the phrases - also fall into
categories, referred to as phrasal categories. For example, the two-word phrases the
government and the officer clearly share enough properties to be included in a category.
Both phrases have a noun as head, both can function as subject or object, and both can
take a plural suffix. Since their head word is a noun (N), they are referred to as noun
phrases (abbreviated NP). Now note that the noun Thailand, although a single word that
doesn't normally take a plural suffix, shares not only key properties of the noun category
but also distributional properties of noun phrases. Thailand has a noun as its head since
it is the only word in the phrase. Moreover, it can function as subject or object and, like
other noun phrases, can be replaced by a pronoun. It can be the object of a preposition
like from, as other noun phrases can, and therefore can be considered to be a one-word
noun phrase.
The phrasal category noun phrase also includes pronouns like they, it, and them. These
pronouns have the same distribution as the phrases the government and Thailand, so
pronouns, a special subcategory of nouns, can be, and almost always are, single-word
noun phrases.
To go one step further in our example, the noun phrase Thailand is itself the object of a
preposition, from, which is the head of the prepositional phrase from Thailand. The
category prepositional phrase (PP) includes such phrases as to Cortina, out of the
village, and with her father. The larger (mother) constituent to which both expelled and
the officers from Thailand belong is expelled the officers from Thailand. For this larger
phrase we can substitute the single intransitive verb resigned:
Like all the other phrase categories, except prepositional phrases, verb phrases (VP) can
consist of just one word, for example, resigned. The sequence expelled the officers from
Thailand is also a verb phrase, one organized around the transitive verb expelled, which
is its head.
The sequence fond of marshmallows is organized around an adjective, the word fond,
which requires a prepositional phrase like of marshmallows to follow it (and to be a
sister of fond on a tree diagram). We can, however, substitute just an adjective for the
phrase fond of marshmallows. Compare these next two examples:
The scoutmaster was fond of marshmallows.
The scoutmaster was obstinate.
The adjective obstinate is not only an adjective but an adjectival phrase (AP), just as
fond of marshmallows is an adjectival phrase. It is also the head and only constituent of
the adjectival phrase.
What exactly is a head? First and most important, the head of a phrase is the word
around which the phrase is organized. This is why the head of a phrase cannot be
omitted. Secondly, the category of the head is the category to which the phrase belongs.
Thirdly, the head word is typically the semantic nucleus of the phrase. Single-word
phrases, in which the word is also the whole phrase, consist of nothing but such a
nucleus. Multiple-word phrases have other categories as constituents, and these
constituents bear grammatical relations within the phrase. In fact, the notion head of a
phrase is itself a grammatical relation, not a word or phrase category.
Let's return now to the sentence The government expelled the officers from Thailand.
Constituent structures for the two interpretations of the sentence can now reflect
categoriality as well as linearity and hierarchy. Just one of the alternative structures is
shown here:
DIAGRAM 4.10
44
The constituent structure above shows the determiner the has as its sister constituent a
unit consisting of the noun officers and the prepositional phrase from Thailand. In the
diagram we have shown this unit as N'. (We will be discussing this category in Chapter
5.) All of these units together form a single, higher-level noun phrase. This higher-level
noun phrase functions as the object.
CHAPTER 5
In traditional grammar, sentences are classified into different types in two ways: first of
all by function, as statements, questions, exclamations and commands; and secondly
according to their structural complexity, as simple or compound. Complex sentences are
made up of a number of simple sentences (which when incorporated as constituents of
larger sentences are, by virtue of this fact, called clauses). Thus: I saw him yesterday
and I shall be seeing him again tomorrow is a complex sentence. Complex sentences are
divided into: (a) those in which the constituent clauses are grammatically co-ordinate,
no one being dependent on the others, but all being, as it were, added together in
sequence, with or without the so-called coordinating conjunctions (and, but, etc.); and
(b) those in which one of the clauses (the ‘main clause’) is ‘modified’ by one or more
subordinate clauses grammatically dependent upon it and generally introduced (in
English) by a subordinating conjunction (if, when, etc.). Subordinate clauses are
subdivided by function as nominal, adjectival, adverbial, etc.; and further as temporal,
conditional, relative, etc. (…).
(Lyons 1972: 178)
Clauses are constructions with one phrase constituent, typically a noun phrase, that bears the
subject relation and another constituent, the verb phrase, bearing the predicate relation. This
construction:
cannot be a clause because it lacks a verb phrase. Here is one example of a clause:
The subject of the clause is Clara and the verb phrase is delayed her graduation. This clause
can stand on its own as a sentence, but could also be embedded inside another clause.
Notice that the embedded clause can be introduced with that. This introducing word, that, is
known as the complementizer. The complementizer was optional in the above example, when
the embedded clause was the object, but it can never be omitted when the embedded clause is
subject of another clause:
That Clara delayed her graduation is unfortunate.
Complementizers never occur when the clause is an independent clause, that is, one capable
of being a full sentence on its own.
We need to mention here one important parameter for clauses that will be dealt with in more
detail later, the finiteness parameter. Clauses can be either finite or nonfinite. In the Clara
clause, with or without the complementizer that, the verb phrase begins with delayed, a verb
marked for past tense. Alternatively, the verb could have been preceded by may, can, would,
will, should, might, must. could, or ought to, special verb forms known as modals. Clauses
that have either modals or verbs indicating past or present tense are known as finite clauses.
What are nonfinite clauses then? They are clauses in which the predicate phrase begins not
with a present or past tense verb or a modal but with a to before the verb. The verb with to is
often called an infinitive verb. Nonfinite clauses are like finite clauses in that they have a
verb phrase and a subject, though the subject is sometimes understood rather than overt. Also
like finite clauses, they can be introduced by a complementizer. But their complementizer is
not that but for. Here is an example:
The subject is again Clara and the predicate phrase is to delay her graduation. The
complementizer for enables us to embed this clause into a larger clause, as in these two
examples:
This results in a viewpoint that classifies sentences into simple and complex, in which the
latter is realized under coordination or subordination relation.
(‘you to want me’ is a non-finite clause embedded in the super-ordinate clause which is the
whole sentence)
Complex sentences are formed by joining a number of simple sentences together. Complex
sentences are classified into two types. Those in which the constituent clauses are co-
ordinate, there is no main - dependent construction within the sentence: no one being
dependent on the others, but all being, as it were, of equal importance and can stand on their
own. The clauses are added together in sequence, follow a logical order as required by the
context, with or without the so-called coordinating conjunctions (and, but, etc.)
(a) We fished all day; we didn’t catch anything.
(b) We fish all day, but we didn’t catch anything.
47
(c) He not only washed his motorbike but (also) polished it (as well / too).
The other type of sentences, on the contrary, is formed by linking simple independent clauses
together, but the constituent clause is not of equal importance. One is subordinate to the other.
(d) Everybody knows that money does not grow on trees.
(e) Holiday resorts which are crowded are not worth staying.
(f) Greenhorns changed completely after he got married.
The subordinate clause is also called embedded clause, however, there is a trend to view
embedded clause as a subclass of subordinate clause. The structure of the subordinate clause,
according to this view, is similar to that of simple sentences. Complex sentences consisting
of main and subordinate clauses need not present problems for either production or
comprehension.
Embedded clauses are different. They function semantically as arguments of predicates (i.e.
the nominal units required by the predicator (either verb group, or noun group, or adjective
group, or preposition) of the sentence. Since embedded clauses also contained predicates and
arguments, problem can arise because the addressee must sort out which arguments go with
which predicates.
It was alleged that two hostages had been ordered to pick up the money.
The above sentence shows that ‘two hostages’ is in the position of the subject of the passive
order, but is not the orderer in the action.
When clauses are linked in a relationship of equality, we say that the relationship is of
coordination relation. Traditional grammar describes complex sentences bearing coordination
relation compound sentences. In the relationship of coordination, both or all clauses have the
same syntactic status. In terms of semantics (meaning), the information presented in one
clause is as important as that presented in the other or others.
The coordinating conjunction which can be used to form this type of sentence are and, and
then, but, for, nor, or, so, yet; either... or; neither ... nor, not only ... but (also/as well/too).
These can be used for the purpose of
- addition: Chris washed his car and polished it.
We were talking and laughing.
- result: I’ve got a terribly fever, so I went to see a doctor.
He fell heavily and broke his arm. (= so)
- condition: Clean the trash, and I’ll pay you 50,000 d. (= if ... then)
- sequence: He finished his exam and fell down in a faint.
- contrast: Paul speaks English, but his wife speaks Japanese.
Tom’s 15 and still sucks his thumb. (despite this)
- choice: Work hard or you’ll fail the exam.
- reason: The boy has to be street vendor, for his family is so poor.
- continuation: The man opened the door, and then
48
When clauses of unequal status are linked, we say that the relationship is one of subordination
relation. In subordination relation sentences, one clause or more clauses are subordinated to
another. The information in the subordinate clause is often presented as backgrounded or
presupposed in relation to the information contained in the sentence. The clause which
includes all subordinate clauses is called the main clause.
(a) Danusa kept quiet because she was afraid.
(b) This is the house that/which was built of mud.
compared with
(c) Wanado knows that Edgar loves Angela.
the difference between (c) and (a) or (b) is that the clause that Edgar loves Angela is a must
argument of the predicate love. To make a complete sentence with the predicate love, we
must need an argument subject and an argument object. Thus, that Edgar loves Angela is
called embedded clause of the complex sentence (c).
Since one of the categories of verbs is finite or nonfinite, subordinate (embedded) clause can
be finite or nonfinite clause.
Wanado knows that Edgar loves Angela. (finite)
Wanado knows what to do. (nonfinite)
We think the ghost appears at midnight. (finite)
We want the ghost to appear at midnight. (nonfinite)
We want to see the ghost at midnight. (nonfinite)
1. Types of Clauses
Finite Nonfinite
1. Adverbial 1. Reduced Adverbial
2. Relative 2. Reduced Relative
3. Noun 3. Gerund
4. Infinitive
Adverbial clauses are typically introduced by what have been traditionally called
subordinating conjunctions and generally fulfill the same functions as AdvPs (...), indicating
time, place, condition, cause, and purpose. They appear in the positions typical of AdvPs
(initial, medial, and final). They’re typically finite, but in some cases, they may be nonfinite.
We provide examples of each of these types with their typical conjunctions. Note that
nonfinite versions of adverbial clauses are elliptical versions of the fuller finite structures.
Time clauses
[a] After you left the party, things really began to swing.
[b] As soon as the mailman came, Terry ran to the door.
[c] Before Reagan was elected, there was more money for schools.
[d] Since the shuttle crashed, NASA has been demoralized.
[e] While he was swinging on the creeper, Tarzan emitted a blood-curdling yell.
[f] While swinging on the creeper, Tarzan emitted a blood-curdling yell. (Nonfinite)
[g] When he was questioned by the police, the suspect demanded to see his lawyers.
[h] When questioned by the police, the suspect demanded to see his lawyers. (Nonfinite)
[i] Before you get into trouble, quit.
[j] Before getting into trouble, quit. (Nonfinite)
Place clauses
[a] Wherever you find cotton, you will find the boll weevil.
[b] Double quotes should be used only where they are appropriate.
[c] Double quotes should be used only where appropriate. (Nonfinite)
Conditional clauses
[a] If you understand this, (then) you will be able to do the exercises.
50
[b] Unless you understand this, you will be unable to do the exercises.
Cause clauses
[a] Because he hoped to elude his pursuers, Fred continued his trek into the mountains.
[b] Since/As funding is scarce, research is hampered.
[c] Being a clever fellow, Fred was able to draw the correct conclusions. (Nonfinite)
[d] Seated by the window, the children could see everything that happened on the street.
(Nonfinite)
Purpose clauses
[a] We packed food for six meals so (that) we could stay out in the forest overnight.
[b] Let us spend a few moments in silence so that/in order that we remember those who
died to preserve our freedom.
Result clauses
[a] She was so stunned that she couldn't speak.
[bl The shooting star moved so quickly that I almost missed it.
Manner clauses
[a] Type this again as I show you a moment ago.
[b] This steak is cooked just how/the way I like it.
[c] I feel as if/as though I’m floating on air.
[d] He sounds as if/as though he is badly injured.
Reason clauses
[a] As/Because/Since there was very little support, the strike was not successful.
[b] Long is trying to find a new private room because he wants to live independently.
Concession clauses
[a] Although/Though/Even though I felt sorry for him, I was secretly pleased that he was
having difficulties.
[b] We decided to travel by plane, even if air fares go up again this year.
[c] No matter where you go, you cannot escape from yourself.
[d] However brilliant you are/may be, you can’t know everything.
4.3 Clauses that function as modifiers of nouns (Relative clauses and Reduced relatives)
Relative clauses (also called adjective clauses) follow the head nouns they modify and may
begin with either that, a wh-word such as who or which, a phrase with a wh-word in it, or no
special word at all. Relative clauses must be divided into two types, restrictive and
nonrestrictive (or appositive) relatives. In written English, appositive relatives are separated
from their head noun by a comma and end with another comma. Restrictive relative aren't set
off by commas. The presence or absence of commas reflects a semantic difference between
these two types, although there are formal differences between them too, which we deal with
below. We begin by illustrating some of the variety of restrictive relatives:
The man that we bought the boat from — skipped town.
The man who(m) we bought the boat from — skipped town.
The man from whom we bought the boat — skipped town.
The man we bought the boat from — skipped town.
The above italicized clauses are of the kind restrictive relative clauses. Restrictive relative
clauses are interpreted as providing information necessary for identifying the referent of the
entire NP. Another kind of relative clause, the nonrestrictive or appositive, only supplies
extra information which isn’t considered necessary to identify the referent on the NP:
Mr. Pickhurst, who(m) we met last week, has gone away.
I hit the brakes, which caused the car to fishtail.
4.3.1. Nonfinite clauses function as relatives
[a] The man to see is Fred Limestone.
[b] The man standing near the entrance is my father.
The italicized are classified into the group of nonfinite clauses as the verbs are in their
nonfinite form and their subjects are covert. Sentence [b] can be interpreted as reduced
relative clause as it’s an elliptical version of:
[c] The man who is standing near the entrance is my father.
Exercises: Make a list of all the conjunctions used as the marker of English adverbial clauses.
Present their meaning(s) and make up sentences to exemplify.
52
1. Sentence Structure
A sentence can be divided into two or more groups, and within each groups the words can be
divided into subgroups, and so on until only single words remain.
This sentence is composed of two main groups: the boys + kicked the ball
which are called constituents
Constituents that can be substituted for one another without loss of grammaticality must
belong to the same syntactic category
+ Old (men and women) are served first. (Young men and young women are not.)
+ (Old men) and women are served first. (Young men are not.)
The claims about the structure of the sentence The boys kicked the ball can be formulated in
terms of ‘rewrite rules’. (read ‘ -->' as ‘rewrite as’ or, less formally, ‘goes to’).
2a. S --> NP VP
2b. NP --> Det N
2c. VP --> V NP
53
(2a - c) are Phrase Structure (PS) rules in the sense that they incorporate claims (specified to
the right of the arrow) about the constituent structures of phrases (specified to the left of the
arrow). PS rules are said to generate structures, where generate is understood to mean ‘make
explicit’. (2a), for example, generates the structure of S by making explicit the information
that S consists of NP and VP. (2b), on the other hand, generates the structure of NP by
making explicit the information that NP consists of Det and N. Finally, (2c) generates the
structure of VP by making explicit the information that VP consists of V and NP. Tree
diagrams and labeled brackets are (visual) devices of representing claims about constituent
structures incorporated in PS rules.
The PS rules (2a - c) were based on sentence in the previous section, reproduced in (2’).
However, their generative capacity goes well beyond (2’), to include all possible sentences in
the language with similar strings. (3a - d) are a few examples of such sentences. They all
resemble (2’) in that they include the same patterns of constituency for each category:
To generate a specific sentence of the set of sentences generated by rules (2a-c), another set
of rules which generate specific lexical items can be added. (2’), for example, is fully
generated by the set of rules in (4):
4a. S --> NP VP
4b. NP --> Det NP
4c. VP --> V NP
4d. Det --> the
4e. N --> boy, ball
4f. V --> kick(ed)
Rules (4a - c) generate phrasal categories one constituent of which is a terminal node.
Terminal nodes are nodes that do not branch and that immediately dominate the lexical item.
For example, the phrasal category VP has the terminal node V as one of its constituents, and
NP has the terminal node N as one of its constituents. S is called the root node. Rules (4d - f),
on the other hand, generate terminal nodes by introducing corresponding lexical items in the
sentence. The structures generated by both sets of rules are called phrase markers.
Obviously, there is also an equally large number of possible sentences which PS rules (4a - c)
cannot generate.
The rules that determine the basic constituent structure of sentences are called Phrase
Structure Rules.
(Art) (AP) N
NP Æ
Pro
VP Æ V NP
VP Æ V NP PP
VP Æ V (NP) (PP)
PP Æ Prep NP
PS rules generally recognize only one level of representation above the terminal node, namely
the phrasal level VP, AP, PP … etc. (XP). This can be clearly seen in (5a-d) which are
abstracted from the corresponding PS rules discussed above
5a. VP --> … V …
5b. NP --> … N …
5c. AP --> … A …
5d. PP --> … P …
Read from right to left, rules (5a-d) encode the generalization that the structural representation
of every category includes a phrasal level, i.e. XP. For example, the structural representation
of V includes VP, the structural representation of N includes NP, and so on. The phrasal level
(XP) is called the maximal projection (of X) in X-bar terminology. Read from left to right,
the rules in (5a-d) convey a different, but related, generalization, namely that every XP has X
as an obligatory constituent. For example, VP has V as an obligatory constituent, NP has N as
an obligatory constituent, and so on. In X-bar terminology, the obligatory constituent of a
maximal projection is called the head (of that maximal projection). This generalization is
related to the previous one in the sense that it actually follows from it. If the structural
representation of every category includes a maximal projection, then every maximal
projection will include the category (i.e. the head) of which it is the maximal projection.
This core property of PS rules can be captured in terms of the schema in (6), where X has the
same categorial value on both sides of the arrow, e.g. if X = V, then XP = VP. (6) is a schema
in the sense that it identifies a property which all members of the class of PS rules in question
have in common. It is understood as a condition on the structural representation of categories
insofar as it specifies the format that such representations must conform to:
6. XP --> … X …
55
Although (6) basically captures a common property of PS rules, there is a sense in which it is
more restrictive than PS rules. Because the latter are rewrite rules, and because systems of
rewrite rules generally make it possible in principle to rewrite a given symbol as one or a
combination of any (number of) symbols, nothing seems to exclude unattested representations
of the type illustrated by the rules in (7a&b). However, these rules are excluded by schema (6)
on the grounds that they do not observe the condition that every maximal projection must
have a head, and that a maximal projection exists insofar as it is a projection of a lexical head:
One of the serious shortcomings of PS rules is the fact that they do not reflect structurally the
distinction between subcategorized and non-subcategorized categories in relation to the head.
The PS rule which generates the NP in (8a), for example, has the form shown in (8b) and
generates the structure shown in (8c):
8c. NP
NP N PP
The PP to the problem is the complement of the noun solution and the NP Mary its (logical)
subject. However, both the PP complement and the NP subject are sisters to the head N and
therefore to each other. This is an undesirable situation, if only because it seems to undermine
the claim that grammatical functions are structurally based. Under a structural definition of
grammatical functions, subjects and complements are expected to have different structural or
grammatical relations. In the structure of the sentence (S), the complement of the verb is a
sister of the verb, but the subject of the sentence is not. The structure of NP needed is one
where the subject and the complement have different structural relations with respect to the
head N. Recall that the notions 'complement' (or object) and ‘sister’ are closely related. The
complement of a head is structurally represented as its sister. In view of this, it seems that it is
the structural relation that the subject has with the head N which needs to be modified in (8).
and the complement of N. The subject is the daughter of NP and sister of N', and the
complement is the daughter of N' and sister of N:
9. NP
Pursuing our attempt to replace PS rules with general schema which act as conditions on the
structural representation of categories, the schema underlying
(9) can be stated as in (10a&b):
(10c) is the abstract structure they generate. X, Y and Z are variables which stand for any
category. In (9), for example, X = N (solution) with the inevitable consequence that X’ = N’
(solution to the problem) and XP = NP (Mary's solution to the problem).
YP = NP (Mary) which is a subject by virtue of being daughter of NP and sister of N’. ZP =
PP (to the problem) which is a complement of N by virtue of being a sister of N.
The intermediate level between the maximal projection XP and the head X is X', called the
single bar projection. The hierarchical relations between the three separate levels of
categorial representation are sometimes represented in terms of the number of bars (or primes)
associated with each level. This is shown in (11a&b), which are purely notational variants of
(10a&b). The hierarchy is from ‘double-bar’ (X” = XP) to ‘single-bar’ (X’ = X’) to ‘zero-bar’
(X° = X) or vice versa. The double-bar projection is the maximal (or phrasal) projection and
X the head. The asterisk associated with the complement category Z” in (14b) means ‘zero or
more occurrences’. It is intended to reflect the (familiar) fact that the presence of
complements, their number and their nature depends on the lexical item in the head position:
Spec(ifier) is a functional term which refers to the category which is the daughter of XP and
the sister of X’. Often the term ‘Spec’ is used interchangeably with the term ‘subject’
especially in relation to categories that are smaller than a clause/sentence. The Spec of an NP,
for example, can be another NP as in (12), a determiner as in (12) or nothing as in (13):
57
Spec N’ N’
Det N PP N
The option of not having a Spec at all is indicated by the asterisk in (11a), which has the same
meaning as in (11b). Recall that determiners are determined by the type of N the NP includes.
English names such as Mary do not take a determiner, unlike common nouns such as solution
which can take either a determiner (12) or a whole NP Spec (9). Later on we will see that
there is a limit on the number of Specifiers that a phrase could have such that a phrase cannot
have more than one Specifier.
X-bar schema applies to all categories in the same way, as indicated by the use of variables in
their formulation. Thus, V is expected to have the representation in (14a), A the representation
in (14b), P the representation in (14c), and so on.
14a. VP 14b. AP
(Spec) V’ (Spec) A’
V (ZP) A (ZP)
14c. PP
(Spec) P’
P (ZP)
By letting the more abstract symbol X” (X-double-bar) stand for either NP or AP, X’ stand
for N’ or A’, and X stand for N or A, we can write a general rule that covers both noun
phrases and adjective phrases:
X” --> (specifier) X’
X’ --> X (complement)
58
where specifier is Art for NPs and Deg for APs. Complement is either PP or S’. This rule
‘schema’ stands for the six rules:
VP --> V S’
S’ --> Comp S
the category on the left side of the arrow is repeated on the right side
You mean (that) you didn’t know (that) I know that she didn’t know that I saw her eat out
with her boyfriend in the Indian restaurant.
Questions:
1. How do you classify types of English sentences?
2. What are Complex sentences? (Review)
Give examples of English complex sentences with:
59
CHAPTER 6
TRANSFORMATIONAL GRAMMAR
I. Introduction
Initially, grammar was considered to consist of a set of Phrase Structure (PS) rules
which generate Phrase Markers called Deep Structures (DS), and a set of
transformational rules which perform various types of operations on these Phrase
Markers to derive appropriately modified Phrase Markers called Surface Structures
(SS). PS rules are ‘rewrite’ rules of basically two types. The ‘context-free’ type of the
form X Æ Y, and the ‘context-sensitive’ type of the form X Æ WYZ, where W and Z
represent the context. The former generate phrasal categories such NP, VP, S… etc., and
the latter introduce lexical items into appropriate contexts in Phrase Markers.
Transformations were largely construction-specific, so that there was a transformation
for passives, a transformation for yes-no questions, and so on. Universal Grammar was
considered to contain a kind of blueprint which prescribes the types of possible rule
systems, and an evaluation metric which restricts the range of possible grammars to the
ones (ideally, one) compatible with the data available to the child.
At a later stage, it became clear that there was a conflict between the desire to provide a
description of further phenomena, that is the desire to achieve descriptive adequacy,
which resulted in the proliferation of rule systems, and the need to constrain this
proliferation, that is the desire to achieve explanatory adequacy. The reaction to this
conflict was basically to derive general principles with broad scope from existing ones
and attribute them to UG. These principles would then serve as conditions on
representations, the application of rules or their output, and perform a restricted range of
operations. As components of UG, these principles also serve to define the notion
‘possible human language’.
The developing theory went though successive stages with distinctive properties called
the Standard Theory, the Extended Theory, Government and Binding Theory, the
Principles and Parameters Theory and the Minimalist Program (or Minimalism). Each of
61
As the theory was developed, its empirical range was widened considerably to include a
fairly broad range of diverse languages. This led to the sharpening of some of the
existing ideas, but most prominently to the formulation of clearer ideas about the
principles responsible for language variation. It turned out that some of the major
aspects of language variation can be accounted for in terms of simple and well-defined
sets of options, technically called parameters, which are largely determined by the
lexical properties of a specific class of categories called functional or inflectional
categories. The comparative work carried out within this framework has been largely
successful in identifying common underlying properties of super-ficially different
languages.
(Jamal Ouhalla 1999: 11-12)
1. Movement rules
Although phrase structure rules interact with the set of complement options permitted by
individual heads to form a very wide range of patterns, there are syntactic phenomena that
they cannot describe in an entirely satisfactory way. This section considers two such
phenomena and discusses the changes that must be made in order to accommodate them.
These sentences have an auxiliary verb to the left of the subject rather than in the specifier
position of the VP, as in (2).
Our phrase structure rules place the auxiliary in the appropriate position in (1), but not in (2).
How does the word order found in the former sentences come about? The question structures
that we are considering are built in two steps. In the first step, the usual phrase structure rules
are used to form a structure in which the Aux occupies its normal position within the VP. This
allows us to express the fact that even in question structures it functions as a specifier, making
more precise the meaning of the verb.
62
NP VP
Det N I V
(Aux)
The second step in the formation of question structures requires a transformation, a special
type of rule that can move an element from one position to another. In the case we are
considering, a transformation known as Inversion moves the Aux from its position within the
VP to a position to the left of the subject. This transformation is called Subject-Auxiliary
Inversion (SAI)
The transformational analysis has at least two advantages. First, we can say that all auxiliaries
function as specifiers of the verb, consistent with the simple analysis in the above sentences.
Those sentences that have an auxiliary verb to the left of the subject simply undergo an ‘extra’
process - the Inversion transformation that moves the auxiliary from its position within the
VP in order to signal a question.
Second, the transformational analysis automatically expresses the fact - known to all speakers
of English - that the sentence Will the boy leave is the question structure corresponding to The
boy will leave. According to the analysis presented here, both sentences have exactly the same
structure after the application of the phrase structure rules. They differ only in that Inversion
has applied to move the auxiliary verb in the question structure.
Let’s look at the interrogative sentence What will the boy find?
The transformational rules that move the auxiliary verb and the Wh-word are specific
examples of a general transformation rule ‘move any constituent’, or move α (move alpha)
rule. For example, this rule may move constituents to the right called postposing, or to the
left, called preposing.
63
Wh- movement & SAI can occur separately, and so the application of one isn’t dependent on
the application of the other. SAI occurs alone in Yes-No interrogatives. Wh-Movement
occurs alone in relative clause and indirect question.
Deletion rule(s)
+ Verb Phrase Deletion Yan can cook, and you can, too.
Insertion rule(s)
The preceding examples show that at least some sentences must be analyzed with the help of
two distinct rule systems - phrase structure rules, which determine the internal structure of
phrasal categories, and transformations, which can modify these tree structures by moving an
element from one position to another. If we think about this in terms of the architecture of
sentence structure, the transformational analysis is claiming that there are two levels of
syntactic structure. The first, called deep structure is formed by the phrase structure rules in
accordance with the head's subcategorization properties. As we will see in the chapter on
semantics, deep structure plays a special role in the interpretation of sentences. The second
level of syntactic structure corresponds to the final syntactic form of the sentence. Called
surface structure, it results from applying whatever transformations are appropriate for the
sentence in question.
Deep Structure: generated by the PS Rules in accordance with the head’s subcategorization
properties (i.e. the first underlying structure to which a transformation applies.)
In contrast, the statement pattern The boy will leave has a surface structure (final syntactic
form) that looks just like its deep structure since no transformations apply.
64
The following diagram depicts the organization of the syntactic component of the grammar as
it has just been outlined.
Transformations
SURFACE STRUCTURE
A kernel clause forms a sentence on its own - i.e. it is not part of some larger syntactic
construction. Thus it is neither subordinate to, nor coordinate with, some other clauses. ...: a
kernel clause can accordingly contain a non-kernel clause within it.
Thus in the following pairs, (i) is the kernel clause but (ii) is not
Questions:
CHAPTER 7
S = Subject - Predicate
Or, a sentence can be seen as comprising of FIVE elements which are Subject, Verb, Object,
Complement and Adverbial. (NB: these are element functions in a sentence.)
S = S V O C A
Different grammarians have different points of view in presenting sentences into patterns.
The most common and current view point describes the structure of English sentences into
seven sentence patterns
(1) SVA S V A
Mary is in the house
(2) SVC S V C
Mary is kind
a nurse
(3) SVO S V O.d
Somebody caught the ball
(4) SVOA S V O.d A
I put the plate on the table
(5) SVOC S V O.d C.o
We have proved him wrong
a fool
(6) SVOO S V O.i O.d
She gives me expensive presents
(7) SV S V
The child laughed
With the appearance of I.C grammar and P.S grammar, the structure of English sentences is
described under a system of P.S rules, in which the sentence structure starts from
S NP + VP
(see Phrase Structure rules)
Depending on the degree or valency of the Verb and/or Predicate of a sentence, the noun
phrases (NPs) have been classified in terms of the role they play in that sentence. Below are
the thematic roles common in current linguistics descriptions assigned to NPs.
67
In an active voice sentence, the agent is typically the subject. A useful test for the agent role
is to see whether the noun phrase can follow the verb order with its original verb following it
in its active infinitive form, as in this sentence:
Nhung’s boyfriend ordered her (Nhung) to send a thousand letters to him (her boyfriend).
The instrument is the thing with which the action is done. The NP argument Penicillin is the
instrument in the following example:
In an active voice clause, the instrument is typically either the object of with or the clause
subject.
The role of theme is the hardest to pin down. The broad definition of the role theme covers
three somewhat different classes of theme.
• The first corresponds to a narrow definition. It is the role of the often inert entity, which is
in a certain state or position or is changing its state or position:
• The third class is that of affected mind-possessing entities. … the term patient is
sometimes used instead of the broader term theme. Patients undergo the action or process
specified by their predicate and are affected by it.
In this grammar we have found it convenient pedagogically to use the term theme to cover
all three classes. The three classes of themes have much in common when compared to the
other roles. Moreover, it becomes too easy to confuse the patient role with the next role to be
described, that of experiencer. In an active voice sentence, the theme is typically the subject
of the verb be or the direct object of a verb.
The experiencer is the one who experiences a mental state or process such as thinking,
knowing, believing, understanding, seeing, hearing, fearing, hoping, being surprised, etc.
Note that the experiencer in the first and third sentences occupies the subject slot; in the
second, it occupies the object slot. In an active voice sentence, the experiencer is typically the
subject.
The term source refers to the location from which someone or something originates and the
term goal to the location that serves or should serve as the destination.
The delegates left Mexico City (source) for Buenos Aires (goal).
The government (agent) took over a billion dollars (theme) from the poor (source).
In an active voice sentence, the goal is typically the indirect object or the object of to, and the
source is typically the object of from.
The benefactive is the role of the individual for whose benefit some action is undertaken:
The benefactive noun phrase is often introduced with the preposition for:
In an active voice sentence, the benefactive is typically the indirect object or the object of for.
Noun phrases which are not arguments of predicates also have thematic roles. What can be
said about, the roles of the boldface noun phrases in the following sentences?
Prepositions are similar semantically to predicates. Like many predicates they take object
arguments, but unlike those predicates, they take no subject arguments. The relation of the
verb approach to the hut in the first example phrase following is very like the relation of the
preposition to to the hut in the second:
Questions:
2. Describe/Present the thematic (semantic) roles played by NPs in the above sentences.
70
CHAPTER 8
One of the concepts that is basic to the Western tradition of grammatical analysis is that of
Subject. Since this is a familiar term, let us take it as the starting point for investigating the
functions in an English clause. Consider the clause:
In accordance with the syntactic principles established by medieval grammarians, which were
themselves based on the grammars of ancient Greece and Rome, each clause contains one
element which can be identified as its Subject; and in this instance, the Subject would be the
duke.
Here are some other clauses with the Subject shown in italic:
It is possible to conclude from these examples that 'Subject' is the label for a grammatical
function of some kind. There seems to be something in common, as regards their status in the
clause, to all the elements we have labeled in this way. But it is not so easy to say exactly
what this is; and it is difficult to find in the grammatical tradition a definitive account of what
the role of Subject means. Instead, various interpretations have grown up around the Subject
notion, ascribing to it a number of rather different functions. These resolve themselves into
three broad definitions, which could be summarized as follows:
These three definitions are obviously not synonymous; they are defining different concepts.
(…).What has happened (…) is that different functions making up the traditional concept of
Subject have been split up among three different constituents of the clause.
71
When these different functions came to be recognized by grammarians as distinct, they were
at first labeled as if they were three different kinds of Subject. (…). The terms that came to be
used in the second half of the nineteenth century, when there was a renewal of interest in
grammatical theory, were:
Î1 Psychological subject - Psychological Subject meant ‘that which is the concern of the
message’. It was called ‘psychological’ because it was what the speaker had in his
mind to start with, when embarking on the production of the clause.
Î2 Grammatical subject - Grammatical Subject meant ‘that of which something is
predicated’. It was called ‘grammatical’ because at that time the construction of
Subject and Predicate was thought of as a purely formal grammatical relationship; it
was seen to determine various other grammatical features, such as the case of the noun
or pronoun that was functioning as Subject, and its concord of person and number
with the verb, but it was not thought to express any particular meaning.
Î3 Logical subject - Logical Subject meant ‘doer of the action’. It was called ‘logical’ in
the sense this term had had from the seventeenth century, that of ‘having to do with
relations between things’, as opposed to ‘grammatical’ relations, which were relations
between symbols.
They are not three kinds of anything; they are three quite different things. The earlier labels
will be replaced by separate ones which relate more specifically to the functions concerned:
The roles of Theme, Subject and Actor can be combined in one element
[…]
(a)
my aunt was given this teapot by the duke
Theme Actor
Subject
72
(b)
this teapot the duke gave to my aunt
Theme Subject
Actor
c)
by the duke my aunt was given this teapot.
Theme Subject
Actor
(i) The Theme functions in the structure of the CLAUSE AS A MESSAGE. A clause has
meaning as a message, a quantum of information; the Theme is the point of departure for the
message. It is the element the speaker selects for ‘grounding’ what he is going on to say.
(ii) The Subject functions in the structure of the CLAUSE AS AN EXCHANGE. A clause has
meaning as an exchange, a transaction between speaker and listener; the Subject is the
warranty of the exchange. It is the element the speaker makes responsible for the validity of
what he is saying.
Theme Rheme
As a general guide, the Theme can be identified as that element which comes in first position
in the clause. We have already indicated that this is not how the category of Theme is defined.
The definition is functional, as it is with all the elements in this interpretation of grammatical
structure. The Theme is one element in a particular structural configuration which, taken as a
whole, organizes the clause as a message; this is the configuration Theme + Rheme. A
message consists of a Theme combined with a Rheme.
The Theme is not necessarily a NOMINAL GROUP, like those above. It may also be an
ADVERBIAL GROUP or PREPOSITIONAL PHRASE.
As a first step we have made two assumptions: that the Theme of a clause consists of just one
structural element, and that that element is represented by just one unit - one nominal
group, adverbial group or prepositional phrase.
1. Thematic Equative
Theme = Rheme
exclamative negative
independent
clause Interrogative yes - no Q?
Wh- Q?
Imperative
1. Clauses as Themes
Theme Rheme
Theme Rheme
Theme Rheme
so why worry
Theme Rheme
76
Theme 1 Rheme 1
Theme Rheme
2. Predicated Themes
Theme Rheme
Theme Rheme
Theme Rheme
Theme Rheme
77
Kiss me.
Pass the salt, please.
Is today Saturday?
When language is used to exchange information, the clause takes on the form of a
proposition, which is the semantic function of the clause. This refers to the speech function
of questions and statements.
I. MOOD
1) The presence of the Mood element, consisting of Subject plus Finite, realizes the feature
‘indicative’.
2) Within the indicative, what is significant is the order of Subject and finite:
In some cases, the Finite element and the lexical verb are fused into a single word:
past give
• Temporal operators
• Modal operators
- 1 Predicator
- 1 or 2 Complements
- up to 7 Adjuncts
• Predicator
1 It specifies time reference other than reference to the time of the speech event.
2 It specifies other aspects and phases like seeming, hoping, trying, …
3 It specifies the voices: active or passive
4 It specifies the process that is predicated of the subject (action, event, mental process,
relation)
• Complement
A complement is an element within the Residue that has the potential of being Subject but is
not. It is typically realized by a nominal group.
80
• Adjunct
An Adjunct is an element that has not got the potential of being subject. It is typically
realized by an adverbial group or a prepositional phrase.
• Wh- interrogative
• Exclamatives
CLAUSE AS REPRESENTATION
Usually when people talk about what a word or a sentence ‘means’, it is this kind of meaning
they have in mind - meaning in the sense of content.
In section 2 we were looking at the clause from the point of view of its interpersonal function,
the part it plays as a form of exchange between speaker and listener. In this section, by
contrast, we shall be concerned with the clause in its experiential function, its guise as a way
of representing patterns of experience. Language enables human beings to build a mental
picture of reality, to make sense of what goes on around them and inside them. Here again the
clause plays a central role, because it embodies a general principle for modeling experience -
namely, the principle that reality is made up of processes.
What is the status of a process, as set up in the grammar of the clause? A Process consists, in
principle, of 3 components
1. Process
What are the different types of process, as construed by the transitivity system in the
grammar?
• Material process
• Mental process
• Relational process
• Behavioral process
• Verbal process
• Existential
2. Circumstantial elements
Above the foot there is one higher constituent, which is the melodic unit of the language.
This is generally referred to as the TONE GROUP, i.e. a number of feet compose of a higher
constituent unit referred to as the TONE GROUP.
The tone group functions as the realization of a unit of information in the discourse. Spoken
discourse takes the form of a sequence of INFORMATION UNITS, one following the other
in unbroken succession with nopause or discontinuity between them.
The information unit is a process of interaction between what is already known or predictable
or what is new or unpredictable. …
Hence, the information unit is a structure made up of two functions, the New and the Given.
Each Information Unit is realized as a PITCH CONTOUR, or TONE, which may be falling,
rising or mixed.
Within the Tone Group, one foot carries the main pitch movement … This feature is known as
TONIC PROMINENCE.
The element that have this Tonic prominence is said to be carrying INFORMATION FOCUS
The typical sequence of informational elements is thus Given followed by the New. But
whereas the end of the New element is marked by Tonic Prominence, there is nothing to mark
where it begins …
I’ll tell you what silver needs to have. It needs to have love.
… a speaker will choose the Theme from within what is Given and locate the focus, the
climax of the New, somewhere within the Rheme.
But although they are related, Given + New and Theme + Rheme are not the same thing …
Theme + Rheme is speaker-oriented, while Given + New is listener-oriented.
Questions:
1. How did people define SUBJECT? How did they classify it?
2. How is SUBJECT defined in terms of Functional Grammar?
3. In which clause do THEME, SUBJECT and ACTOR function?
CLAUSE AS MESSAGE
4. How is THEME identified in a clause (structure)?
5. Which word group can play the role of THEME?
6. How is THEME related to MOOD?
CLAUSE AS EXCHANGE
7. What are the elements of the clause? What is its grammatical system?
8. What are the elements of ‘the MOOD’?
9. What are the elements of the RESIDUE?
10. Typical types of MOOD and RESIDUE.
CLAUSE AS REPRESENTATION
11. What are the structural elements of the clause as representation?
12. What are the types of process?
13. What are the types of circumstances?
86
CHAPTER 9
1 Polar Question
Q S’
COMP S
I NP PRED. P
(+ Finite) N’ I VP
(Present) N V NP
Q S’
COMP S
NP PRED. P
I N’ I VP
+ Finite N V PP
Present
Aux
2 Wh- Question
Q S’
NP COMP S
N’ N’ I NP PRED.P
(+Finite)
Modal
N N N’ I VP
V NP
Q S’
N’ COMP S
N I NP PRED. P
(+Finite)
N I VP
V NP AdvP PP
Adv P NP
N N
3 Complex sentences
3.1 Complex sentences - Embedded Clause
NP VP
DET N’ V S’
N COM S
NP VP
(V-prt)
N N
NP VP
Det N’ V PP
N S’ P NP
COMP S N N
NP VP
V PP
P N’
DET NP
N N
(Φ)
S’
COMP S
NP VP
V NP S’
N’ DET N’ COMP S
NP VP
N N N’ V NP
N DET N’
N N
NP VP
N V S’
COMP S
NP PRED.P
DET N’ I VP
N - Finite V
S’1 S’2
Comp S Comp S
NP VP Conj NP VP
N’ V NP N’ V NP
DET N’ DET N’
N N Pro N
S’1 S’2
COMP S COM S
Conj NP VP Conj NP VP
V Adj V NP
Bibliography
Delahunty G.P. and Garvey, J.J. (1994). Language, Grammar, and Communication.
McGraw-Hill, Inc, New York.
Jacobs, R.A. (1995). English Syntax - A Grammar for English Language Professionals.
Oxford University Press, New York.
Richards, J.C. et al. (1993) Dictionary of Language Teaching and Applied Linguistics.
Longman Group UK Limited, England.
114